Ibm total storage productivity center v2.3 getting started sg246490

Front cover

IBM TotalStorage
Productivity Center V2.3:
Getting Started
Effectively use the IBM TotalStorage
Productivity Center

Learn to install and customize the IBM
TotalStorage Productivity Center

Understand the IBM TotalStorage
Open Software Family

Mary Lovelace
Larry Mc Gimsey
Ivo Gomilsek
Mary Anne Marquez

ibm.com/redbooks

International Technical Support Organization

IBM TotalStorage Productivity Center V2.3:
Getting Started

December 2005

SG24-6490-01

Note: Before using this information and the product it supports, read the information in “Notices” on
page xiii.

Second Edition (December 2005)

This edition applies to Version 2, Release 3 of IBM TotalStorage Productivity Center (product number
5608-UC1, 5608-UC3, 5608-UC4, 5608-UC5.

© Copyright International Business Machines Corporation 2005. All rights reserved.
Note to U.S. Government Users Restricted Rights -- Use, duplication or disclosure restricted by GSA ADP Schedule
Contract with IBM Corp.

Contents

Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv
The team that wrote this redbook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv
Become a published author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii
Comments welcome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii

Part 1. IBM TotalStorage Productivity Center foundation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Chapter 1. IBM TotalStorage Productivity Center overview . . . . . . . . . . . . . . . . . . . . . . 3
1.1 Introduction to IBM TotalStorage Productivity Center . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.1 Standards organizations and standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 IBM TotalStorage Open Software family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 IBM TotalStorage Productivity Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3.1 Data subject matter expert: TotalStorage Productivity Center for Data . . . . . . . . . 7
1.3.2 Fabric subject matter expert: Productivity Center for Fabric . . . . . . . . . . . . . . . . . . 9
1.3.3 Disk subject matter expert: TotalStorage Productivity Center for Disk . . . . . . . . . 12
1.3.4 Replication subject matter expert: Productivity Center for Replication . . . . . . . . . 14
1.4 IBM TotalStorage Productivity Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.4.1 Productivity Center for Disk and Productivity Center for Replication . . . . . . . . . . 17
1.4.2 Event services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.5 Taking steps toward an On Demand environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Chapter 2. Key concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.1 IBM TotalStorage Productivity Center architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.1.1 Architectural overview diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.1.2 Architectural layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.1.3 Relationships between the managers and components . . . . . . . . . . . . . . . . . . . . 31
2.1.4 Collecting data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.2 Standards used in IBM TotalStorage Productivity Center . . . . . . . . . . . . . . . . . . . . . . . 34
2.2.1 ANSI standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.2.2 Web-Based Enterprise Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.2.3 Storage Networking Industry Association . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.2.4 Simple Network Management Protocol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.2.5 Fibre Alliance MIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.3 Service Location Protocol (SLP) overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.3.1 SLP architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.3.2 Common Information Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.4 Component interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.4.1 CIMOM discovery with SLP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.4.2 How CIM Agent works. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.5 Tivoli Common Agent Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2.5.1 Tivoli Agent Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.5.2 Common Agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

Part 2. Installing the IBM TotalStorage Productivity Center base product suite . . . . . . . . . . . . . . . . . 55

Chapter 3. Installation planning and considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . 57

© Copyright IBM Corp. 2005. All rights reserved. iii

3.1 Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.2 Installation prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.2.1 TCP/IP ports used by TotalStorage Productivity Center . . . . . . . . . . . . . . . . . . . . 59
3.2.2 Default databases created during the installation . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.3 Our lab setup environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.4 Pre-installation check list. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.5 User IDs and security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.5.1 User IDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.5.2 Increasing user security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.5.3 Certificates and key files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.5.4 Services and service accounts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.6 Starting and stopping the managers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.7 Windows Management Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.8 World Wide Web Publishing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.9 Uninstalling Internet Information Services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.10 Installing SNMP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.11 IBM TotalStorage Productivity Center for Fabric. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.11.1 The computer name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.11.2 Database considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.11.3 Windows Terminal Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.11.4 Tivoli NetView . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.11.5 Personal firewall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.11.6 Changing the HOSTS file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.12 IBM TotalStorage Productivity Center for Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.12.1 Server recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.12.2 Supported subsystems and databases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.12.3 Security considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
3.12.4 Creating the DB2 database. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Chapter 4. Installing the IBM TotalStorage Productivity Center suite . . . . . . . . . . . . . 83
4.1 Installing the IBM TotalStorage Productivity Center . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.1.1 Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.2 Prerequisite Software Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.2.1 Best practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.2.2 Installing prerequisite software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.3 Suite installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
4.3.1 Best practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
4.3.2 Installing the TotalStorage Productivity Center suite . . . . . . . . . . . . . . . . . . . . . 110
4.3.3 IBM TotalStorage Productivity Center for Disk and Replication Base. . . . . . . . . 125
4.3.4 IBM TotalStorage Productivity Center for Disk . . . . . . . . . . . . . . . . . . . . . . . . . . 140
4.3.5 IBM TotalStorage Productivity Center for Replication. . . . . . . . . . . . . . . . . . . . . 146
4.3.6 IBM TotalStorage Productivity Center for Fabric. . . . . . . . . . . . . . . . . . . . . . . . . 157
4.3.7 IBM TotalStorage Productivity Center for Data . . . . . . . . . . . . . . . . . . . . . . . . . . 171

Chapter 5. CIMOM install and configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
5.2 Planning considerations for Service Location Protocol . . . . . . . . . . . . . . . . . . . . . . . . 192
5.2.1 Considerations for using SLP DAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
5.2.2 SLP configuration recommendation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
5.3 General performance guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
5.4 Planning considerations for CIMOM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
5.4.1 CIMOM configuration recommendations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
5.5 Installing CIM agent for ESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

iv IBM TotalStorage Productivity Center V2.3: Getting Started

5.5.1 ESS CLI Install . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
5.5.2 DS CIM Agent install . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
5.5.3 Post Installation tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
5.6 Configuring the DS CIM Agent for Windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
5.6.1 Registering DS Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
5.6.2 Registering ESS Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
5.6.3 Register ESS server for Copy services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
5.6.4 Restart the CIMOM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
5.6.5 CIMOM user authentication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
5.7 Verifying connection to the ESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
5.7.1 Problem determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
5.7.2 Confirming the ESS CIMOM is available . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
5.7.3 Setting up the Service Location Protocol Directory Agent . . . . . . . . . . . . . . . . . 221
5.7.4 Configuring TotalStorage Productivity Center for SLP discovery . . . . . . . . . . . . 223
5.7.5 Registering the DS CIM Agent to SLP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
5.7.6 Verifying and managing CIMOM’s availability. . . . . . . . . . . . . . . . . . . . . . . . . . . 224
5.8 Installing CIM agent for IBM DS4000 family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
5.8.1 Verifying and Managing CIMOM availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
5.9 Configuring CIMOM for SAN Volume Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
5.9.1 Adding the SVC TotalStorage Productivity Center for Disk user account. . . . . . 235
5.9.2 Registering the SAN Volume Controller host in SLP . . . . . . . . . . . . . . . . . . . . . 241
5.10 Configuring CIMOM for TotalStorage Productivity Center for Disk summary . . . . . . 241
5.10.1 SLP registration and slptool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
5.10.2 Persistency of SLP registration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
5.10.3 Configuring slp.reg file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243

Part 3. Configuring the IBM TotalStorage Productivity Center. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245

Chapter 6. Configuring IBM TotalStorage Productivity Center for Disk. . . . . . . . . . . 247
6.1 Productivity Center for Disk Discovery summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
6.2 SLP DA definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
6.2.1 Verifying and managing CIMOMs availability . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
6.3 Disk and Replication Manager remote GUI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
6.3.1 Installing Remote Console for Performance Manager function. . . . . . . . . . . . . . 270
6.3.2 Launching Remote Console for TotalStorage Productivity Center . . . . . . . . . . . 277

Chapter 7. Configuring TotalStorage Productivity Center for Replication . . . . . . . . 279
7.1 Installing a remote GUI and CLI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280

Chapter 8. Configuring IBM TotalStorage Productivity Center for Data . . . . . . . . . . 289
8.1 Configuring the CIM Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
8.1.1 CIM and SLP interfaces within Data Manager . . . . . . . . . . . . . . . . . . . . . . . . . . 290
8.1.2 Configuring CIM Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
8.1.3 Setting up a disk alias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
8.2 Setting up the Web GUI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
8.2.1 Using IBM HTTP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
8.2.2 Using Internet Information Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
8.2.3 Configuring the URL in Fabric Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
8.3 Installing the Data Manager remote console. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
8.4 Configuring Data Manager for Databases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
8.5 Alert Disposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316

Chapter 9. Configuring IBM TotalStorage Productivity Center for Fabric . . . . . . . . . 319
9.1 TotalStorage Productivity Center component interaction . . . . . . . . . . . . . . . . . . . . . . 320

Contents v

9.1.1 IBM TotalStorage Productivity Center for Disk and Replication Base. . . . . . . . . 320
9.1.2 SNMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
9.1.3 Tivoli Provisioning Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
9.2 Post-installation procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
9.2.1 Installing Productivity Center for Fabric – Agent . . . . . . . . . . . . . . . . . . . . . . . . . 321
9.2.2 Installing Productivity Center for Fabric – Remote Console . . . . . . . . . . . . . . . . 331
9.3 Configuring IBM TotalStorage Productivity Center for Fabric . . . . . . . . . . . . . . . . . . . 342
9.3.1 Configuring SNMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
9.3.2 Configuring the outband agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
9.3.3 Checking inband agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
9.3.4 Performing an initial poll and setting up the poll interval . . . . . . . . . . . . . . . . . . . 349

Chapter 10. Deployment of agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
10.1 Installing the agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
10.2 Data Agent installation using the installer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
10.3 Deploying the agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361

Part 4. Using the IBM TotalStorage Productivity Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373

Chapter 11. Using TotalStorage Productivity Center for Disk. . . . . . . . . . . . . . . . . . . 375
11.1 Productivity Center common base: Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
11.2 Launching TotalStorage Productivity Center. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
11.3 Exploiting Productivity Center common base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
11.3.1 Launch Device Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
11.4 Performing volume inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
11.5 Changing the display name of a storage device . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
11.6 Working with ESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
11.6.1 ESS Volume inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
11.6.2 Assigning and unassigning ESS Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
11.6.3 Creating new ESS volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
11.6.4 Launch device manager for an ESS device . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
11.7 Working with DS8000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
11.7.1 DS8000 Volume inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
11.7.2 Assigning and unassigning DS8000 Volumes . . . . . . . . . . . . . . . . . . . . . . . . . 392
11.7.3 Creating new DS8000 volumes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
11.7.4 Launch device manager for an DS8000 device . . . . . . . . . . . . . . . . . . . . . . . . 394
11.8 Working with SAN Volume Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
11.8.1 Working with SAN Volume Controller MDisks. . . . . . . . . . . . . . . . . . . . . . . . . . 396
11.8.2 Creating new MDisks on supported storage devices . . . . . . . . . . . . . . . . . . . . 399
11.8.3 Create and view SAN Volume Controller VDisks . . . . . . . . . . . . . . . . . . . . . . . 402
11.9 Working with DS4000 family or FAStT storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
11.9.1 Working with DS4000 or FAStT volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
11.9.2 Creating DS4000 or FAStT volumes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
11.9.3 Assigning hosts to DS4000 and FAStT Volumes . . . . . . . . . . . . . . . . . . . . . . . 413
11.9.4 Unassigning hosts from DS4000 or FAStT volumes. . . . . . . . . . . . . . . . . . . . . 414
11.9.5 Volume properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
11.10 Event Action Plan Builder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
11.10.1 Applying an Event Action Plan to a managed system or group . . . . . . . . . . . 421
11.10.2 Exporting and importing Event Action Plans . . . . . . . . . . . . . . . . . . . . . . . . . . 423

Chapter 12. Using TotalStorage Productivity Center Performance Manager . . . . . . 427
12.1 Exploiting Performance Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
12.1.1 Performance Manager GUI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
12.1.2 Performance Manager data collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429

vi IBM TotalStorage Productivity Center V2.3: Getting Started

12.1.3 Using IBM Director Scheduler function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
12.1.4 Reviewing data collection task status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
12.1.5 Managing Performance Manager Database . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
12.1.6 Performance Manager gauges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
12.1.7 ESS thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
12.1.8 Data collection for SAN Volume Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
12.1.9 SAN Volume Controller thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461
12.1.10 Data collection for the DS6000 and DS8000 . . . . . . . . . . . . . . . . . . . . . . . . . 463
12.1.11 DS6000 and DS8000 thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
12.2 Exploiting gauges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
12.2.1 Before you begin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
12.2.2 Creating gauges: an example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
12.2.3 Zooming in on the specific time period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
12.2.4 Modify gauge to view array level metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
12.2.5 Modify gauge to review multiple metrics in same chart. . . . . . . . . . . . . . . . . . . 474
12.3 Performance Manager command line interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
12.3.1 Performance Manager CLI commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
12.3.2 Sample command outputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
12.4 Volume Performance Advisor (VPA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
12.4.1 VPA introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
12.4.2 The provisioning challenge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
12.4.3 Workload characterization and workload profiles . . . . . . . . . . . . . . . . . . . . . . . 479
12.4.4 Workload profile values. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479
12.4.5 How the Volume Performance Advisor makes decisions . . . . . . . . . . . . . . . . . 480
12.4.6 Enabling the Trace Logging for Director GUI Interface . . . . . . . . . . . . . . . . . . . 481
12.4.7 Getting started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482
12.4.8 Creating and managing workload profiles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508

Chapter 13. Using TotalStorage Productivity Center for Data . . . . . . . . . . . . . . . . . . 521
13.1 TotalStorage Productivity Center for Data overview . . . . . . . . . . . . . . . . . . . . . . . . . 522
13.1.1 Business purpose of TotalStorage Productivity Center for Data. . . . . . . . . . . . 522
13.1.2 Components of TotalStorage Productivity Center for Data . . . . . . . . . . . . . . . . 522
13.1.3 Security considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523
13.2 Functions of TotalStorage Productivity Center for Data . . . . . . . . . . . . . . . . . . . . . . 523
13.2.1 Basic menu displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524
13.2.2 Discover and monitor Agents, disks, filesystems, and databases . . . . . . . . . . 526
13.2.3 Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529
13.2.4 Alerts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532
13.2.5 Chargeback: Charging for storage usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
13.3 OS Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
13.3.1 Navigation tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
13.3.2 Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535
13.3.3 Discovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
13.3.4 Pings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
13.3.5 Probes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
13.3.6 Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 547
13.3.7 Scans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
13.4 OS Alerts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
13.4.1 Alerting navigation tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 558
13.4.2 Computer Alerts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560
13.4.3 Filesystem Alerts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 562
13.4.4 Directory Alerts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
13.4.5 Alert logs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564

Contents vii

13.5 Policy management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
13.5.1 Quotas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
13.5.2 Network Appliance Quotas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
13.5.3 Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
13.5.4 Filesystem extension and LUN provisioning . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
13.5.5 Scheduled Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582
13.6 Database monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583
13.6.1 Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584
13.6.2 Probes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
13.6.3 Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 586
13.6.4 Scans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
13.7 Database Alerts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588
13.7.1 Instance Alerts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588
13.7.2 Database-Tablespace Alerts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588
13.7.3 Table Alerts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
13.7.4 Alert log . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
13.8 Databases policy management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
13.8.1 Network Quotas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590
13.8.2 Instance Quota . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
13.8.3 Database Quota . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
13.9 Database administration samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
13.9.1 Database up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
13.9.2 Database utilization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
13.9.3 Need for reorganization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
13.10 Data Manager reporting capabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
13.10.1 Major reporting categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
13.11 Using the standard reporting functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
13.11.1 Asset Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
13.11.2 Storage Subsystems Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
13.11.3 Availability Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
13.11.4 Capacity Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
13.11.5 Usage Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607
13.11.6 Usage Violation Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
13.11.7 Backup Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 627
13.12 TotalStorage Productivity Center for Data ESS Reporting . . . . . . . . . . . . . . . . . . . 634
13.12.1 ESS Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 634
13.13 IBM Tivoli Storage Resource Manager top 10 reports . . . . . . . . . . . . . . . . . . . . . . 653
13.13.1 ESS used and free storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
13.13.2 ESS attached hosts report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
13.13.3 Computer Uptime Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
13.13.4 Growth in storage used and number of files . . . . . . . . . . . . . . . . . . . . . . . . . . 659
13.13.5 Incremental backup trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
13.13.6 Database reports against DBMS size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
13.13.7 Database instance storage report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
13.13.8 Database reports size by instance and by computer . . . . . . . . . . . . . . . . . . . 667
13.13.9 Locate the LUN on which a database is allocated . . . . . . . . . . . . . . . . . . . . . 669
13.13.10 Finding important files on your systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 672
13.14 Creating customized reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683
13.14.1 System Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683
13.14.2 Reports owned by a specific username . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 686
13.14.3 Batch Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 688
13.15 Setting up a schedule for daily reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 697
13.16 Setting up a reports Web site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 698

viii IBM TotalStorage Productivity Center V2.3: Getting Started

13.17 Charging for storage usage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 700

Chapter 14. Using TotalStorage Productivity Center for Fabric . . . . . . . . . . . . . . . . . 703
14.1 NetView navigation overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704
14.1.1 NetView interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704
14.1.2 Maps and submaps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704
14.1.3 NetView window structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704
14.1.4 NetView Explorer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 705
14.1.5 NetView Navigation Tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
14.1.6 Object selection and NetView properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
14.1.7 Object symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709
14.1.8 Object status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709
14.1.9 Status propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 711
14.1.10 NetView and Productivity Center for Fabric integration . . . . . . . . . . . . . . . . . 711
14.2 Walk-through of Productivity Center for Fabric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712
14.2.1 Device Centric view. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713
14.2.2 Host Centric view . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714
14.2.3 SAN view . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714
14.2.4 Launching element managers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
14.2.5 Explore view . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
14.3 Topology views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
14.3.1 SAN view . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 727
14.3.2 Device Centric View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731
14.3.3 Host Centric View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732
14.3.4 iSCSI discovery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733
14.3.5 MDS 9000 discovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734
14.4 SAN menu options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 735
14.4.1 SAN Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 735
14.5 Application launch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739
14.5.1 Native support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 740
14.5.2 NetView support for Web interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 740
14.5.3 Launching TotalStorage Productivity Center for Data. . . . . . . . . . . . . . . . . . . . 742
14.5.4 Other menu options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 742
14.6 Status cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 743
14.7 Practical cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745
14.7.1 Cisco MDS 9000 discovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745
14.7.2 Removing a connection on a device running an inband agent . . . . . . . . . . . . . 747
14.7.3 Removing a connection on a device not running an agent . . . . . . . . . . . . . . . . 750
14.7.4 Powering off a switch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
14.7.5 Running discovery on a RNID-compatible device. . . . . . . . . . . . . . . . . . . . . . . 756
14.7.6 Outband agents only . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758
14.7.7 Inband agents only . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760
14.7.8 Disk devices discovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762
14.7.9 Well placed agent strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764
14.8 Netview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 766
14.8.1 Reporting overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
14.8.2 SNMP and MIBs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
14.9 NetView setup and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769
14.9.1 Advanced Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769
14.9.2 Copy Brocade MIBs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 770
14.9.3 Loading MIBs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
14.10 Historical reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
14.10.1 Creating a Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775

Contents ix

14.10.2 Database maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
14.10.3 Troubleshooting the Data Collection daemon . . . . . . . . . . . . . . . . . . . . . . . . . 784
14.10.4 NetView Graph Utility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784
14.11 Real-time reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786
14.11.1 MIB Tool Builder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787
14.11.2 Displaying real-time data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791
14.11.3 SmartSets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794
14.11.4 SmartSets and Data Collections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 802
14.11.5 Seed file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 805
14.12 Productivity Center for Fabric and iSCSI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810
14.13 What is iSCSI? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811
14.14 How does iSCSI work? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811
14.15 Productivity Center for Fabric and iSCSI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812
14.15.1 Functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813
14.15.2 iSCSI discovery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813
14.16 ED/FI - SAN Error Predictor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
14.16.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
14.16.2 Error processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816
14.16.3 Configuration for ED/FI - SAN Error Predictor . . . . . . . . . . . . . . . . . . . . . . . . 818
14.16.4 Using ED/FI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 820
14.16.5 Searching for the faulted device on the topology map . . . . . . . . . . . . . . . . . . 822
14.16.6 Removing notifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825

Chapter 15. Using TotalStorage Productivity Center for Replication. . . . . . . . . . . . . 827
15.1 TotalStorage Productivity Center for Replication overview . . . . . . . . . . . . . . . . . . . . 828
15.1.1 Supported Copy Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 828
15.1.2 Replication session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 830
15.1.3 Storage group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 831
15.1.4 Storage pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 831
15.1.5 Relationship of group, pool, and session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832
15.1.6 Copyset and sequence concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 833
15.2 Exploiting Productivity Center for replication. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834
15.2.1 Before you start. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834
15.2.2 Adding a replication device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834
15.2.3 Creating a storage group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 838
15.2.4 Modifying a storage group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
15.2.5 Viewing storage group properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842
15.2.6 Deleting a storage group. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843
15.2.7 Creating a storage pool. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844
15.2.8 Modifying a storage pool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 847
15.2.9 Deleting a storage pool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 848
15.2.10 Viewing storage pool properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849
15.2.11 Creating storage paths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 850
15.2.12 Point-in-Time Copy - creating a session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 852
15.2.13 Creating a session - verifying source-target relationship . . . . . . . . . . . . . . . . 856
15.2.14 Continuous Synchronous Remote Copy - creating a session. . . . . . . . . . . . . 861
15.2.15 Managing a Point-in-Time copy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 866
15.2.16 Managing a Continuous Synchronous Remote Copy . . . . . . . . . . . . . . . . . . . 873
15.3 Using Command Line Interface (CLI) for replication . . . . . . . . . . . . . . . . . . . . . . . . . 884
15.3.1 Session details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
15.3.2 Starting a session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 888
15.3.3 Suspending a session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892
15.3.4 Terminating a session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893

x IBM TotalStorage Productivity Center V2.3: Getting Started

Chapter 16. Hints, tips, and good-to-knows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899
16.1 SLP configuration recommendation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900
16.1.1 SLP registration and slptool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901
16.2 Tivoli Common Agent Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901
16.2.1 Locations of configured user IDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901
16.2.2 Resource Manager registration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902
16.2.3 Tivoli Agent Manager status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902
16.2.4 Registered Fabric Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904
16.2.5 Registered Data Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 906
16.3 Launchpad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 906
16.3.1 Launchpad installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 907
16.3.2 Launchpad customization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 909
16.4 Remote consoles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 911
16.5 Verifying whether a port is in use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 911
16.6 Manually removing old CIMOM entries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 911
16.7 Collecting logs for support. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 917
16.7.1 IBM Director logfiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 917
16.7.2 Using Event Action Plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 921
16.7.3 Following Discovery using Windows raswatch utility . . . . . . . . . . . . . . . . . . . . 921
16.7.4 DB2 database checking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 922
16.7.5 IBM WebSphere tracing and logfile browsing . . . . . . . . . . . . . . . . . . . . . . . . . . 927
16.8 SLP and CIM Agent problem determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 928
16.8.1 Enabling SLP tracing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 929
16.8.2 Device registration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930
16.9 Replication Manager problem determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930
16.9.1 Diagnosing an indications problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 931
16.9.2 Restarting the replication environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 931
16.10 Enabling trace logging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 931
16.10.1 Enabling WebSphere Application Server trace . . . . . . . . . . . . . . . . . . . . . . . . 932
16.11 ESS user authentication problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940
16.12 SVC Data collection task failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940

Chapter 17. Database management and reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . 943
17.1 DB2 database overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 944
17.2 Database purging in TotalStorage Productivity Center . . . . . . . . . . . . . . . . . . . . . . . 944
17.2.1 Performance manager database panel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945
17.3 IBM DB2 tool suite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948
17.3.1 Command Line Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948
17.3.2 Development Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 950
17.3.3 General Administration Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 950
17.3.4 Monitoring Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 951
17.4 DB2 Command Center overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 952
17.4.1 Command Center navigation example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 952
17.5 DB2 Command Center custom report example . . . . . . . . . . . . . . . . . . . . . . . . . . . . 956
17.5.1 Extracting LUN data report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 956
17.5.2 Command Center report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 959
17.6 Exporting collected performance data to a file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 976
17.6.1 Control Center. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 976
17.6.2 Data extraction tools, tips and reporting methods. . . . . . . . . . . . . . . . . . . . . . . 979
17.7 Database backup and recovery overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 984
17.8 Backup example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 988

Appendix A. Worksheets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 991
User IDs and passwords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 992

Contents xi

Server information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 992
User IDs and passwords for key files and installation. . . . . . . . . . . . . . . . . . . . . . . . . . 993
Storage device information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 994
IBM TotalStorage Enterprise Storage Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 994
IBM FAStT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 995
IBM SAN Volume Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996

Related publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 997
IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 997
Other publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 997
Online resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 997
How to get IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 998
Help from IBM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 998

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 999

xii IBM TotalStorage Productivity Center V2.3: Getting Started

Notices

This information was developed for products and services offered in the U.S.A.

IBM may not offer the products, services, or features discussed in this document in other countries. Consult
your local IBM representative for information on the products and services currently available in your area. Any
reference to an IBM product, program, or service is not intended to state or imply that only that IBM product,
program, or service may be used. Any functionally equivalent product, program, or service that does not
infringe any IBM intellectual property right may be used instead. However, it is the user's responsibility to
evaluate and verify the operation of any non-IBM product, program, or service.

IBM may have patents or pending patent applications covering subject matter described in this document. The
furnishing of this document does not give you any license to these patents. You can send license inquiries, in
writing, to:
IBM Director of Licensing, IBM Corporation, North Castle Drive Armonk, NY 10504-1785 U.S.A.

The following paragraph does not apply to the United Kingdom or any other country where such provisions are
inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION PROVIDES THIS
PUBLICATION "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED,
INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF NON-INFRINGEMENT,
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer of
express or implied warranties in certain transactions, therefore, this statement may not apply to you.

This information could include technical inaccuracies or typographical errors. Changes are periodically made
to the information herein; these changes will be incorporated in new editions of the publication. IBM may make
improvements and/or changes in the product(s) and/or the program(s) described in this publication at any time
without notice.

Any references in this information to non-IBM Web sites are provided for convenience only and do not in any
manner serve as an endorsement of those Web sites. The materials at those Web sites are not part of the
materials for this IBM product and use of those Web sites is at your own risk.

IBM may use or distribute any of the information you supply in any way it believes appropriate without incurring
any obligation to you.

Information concerning non-IBM products was obtained from the suppliers of those products, their published
announcements or other publicly available sources. IBM has not tested those products and cannot confirm the
accuracy of performance, compatibility or any other claims related to non-IBM products. Questions on the
capabilities of non-IBM products should be addressed to the suppliers of those products.

This information contains examples of data and reports used in daily business operations. To illustrate them
as completely as possible, the examples include the names of individuals, companies, brands, and products.
All of these names are fictitious and any similarity to the names and addresses used by an actual business
enterprise is entirely coincidental.

COPYRIGHT LICENSE:
This information contains sample application programs in source language, which illustrates programming
techniques on various operating platforms. You may copy, modify, and distribute these sample programs in
any form without payment to IBM, for the purposes of developing, using, marketing or distributing application
programs conforming to the application programming interface for the operating platform for which the sample
programs are written. These examples have not been thoroughly tested under all conditions. IBM, therefore,
cannot guarantee or imply reliability, serviceability, or function of these programs. You may copy, modify, and
distribute these sample programs in any form without payment to IBM for the purposes of developing, using,
marketing, or distributing application programs conforming to IBM's application programming interfaces.

© Copyright IBM Corp. 2005. All rights reserved. xiii

Trademarks
The following terms are trademarks of the International Business Machines Corporation in the United States,
other countries, or both:
AIX® iSeries™ Sequent®
Cloudscape™ MVS™ ThinkPad®
DB2® Netfinity® Tivoli Enterprise™
DB2 Universal Database™ NetView® Tivoli Enterprise Console®
e-business on demand™ OS/390® Tivoli®
Enterprise Storage Server® Predictive Failure Analysis® TotalStorage®
Eserver® pSeries® WebSphere®
Eserver® QMF™ xSeries®
FlashCopy® Redbooks™ z/OS®
IBM® Redbooks (logo) ™ zSeries®
ibm.com® S/390® 1-2-3®

The following terms are trademarks of other companies:

Java and all Java-based trademarks and logos are trademarks or registered trademarks of Sun Microsystems,
Inc. in the United States, other countries, or both.

Microsoft, Windows, Windows NT, and the Windows logo are trademarks of Microsoft Corporation in the
United States, other countries, or both.

Intel, Intel Inside (logos), MMX, and Pentium are trademarks of Intel Corporation in the United States, other
countries, or both.

UNIX is a registered trademark of The Open Group in the United States and other countries.

Linux is a trademark of Linus Torvalds in the United States, other countries, or both.

Other company, product, and service names may be trademarks or service marks of others.

xiv IBM TotalStorage Productivity Center V2.3: Getting Started

Preface

IBM® TotalStorage® Productivity Center is a suite of infrastructure management software
that can centralize, automate, and simplify the management of complex and heterogeneous
storage environments. It can help reduce the effort of managing complex storage
infrastructures, improve storage capacity utilization, and improve administration efficiency.
IBM TotalStorage Productivity Center allows you to respond to on demand storage needs and
brings together, in a single point, the management of storage devices, fabric, and data.

This IBM Redbook is intended for administrators and users who are installing and using IBM
TotalStorage Productivity Center V2.3. It provides an overview of the product components
and functions. We describe the hardware and software environment required, provide a
step-by-step installation procedure, and offer customization and usage hints and tips.

This book is not a replacement for the existing IBM Redbooks™, or product manuals, that
detail the implementation and configuration of the individual products that make up the IBM
TotalStorage Productivity Center, or the products as they may have been called in previous
versions. We refer to those books as appropriate throughout this book.

The team that wrote this redbook
This redbook was produced by a team of specialists from around the world working at the
International Technical Support Organization (ITSO), San Jose Center.

Mary Lovelace is a Consulting IT Specialist at the ITSO in San Jose, California. She has
more than 20 years of experience with IBM in large systems, storage and Storage Networking
product education, system engineering and consultancy, and systems support.

Larry Mc Gimsey is a consulting IT Architect working in Managed Storage Services delivery
supporting worldwide SAN storage customers. He has over 30 years experience in IT. He
joined IBM 6 years ago as a result of an outsourcing engagement. Most of his experience
prior to joining IBM was in mainframe systems support. It included system programming,
performance management, capacity planning, system automation and storage management.
Since joining IBM, Larry has been working with large SAN environments. He currently works
with Managed Storage Services offering and delivery teams to define the architecture used to
deliver worldwide storage services.

Ivo Gomilsek is an IT Specialist for IBM Global Services, Slovenia, supporting the Central
and Eastern European Region in architecting, deploying, and supporting SAN/storage/DR
solutions. His areas of expertise include SAN, storage, HA systems, xSeries® servers,
network operating systems (Linux, MS Windows, OS/2®), and Lotus® Domino™ servers. He
holds several certifications from various vendors (IBM, Red Hat, Microsoft). Ivo has
contributed to various other redbooks on Tivoli products, SAN, Linux/390, xSeries, and Linux.

Mary Anne Marquez is the team lead for tape performance at IBM Tucson. She has
extensive knowledge in setting up a TotalStorage Productivity Center environment for use
with Copy Services and Performance Management, as well as debugging the various
components of TotalStorage Productivity Center including WebSphere, ICAT, and the CCW
interface for ESS. In addition to TPC, Mary Anne has experience with the native Copy
Services tools on ESS model-800 and DS8000. She has authored several performance
white papers.

© Copyright IBM Corp. 2005. All rights reserved. xv

Thanks to the following people for their contributions to this project:

Sangam Racherla
Yvonne Lyon
ITSO, San Jose Center

Bob Haimowitz
ITSO, Raleigh Center

Diana Duan
Tina Dunton
Nancy Hobbs
Paul Lee
Thiha Than
Miki Walter
IBM San Jose

Martine Wedlake
IBM Beaverton

Ryan Darris
IBM Tucson

Doug Dunham
Tivoli Storage SWAT Team

Mike Griese
Technical Support Marketing Lead, Rochester

Curtis Neal
Scott Venuti
Open System Demo Center, San Jose

xvi IBM TotalStorage Productivity Center V2.3: Getting Started

Become a published author
Join us for a two- to six-week residency program! Help write an IBM Redbook dealing with
specific products or solutions, while getting hands-on experience with leading-edge
technologies. You'll team with IBM technical professionals, Business Partners and/or
customers.

Your efforts will help increase product acceptance and customer satisfaction. As a bonus,
you'll develop a network of contacts in IBM development labs, and increase your productivity
and marketability.

Find out more about the residency program, browse the residency index, and apply online at:
ibm.com/redbooks/residencies.html

Comments welcome
Your comments are important to us!

We want our Redbooks to be as helpful as possible. Send us your comments about this or
other Redbooks in one of the following ways:
Use the online Contact us review redbook form found at:
ibm.com/redbooks
Send your comments in an email to:
redbook@us.ibm.com
Mail your comments to:
IBM Corporation, International Technical Support Organization
Dept. QXXE Building 80-E2
650 Harry Road
San Jose, California 95120-6099

Preface xvii

xviii IBM TotalStorage Productivity Center V2.3: Getting Started

Part 1

Part 1 IBM TotalStorage
Productivity Center
foundation
In this part of the book we introduce the IBM TotalStorage Productivity Center:
Chapter 1, “IBM TotalStorage Productivity Center overview” on page 3, contains an
overview of the components of IBM TotalStorage Productivity Center.
Chapter 2, “Key concepts” on page 27, provides information about the communication,
protocols, and standards organization that is the foundation of understanding the IBM
TotalStorage Productivity Center.

© Copyright IBM Corp. 2005. All rights reserved. 1

2 IBM TotalStorage Productivity Center V2.3: Getting Started

1

Chapter 1. IBM TotalStorage Productivity
Center overview
IBM TotalStorage Productivity Center is software, part of the IBM TotalStorage open software
family, designed to provide a single point of control for managing both IBM and non-IBM
networked storage devices that implement the Storage Management Initiative Specification
(SMI-S), including the IBM TotalStorage SAN Volume Controller (SVC), IBM TotalStorage
Enterprise Storage Server (ESS), IBM TotalStorage Fibre Array Storage Technology (FAStT),
IBM TotalStorage DS4000, IBM TotalStorage DS6000, and IBM TotalStorage DS8000 series.

TotalStorage Productivity Center is a solution for customers with storage management
requirements, who want to reduce the complexities and costs of storage management,
including management of SAN-based storage, while consolidating control within a consistent
graphical user interface.

This chapter provides an overview of the entire IBM TotalStorage Open Software Family.


1.1 Introduction to IBM TotalStorage Productivity Center
The IBM TotalStorage Productivity Center consists of software components which enable
storage administrators to monitor, configure, and manage storage devices and subsystems
within a SAN environment.

The TotalStorage Productivity Center is based on the recent standard issued by the Storage
Networking Industry Association (SNIA). The standard addresses the interoperability of
storage hardware and software within a SAN.

1.1.1 Standards organizations and standards
Today, there are at least 10 organizations involved in creating standards for storage, storage
management, SAN management, and interoperability. Figure 1-1 shows the key
organizations involved in developing and promoting standards relating to storage, storage
management, and SAN management, and the relevant standards for which they are
responsible.

Figure 1-1 SAN management standards bodies

Key standards for Storage Management are:
Distributed Management Task Force (DMTF) Common Information Model (CIM)
Standards. This includes the CIM Device Model for Storage, which at the time of writing
was Version 2.7.2 for the CIM schema.
Storage Networking Industry Association (SNIA) Storage Management Initiative
Specification (SMI-S).


1.2 IBM TotalStorage Open Software family
The IBM TotalStorage Open Software Family, is designed to provide a full range of
capabilities, including storage infrastructure management, Hierarchical Storage Management
(HSM), archive management, and recovery management.

The On Demand storage environment is shown in Figure 1-2. The hardware infrastructure is
a complete range of IBM storage hardware and devices providing flexibility in choice of
service quality and cost structure. On top of the hardware infrastructure is the virtualization
layer. The storage virtualization is infrastructure software designed to pool storage assets,
enabling optimized use of storage assets across the enterprise and the ability to modify the
storage infrastructure with minimal or no disruption to application services. The next layer is
composed of storage infrastructure management to help enterprises understand and
proactively manage their storage infrastructure in the on demand world; hierarchical storage
management to help control growth; archive management to manage cost of storing huge
quantities of data; recovery management to ensure recoverability of data. The top layer is
storage orchestration which automates work flows to help eliminate human error.

Figure 1-2 Enabling customer to move toward On Demand

Chapter 1. IBM TotalStorage Productivity Center overview 5

Previously we discussed the next steps or entry points into an On Demand environment. The
IBM software products which represent these entry points and which comprise the IBM
TotalStorage Open Software Family is shown in Figure 1-3.

Figure 1-3 IBM TotalStorage Open Software Family

1.3 IBM TotalStorage Productivity Center
The IBM TotalStorage Productivity Center is an open storage infrastructure management
solution designed to help reduce the effort of managing complex storage infrastructures, to
help improve storage capacity utilization, and to help improve administrative efficiency. It is
designed to enable an agile storage infrastructure that can respond to On Demand storage
needs.

The IBM TotalStorage Productivity Center offering is a powerful set of tools designed to help
simplify the management of complex storage network environments. The IBM TotalStorage
Productivity Center consists of TotalStorage Productivity Center for Disk, TotalStorage
Productivity Center for Replication, TotalStorage Productivity Center for Data (formerly Tivoli
Storage Resource Manager), and TotalStorage Productivity Center for Fabric (formerly Tivoli
SAN Manager).


Taking a closer look at storage infrastructure management (see Figure 1-4), we focus on four
subject matter experts to empower the storage administrators to effectively do their work.
Data subject matter expert
San Fabric subject matter expert
Disk subject matter expert
Replication subject matter expert

Figure 1-4 Centralized, automated storage infrastructure management

1.3.1 Data subject matter expert: TotalStorage Productivity Center for Data
The Data subject matter expert has intimate knowledge of how storage is used, for example
whether the data is used by a file system or a database application. Figure 1-5 on page 8
shows the role of the Data subject matter expert which is filled by the TotalStorage
Productivity Center for Data (formerly the IBM Tivoli Storage Resource Manager).


Figure 1-5 Monitor and Configure the Storage Infrastructure Data area

Heterogeneous storage infrastructures, driven by growth in file and database data, consume
increasing amounts of administrative time, as well as actual hardware resources. IT
managers need ways to make their administrators more efficient and more efficiently utilize
their storage resources. Tivoli Storage Resource Manager gives storage administrators the
automated tools they need to manage their storage resources more cost-effectively.

TotalStorage Productivity Center for Data allows you to identify different classes of data,
report how much space is being consumed by these different classes, and take appropriate
actions to keep the data under control.

Features of the TotalStorage Productivity Center for Data are:
Automated identification of the storage resources in an infrastructure and analysis of how
effectively those resources are being used.
File-system and file-level evaluation uncovers categories of files that, if deleted or
archived, can potentially represent significant reductions in the amount of data that must
be stored, backed up and managed.
Automated control through policies that are customizable with actions that can include
centralized alerting, distributed responsibility and fully automated response.
Predict future growth and future at-risk conditions with historical information.

Through monitoring and reporting, TotalStorage Productivity Center for Data helps the
storage administrator prevent outages in the storage infrastructure. Armed with timely
information, the storage administrator can take action to keep storage and data available to
the application. TotalStorage Productivity Center for Data also helps to make the most
efficient use of storage budgets, by allowing administrators to use their existing storage more
efficiently, and more accurately predict future storage growth.


TotalStorage Productivity Center for Data monitors storage assets, capacity, and usage
across an enterprise. TotalStorage Productivity Center for Data can look at:
Storage from a host perspective: Manage all the host-attached storage, capacity and
consumption attributed to file systems, users, directories, and files
Storage from an application perspective: Monitor and manage the storage activity inside
different database entities including instance, tablespace, and table
Storage utilization and provide chargeback information.

Architecture
The TotalStorage Productivity Center for Data server system manages a number of Agents,
which can be servers with storage attached, NAS systems, or database application servers.
Information is collected from the Agents and stored in a database repository. The stored
information can then be displayed from a native GUI client or browser interface anywhere in
the network. The GUI or browser interface gives access to the other functions of TotalStorage
Productivity Center for Data, including creating and customizing of a large number of different
types of reports and setting up alerts.

With TotalStorage Productivity Center for Data, you can:
Monitor virtually any host
Monitor local, SAN-attached and Network Attached Storage from a browser anywhere on
the network

For more information refer to the redbook IBM Tivoli Storage Resource Manager: A Practical
Introduction, SG24-6886.

1.3.2 Fabric subject matter expert: Productivity Center for Fabric
The storage infrastructure management for Fabric covers the Storage Area Network (SAN).
To handle and manage SAN events you need a comprehensive tool. The tool must have a
single point of operation and it tool must be able to perform all the tasks from the SAN. This
role is filled by the TotalStorage Productivity Center for Fabric (formerly the IBM Tivoli SAN
Manager) which is a part of the IBM TotalStorage Productivity Center.

The Fabric subject matter expert is the expert in the SAN. Its role is:
Discovery of fabric information
Provide the ability to specify fabric policies
– What HBAs to use for each host and for what purpose
– Objectives for zone configuration (for example, shielding host HBAs from one another
and performance)
Automatically modify the zone configuration

TotalStorage Productivity Center for Fabric provides real-time visual monitoring of SANs,
including heterogeneous switch support, and is a central point of control for SAN
configuration (including zoning). It automates the management of heterogeneous storage
area networks, resulting in”
Improved Application Availability
– Predicting storage network failures before they happen enabling preventative
maintenance
– Accelerate problem isolation when failures do happen


Optimized Storage Resource Utilization by reporting on storage network performance
Enhanced Storage Personnel Productivity - Tivoli SAN Manager creates a single point of
control, administration and security for the management of heterogeneous storage
networks

Figure 1-6 describes the requirements that must be addressed by the Fabric subject matter
expert.

Figure 1-6 Monitor and Configure the Storage Infrastructure Fabric area

TotalStorage Productivity Center for Fabric monitors and manages switches and hubs,
storage and servers in a Storage Area Network. TotalStorage Productivity Center for Fabric
can be used for both online monitoring and historical reporting. TotalStorage Productivity
Center for Fabric:
Manages fabric devices (switches) through outband management.
Discovers many details about a monitored server and its local storage through an Agent
loaded onto a SAN-attached host (Managed Host).
Monitors the network and collects events and traps
Launches vendor-provided specific SAN element management applications from the
TotalStorage Productivity Center for Fabric Console.
Discovers and manages iSCSI devices.
Provides a fault isolation engine for SAN problem determination (ED/FI - SAN Error
Predictor)

TotalStorage Productivity Center for Fabric is compliant with the standards relevant to SAN
storage and management.


TotalStorage Productivity Center for Fabric components
The major components of the TotalStorage Productivity Center for Fabric include:
A manager or server, running on a SAN managing server
Agents, running on one or more managed hosts
Management console, which is by default on the Manager system, plus optional additional
remote consoles
Outband agents - consisting of vendor-supplied MIBs for SNMP

There are two additional components which are not included in the TotalStorage Productivity
Center.
IBM Tivoli Enterprise™ Console (TEC) which is used to receive TotalStorage Productivity
Center for Fabric generated events. Once forwarded to TEC, These can then be
consolidated with events from other applications and acted on according to enterprise
policy.
IBM Tivoli Enterprise Data Warehouse (TEDW) is used to collect and analyze data
gathered by the TotalStorage Productivity Center for Fabric. The Tivoli Data Enterprise
Warehouse collects, organizes, and makes data available for the purpose of analysis in
order to give management the ability to access and analyze information about its
business.

The TotalStorage Productivity Center for Fabric functions are distributed across the Manager
and the Agent.

TotalStorage Productivity Center for FabricServer
Performs initial discovery of environment:
– Gathers and correlates data from agents on managed hosts
– Gathers data from SNMP (outband) agents
– Graphically displays SAN topology and attributes
Provides customized monitoring and reporting through NetView®
Reacts to operational events by changing its display
(Optionally) forwards events to Tivoli Enterprise Console® or SNMP managers

TotalStorage Productivity Center for Fabric Agent
Gathers information about:
SANs by querying switches and devices for attribute and topology information
Host-level storage, such as file systems and LUNs
Event and other information detected by HBAs
Forwards topology and event information to the Manager

Discover SAN components and devices
TotalStorage Productivity Center for Fabric uses two methods to discover information about
the SAN - outband discovery, and inband discovery.

Outband discovery is the process of discovering SAN information, including topology and
device data, without using the Fibre Channel data paths. Outband discovery uses SNMP
queries, invoked over IP network. Outband management and discovery is normally used to
manage devices such as switches and hubs which support SNMP.


In outband discovery, all communications occur over the IP network:
TotalStorage Productivity Center for Fabric requests information over the IP network from
a switch using SNMP queries on the device.
The device returns the information toTotalStorage Productivity Center for Fabric, also over
the IP network.

Inband discovery is the process of discovering information about the SAN, including
topology and attribute data, through the Fibre Channel data paths. In inband discovery, both
the IP and Fibre Channel networks are used:
TotalStorage Productivity Center for Fabric requests information (via the IP network) from
a Tivoli SAN Manager agent installed on a Managed Host.
That agent requests information over the Fibre Channel network from fabric elements and
end points in the Fibre Channel network.
The agent returns the information to TotalStorage Productivity Center for Fabric over the
IP network.
TotalStorage Productivity Center for Fabric collects, co-relates and displays information
from all devices in the storage network, using both the IP network and the Fibre Channel
network. If the Fibre Channel network is unavailable for any reason, monitoring can still
continue over the IP network.

TotalStorage Productivity Center for Fabric benefits
TotalStorage Productivity Center for Fabric discovers the SAN infrastructure, and monitors
the status of all the discovered components. Through Tivoli NetView, the administrator can
provide reports on faults on components (either individually or in groups, or “smartsets”, of
components). This will help them increase data availability for applications so the company
can either be more efficient, or maximize the opportunity to produce revenue.

TotalStorage Productivity Center for Fabric helps the storage administrator:
Prevent faults in the SAN infrastructure through reporting and proactive maintenance, and
Identify and resolve problems in the storage infrastructure quickly, when a problem
Supported devices for TotalStorage Productivity Center for Fabric
Provide fault isolation of SAN links.

For more information about the TotalStorage Productivity Center for Fabric, refer to IBM Tivoli
Storage Area Network Manager: A Practical Introduction, SG24-6848.

1.3.3 Disk subject matter expert: TotalStorage Productivity Center for Disk
The Disk subject matter expert’s job allows you to manage the disk systems. It will discover
and classify all disk systems that exist and draw a picture of all discovered disk systems. The
Disk subject matter expert provides the ability to monitor, configure, create disks and do LUN
masking of disks. It also does performance trending and performance threshold I/O analysis
for both real disks and virtual disks. It also does automated status and problem alerts via
SNMP. This role is filled by the TotalStorage Productivity Center for Disk (formerly the IBM
TotalStorage Multiple Device Manager Performance Manager component).

The requirements addressed by the Disk subject matter expert are shown in Figure 1-7 on
page 13. The disk systems monitoring and configuration needs must be covered by a
comprehensive management tool like the TotalStorage Productivity Center for Disk.


Figure 1-7 Monitor and configure the Storage Infrastructure Disk area

The TotalStorage Productivity Center for Disk provides the raw capabilities of initiating and
scheduling performance data collection on the supported devices, of storing the received
performance statistics into database tables for later use, and of analyzing the stored data and
generating reports for various metrics of the monitored devices. In conjunction with data
collection, the TotalStorage Productivity Center for Disk is responsible for managing and
monitoring the performance of the supported storage devices. This includes the ability to
configure performance thresholds for the devices based on performance metrics, the
generation of alerts when these thresholds are exceeded, the collection and maintenance of
historical performance data, and the creation of gauges, or performance reports, for the
various metrics to display the collected historical data to the end user. The TotalStorage
Productivity Center for Disk enables you to perform sophisticated performance analysis for
the supported storage devices.

Functions
TotalStorage Productivity Center for Disk provides the following functions:
Collect data from devices
The Productivity Center for Disk collects data from the IBM TotalStorage Enterprise
Storage Server (ESS), SAN Volume Controller (SVC), DS400 family and SMI-S enabled
devices. Each Performance Collector collects performance data from one or more storage
groups, all of the same device type (for example, ESS or SAN Volume Controller). Each
Performance Collection has a start time, a stop time, and a sampling frequency. The
performance sample data is stored in DB2® database tables.
Configure performance thresholds
You can use the Productivity Center for Disk to set performance thresholds for each device
type. Setting thresholds for certain criteria enables Productivity Center for Disk to notify
you when a certain threshold has been exceeded, so that you to take action before a
critical event occurs.


You can specify what action should be taken when a threshold-exceeded condition occurs.
The action may be to log the occurrence or to trigger an event. The threshold settings can
vary by individual device.
Monitor performance metrics across storage subsystems from a single console
Receive timely alerts to enable event action based on customer policies
View performance data from the Productivity Center for Disk database
You can view performance data from the Productivity Center for Disk database in both
graphical and tabular forms.
The Productivity Center for Disk allows a TotalStorage Productivity Center user to access
recent performance data in terms of a series of values of one or more metrics, associated
with a finite set of components per device. Only recent performance data is available for
gauges. Data that has been purged from the database cannot be viewed. You can define
one or more gauges by selecting certain gauge properties and saving them for later
referral. Each gauge is identified through a user-specified name, and once defined, a
gauge can be “started”, which means it is then displayed in a separate window of the
TotalStorage Productivity Center GUI. You can have multiple gauges active at the same
time.
Gauge definition will be accomplished through a wizard, to aid in entering a valid set of
gauge properties. Gauges are saved in the Productivity Center for Disk database and
retrieved upon request. When you request data pertaining to a defined gauge, the
Performance Manager builds a query to the database, retrieves and formats the data and
returns it to you. Once started, a gauge is displayed in its own window, and displays all
available performance data for the specified initial date/time range. The date/time range
can be changed after the initial gauge widow is displayed.
Focus on storage optimization through identification of best LUN
The Volume Performance Advisor is an automated tool to help the storage administrator
pick the best possible placement of a new LUN to be allocated, that is, the best placement
from a performance perspective. It also uses the historical performance statistics collected
from the supported devices, to locate unused storage capacity on the SAN that exhibits
the best (estimated) performance characteristics. Allocation optimization involves several
variables which are user controlled, such as required performance level and the time of
day/week/month of prevalent access. This function is fully integrated with the Device
Manager function, this is so that when a new LUN is added, for example, to the ESS, the
Performance Manager can seamlessly select the best possible LUN.

For detailed information about how to use the functions of the TotalStorage Productivity
Center for Disk refer to Chapter 11, “Using TotalStorage Productivity Center for Disk” on
page 375.

1.3.4 Replication subject matter expert: Productivity Center for Replication
The Replication subject matter expert’s job is to provide a single point of control for all
replication activities. This role is filled by the TotalStorage Productivity Center for Replication.
Given a set of source volumes to be replicated, the Productivity Center for Replication will find
the appropriate targets, perform all the configuration actions required, and ensure the source
and target volumes relationships are set up. Given a set of source volumes that represent an
application, the Productivity Center for Replication will group these in a consistency group,
give that consistency group a name, and allow you to start replication on the application.


Productivity Center for Replication will start up all replication pairs and monitor them to
completion. If any of the replication pairs fail, meaning the application is out of sync, the
Productivity Center for Replication will suspend them until the problem is resolved, resync
them and resume the replication. The Productivity Center for Replication provides complete
management of the replication process.

The requirements addressed by the Replication subject matter expert are shown Figure 1-8.
Replication in a complex environment needs to be addressed by a comprehensive
management tool like the TotalStorage Productivity Center for Replication.

Figure 1-8 Monitor and Configure the Storage Infrastructure Replication area

Functions
Data replication is the core function required for data protection and disaster recovery. It
provides advanced copy services functions for supported storage subsystems on the SAN.

Replication Manager administers and configures the copy services functions and monitors
the replication actions. Its capabilities consist of the management of two types of copy
services: the Continuous Copy (also known as Peer-to-Peer, PPRC, or Remote Copy), and
the Point-in-Time Copy (also known as FlashCopy®). At this time TotalStorage Productivity
Center for Replication supports the IBM TotalStorage ESS.

Productivity Center for Replication includes support for replica sessions, which ensures that
data on multiple related heterogeneous volumes is kept consistent, provided that the
underlying hardware supports the necessary primitive operations. Productivity Center for
Replication also supports the session concept, such that multiple pairs are handled as a
consistent unit, and that Freeze-and-Go functions can be performed when errors in mirroring
occur. Productivity Center for Replication is designed to control and monitor the copy services
operations in large-scale customer environments.


Productivity Center for Replication provides a user interface for creating, maintaining, and
using volume groups and for scheduling copy tasks. The User Interface populates lists of
volumes using the Device Manager interface. Some of the tasks you can perform with
Productivity Center for Replication are:
Create a replication group. A replication group is a collection of volumes grouped together
so that they can be managed concurrently.
Set up a Group for replication.
Create, save, and name a replication task.
Schedule a replication session with the user interface:
– Create Session Wizard.
– Select Source Group.
– Select Copy Type.
– Select Target Pool.
– Save Session.
Start a replication session

A user can also perform these tasks with the Productivity Center for Replication
command-line interface.

For more information about the Productivity Center for Replication functions refer to
Chapter 15, “Using TotalStorage Productivity Center for Replication” on page 827.

1.4 IBM TotalStorage Productivity Center
All the subject matter experts, for Data, Fabric, Disk, and Replication are components of the
IBM TotalStorage Productivity Center.

The IBM TotalStorage Productivity Center is the first offering to be delivered as part of the
IBM TotalStorage Open Software Family. The IBM TotalStorage Productivity Center is an
open storage infrastructure management solution designed to help reduce the effort of
managing complex storage infrastructures, to help improve storage capacity utilization, and to
help improve administrative efficiency. It is designed to enable an agile storage infrastructure
that can respond to on demand storage needs.

The IBM TotalStorage Productivity Center allows you to manage your storage infrastructure
using existing storage management products — Productivity Center for Data, Productivity
Center for Fabric, Productivity Center for Disk and Productivity Center for Replication — from
one physical place.

The IBM TotalStorage Productivity Center components can be launched from the IBM
TotalStorage Productivity Center launch pad as shown in Figure 1-9 on page 17.


Figure 1-9 IBM TotalStorage Productivity Center Launch Pad

The IBM TotalStorage Productivity Center establishes the foundation for IBM’s e-business On
Demand technology. We need the function in an On Demand environment to provide IT
resources On Demand - when the resources are needed by an application to support the
customers business process. Of course, we are able to provide resources or remove
resources today but the question is how. The process is expensive and time consuming.

The IBM TotalStorage Productivity Center is the basis for the provisioning of storage
resources to make the e-business On Demand environment a reality. In the future there will
be more automation required to handle the hugh amount work in the provisioning area, more
automation like the BM TotalStorage Productivity Center launch pad provides. Automation
means workflow. Workflow is the key to getting work automated. IBM has a long history and
investment in building workflow engines and work flows. Today IBM is using the IBM Tivoli
Intelligent Orchestrator and IBM Tivoli Provisioning Manager to satisfy the resource requests
in the e-business on demand™ environment in the server arena.

The IBM Tivoli Intelligent Orchestrator and The IBM Tivoli Provisioning Manager provide the
provisioning in the e-business On Demand environment.

1.4.1 Productivity Center for Disk and Productivity Center for Replication
The Productivity Center for Disk and Productivity Center for Replication is software that has
been designed to enable administrators to manage SANs and storage from a single console
(Figure 1-10 on page 18). This software solution is designed specifically for managing
networked storage components based on the SMI-S, including:
IIBM TotalStorage SAN Volume Controller
IBM TotalStorage Enterprise Storage Server (ESS)
IBM TotalStorage Fibre Array Storage Technology (FAStT)
IBM TotalStorage DS4000 series
SMI enabled device


Figure 1-10 Managing multiple devices

Productivity Center for Disk and Productivity Center for Replication are built on IBM Director,
a comprehensive server management solution. Using Director with the multiple device
management solution enables administrators to consolidate the administration of IBM storage
subsystems and provide advanced storage management functions (including replication and
performance management) across multiple IBM storage subsystems. It interoperates with
SAN Management and Enterprise System Resource Manager (ESRM) products from IBM,
includingTotalStorage Productivity Center for Data and SAN Management products from
other vendors.

In a SAN environment, multiple devices work together to create a storage solution. The
Productivity Center for Disk and Productivity Center for Replication provides integrated
administration, optimization, and replication features for interacting SAN devices, including
the SAN Volume Controller and DS4000 Family devices. It provides an integrated view of the
underlying system so that administrators can drill down through the virtualized layers to easily
perform complex configuration tasks and more productively manage the SAN infrastructure.

Because the virtualization layers support advanced replication configurations, the Productivity
Center for Disk and Productivity Center for Replication products offer features that simplify
the configuration, monitoring, and control of disaster recovery and data migration solutions. In
addition, specialized performance data collection, analysis, and optimization features are
provided. As the SNIA standards mature, the Productivity Center view will be expanded to
include CIM-enabled devices from other vendors, in addition to IBM storage. Figure 1-11 on
page 19 provides an overview of Productivity Center for Disk and Productivity Center for
Replication.


IBM TotalStorage Productivity Center

Performance Replication
Manager Manager

Device Manager

IBM Director

WebSphere Application Server DB2

Figure 1-11 Productivity Center overview

The Productivity Center for Disk and Productivity Center for Replication provides support for
configuration, tuning, and replication of the virtualized SAN. As with the individual devices,
the Productivity Center for Disk and Productivity Center for Replication layers are open and
can be accessed via a GUI, CLI, or standards-based Web Services. Productivity Center for
Disk and Productivity Center for Replication provide the following functions:
Device Manager - Common function provided when you install the base prerequisite
products for either Productivity Center for Disk or Productivity Center for Replication
Performance Manager - provided by Productivity Center for Disk
Replication Manager - provided by Productivity Center for Replication

Device Manager
The Device Manager is responsible for the discovery of supported devices; collecting asset,
configuration, and availability data from the supported devices; and providing a limited
topography view of the storage usage relationships between those devices.

The Device Manager builds on the IBM Director discovery infrastructure. Discovery of storage
devices adheres to the SNIA SMI-S specification standards. Device Manager uses the
Service Level Protocol (SLP) to discover SMI-S enabled devices. The Device Manager
creates managed objects to represent these discovered devices. The discovered managed
objects are displayed as individual icons in the Group Contents pane of the IBM Director
Console as shown in Figure 1-12 on page 20.


Figure 1-12 IBM Director Console

Device Manager provides a subset of configuration functions for the managed devices,
primarily LUN allocation and assignment. Its function includes certain cross-device
configuration, as well as the ability to show and traverse inter-device relationships. These
services communicate with the CIM Agents that are associated with the particular devices to
perform the required configuration. Devices that are not SMI-S compliant are not supported.
The Device Manager also interacts and provides some SAN management functionality when
IBM Tivoli SAN Manager is installed.

The Device Manager health monitoring keeps you aware of hardware status changes in the
discovered storage devices. You can drill down to the status of the hardware device, if
applicable. This enables you to understand which components of a device are malfunctioning
and causing an error status for the device.

SAN Management
When a supported SAN Manager is installed and configured, the Device Manager leverages
the SAN Manager to provide enhanced function. Along with basic device configuration
functions such as LUN creation, allocation, assignment, and deletion for single and multiple
devices, basic SAN management functions such as LUN discovery, allocation, and zoning are
provided in one step. IBM TotalStorage Productivity Center for Fabric (formerly IBM Tivoli
SAN Manager) is currently the supported SAN Manager.

The set of SAN Manager functions that will be exploited are:
The ability to retrieve the SAN topology information, including switches, hosts, ports, and
storage devices
The ability to retrieve and to modify the zoning configuration on the SAN
The ability to register for event notification, to ensure that Productivity Center for Disk is
aware when the topology or zoning changes as new devices are discovered by the SAN
Manager, and when hosts' LUN configurations change


Performance Manager function
The Performance Manager function provides the raw capabilities of initiating and scheduling
performance data collection on the supported devices, of storing the received performance
statistics into database tables for later use, and of analyzing the stored data and generating
reports for various metrics of the monitored devices. In conjunction with data collection, the
Performance Manager is responsible for managing and monitoring the performance of the
supported storage devices. This includes the ability to configure performance thresholds for
the devices based on performance metrics, the generation of alerts when these thresholds
are exceeded, the collection and maintenance of historical performance data, and the
creation of gauges, or performance reports, for the various metrics to display the collected
historical data to the end user. The Performance Manager enables you to perform
sophisticated performance analysis for the supported storage devices.

Functions
Collect data from devices
The Performance Manager collects data from the IBM TotalStorage Enterprise Storage
Server (ESS), IBM TotalStorage SAN Volume Controller (SVC), IBM TotalStorage DS4000
series, IBM TotalStorage DS6000 and IBM TotalStorage DS8000 series and SMI-S
enabled devices. The performance collection task collects performance data from one or
more storage groups, all of the same device type (for example, ESS or SVC). Each
performance collection task has a start time, a stop time, and a sampling frequency. The
performance sample data is stored in DB2 database tables.
Configure performance thresholds
You can use the Performance Manager to set performance thresholds for each device
type. Setting thresholds for certain criteria enables Performance Manager to notify you
when a certain threshold has been exceeded, so that you can take action before a critical
event occurs.
You can specify what action should be taken when a threshold-exceeded condition occurs.
The action may be to log the occurrence or to trigger an event. The threshold settings can
vary by individual device.

The eligible metrics for threshold checking are fixed for each storage device. If the threshold
metrics are modified by the user, the modifications are accepted immediately and applied to
checking being performed by active performance collection tasks. Examples of threshold
metrics include:
Disk utilization value
Average cache hold time
Percent of sequential I/Os
I/O rate
NVS full value
Virtual disk I/O rate
Managed disk I/O rate

There is a user interface that supports threshold settings, enabling a user to:
Modify a threshold property for a set of devices of like type.
Modify a threshold property for a single device.
– Reset a threshold property to the IBM-recommended value (if defined) for a set of
devices of like type.
IBM-recommended critical and warning values will be provided for all thresholds
known to indicate potential performance problems for IBM storage devices.


– Reset a threshold property to the IBM-recommended value (if defined) for a single
device.
Show a summary of threshold properties for all of the devices of like type.
View performance data from the Performance Manager database.

Gauges
The Performance Manager supports a performance-type gauge. The performance-type
gauge presents sample-level performance data. The frequency at which performance data is
sampled on a device depends on the sampling frequency that you specify when you define
the performance collection task. The maximum and minimum values of the sampling
frequency depend on the device type. The static display presents historical data over time.
The refreshable display presents near real-time data from a device that is currently collecting
performance data.

The Performance Manager enables a Productivity Center for Disk user to access recent
performance data in terms of a series of values of one or more metrics associated with a finite
set of components per device. Only recent performance data is available for gauges. Data
that has been purged from the database cannot be viewed. You can define one or more
gauges by selecting certain gauge properties and saving them for later referral. Each gauge
is identified through a user-specified name and, when defined, a gauge can be started, which
means that it is then displayed in a separate window of the Productivity Center GUI. You can
have multiple gauges active at the same time.

Gauge definition is accomplished through a wizard to aid in entering a valid set of gauge
properties. Gauges are saved in the Productivity Center for Disk database and retrieved upon
request. When you request data pertaining to a defined gauge, the Performance Manager
builds a query to the database, retrieves and formats the data, and returns it to you. When
started, a gauge is displayed in its own window, and it displays all available performance data
for the specified initial date/time range. The date/time range can be changed after the initial
gauge window is displayed.

For performance-type gauges, if a metric selected for display is associated with a threshold
enabled for checking, the current threshold properties are also displayed in the gauge window
and are updated each time the gauge data is refreshed.

Database services for managing the collected performance data
The performance data collected from the supported devices is stored in a DB2 database.
Database services are provided that enable you to manage the potential volumes of data.
Database purge function
A database purge function deletes older performance data samples and, optionally, the
associated exception data. Flexibility is built into the purge function, and it enables you to
specify the data to purge, allowing important data to be maintained for trend purposes.
You can specify to purge all of the sample data from all types of devices older than a
specified number of days. You can specify to purge the data associated with a particular
type of device.
If threshold checking was enabled at the time of data collection, you can exclude data that
exceeded at least one threshold value from being purged.
You can specify the number of days that data is to remain in the database before being
purged. Sample data and, optionally, exception data older than the specified number of
days will be purged.
A reorganization function is performed on the database tables after the sample data is
deleted from the respective database tables.


Database information function
Due to the amount of data collected by the Performance Manager function provided by
Productivity Center for Disk, the database should be monitored to prevent it from running
out of space. The database information function returns the database % full. This function
can be invoked from either the Web user interface or the CLI.

Volume Performance Advisor
The advanced performance analysis provided by Productivity Center for Disk is intended to
address the challenge of allocating more storage in a storage system so that the users of the
newly allocated storage achieve the best possible performance.

The Volume Performance Advisor is an automated tool that helps the storage administrator
pick the best possible placement of a new LUN to be allocated (that is, the best placement
from a performance perspective). It also uses the historical performance statistics collected
from the supported devices to locate unused storage capacity on the SAN that exhibits the
best (estimated) performance characteristics. Allocation optimization involves several
variables that are user-controlled, such as required performance level and the time of
day/week/month of prevalent access. This function is fully integrated with the Device Manager
function so that, for example, when a new LUN is added to the ESS, the Device Manager can
seamlessly select the best possible LUN.

Replication Manager function
Data replication is the core function required for data protection and disaster recovery. It
provides advanced copy services functions for supported storage subsystems on the SAN.

Productivity Center for Replication administers and configures the copy services functions
and monitors the replication actions. Its capabilities consist of the management of two types
of copy services: the Continuous Copy (also known as Peer-to-Peer, PPRC, or Remote
Copy), and the Point-in-Time Copy (also known as FlashCopy). Currently replication functions
are provided for the IBM TotalStorage ESS.

Productivity Center for Replication includes support for replica sessions, which ensures that
data on multiple related heterogeneous volumes is kept consistent, provided that the
underlying hardware supports the necessary primitive operations. Multiple pairs are handled
as a consistent unit, Freeze-and-Go functions can be performed when errors in mirroring
occur. Productivity Center for Replication is designed to control and monitor the copy services
operations in large-scale customer environments.

Productivity Center for Replication is controlled by applying predefined policies to Groups and
Pools, which are groupings of LUNs that are managed by the Replication Manager. It
provides the ability to copy a Group to a Pool, in which case it creates valid mappings for
source and target volumes and optionally presents them to the user for verification that the
mapping is acceptable. In this case, it manages Pool membership by removing target
volumes from the pool when they are used, and by returning them to the pool only if the target
is specified as being discarded when it is deleted.

1.4.2 Event services
At the heart of any systems management solution is the ability to alert the system
administrator in the event of a system problem. IBM Director provides a method of alerting
called Event Action Plans, which enables the definition of event triggers independently from
actions that might be taken.


An event is an occurrence of a predefined condition relating to a specific managed object that
identifies a change in a system process or a device. The notification of that change can be
generated and tracked (for example, notification that a Productivity Center component is not
available). Productivity Center for Disk and Productivity Center for Replication take full
advantage of, and build upon, the IBM Director Event Services.

The IBM Director includes sophisticated event-handling support. Event Action Plans can be
set up that specify what steps, if any, should be taken when particular events occur in the
environment. Director Event Management encompasses the following concepts:
Events can be generated by any managed object. IBM Director receives such events and
calls appropriate internal event handlers that have been registered.
Actions are user-configured steps to be taken for a particular event or type of event. There
can be zero or more actions associated with a particular action plan. System
administrators can create their own actions by customizing particular predefined actions.
Event Filters are a set of characteristics or criteria that determine whether an incoming
event should be acted on.
Event Action Plans are associations of one or more event filters with one or more actions.
Event Action Plans become active when you apply them to a system or a group of
systems.

The IBM Director Console includes an extensive set of GUI panels, called the Event Action
Plan Builder, that enable the user to create action plans and event filters. Event Filters can be
configured using the Event Action Plan Builder and set up with a variety of criteria, such as
event types, event severities, day and time of event occurrence, and event categories. This
allows control over exactly what action plans are invoked for each specific event.

Productivity Center provides extensions to the IBM Director event management support.
It takes full advantage of the IBM Director built-in support for event logging and viewing.
It generates events that will be externalized. Action plans can be created based on filter
criteria for these events. The default action plan is to log all events in the event log.
It creates additional event families, and event types within those families, that will be listed
in the Event Action Plan Builder.
Event actions that enable Productivity Center functions to be exploited from within action
plans will be provided. An example is the action to indicate the amount of historical data to
be kept.

1.5 Taking steps toward an On Demand environment
So what is an On Demand operating environment? It is not a specific set of hardware and
software. Rather, it is an environment that supports the needs of the business, allowing it to
become and remain responsive, variable, focused, and resilient.

An On Demand operating environment unlocks the value within the IT infrastructure to be
applied to solving business problems. It is an integrated platform, based on open standards,
to enable rapid deployment and integration of business applications and processes.
Combined with an environment that allows true virtualization and automation of the
infrastructure, it enables delivery of IT capability On Demand.


An On Demand operating environment must be:
Flexible
Self-managing
Scalable
Economical
Resilient
Based on open standards

The move to an On Demand storage environment is an evolving one, it does not happen all at
once. There are several next steps that you may take to move to the On Demand
environment.
Constant changes to the storage infrastructure (upgrading or changing hardware for
example) can be addressed by virtualization which provides flexibility by hiding the
hardware and software from users and applications.
Empower administrators with automated tools for managing heterogeneous storage
infrastructures. and eliminate human error.
Control storage growth with automated identification and movement of low-activity or
inactive data to a hierarchy of lower-cost storage.
Manage cost associated with capturing point-in-time copies of important data for
regulatory or bookkeeping requirements by maintaining this inactive data in a hierarchy of
lower-cost storage.
Ensure recoverability through the automated creation, tracking and vaulting of reliable
recovery points for all enterprise data.
The ultimate goal to eliminate human errors by preparing for Infrastructure Orchestration
software that can be used to automate workflows.

No matter which steps you take to an On Demand environment there will be results. The
results will be improved application availability, optimized storage resource utilization, and
enhanced storage personnel productivity.


2

Chapter 2. Key concepts
There are certain industry standards and protocols that are the basis of the IBM TotalStorage
Productivity Center. The understanding of these concepts is important for installing and
customizing the IBM TotalStorage Productivity Center.

In this chapter, we describe the standards on which the IBM TotalStorage Productivity Center
is built, as well as the methods of communication used to discover and manage storage
devices. We also discuss communication between the various components of the IBM

To help you understand these concepts, we provide diagrams to show the relationship and
interaction of the various elements in the IBM TotalStorage Productivity Center environment.


2.1 IBM TotalStorage Productivity Center architecture
This chapter provides an overview of the components and functions that are included in the
IBM TotalStorage Productivity Center.

2.1.1 Architectural overview diagram
The architectural overview diagram in Figure 2-1 helps to illustrate the governing ideas and
building blocks of the product suite which makes up the IBM TotalStorage Productivity Center.
It provides a logical overview of the main conceptual elements and relationships in the
architecture, components, connections, users, and external systems.

Figure 2-1 IBM TotalStorage Productivity Center architecture overview diagram

IBM TotalStorage Productivity Center and Tivoli Provisioning Manager are presented as
building blocks in the diagram. Both of the products are not a single application but a complex
environment by themselves.

The diagram also shows the different methods used to collect information from multiple
systems to give an administrator the necessary views on the environment, for example:
Software clients (agents)
Standard interfaces and protocols (for example, Simple Network Management Protocol
(SNMP), Common Information Model (CIM) Agent)
Proprietary interfaces (for only a few devices)

In addition to the central data collection, Productivity Center provides a single point of control
for a storage administrator, even though each manager still comes with its own interface.
A program called the Launchpad is provided to start the individual applications from a central
dashboard.


The Tivoli Provisioning Manager relies on Productivity Center to make provisioning possible.

2.1.2 Architectural layers
The IBM TotalStorage Productivity Center architecture can be broken up in three layers as
shown in Figure 2-2. Layer one represents a high level overview. There is only IBM
TotalStorage Productivity Center instance in the environment. Layers two and three drill down
into the TotalStorage Productivity Center environment so you can see the managers and the
prerequisite components.

Figure 2-2 Architectural layers

Layer two consists of the individual components that are part of the product suite:
IBM TotalStorage Productivity Center for Disk
IBM TotalStorage Productivity Center for Replication
IBM TotalStorage Productivity Center for Fabric
IBM TotalStorage Productivity Center for Data

Throughout this redbook, these products are referred to as managers or components.

Layer three includes all the prerequisite components, for example IBM DB2, IBM WebSphere,
IBM Director, IBM Tivoli NetView, and Tivoli Common Agent Services.

IBM TotalStorage Productivity Center for Fabric can be installed on a full version of
WebSphere Application Server or on the embedded WebSphere Application Server, which is
shipped with Productivity Center for Fabric. Installation on a full version of WebSphere
Application Server is used when other components of TotalStorage Productivity Center are
installed on the same logical server. IBM TotalStorage Productivity Center for Fabric can
utilize an existing IBM Tivoli Netview installation or can be installed along with it.

Note: Each of the manager and prerequisite components can be drilled down even further,
but in this book we go into this detail only where necessary. The only exception is Tivoli
Common Agent Services, which is a new underlying service in the Tivoli product family.

Terms and definitions
When you look at the diagram in Figure 2-2, you see that each layer has a different name.
The following sections explain each of these names as well as other terms commonly used in
this book.

Chapter 2. Key concepts 29

Product
A product is something that is available to be ordered. The individual products that are
included in IBM TotalStorage Productivity Center are introduced in Chapter 1, “IBM
TotalStorage Productivity Center overview” on page 3.

Components
Products (licensed software packages) and prerequisite software applications are in general
called components. Some of the components are internal, meaning that, from the installation
and configuration point of view, they are somewhat transparent. External components have to
be separately installed. We usually use the term components for the following applications:
IBM Director (external, used by Disk and Replication Manager)
IBM DB2 (external, used by all managers)
IBM WebSphere Application Server (external, used by Disk and Replication Manager,
used by Fabric Manager if installed on the same logical server)
Embedded WebSphere Application Server (internal, used by Fabric Manager)
Tivoli NetView (internal, used by Fabric Manager)
Tivoli Common Agent Services (external, used by Data and Fabric Manager)

Not all of the internal components are always shown in the diagrams and lists in this book.

The term subcomponent is used to emphasize that a certain component (the subcomponent)
belongs to or is used by another component. For example, a Resource Manager is a
subcomponent of the Fabric or Data Manager.

Managers
The managers are the central components of the IBM TotalStorage Productivity Center
environment. They may share some of the prerequisite components. For example, IBM DB2
and IBM WebSphere are used by different managers.
In this book, we sometimes use the following terms:
Disk Manager for Productivity Center for Disk
Replication Manager for Productivity Center for Replication
Data Manager for Productivity Center for Data
Fabric Manager for Productivity Center for Fabric

In addition, we use the term manager for the Agent Manager for Tivoli Agent Manager
component, because the name of the component already includes that term.

Agents
The agents are not shown in the diagram in Figure 2-2 on page 29, but they have an
important role in the IBM TotalStorage Productivity Center environment. There are two types
of agents: Common Information Model (CIM) Agents and agents that belong to one of the
managers:
CIM Agents: Agents that offer a CIM interface for management applications, for example,
for IBM TotalStorage DS8000 and DS6000 series storage systems, IBM TotalStorage
Enterprise Storage Server (ESS), SAN (Storage Area Network) Volume Controller, and
DS4000 Storage Systems formerly known as FAStT (Fibre Array Storage Technology)
Storage Systems
Agents that belong to one of the managers:
– Data Agents: Agents to collect data for the Data Manager
– Fabric Agents: Agents that are used by the Fabric Manager for inband SAN data
discovery and collection


In addition to these agents, the Service Location Protocol (SLP) also use the term agent for
these components:
User Agent
Service Agent
Directory Agent

Elements
We use the generic term element whenever we do not differentiate between components and
managers.

2.1.3 Relationships between the managers and components
An IBM TotalStorage Productivity Center environment includes many elements and is
complex. This section tries to explain how all the elements work together to form a center for
storage administration. Figure 2-3 shows the communication between the elements and how
they relate to each other.

Each gray box in the diagram represents one machine. The dotted line within a machine
separates two distinct managers of the IBM TotalStorage Productivity Center.

Figure 2-3 Manager and component relationship diagram

All these components can also run on one machine. In this case all managers and IBM
Director will share the same DB2 installation and all managers and IBM Tivoli Agent Manager
will share the same WebSphere installation.


2.1.4 Collecting data
Multiple methods are used within the different components to collect data from the devices in
your environment. In this version of the product, the information is stored in different
databases (see Table 3-6 on page 62) that are not shared between the individual
components.

Productivity Center for Disk and Productivity Center for Replication
Productivity Center for Disk and Productivity Center for Replication use the Storage
Management Initiative - Specification (SMI-S) standard (see “Storage Management Initiative -
Specification” on page 35) to collect information about subsystems. For devices that are not
CIM ready, this requires the installation of a proxy application (CIM Agent or CIM Object
Manager (CIMOM)). It does not use its own agent such as the Data Manager and Fabric
Manager.

IBM TotalStorage Productivity Center for Fabric uses two methods to collect information:
inband and outband discovery. You can use either method or you can use both at the same
time to obtain the most complete picture of your environment. Using just one of the methods
will give you incomplete information, but topology information will be available in both cases.

Outband discovery is the process of discovering SAN information, including topology and
device data, without using the Fibre Channel data paths. Outband discovery uses SNMP
queries, invoked over IP network. Outband management and discovery is normally used to
manage devices such as switches and hubs that support SNMP.

Inband discovery is the process of discovering information about the SAN, including
topology and attribute data, through the Fibre Channel data paths. Inband discovery uses the
following general process:
The Agent sends commands through its Host Bus Adapters (HBA) and the Fibre Channel
network to gather information about the switches.
The switch returns the information through the Fibre Channel network and the HBA to the
Agent.
The Agent queries the endpoint devices using RNID and SCSI protocols.
The Agent returns the information to the Manager over the IP network.
The Manager then responds to the new information by updating the database and
redrawing the topology map if necessary.

Internet SCSI (iSCSI) Discovery is an Internet Protocol (IP)-based storage networking
standard for linking data storage. It was developed by the Internet Engineering Task Force
(IETF). iSCSI can be used to transmit data over LANs and WANs.


The discovery paths are shown in parentheses in the diagram in Figure 2-4.

Figure 2-4 Fabric Manager inband and outband discovery paths

Within the IBM TotalStorage Productivity Center, the data manager is used to collect
information about logical drives, file systems, individual files, database usage, and more.
Agents are installed on the application servers and perform a regular scan to report back the
information. To report on a subsystem level, a SMI-S interface is also built in. This information
is correlated with the data that is gathered from the agents to show the LUNs that a host is
using (an agent must be installed on that host).

In contrast to Productivity Center for Disk and Productivity Center for Replication, the SMI-S
interface in Productivity Center for Data is only used to retrieve information, but not to
configure a device.

Restriction: The SLP User Agent integrated into the Data Manager uses SLP Directory
Agents and Service Agents to find services in the local subnet. To discover CIM Agents
from remote networks, they have to be registered to either the Directory Agent or Service
Agent, which is located in the local subnet unless routers are configured to also route
multicast packets.

You need to add each CIM Agent (that is not discovered) manually to the Data Manager;
refer to “Configuring the CIM Agents” on page 290.


2.2 Standards used in IBM TotalStorage Productivity Center
This section presents an overview of the standards that are used within IBM TotalStorage
Productivity Center by the different components. SLP and CIM are described in detail since
they are new concepts to many people that work with IBM TotalStorage Productivity Center
and are important to understand.

Vendor specific tools are available to manage devices in the SAN, but these proprietary
interfaces are not used within IBM TotalStorage Productivity Center. The only exception is the
application programming interface (API) that Brocade has made available to manage their
Fibre Channel switches. This API is used within IBM TotalStorage Productivity Center for
Fabric.

2.2.1 ANSI standards
Several standards have been published for the inband management of storage devices, for
example, SCSI Enclosure Services (SES).

T11 committee
Since the 1970s, the objective of the ANSI T11 committee is to define interface standards for
high-performance and mass storage applications. Since that time, the committee has
completed work on three projects:
High-Performance Parallel Interface (HIPPI)
Intelligent Peripheral Interface (IPI)
Single-Byte Command Code Sets Connection (SBCON)

Currently the group is working on Fibre Channel (FC) and Storage Network Management
(SM) standards.

Fibre Channel Generic Services
The Fibre Channel Generic Services (FC-GS-3) Directory Service and the Management
Service are being used within IBM TotalStorage Productivity Center for the SAN
management. The availability and level of function depends on the implementation by the
individual vendor. IBM TotalStorage Productivity Center for Fabric uses this standard.

2.2.2 Web-Based Enterprise Management
Web-Based Enterprise Management (WBEM) is an initiative of the Distributed Management
Task Force (DTMF) with the objective to enable the management of complex IT
environments. It defines a set of management and Internet standard technologies to unify the
management of complex IT environments.

The three main conceptual elements of the WBEM initiative are:
Common Information Model (CIM)
CIM is a formal object-oriented modeling language that is used to describe the
management aspects of systems. See also “Common Information Model” on page 47.
xmlCIM
This is a grammar to describe CIM declarations and messages used by the CIM protocol.
Hypertext Transfer Protocol (HTTP)
HTTP is used as a way to enable communication between a management application and
a device that both use CIM.


The WBEM architecture defines the following elements:
CIM Client
The CIM Client is a management application similar to IBM TotalStorage Productivity
Center that uses CIM to manage devices. A CIM Client can reside anywhere in the
network, because it uses HTTP to talk to CIM Object Managers and Agents.
CIM Managed Object
A CIM Managed Object is a hardware or software component that can be managed by a
management application using CIM.
CIM Agent
The CIM Agent is embedded into a device or it can be installed on the server using the
CIM provider as the translator of device’s proprietary commands to CIM calls, and
interfaces with the management application (the CIM Client). The CIM Agent is linked to
one device.
CIM Provider
A CIM Provider is the element that translates CIM calls to the device-specific commands.
It is like a device driver. A CIM Provider is always closely linked to a CIM Object Manager
or CIM Agent.
CIM Object Manager
A CIM Object Manager (CIMOM) is a part of the CIM Server that links the CIM Client to
the CIM Provider. It enables a single CIM Agent to talk to multiple devices.
CIM Server
A CIM Server is the software that runs the CIMOM and the CIM provider for a set of
devices. This approach is used when the devices do not have an embedded CIM Agent.
This term is often not used. Instead people often use the term CIMOM when they really
mean the CIM Server.

2.2.3 Storage Networking Industry Association
The Storage Networking Industry Association (SNIA) defines standards that are used within
IBM TotalStorage Productivity Center. You can find more information on the Web at:
http://www.snia.org

Fibre Channel Common HBA API
The Fibre Channel Common HBA API is used as a standard for inband storage management.
It acts as a bridge between a SAN management application like Fabric Manager and the Fibre
Channel Generic Services. IBM TotalStorage Productivity Center for Fabric Agent uses this
standard.

Storage Management Initiative - Specification
SNIA has fully adopted and enhanced the CIM for Storage Management in its SMI-S.
SMI-S was launched in mid-2002 to create and develop a universal open interface for
managing storage devices including storage networks.


The idea behind SMI-S is to standardize the management interfaces so that management
applications can use these and provide cross device management. This means that a newly
introduced device can be immediately managed as it conforms to the standards. SMI-S
extends CIM and WBEM with the following features:
A single management transport
Within the WBEM architecture, the CIM-XML over HTTP protocol was selected for this
transport in SMI-S.
A complete, unified, and rigidly specified object model
SMI-S defines profiles and recipes within the CIM that enables a management client to
reliably use a component vendor’s implementation of the standard, such as the control of
LUNs and zones in the context of a SAN.
Consistent use of durable names
As a storage network configuration evolves and is re-configured, key long-lived resources,
such as disk volumes, must be uniquely and consistently identified over time.
Rigorously documented client implementation considerations
SMI-S provides client developers with vital information for traversing CIM classes within a
device or subsystem and between devices and subsystems such that complex storage
networking topologies can be successfully mapped and reliably controlled.
An automated discovery system
SMI-S compliant products, when introduced in a SAN environment, automatically
announce their presence and capabilities to other constituents using SLP (see 2.3.1, “SLP
architecture” on page 38).
Resource locking
SMI-S compliant management applications from multiple vendors can exist in the same
storage device or SAN and cooperatively share resources through a lock manager.

The models and protocols in the SMI-S implementation are platform-independent, enabling
application development for any platform, and enabling them to run on different platforms. The
SNIA also provides interoperability tests which help vendors to test their applications and
devices if they conform to the standard.

Managers or components that use this standard include:

2.2.4 Simple Network Management Protocol
The SNMP is an Internet Engineering Task Force (IETF) protocol for monitoring and
managing systems and devices in a network. Functions supported by the SNMP protocol are
the request and retrieval of data, the setting or writing of data, and traps that signal the
occurrence of events. SNMP is a method that enables a management application to query
information from a managed device.

The managed device has software running that sends and receives the SNMP information.
This software module is usually called the SNMP agent.


Device management
An SNMP manager can read information from an SNMP agent to monitor a device. Therefore
the device needs to be polled on an interval basis. The SNMP manager can also change the
configuration of a device, by setting certain values to corresponding variables.

Managers or components that use these standards include the IBM TotalStorage Productivity
Center for Fabric.

Traps
A device can also be set up to send a notification to the SNMP manager (this is called a trap)
to asynchronously inform this SNMP manager of a status change. Depending on the existing
environment and organization, it is likely that your environment already has an SNMP
management application in place.

The managers or components that use this standard are:
IBM TotalStorage Productivity Center for Fabric (sending and receiving of traps)
IBM TotalStorage Productivity Center for Data can be set up to send traps, but does not
receive traps
IBM TotalStorage Productivity Center for Disk and IBM TotalStorage Productivity Center
for Replication events can be sent as SNMP traps by utilizing the IBM Director
infrastructure.

Management Information Base
SNMP use a hierarchical structured Management Information Base (MIB) to define the
meaning and the type of a particular value. An MIB defines managed objects that describe
the behavior of the SNMP entity, which can be anything from a IP router to a storage
subsystem. The information is organized in a tree structure.

Note: For more information about SNMP, refer to TCP/IP Tutorial and Technical Overview,
GG24-3376.

IBM TotalStorage Productivity Center for Data MIB file
For users planning to use the IBM TotalStorage Productivity Center for Data SNMP trap alert
notification capabilities, an SNMP MIB is included in the server installation. You can find the
SNMP MIB in the file tivoli_install_directory/snmp/tivoliSRM.MIB.

The MIB is provided for use by your SNMP management console software. Most SNMP
management station products provide a program called an MIB compiler that can be used to
import MIBs. This allows you to better view Productivity Center for Data generated SNMP
traps from within your management console software. Refer to your management console
software documentation for instructions on how to compile or import a third-party MIB.

2.2.5 Fibre Alliance MIB
The Fibre Alliance has defined an MIB for the management of storage devices. The Fibre
Alliance is presenting the MIB to the IETF standardization. The intention of putting together
this MIB was to have one MIB that covers most (if not all) of the attributes of storage devices
from multiple vendors. The idea was to have only one MIB that is loaded onto an SNMP
manager, one MIB file for each component. However, this requires that all devices comply
with that standard MIB, which is not always the case.


Note: This MIB is not part of IBM TotalStorage Productivity Center.

To learn more about Fibre Alliance and MIB, refer to the following Web sites:
http://www.fibrealliance.org
http://www.fibrealliance.org/fb/mib_intro.htm

2.3 Service Location Protocol (SLP) overview
The SLP is an IETF standard, documented in Request for Comments (RFCs) 2165, 2608,
2609, 2610, and 2614. SLP provides a scalable framework for the discovery and selection of
network services.

SLP enables the discovery and selection of generic services, which can range in function
from hardware services such as those for printers or fax machines, to software services such
as those for file servers, e-mail servers, Web servers, databases, or any other possible
services that are accessible through an IP network.

Traditionally, to use a particular service, an end-user or client application needs to supply the
host name or network IP address of that service. With SLP, however, the user or client no
longer needs to know individual host names or IP addresses (for the most part). Instead, the
user or client can search the network for the desired service type and an optional set of
qualifying attributes.

For example, a user can specify to search for all available printers that support PostScript,
based on the given service type (printers), and the given attributes (PostScript). SLP
searches the user’s network for any matching services and returns the discovered list to the
user.

2.3.1 SLP architecture
The SLP architecture includes three major components, a Service Agent (SA), a User Agent
(UA), and a Directory Agent (DA). The SA and UA are required components in an SLP
environment, where the SLP DA is optional.

The SMI-S specification introduces SLP as the method for the management applications (the
CIM clients) to locate managed objects. In SLP, an SA is used to report to UAs that a service
that has been registered with the SA is available.

The following sections describe each of these components.

Service Agent (SA)
The SLP SA is a component of the SLP architecture that works on behalf of one or more
network services to broadcast the availability of those services by using broadcasts. The SA
replies to external service requests using IP unicasts to provide the requested information
about the registered services, if it is available.


The SA can run in the same process or in a different process as the service itself. In either
case, the SA supports registration and de-registration requests for the service (as shown in
the right part of Figure 2-5). The service registers itself with the SA during startup, and
removes the registration for itself during shutdown. In addition, every service registration is
associated with a life-span value, which specifies the time that the registration will be active.
In the left part of the diagram, you can see the interaction between a UA and the SA.

Figure 2-5 SLP SA interactions (without SLP DA)

A service is required to reregister itself periodically, before the life-span of its previous
registration expires. This ensures that expired registration entries are not kept. For instance, if
a service becomes inactive without removing the registration for itself, that old registration is
removed automatically when its life span expires. The maximum life span of a registration is
65535 seconds (about 18 hours).

User Agent (UA)
The SLP UA is a process working on the behalf of the user to establish contact with some
network service. The UA retrieves (or queries for) service information from the Service
Agents or Directory Agents.

The UA is a component of SLP that is closely associated with a client application or a user
who is searching for the location of one or more services in the network. You can use the SLP
UA by defining a service type that you want the SLP UA to locate. The SLP UA then retrieves
a set of discovered services, including their service Uniform Resource Locator (URL) and any
service attributes. You can then use the service’s URL to connect to the service.

The SLP UA locates the registered services, based on a general description of the services
that the user or client application has specified. This description usually consists of a service
type, and any service attributes, which are matched against the service URLs registered in
the SLP Service Agents.

The SLP UA usually runs in the same process as the client application, although it is not
necessary to do so. The SLP UA processes find requests by sending out multicast messages
to the network and targeting all SLP SAs within the multicast range with a single User
Datagram Protocol (UDP) message. The SLP UA can, therefore, discover these SAs with a
minimum of network overhead. When an SA receives a service request, it compares its own
registered services with the requested service type and any service attributes, if specified,
and returns matches to the UA using a unicast reply message.


The SLP UA follows the multicast convergence algorithm and sends repeated multicast
messages until no new replies are received. The resulting set of discovered services,
including their service URL and any service attributes, are returned to the client application or
user. The client application or user is then responsible for contacting the individual services,
as needed, using the service’s URL (see Figure 2-6).

Figure 2-6 SLP UA interactions without SLP DA

An SLP UA is not required to discover all matching services that exist in the network, but only
enough of them to provide useful results. This restriction is mainly due to the transmission
size limits for UDP packets. They can be exceeded when there are many registered services
or when the registered services have lengthy URLs or a large number of attributes.

However, in most modern SLP implementations, the UAs can recognize truncated service
replies and establish TCP connections to retrieve all of the information of the registered
services. With this type of UA and SA implementation, the only exposure that remains is when
there are too many SAs within the multicast range. This can cut short the multicast
convergence mechanism. This exposure can be mitigated by the SLP administrator by setting
up one or more SLP DAs.

Directory Agent
The SLP DA is an optional component of SLP that collects and caches network service
broadcasts. The DA is primarily used to simplify SLP administration and to improve SLP
performance.

You can consider the SLP DA as an intermediate tier in the SLP architecture. It is placed
between the UAs and the SAs so that both UAs and SAs communicate only with the DA
instead of with each other. This eliminates a large portion of the multicast request or reply
traffic in the network. It also protects the SAs from being overwhelmed by too many service
requests if there are many UAs in the environment.


Figure 2-7 shows the interactions of the SLP UAs and SAs in an environment with SLP DAs.

Figure 2-7 SLP User Agent interactions with User Agent and Service Agent

When SLP DAs are present, the behavior of both SAs and UAs changes significantly. When
an SA is first initializing, it performs a DA discovery using a multicast service request. It also
specifies the special, reserved service type service:directory-agent. This process is also
called active DA discovery. It is achieved through the same mechanism as any other
discovery using SLP.

Similarly, in most cases, an SLP UA also performs active DA discovery using multicasting
when it first starts. However, if the SLP UA is statically configured with one or more DA
addresses, it uses those addresses instead. If it is aware of one or more DAs, either through
static configuration or active discovery, it sends unicast service requests to those DAs instead
of multicasting to SAs. The DA replies with unicast service replies, providing the requested
service URLs and attributes. Figure 2-8 shows the interactions of UAs and SAs with DAs,
during active DA discovery.

Figure 2-8 SLP Directory Agent discovery interactions


The SLP DA functions similarly to an SLP SA, receiving registration and deregistration
requests, and responding to service requests with unicast service replies. There are a couple
of differences, where DAs provide more functionality than SAs. One area, mentioned
previously, is that DAs respond to service requests of the service:directory-agent service type
with a DA advertisement response message, passing back a service URL containing the DA’s
IP address. This allows SAs and UAs to perform active discovery on DAs.

One other difference is that when a DA first initializes, it sends a multicast DA advertisement
message to advertise its services to any existing SAs (and UAs) that may already be active in
the network. UAs can optionally listen for, and SAs are required to listen for, such
advertisement messages. This listening process is also sometimes called passive DA
discovery. When the SA finds a new DA through passive DA discovery, it sends registration
requests for all its currently registered services to that new DA.

Figure 2-9 shows the interactions of DAs with SAs and UAs, during passive DA discovery.

Figure 2-9 Service Location Protocol passive DA discovery

Why use an SLP DA?
The primary reason to use DAs is to reduce the amount of multicast traffic involved in service
discovery. In a large network with many UAs and SAs, the amount of multicast traffic involved
in service discovery can become so large that network performance degrades. By deploying
one or more DAs, UAs must unicast to DAs for service and SAs must register with DAs using
unicast. The only SLP-registered multicast in a network with DAs is for active and passive DA
discovery.

SAs register automatically with any DAs they discover within a set of common scopes.
Consequently, DAs within the UAs scopes reduce multicast. By eliminating multicast for
normal UA request, delays and timeouts are eliminated.

DAs act as a focal point for SA and UA activity. Deploying one or several DAs for a collection
of scopes provides a centralized point for monitoring SLP activity. You can deploy any number
of DAs for a particular scope or scopes, depending on the need to balance the load.

In networks without multicasting enabled, you can configure SLP to use broadcast. However,
broadcast is inefficient, because it requires each host to process the message. Broadcast
also does not normally propagate across routers. As a result, in a network without multicast,
DAs can be deployed on multihomed hosts to bridge SLP advertisements between the
subnets.


When to use DAs
Use DAs in your enterprise when any of the following conditions are true:
Multicast SLP traffic exceeds 1% of the bandwidth on your network, as measured by
snoop.
UA clients experience long delays or timeouts during multicast service request.
You want to centralize monitoring of SLP service advertisements for particular scopes on
one or several hosts.
Your network does not have multicast enabled and consists of multiple subnets that must
share services.

SLP communication
SLP uses three methods to send messages across an IP network: unicast, broadcast, or
multicast. Data can be sent to one single destination (unicast) or to multiple destinations that
are listening at the same time (multicast). The difference between a multicast and a broadcast
is quite important. A broadcast addresses all stations in a network. Multicast messages are
only used by those stations that are members of a multicast group (that have joined a
multicast group).

Unicast
The most common communication method, unicast, requires that a sender of a message
identifies one and only one target of that message. The target IP address is encoded within
the message packet, and is used by the routers along the network path to route the packet to
the proper destination.

If a sender wants to send the same message to multiple recipients, then multiple messages
must be generated and placed in the network, one message per recipient. When there are
many potential recipients for a particular message, then this places an unnecessary strain on
the network resources, since the same data is duplicated many times, where the only
difference is the target IP address encoded within the messages.

Broadcast
In cases where the same message must be sent to many targets, broadcast is a much better
choice than unicast, since it puts much less strain in the network. Broadcasting uses a special
IP address, 255.255.255.255, which indicates that the message packet is intended to be sent
to all nodes in a network. As a result, the sender of a message needs to generate only a
single copy of that message, and can still transmit it to multiple recipients, that is to all
members of the network.

The routers multiplex the message packet, as it is sent along all possible routes in the
network to reach all possible destinations. This puts much less strain on the network
bandwidth, since only a single message stream enters the network, as opposed to one
message stream per recipient. However, it puts much more strain on the individual nodes
(and routers) in the network, since every node receives the message, even though most likely
not every node is interested in the message. This means that those members of the network
that were not the intended recipients, who receive the message anyway, must receive the
unwanted message and discard it. Due to this inefficiency, in most network configurations,
routers are configured to not forward any broadcast traffic. This means that any broadcast
messages can only reach nodes on the same subnet as the sender.

Multicast
The ability of the SLP to automatically discover services that are available in the network,
without a lot of setup or configuration, depends in a large part on the use of IP multicasting. IP
multicasting is a broad subject in itself, and only a brief and simple overview is provided here.


Multicasting can be thought of as more sophisticated broadcast, which aims to solve some of
the inefficiencies inherent in the broadcasting mechanism. With multicasting, again the
sender of a message has to generate only a single copy of the message, saving network
bandwidth. However unlike broadcasting, with multicasting, not every member of the network
receives the message. Only those members who have explicitly expressed an interest in the
particular multicast stream receive the message.

Multicasting introduces a concept called a multicast group, where each multicast group is
associated with a specific IP address. A particular network node (host) can join one or more
multicast groups, which notifies the associated router or routers that there is an interest in
receiving multicast streams for those groups. When the sender, who does not necessarily
have to be part of the same group, sends messages to a particular multicast group, that
message is routed appropriately to only those subnets, which contain members of that
multicast group. This avoids flooding the entire network with the message, as is the case for
broadcast traffic.

Multicast addresses
The Internet Assigned Numbers Authority (IANA), which controls the assignment of IP
addresses, has assigned the old Class D IP address range to be used for IP multicasting. Of
this entire range, which extends from 224.0.0.0 to 239.255.255.255, the 224.0.0.* addresses
are reserved for router management and communication. Some of the 224.0.1.* addresses
are reserved for particular standardized multicast applications. Each of the remaining
addresses corresponds to a particular general purpose multicast group.

The Service Location Protocol uses address 239.255.255.253 for all its multicast traffic. The
port number for SLP is 427, for both unicast and multicast.

Configuration recommendations
Ideally, after IBM TotalStorage Productivity Center is installed, it would discover all storage
devices that it can physically reach over the IP network. However in most situations, this is
not the case. This is primarily due to the previously mentioned limitations of multicasting and
the fact that the majority of routers have multicasting disabled by default. As a result, in most
cases without any additional configuration, IBM TotalStorage Productivity Center discovers
only those storage devices that reside in its own subnet, but no more. The following sections
provide some configuration recommendations to enable TotalStorage Productivity Center to
discover a larger set of storage devices.

Router configuration
The vast majority of the intelligence that allows multicasting to work is implemented in the
router operating system software. As a result, it is necessary to properly configure the routers
in the network to allow multicasting to work effectively. Unfortunately, there is a dizzying array
of protocols and algorithms which can be used to configure particular routers to enable
multicasting. These are the most common ones:
Internet Group Management Protocol (IGMP) is used to register individual hosts in
particular multicast groups, and to query group membership on particular subnets.
Distance Vector Multicast Routing Protocol (DVMRP) is a set of routing algorithms that
use a technique called Reverse Path Forwarding to decide how multicast packets are to
be routed in the network.
Protocol-Independent Multicast (PIM) comes in two varieties: dense mode (PIM-DM) and
sparse mode (PIM-SM). They are optimized to networks where either a large percentage
of nodes require multicast traffic (dense), or a small percentage require the traffic
(sparse).


Multicast Open Shortest Path First (MOSPF) is an extension of OSPF, a “link-state”
unicast routing protocol that attempts to find the shortest path between any two networks
or subnets to provide the most optimal routing of packets.

The routers of interest are all those which are associated with subnets that contain one or
more storage devices which are to be discovered and managed by TotalStorage Productivity
Center. You can configure the routers in the network to enable multicasting in general, or at
least to allow multicasting for the SLP multicast address, 239.255.255.253, and port, 427.
This is the most generic solution and permits discovery to work the way that it was intended
by the designers of SLP.

To properly configure your routers for multicasting, refer to your router manufacturer’s
reference and configuration documentation. Although older hardware may not support
multicasting, all modern routers do. However, in most cases, multicast support is disabled by
default, which means that multicast traffic is sent only among the nodes of a subnet but is not
forwarded to other subnets. For SLP, this means that service discovery is limited to only those
agents which reside in the same subnet.

Firewall configuration
In the case where one or more firewalls are used between TotalStorage Productivity Center
and the storage devices that are to be managed, the firewalls need to be configured to pass
traffic in both directions, as SLP communication is two way. This means that when
TotalStorage Productivity Center, for example, queries an SLP DA that is behind a firewall for
the registered services, the response will not use an already opened TCP/IP session but will
establish another connection in the direction from the SLP DA to the TotalStorage
Productivity Center. For this reason, port 427 should be opened in both directions, otherwise
the response will not be received and TotalStorage Productivity Center will not recognize
services offered by this SLP DA.

SLP DA configuration
If router configuration is not feasible, another technique is to use SLP DAs to circumvent the
multicast limitations. Since with statically configured DAs, all service requests are unicast
instead of multicast by the UA, it is possible to simply configure one DA for each subnet that
contains storage devices which are to be discovered by TotalStorage Productivity Center.

One DA is sufficient for each of such subnets, although more can be configured without harm,
perhaps for reasons of fault tolerance. Each of these DAs can discover all services within its
own subnet, but no other services outside its own subnet. To allow Productivity Center to
discover all of the devices, you must statically configure it with the addresses of each of these
DAs. You accomplish this using the IBM Director GUI’s Discovery Preference panel. From the
MDM SLP Configuration tab, you can enter a list of DA addresses.

As described previously, Productivity Center unicasts service requests to each of these
statically configured DAs, but also multicasts service requests on the local subnet on which
Productivity Center is installed. Figure 2-10 on page 46 displays a sample environment where
DAs have been used to bridge the multicast gap between subnets in this manner.

Note: At this time, you cannot set up IBM TotalStorage Productivity Center for Data to use
remote DAs such as Productivity Center for Disk and Productivity Center for Replication.
You need to define all remote CIM Agents by creating a new entry in the CIMOM Login
panel or you can register remote services in DA which resides in local subnet. Refer to
“Configuring the CIM Agents” on page 290 for detailed information.


Figure 2-10 Recommended SLP configuration

You can easily configure an SLP DA by changing the configuration of the SLP SA included as
part of an existing CIM Agent installation. This causes the program that normally runs as an
SLP SA to run as an SLP DA instead. The procedure to perform this configuration is
explained in 6.2, “SLP DA definition” on page 248. Note that the change from SA to DA does
not affect the CIMOM service of the subject CIM Agent, which continues to function as
normal, sending registration and de-registration commands to the DA directly.

SLP configuration with services outside local subnet

SLA DA and SA can also be configured to cache CIM services information from non-local
subnets. Usually CIM Agents or CIMOMs will have local SLP SA function. When there is a
need to discover CIM services outside the local subnet and the network configuration does
not permit the use of SLP DA in each of them (for example, firewall rules do not allow two way
communication on port 427), remote services can be registered on the SLP DA in the local
subnet.

This configuration can be done by using slptool, which is part of SLP installation packages.
Such registration is not persistent across system restarts. To achieve persistent registration of
services outside of the local subnet, these services need to be defined in the registration file
used by SLP DA at startup.

Refer to 5.7.3, “Setting up the Service Location Protocol Directory Agent” on page 221 for
information on setting up the slp.reg file.


2.3.2 Common Information Model
The CIM Agent provides a means by which a device can be managed by common building
blocks rather than proprietary software. If a device is CIM-compliant, software that is also
CIM-compliant can manage the device. Vendor applications can benefit from adopting the
common information model because they can manage CIM-compliant devices in a common
way, rather than using device-specific programming interfaces. Using CIM, you can perform
tasks in a consistent manner across devices and vendors.

CIM uses schemas as a kind of class library to define objects and methods. The schemas
can be categorized into three types:
Core schema: Defines classes and relationships of objects
Common schema: Defines common components of systems
Extension schema: Entry point for vendors to implement their own schema

The CIM/WBEM architecture defines the following elements:
Agent code or CIM Agent
An open-systems standard that interprets CIM requests and responses as they transfer
between the client application and the device. The Agent is embedded into a device, which
can be hardware or software.
CIM Object Manager
The common conceptual framework for data management that receives, validates, and
authenticates the CIM requests from the client application. It then directs the requests to
the appropriate component or a device provider such as a CIM Agent.
Client application or CIM Client
A storage management program, such as TotalStorage Productivity Center, that initiates
CIM requests to the CIM Agent for the device. A CIM Client can reside anywhere in the
network, because it uses HTTP to talk to CIM Object Managers and Agents.
Device or CIM Managed Object
A Managed Object is a hardware or software component that can be managed by a
management application by using CIM, for example, a IBM SAN Volume Controller.
Device provider
A device-specific handler that serves as a plug-in for the CIMOM. That is, the CIMOM
uses the handler to interface with the device.

Note: The terms CIM Agent and CIMOM are often used interchangeably. At this time, few
devices come with an integrated CIM Agent. Most devices need a external CIMOM for CIM
to enable management applications (CIM Clients) to talk to the device.

For ease of installation, IBM provides an Integrated Configuration Agent Technology
(ICAT), which is a bundle that includes the CIMOM, the device provider, and an SLP SA.

Integrating legacy devices into the CIM model
Since these standards are still evolving, we cannot expect that all devices will support the
native CIM interface. Because of this, the SMI-S is introducing CIM Agents and CIM Object
Managers. The agents and object managers bridge proprietary device management to device
management models and protocols used by SMI-S. The agent is used for one device and an
object manager for a set of devices. This type of operation is also called proxy model and is
shown in Figure 2-11 on page 48.


The CIM Agent or CIMOM translates a proprietary management interface to the CIM
interface. The CIM Agent for the IBM TotalStorage ESS includes a CIMOM inside it.

In the future, more and more devices will be native CIM compliant, and will therefore have a
built-in Agent as shown in the “Embedded Model” in Figure 2-11. When widely adopted,
SMI-S will streamline the way that the entire storage industry deals with management.
Management application developers will no longer have to integrate incompatible
feature-poor interfaces into their products. Component developers will no longer have to push
their unique interface functionality to application developers. Instead, both will be better able
to concentrate on developing features and functions that have value to end-users. Ultimately,
faced with reduced costs for management, end-users will be able to adopt storage-networking
technology faster and build larger, more powerful networks.

CIM Client
Management
Application
0..n

CIMxml
CIM operations over http [TCP/IP]

Agent Object Manager
0..n Provider 0..n
1
Proprietary 1
1 n
Proprietary
Agent
Device or Device or
Subsystem Device or
Subsystem 0..n Subsystem
Proxy Model Embedded Model Proxy Model

Figure 2-11 CIM Agent and Object Manager overview

CIM Agent implementation
When a CIM Agent implementation is available for a supported device, the device may be
accessed and configured by management applications using industry-standard
XML-over-HTTP transactions. This interface enables IBM TotalStorage Productivity Center
for Data, IBM TotalStorage Productivity Center for Disk, IBM TotalStorage Productivity Center
for Replication, IBM Director, and vendor tools to manage the SAN infrastructure more
effectively.

By implementing a standard interface over all devices, an open environment is created in
which tools from a variety of vendors can work together. This reduces the cost of developing
integrated management applications, installing and configuring management applications,
and managing the SAN infrastructure. Figure 2-12 on page 49 shows an overview of the CIM
Agent.


Figure 2-12 CIM Agent overview

The CIM Agent includes a CIMOM, which adapts various devices using a plug-in called a
provider. The CIM Agent can work as a proxy or can be embedded in storage devices. When
the CIM Agent is installed as a proxy, the IBM CIM Agent can be installed on the same server
that supports the device user interface.

CIM Object Manager
The SNIA SMI-S standard designates that either a proxy or an embedded agent may be used
to implement CIM. In each case, the CIM objects are supported by a CIM Object Manager.
External applications communicate with CIM through HTTP to exchange XML messages that
are used to configure and manage the device.

In a proxy configuration, the CIMOM runs outside of the device and can manage multiple
devices. In this case, a provider component is installed into the CIMOM to enable the CIMOM
to manage specific devices such as the ESS or SAN Volume Controller. The providers adapt
the CIMOM to work with different devices and subsystems. In this way, a single CIMOM
installation can be used to access more than one device type and more than one device of
each type on a subsystem.

The CIMOM acts as a catcher for requests that are sent from storage management
applications. The interactions between the catcher and sender use the language and models
defined by the SMI-S standard. This enables storage management applications, regardless of
vendor, to query status and perform command and control using XML-based CIM
interactions.

2.4 Component interaction
This section provides an overview of the interactions between the different components by
using standardized management methods and protocols.

2.4.1 CIMOM discovery with SLP
The SMI-S specification introduces SLP as the method for the management applications (the
CIM clients) to locate managed objects. SLP is explained in more detail in 2.3, “Service
Location Protocol (SLP) overview” on page 38. Figure 2-13 on page 50 shows the interaction
between CIMOMs and SLP components.


Lock CIM Client Directory
Manager Management Manager
Application
SA
SA UA 0..n DA
0..n 0..n

SLP TCP/IP

CIMxml
CIM operations over http [TCP/IP]

SA
SA
Agent Object Manager
0..n Provider 0..n
1
1
Proprietary SA 1
Proprietary
n
Agent
Device or Device or
Subsystem Device or
Subsystem 0..n Subsystem
Proxy Model Embedded Model Proxy Model

Figure 2-13 SMI-S extensions to WBEM/CIM

2.4.2 How CIM Agent works
The CIM Agent typically works as explained in the following sequence and as shown in
Figure 2-14 on page 51:
1. The client application locates the CIMOM by calling an SLP directory service.
2. The CIMOM is invoked.
3. The CIMOM registers itself to the SLP and supplies its location, IP address, port number,
and the type of service it provides.
4. With this information, the client application starts to directly communicate with the CIMOM.
5. The client application sends CIM requests to the CIMOM. As requests arrive, the CIMOM
validates and authenticates each request.
6. The CIMOM directs the requests to the appropriate functional component of the CIMOM
or to a device provider.
7. The provider makes calls to a device-unique programming interface on behalf of the
CIMOM to satisfy client application requests.
8. — 10. The client application requests are made.


Figure 2-14 CIM Agent work flow

2.5 Tivoli Common Agent Services
The Tivoli Common Agent Services is a new concept with the goal to provide a set of
functions for the management of agents that will be common to all Tivoli products. At the time
of this writing, IBM TotalStorage Productivity Center for Fabric and IBM TotalStorage
Productivity Center for Data are the first applications that use this new concept. See
Figure 2-15 on page 52 for an overview of the three elements in the Tivoli Common Agent
Services infrastructure.

In each of the planning and installation guides of the Productivity Center for Fabric and
Productivity Center for Data, there is a chapter that provides information about the benefits,
system requirements and sizing, security considerations, and the installation procedures.

The Agent Manager is the central network element, that together with the distributed
Common Agents, builds an infrastructure which is used by other applications to deploy and
manage an agent environment. Each application uses a Resource Manager that is built into
the application server (Productivity Center for Data or Productivity Center for Fabric) to
integrate in this environment.

Note: You can have multiple Resource Managers of the same type using a single Agent
Manager. This may be necessary to scale the environment when, for example, one Data
Manager cannot handle the load any more.

The Agents will be managed by only one of the Data Managers as in this example.


Figure 2-15 Tivoli Common Agent Services

The Common Agent provides the platform for the application specific agents. Depending on
the tasks for which a subagent is used, the Common Agent is installed on the customers’
application servers, desktop PCs, or notebooks.

Note: In different documentation, Readme files, directory and file names, you also see the
terms Common Endpoint, Endpoint, or simply EP. This always refers to the Common
Agent, which is part of the Tivoli Common Agent Services.

The Common Agent talks to the application specific subagent, with the Agent Manager and
the Resource Manager, but the actual system level functions are invoked by the subagent.
The information that the subagent collects is sent directly to the Resource Manager by using
the application’s native protocol. This is enabled to have down-level agents in the same
environment, as the new agents that are shipped with the IBM TotalStorage Productivity
Center.

Certificates are used to validate if a requester is allowed to establish a communication. Demo
keys are supplied to quickly set up and configure a small environment, since every installation
CD uses the same certificates, this is not secure. If you want to use Tivoli Common Agent
Services in a production environment, we recommend that you use your own keys that can be
created during the Tivoli Agent Manager installation.

One of the most important certificates is stored in the agentTrust.jks file. The certificate can
also be created during the installation of Tivoli Agent Manager. If you do not use the demo
certificates, you need to have this file available during the installation of the Common Agent
and the Resource Manager.

This file is locked with a password (the agent registration password) to secure the access to
the certificates. You can use the ikeyman utility in the javajre subdirectory to verify your
password.


2.5.1 Tivoli Agent Manager
The Tivoli Agent Manager requires a database to store information in what is called the
registry. Currently there are three options for installing the database: using IBM
Cloudscape™ (provided on the installation CD), a local DB2 database, or a remote DB2
database. Since the registry does not contain much information, using the Cloudscape
database is OK. In our setup described later in the book, we chose a local DB2 database,
because the DB2 database was required for another component that was installed on the
same machine.

WebSphere Application Server is the second prerequisite for the Tivoli Agent Manager. This
is installed if you use the Productivity Center Suite Installer or if you choose to use the Tivoli
Agent Manager installer. We recommend that you do not install WebSphere Application
Server manually.

Three dedicated ports are used by the Agent Manager (9511-9513). Port 9511 is the most
important port because you have to enter this port during the installation of a Resource
Manager or Common Agent, if you choose to change the defaults.

When the WebSphere Application Server is being installed, make sure that the Microsoft
Internet Information Server (IIS) is not running, or even better that it is not installed. Port 80 is
used by the Tivoli Agent Manager for the recovery of agents that can no longer communicate
with the manager, because of lost passwords or certificates. This Agent Recovery Service is
located by a DNS entry with the unqualified host name of TivoliAgentRecovery.

Periodically, check the Agent Manager log for agents that are unable to communicate with the
Agent Manager server. The recovery log is in the %WAS_INSTALL_ROOT%AgentManager
logsSystemOut.log file. Use the information in the log file to determine why the agent could
not register and then take corrective action.

During the installation, you also have to specify the agent registration password and the
Agent Registration Context Root. The password is stored in the AgentManager.properties file
on the Tivoli Agent Manager. This password is also used to lock the agentTrust.jks certificate
file.

Important: A detailed description about how to change the password is available in the
corresponding Resource Manager Planning and Installation Guide. Since this involves
redistributing the agentTrust.jks files to all Common Agents, we encourage you to use your
own certificates from the beginning.

To control the access from the Resource Manager to the Common Agent, certificates are
used to make sure that only an authorized Resource Manager can install and run code on a
computer system. This certificate is stored in the agentTrust.jks and locked with the agent
registration password.

2.5.2 Common Agent
As mentioned earlier, the Common Agent is used as a platform for application specific
agents. These agents sometimes are called subagents.

The subagents can be installed using two different methods:
Using an application specific installer
From a central location once the Common Agent is installed


When you install the software, the agent has to register with the Tivoli Agent Manager. During
this procedure, you need to specify the registration port on the manager (by default 9511).
Furthermore, you need to specify an agent registration password. This registration is
performed by the Common Agent, which is installed automatically if not already installed.

If the subagent is deployed from a central location, the port 9510 is by default used by the
installer (running on the central machine), to communicate with the Common Agent to
download and install the code. When this method is used, no password or certificate is
required, because these were already provided during the Common Agent installation on the
machine.

If you choose to use your own certificate during the Tivoli Agent Manager installation, you
need to supply it for the Common Agent installation.


Part 2

Part 2 Installing the
IBM TotalStorage
Productivity Center
base product suite
In this part of the book we provide information to help you successfully install the prerequisite
products that are required before you can install the IBM TotalStorage Productivity Center
product suite. This includes installing:
DB2
IBM Director
WebSphere Application Server
Tivoli Agent Manager


3

Chapter 3. Installation planning and
considerations

IBM TotalStorage Productivity Center is made up of several products which can be installed
individually, as a complete suite, or any combination in between. By installing multiple
products, a synergy is created which allows the products to interact with each other to provide
a more complete solution to help you meet your business storage management objectives.

This chapter contains information that you will need before beginning the installation. It also
discusses the supported environments and pre-installation tasks.


3.1 Configuration
You can install the storage management components of IBM TotalStorage Productivity
Center on a variety of platforms. However, for the IBM TotalStorage Productivity Center suite,
when all four manager components are installed on the same system, the only common
platforms for the managers are:
Windows 2000 Server with Service Pack 4
Windows 2000 Advanced Server
Windows 2003 Enterprise Edition

Note: Refer to the following Web site for the updated support summaries, including specific
software, hardware, and firmware levels supported:
http://www.storage.ibm.com/software/index.html

If you are using the storage provisioning workflows, you must install IBM TotalStorage
Productivity Center for Disk and TotalStorage Productivity Center for Replication and IBM
TotalStorage Productivity Center for Fabric on the same machine.

Because of processing requirements, we recommend that you install IBM Tivoli Provisioning
Manager on a separate Windows machine.

3.2 Installation prerequisites
This section lists the minimum prerequisites for installing IBM TotalStorage Productivity
Center.

Hardware
The following hardware is required:
Dual Pentium® 4 or Intel® Xeon 2.4 GHz or faster processors
4 GB of DRAM
Network connectivity
Subsystem Device Driver (SDD), for IBM TotalStorage Productivity Center for Fabric
(optional)
5 GB available disk space.

Database
You must comply with the following database requirements:
The installation of DB2 Version 8.2 is part of the Prerequisite Software Installer and is
required by all the managers.
Other databases that are supported are:
– For IBM TotalStorage Productivity Center for Fabric:
• IBM Cloudscape 5.1.60 (provided on the CD)
– For IBM TotalStorage Productivity Center for Data:
• Microsoft SQL Server Version 7.0, 2000
• Oracle 8i, 9i, 9i V2
• Sybase SQL Server (Adaptive Server Enterprise) Version 12.5 or higher
• IBM Cloudscape 5.1.60 (provided on the CD)


3.2.1 TCP/IP ports used by TotalStorage Productivity Center
This section provides an overview of the TCP/IP ports used by IBM TotalStorage Productivity
Center.

TCP/IP ports used by Disk and Replication Manager
The IBM TotalStorage Productivity Center for Disk and IBM TotalStorage Productivity Center
for Replication Manager installation program preconfigures the TCP/IP ports used by
WebSphere. Table 3-1 lists the values that correspond to the WebSphere ports.

Table 3-1 TCP/IP ports for IBM TotalStorage Productivity Center for Disk and Replication Base
Port value WebSphere ports

427 SLP port

2809 Bootstrap port

9080 HTTP Transport port

9443 HTTPS Transport port

9090 Administrative Console port

9043 Administrative Console Secure Server port

5559 JMS Server Direct Address port

5557 JMS Server Security port 5

5558 JMS Server Queued Address port

8980 SOAP Connector Address port

7873 DRS Client Address port

TCP/IP ports used by Agent Manager
The Agent Manager uses the TCP/IP ports listed in Table 3-2.

Table 3-2 TCP/IP ports for Agent Manager
Port value Usage

9511 Registering agents and resource managers

9512 Providing configuration updates
Renewing and revoking certificates
Querying the registry for agent information
Requesting ID resets

9513 Requesting updates to the certificate revocation list
Requesting Agent Manager information
Downloading the truststore file

80 Agent recovery service

Chapter 3. Installation planning and considerations 59

TCP/IP ports used by IBM TotalStorage Productivity Center for Fabric
The Fabric Manager uses the default TCP/IP ports listed in Table 3-3.

Table 3-3 TCP/IP ports for IBM TotalStorage Productivity Center for Fabric
Port value Usage

8080 NetView Remote Web console

9550 HTTP port

9551 Reserved

9552 Reserved

9553 Cloudscape server port

9554 NVDAEMON port

9555 NVREQUESTER port

9556 SNMPTrapPort port on which to get events forwarded from Tivoli NetView

9557 Reserved

9558 Reserved

9559 Tivoli NetView Pager daemon

9560 Tivoli NetView Object Database daemon

9661 Tivoli NetView Topology Manager daemon

9562 Tivoli NetView Topology Manager socket

9563 Tivoli General Topology Manager

9564 Tivoli NetView OVs_PMD request services

9565 Tivoli NetView OVs_PMD management services

9565 Tivoli NetView trapd socket

9567 Tivoli NetView PMD service

9568 Tivoli NetView General Topology map service

9569 Tivoli NetView Object Database event socket

9570 Tivoli NetView Object Collection facility socket

9571 Tivoli NetView Web Server socket

9572 Tivoli NetView SnmpServer


Fabric Manager remote console TCP/IP default ports
The Fabric Manager uses the ports in Table 3-4 for its remote console.

Table 3-4 TCP/IP ports for IBM TotalStorage Productivity Center for Fabric remote console
Port value Usage

9560 HTTP port 9561 Reserved

9561 Reserved

9562 ASF Jakarta Tomcat’s Local Server port

9563 Tomcat’s warp port

9564 NVDAEMON port

9565 NVREQUESTER port

9569 Tivoli NetView Pager daemon

9570 Tivoli NetView Object Database daemon

9571 Tivoli NetView Topology Manager daemon

9572 Tivoli NetView Topology Manager socket

9573 Tivoli General Topology Manager

9574 Tivoli NetView OVs_PMD request services

9575 Tivoli NetView OVs_PMD management services

9576 Tivoli NetView trapd socket

9577 Tivoli NetView PMD service

9578 Tivoli NetView General Topology map service

9579 Tivoli NetView Object Database event socket

9580 Tivoli NetView Object Collection facility socket

9581 Tivoli NetView Web Server socket

9582 Tivoli NetView SnmpServer

Fabric Agents TCP/IP ports
The Fabric Agents use the TCP/IP ports listed in Table 3-5.

Table 3-5 TCP/IP ports for IBM TotalStorage Productivity Center for Fabric Agents
Port value Usage

9510 Common agent

9514 Used to restart the agent

9515 Used to restart the agent


3.2.2 Default databases created during the installation
During the installation of IBM TotalStorage Productivity Center, we recommend that you use
DB2 as the preferred database type. Table 3-6 lists all the default databases that the installer
creates during the installation.

Table 3-6 Default DB2 databases
Application Default database name (DB2)

IBM Director No default; we created database, IBMDIR

Tivoli Agent Manager IBMCDB

IBM TotalStorage Productivity Center for Disk DMCOSERV
and Replication Base

IBM TotalStorage Productivity Center for Disk PMDATA

IBM TotalStorage Productivity Center for ESSHWL
Replication hardware subcomponent

IBM TotalStorage Productivity Center for ELEMCAT
Replication element catalog

IBM TotalStorage Productivity Center for REPMGR
Replication replication manager

IBM TotalStorage Productivity Center for SVCHWL
Replication SVC hardware subcomponent

IBM TotalStorage Productivity Center for Fabric ITSANM

IBM TotalStorage Productivity Center for Data No default; we created Database, TPCDATA

3.3 Our lab setup environment
This section gives a brief overview of what our lab setup environment looked like and what we
used to document the installation.

Server hardware used
We used four IBM Eserver xSeries servers with:
2 x 2.4 GHz CPU per system
4 GB Memory per system
73 GB HDD per system
Windows 2000 with Service Pack 4

System 1
The name of our first system was Colorado. The following applications were installed on this
system:
DB2
IBM Director
WebSphere Application Server
WebSphere Application Server update
Tivoli Agent Manager
IBM TotalStorage Productivity Center for Disk and Replication Base



System 2
The name of our second system was Gallium. The following applications were installed on
this server:
Data Agent

System 3
The name of our third system was PQDISRV. The following applications were installed on this
server:
DB2
Application software

Systems used for CIMOM servers
We used four xSeries servers for our Common Information Model Object Manager (CIMOM)
servers. They consisted of:
2 GHz CPU per system
2 GB Memory per system
40 GB HDD per system
Windows 2000 server with Service Pack 4

CIMOM system 1
Our first CIMOM server was named TPCMAN. On this server, we installed
ESS CLI
ESS CIMOM
LSI Provider (FAStT CIMOM)

CIMOM system 3
Our third CIMOM system was named SVCCON. We installed the following applications on
this server:
SAN Volume Controller (SVC) Console
SVC CIMOM

Networking
We used the following switches for networking:
IBM Ethernet 10/100 24 Port switch
2109 F16 Fiber switch

Storage devices
We employed the following storage devices:
IBM TotalStorage Enterprise Storage Server (ESS)
800 and F20
DS8000
DS6000
DS4000
IBM SVC

Figure 3-1 on page 64 shows a diagram of our lab setup environment.


.
ESS Management SVCCCONN MARYLAMB TPCMAN
XXX.YYY.6.29 Console SVC CIMOM ESS CIMOM FAStT CIMOM
XXX.YYY.6.26 W2K W2K W2K W2K
SVC Cluster
XXX.YYY.140.14
XXX.YYY.140.15
XXX.YYY.ZZZ.25 XXX.YYY.ZZZ.34 XXX.YYY.ZZZ.35 XXX.YYY.ZZZ.73

Ethernet Switch
XXX.YYY.ZZZ.10

Colorado Server Gallium Server
W2K W2K PQDISRV Server Faroe Server
W2K W2K
-> IBM TotalStorage Productivity Center for -> Tivoli Agent Manager
Disk, and Replication -> Application Server ->Application Server
-> IBM TotalStorage Productivity Center for -> IBM TotalStorage Productivity
Fabric Center for Data

XXX.YYY.ZZZ.49 XXX.YYY.ZZZ.36 XXX.YYY.ZZZ.100 XXX.YYY.ZZZ.69

2109-F16 Fiber
Switch
XXX.YYY.ZZZ.201
FAStT 700
XXX.YYY.ZZZ.202
XXX.YYY.ZZZ.203

Figure 3-1 Lab setup environment

3.4 Pre-installation check list
You need to complete the following tasks in preparation for installing the IBM TotalStorage
Productivity Center. Print the tables in Appendix A, “Worksheets” on page 991, to keep track
of the information you will need during the installation, such as user names, ports, IP
addresses, and locations of servers and managed devices.
1. Determine which elements of the TotalStorage Productivity Center you will install.
2. Uninstall Internet Information Services.
3. Grant the following privileges to the user account that will be used to install the
TotalStorage Productivity Center:
– Act as part of the operating system
– Create a token object
– Increase quotas
– Replace a process-level token
– Logon as a service
4. Install and configure Simple Network Management Protocol (SNMP) (Fabric requirement).
5. Identify any firewalls and obtain the required authorization.
6. Obtain the static IP addresses that will be used for the TotalStorage Productivity Center
servers.


3.5 User IDs and security
This section discusses the user IDs that are used during the installation and those that are
used to manage and work with TotalStorage Productivity Center. It also explains how you can
increase the basic security of the different components.

3.5.1 User IDs
This section lists and explains the user IDs used in a IBM TotalStorage Productivity Center
environment. For some of the IDs, refer to Table 3-8 for a link to additional information that is
available in the manuals.

Suite Installer user
We recommend that you use the Windows Administrator or a dedicated user for the
installation of TotalStorage Productivity Center. That user ID should have the user rights
listed in Table 3-7.

Table 3-7 Requirements for the Suite Installer user
User rights/policy Used for

Act as part of the operating system DB2
Productivity Center for Disk
Fabric Manager

Create a token object DB2

Increase quotas DB2

Replace a process-level token DB2

Log on as a service DB2

Debug programs Productivity Center for Disk

Table 3-8 shows the user IDs that are used in a TotalStorage Productivity Center
environment. It provides information about the Windows group to which the user ID must
belong, whether it is a new user ID that is created during the installation, and when the user
ID is used.

Table 3-8 User IDs used in a IBM TotalStorage Productivity Center environment
Element User ID New user Type Group or Usage
groups

Suite Installer Administrator No

DB2 db2admina Yes, Windows DB2 management and Windows
will be Service Account
created

IBM Director tpcadmina No Windows DirAdmin or Windows Service Account
(see also “IBM DirSuper
Director” on
page 67)


Element User ID New user Type Group or Usage
groups

Resource Manager managerb No, Tivoli N/A, internal Used during the registration of a
default Agent user Resource Manager to the Agent
user Manager Manager

Common Agent AgentMgrb No Tivoli N/A, internal Used to authenticate agents and
(see also “Common Agent user lock the certificate key files
Agent” on page 67) Manager

Common Agent itcauserb Yes, Windows Windows Windows Service Account
will be
created

TotalStorage tpccimoma Yes, will Windows DirAdmin This ID is used to accomplish
Productivity Center be connectivity with the managed
universal user created devices. For example, this ID has to
be set up on the CIM Agents.

Tivoli NetView c
Windows See “Fabric Manager User IDs” on
page 68

IBM WebSphere a
page 68

Host Authentication a
page 68
a. This account can have any name you choose.
b. This account name cannot be changed during the installation.
c. The DB2 administrator user ID and password are used here. See “Fabric Manager User IDs” on page 68.

Granting privileges
Grant privileges to the user ID used to install the IBM TotalStorage Productivity Center for
Disk and Replication Base, IBM TotalStorage Productivity Center for Disk, and the IBM
TotalStorage Productivity Center for Replication.

These user rights are governed by the local security policy and are not initially set as the
defaults for administrators. They may not be in effect when you log on as the local
administrator. If the IBM TotalStorage Productivity Center installation program does not
detect the required user rights for the logged on user name, the program can optionally set
them. The program can set the local security policy settings to assign these user rights.
Alternatively, you can manually set them prior to performing the installation.

To manually set these privileges, follow these steps:
1. Click Start →Settings →Control Panel.
2. Double-click Administrative Tools.
3. Double-click Local Security Policy.
4. The Local Security Settings window opens. Expand Local Policies. Then double-click
User Rights Assignments to see the policies in effect on your system. For each policy
added to the user, perform the following steps:
a. Highlight the policy to be selected.
b. Double-click the policy and look for the user’s name in the Assigned To column of the
Local Security Policy Setting window to verify the policy setting. Ensure that the Local
Policy Setting and the Effective Policy Setting options are selected.


c. If the user name does not appear in the list for the policy, you must add the policy to the
user. Perform the following steps to add the user to the list:
i. In the Local Security Policy Setting window, click Add.
ii. In the Select Users or Groups window, under the Name column, highlight the user
of group.
iii. Click Add to place the name in the lower window.
iv. Click OK to add the policy to the user or group.
5. After you set these user rights, either by using the installation program or manually, log off
the system and then log on again for the user rights to take effect.
6. Restart the installation program to continue with the IBM TotalStorage Productivity Center
for Disk and Replication Base.

TotalStorage Productivity Center communication user
The communication user account is used for authentication between several different
elements of the environment. For example, if WebSphere Application Server is installed with
the Suite Installer, its Administrator ID is the communication users.

IBM Director
With Version 4.1, you no longer need to create an “internal” user account. All user IDs must
be operating system accounts and members of one of the following groups:
DirAdmin or DirSuper groups (Windows), diradmin, or dirsuper groups (Linux)
Administrator or Domain Administrator groups (Windows), root (Linux)

In addition, a host authentication password is used to allow managed hosts and remote
consoles to communicate with IBM Director.

Resource Manager
The user ID and password (default is manager and password) for the Resource Manager is
stored in the AgentManagerconfigAuthorization.xml file on the Agent Manager. Since this is
used only during the initial registration of a new Resource Manager, there is no problem with
changing the values at any time. You can find a detailed procedure on how to change this in
the Installation and Planning Guides of the corresponding manager.

You can have multiple Resource Manager user IDs if you want to separate the administrators
for the different managers, for example for IBM TotalStorage Productivity Center for Data and
IBM TotalStorage Productivity Center for Fabric.

Common Agent
Each time the Common Agent is started, this context and password are used to validate the
registration of the agent with the Tivoli Agent Manager. Furthermore the password is used to
lock the certificate key files (agentTrust.jks).

The default password is changeMe, but you should change the password when you install the
Tivoli Agent Manager. The Tivoli Agent Manager stores this password in the
AgentManager.properties file.

If you start with the defaults, but want to change the password later, all the agents have to be
changed. A procedure to change the password is available in the Installation and Planning
Guides of the corresponding managers (at this time Data or Fabric). Since the password is
used to lock the certificate files, you must also apply this change to Resource Managers.


The Common Agent user ID AgentMgr is not a user ID, but rather the context in which the
agent is registered at the Tivoli Agent Manager. There is no need to change this, so we
recommend that you accept the default.

TotalStorage Productivity Center universal user
The account used to accomplish connectivity with managed devices has to be part of the
DirAdmin (Windows) or diradmin (Linux) group. This user ID communicates with CIMOMs
during install and post install. It also communicates with WebSphere.

Fabric Manager User IDs
During the installation of IBM TotalStorage Productivity Center for Fabric, you can select if
you want to use individual passwords for such subcomponents as DB2, IBM WebSphere,
NetView and the Host Authentication. You can also choose to use the DB2 administrator’s
user ID and password to make the configuration simpler. Figure 4-117 on page 164 shows
the window where you can choose the options.

3.5.2 Increasing user security
The goal of increasing security is to have multiple roles available for the various tasks that can
be performed. Each role is associated with a certain group. The users are only added to
those groups that they need to be part of to fulfill their work.

Not all components have the possibility to increase the security. Others methods require
some degree of knowledge about the specific components to perform the configuration
successfully.

During the installation of Productivity Center for Data, you can enter the name of a Windows
group. Every user within this group is allowed to manage Productivity Center for Data. Other
users may only start the interface and look at it.

You can add or change the name of that group later by editing the server.config file and
restarting Productivity Center for Data.

Productivity Center for Data does not support the following domain login formats for logging
into its server component:
(domain name)/(username)
(username)@(domain)

Because it does not support these formats, you must set up users in a domain account that
can log into the server. Perform the following steps before you install Productivity Center for
Data in your environment:
1. Create a Local Admin Group.
2. Create a Domain Global Group.
3. Add the Domain Global Group to the Local Admin Group.

Productivity Center for Data looks up the SID for the domain user when the login occurs. You
only need to specify a user name and password.


3.5.3 Certificates and key files
Within a TotalStorage Productivity Center environment, several applications use certificates
to ensure security: Productivity Center for Disk, Productivity Center for Replication, and Tivoli
Agent Manager.

Productivity Center for Disk and Replication certificates
The WebSphere Application Server that is part of Productivity Center for Disk and Replication
uses certificates for Secure Sockets Layer (SSL) communication. During the installation, key
files can be generated as self-signed certificates, but you must enter a password for each file
to lock it. The default file names are:
MDMServerKeyFile.jks
MDServerTrusFile.jks

The default directory for the key file on the Agent Manager is C:IBMmdmdmkeys.

Tivoli Agent Manager certificates
The Agent Manager comes with demonstration certificates that you can use. However, you
can also create new certificates during the installation of Agent Manager (see Figure 4-26 on
page 104).

If you choose to create new files, the password that you enter on the panel, as shown in
Figure 4-27 on page 105, as the Agent registration password is used to lock the
agentTrust.jks key file. The default directory for that key file on the Agent Manager is
C:Program FilesIBMAgentManagercerts.

There are more key files in that directory, but during the installation and first steps, the
agentTrust.jks file is the most important one. This is only important if you allow the installer to
create your keys.

3.5.4 Services and service accounts
The managers and components that belong to the TotalStorage Productivity Center are
started as Windows Services. Table 3-9 provides an overview of the most important services.
To keep it simple, we did not include all the DB2 services in the table.

Table 3-9 Services and service accounts
Element Service name Service account Comment

DB2 db2admin The account needs to be part of
Administrators and DB2ADMNS.

IBM Director IBM Director Server Administrator You need to modify the account to be part
of one of the groups: DirAdmin or DirSuper.

Agent Manager IBM WebSphere Application LocalSystem You need to set this service to start
Server V5 — Tivoli Agent automatically, after the installation.
Manager

Common Agent IBM Tivoli Common Agent — itcauser
C:Program Filestivoliep

Productivity Center IBM TotalStorage Productivity TSRMsrv1
for Data Center for Data server

Productivity Center IBM WebSphere Application LocalSystem
for Fabric Server V5 — Fabric Manager


Element Service name Service account Comment

Tivoli NetView Tivoli NetView Service NetView
Service

3.6 Starting and stopping the managers
To start, stop or restart one of the managers or components, you use the Windows control
panel. Table 3-10 shows a list of the services.

Table 3-10 Services used for TotalStorage Productivity Center
Element Service name Service account

DB2 db2admin

IBM Director IBM Director Server Administrator

Agent Manager IBM WebSphere Application Server V5 - Tivoli Agent Manager LocalSystem

Common Agent IBM Tivoli Common Agent — C:Program Filestivoliep itcauser

Productivity Center for Data IBM TotalStorage Productivity Center for Data Server TSRMsrv1

Productivity Center for Fabric IBM WebSphere Application Server V5 - Fabric Manager LocalSystem

Tivoli NetView Service Tivoli NetView Service NetView

3.7 Windows Management Instrumentation
Before beginning the Prerequisite Software installation, the Windows Management
Instrumentation service must first be stopped and disabled. To disable the service, follow the
steps below.
1. Go to Start → Settings → Control Panel → Administrative Tools → Services.
2. Scroll down and double-click the Windows Management Instrumentation service (see
Figure 3-2 on page 71).


Figure 3-2 Windows Management Instrumentation service

3. In the Windows Management Instrumentation Properties window, go down to Service
status and click the Stop button (Figure 3-3). Wait for the service to stop.

Figure 3-3 Stopping Windows Management Instrumentation


4. After the service is stopped, in the Windows Management Instrumentation Properties
window, change the Startup type to Disabled (Figure 3-4) and click OK.

Figure 3-4 Disabled Windows Management Instrumentation

5. After disabling the service, it may start again. If so, go back and stop the service again.
The service should now be stopped and disabled as shown in Figure 3-5.

Figure 3-5 Windows Management Instrumentation successfully disabled

Important: After the Prerequisite Software installation completes. You must enable the
Windows Management Instrumentation service before installing the suite. To enable the
service, change the Startup type from Disabled (see Figure 3-4) to Automatic.


3.8 World Wide Web Publishing
As with the Windows Management Instrumentation service, the World Wide Web Publishing
service must also be stopped and disabled before starting the Prerequisite Software Installer.
To stop the World Wide Web Publishing service, simply follow the same steps in section
Figure 3.7 on page 70. This service can remain disabled.

3.9 Uninstalling Internet Information Services
Make sure Internet Information Services (IIS) is not installed on the server. If it is installed,
uninstall it using the following procedure.
1. Click Start → Settings → Control Panel.
2. Click Add/Remove Programs.
3. In the Add or Remove Programs window, click Add/Remove Windows Components.
4. In the Windows Components panel, deselect IIS.

3.10 Installing SNMP
Before you install the components of the TotalStorage Productivity Center, install and
configure SNMP.
1. Click Start → Settings → Control Panel.
2. Click Add/Remove Programs.
3. In the Add or Remove Programs window, click Add/Remove Windows Components.
4. Double-click Management and Monitoring Tools.
5. In the Windows Components panel, select Simple Network Management Protocol and
click OK.
6. Close the panels and accept the installation of the components.
7. The Windows installation CD or installation files are required. Make sure that the SNMP
services are configured as explained in these steps:
a. Right-click My Computer and select Manage.
b. In the Computer Management window, click Services and Applications.
c. Double-click Services.
8. Scroll down to and double-click SNMP Service.
9. In the SNMP Service Properties window, follow these steps:
10.Click the Traps tab (see Figure 3-6 on page 74).


d. Make sure that the public name is available.

Figure 3-6 Traps tab in the SNMP Service Properties window

e. Click the Security tab (see Figure 3-7).
f. Select Accept SNMP packets from any host.
g. Click OK.

Figure 3-7 SNMP Security Properties window

11.After you set the public community name, restart the SNMP community service.


3.11 IBM TotalStorage Productivity Center for Fabric
Prior to installing IBM TotalStorage Productivity Center for Fabric, there are planning
considerations and prerequisite tasks that you need to complete.

3.11.1 The computer name
IBM TotalStorage Productivity Center for Fabric requires fully qualified host names for the
manager, managed hosts, and the remote console. To verify your computer name on
Windows, follow this procedure.
1. Right-click the My Computer icon on your desktop and select Properties.
2. The System Properties window opens.
a. Click the Network Identification tab. Click Properties.
b. The Identification Changes panel opens.
i. Verify that your computer name is entered correctly. This is the name that the
computer is identified as in the network.
ii. Verify that the full computer name is a fully qualified host name. For example,
user1.sanjose.ibm.com is a fully qualified host name.
iii. Click More.
c. The DNS Suffix and NetBIOS Computer Name panel opens. Verify that the Primary
DNS suffix field displays a domain name.

Important: The fully qualified host name must match the HOSTS file name (including
case-sensitive characters).

3.11.2 Database considerations
When you install IBM TotalStorage Productivity Center for Fabric, a DB2 database is
automatically created if you specified the DB2 database. The default database name is
TSANMDB.

If you installed IBM TotalStorage Productivity Center for Fabric previously, are using a DB2
database, and want to save the information in the database before you re-install the
manager, you must use DB2 commands to back up the database. The default name for the
IBM TotalStorage Productivity Center for Fabric DB2 database is TSANMDB. The database
name for Cloudscape is also TSANMDB. You cannot change this database name.

If you are installing the manager on more than one machine in a Windows domain, the
managers on different machines may end up sharing the same DB2 database. To avoid this
situation, you must either use different database names or different DB2 user names when
installing the manager on different machines.

3.11.3 Windows Terminal Services
You cannot use the Windows Terminal Services to access a machine that is running the IBM
TotalStorage Productivity Center for Fabric console (either the manager or remote console
machine). Any TotalStorage Productivity Center for Fabric dialogs launched from the SAN
menu in Tivoli NetView appear on the manager or remote console machine only. The dialogs
do not appear in the Windows Terminal Services session.


3.11.4 Tivoli NetView
IBM TotalStorage Productivity Center for Fabric also installs Tivoli NetView 7.1.3. If you
already have Tivoli NetView 7.1.1 installed, IBM TotalStorage Productivity Center for Fabric
upgrades it to version 7.1.3. If you have a Tivoli NetView release earlier than Version 7.1.1,
IBM TotalStorage Productivity Center for Fabric prompts you to uninstall Tivoli NetView before
you install this product.

If you have Tivoli NetView 7.1.3 installed, ensure that the following applications are stopped.
You can check for Tivoli NetView by opening the Tivoli NetView console icon on your desktop.
Web Console
Web Console Security
MIB Loader
MIB Browser
Netmon Seed Editor
Tivoli Event Console Adapter

Important: Ensure that the Windows 2000 Terminal Services is not running. Go to the
Services panel and check for Terminal Services.

User IDs and password considerations
TotalStorage Productivity Center for Fabric only supports local user IDs and groups. It does
not support domain user IDs and groups.

Cloudscape database
If you install TotalStorage Productivity Center for Fabric and specify the Cloudscape
database, you need the following user IDs and passwords:
Agent manager name or IP address and password
Common agent password to register with the Agent Manager
Resource manager user ID and password to register with the Agent Manager
WebSphere administrative user ID and password host authentication password
Tivoli NetView password only

DB2 database
If you install IBM TotalStorage Productivity Center for Fabric and specify the DB2 database,
you need the following user IDs and passwords:
Agent manager name or IP address and password
Common agent password to register with the Agent Manager
Resource manager user ID and password to register with the Agent Manager
DB2 administrator user ID and password
DB2 user ID and password
WebSphere administrative user ID and password
Host authentication password only
Tivoli NetView password only

Note: If you are running Windows 2000, when the IBM TotalStorage Productivity Center
for Fabric installation program asks for an existing user ID for WebSphere, that user ID
must act as part of the operating system user.

WebSphere
To change the WebSphere user ID and password, follow this procedure:
1. Open the install_locationappswaspropertiessoap.client.props file.


2. Modify the following entries:
– com.ibm.SOAP. login Userid=user_ID (enter a value for user_ID)
– com.ibm.SOAP. login Password=password (enter a value for password)
3. Save the file.
4. Run the following script:
ChangeWASAdminPass.bat user_ID password install_dir
Here user_ID is the WebSphere user ID and password is the password. install_dir is the
directory where the manager is installed and is optional. For example, install_dir is
c:Program FilesIBMTPCFabricmanagerbinW32-ix86.

3.11.5 Personal firewall
If you have a software firewall on your system, disable the firewall while installing the Fabric
Manager. The firewall causes Tivoli NetView installation to fail. You can enable the firewall
after you install the Fabric Manager.

Security considerations
You set up security by using certificates. There are demonstration certificates or you can
generate new certificates. This option is specified when you installed the Agent Manager. See
Figure 4-26 on page 104. We recommend that you generate new certificates.

If you used the demonstration certificates, continue with the installation.

If you generated new certificates, follow this procedure:
1. Copy the manager CD image to your computer.
2. Copy the agentTrust.jks file from the Agent Manager (AgentManager/certs directory) to
the /certs directory of the manager CD image. This overwrites the existing agentTrust.jks
file.
3. You can write a new CD image with the new file or keep this image on your computer and
point the Suite Installer to the directory when requested.

3.11.6 Changing the HOSTS file
When you install Service Pack 3 for Windows 2000 on your computers, follow these steps to
avoid addressing problems with IBM TotalStorage Productivity Center for Fabric. The
problem is caused by the address resolution protocol, which returns the short name and not
the fully qualified host name.

You can avoid this problem by changing the entries in the corresponding host tables on the
Domain Name System (DNS) server and on the local computer. The fully qualified host name
must be listed before the short name as shown in Example 3-1. See 3.11.1, “The computer
name” on page 75, for details about determining the host name.

To correct this problem, you have to edit the HOSTS file. The HOSTS file is in the
%SystemRoot%system32driversetc directory.

Example 3-1 Sample HOSTS file
# This is a sample HOSTS file used by Microsoft TCP/IP for Windows.
#
# This file contains the mappings of IP addresses to host names. Each
# entry should be kept on an individual line. The IP address should
# be placed in the first column followed by the corresponding host name.


# The IP address and the host name should be separated by at least one
# space.
#
# Additionally, comments (such as these) may be inserted on individual
# lines or following the machine name denoted by a '#' symbol.
#
# For example:
#
# 102.54.94.97 rhino.acme.com # source server
# 38.25.63.10 x.acme.com # x client host

127.0.0.1 localhost
#
192.168.123.146 jason.groupa.mycompany.com jason
192.168.123.146 jason jason.groupa.mycompany.com

Note: Host names are case sensitive, which is limitation within WebSphere. Check your
host name.

3.12 IBM TotalStorage Productivity Center for Data
Prior to installing IBM TotalStorage Productivity Center for Data, there are planning
considerations and prerequisite tasks that you need to complete.

3.12.1 Server recommendations
The IBM TotalStorage Productivity Center for Data server component acts as a traffic officer
for directing information and handling requests from the agent and UI components installed
within an environment. You need to install at least one server within your environment.

We recommend that you do not manage more than 1000 agents with a single server. If you
need to install more than 1000 agents, we suggest that you install an additional server for
those agents to maintain optimal performance.

3.12.2 Supported subsystems and databases
This section contains the subsystems, file system formats, and databases that the
TotalStorage Productivity Center for Data supports.

Storage subsystem support
Data Manager currently supports the monitoring and reporting of the following storage
subsystems:
Hitachi Data Systems
HP StorageWorks
IBM FAStT 200, 600, 700, and 900 with an SMI-S 1.0 compliant CIM interface
SAN Volume Controller Console Version 1.1.0.2, 1.1.0.9, 1.2.0.5, 1.2.0.6 (1.3.2 Patch
available), 1.2.1.x, 1.2.0.6,
SAN Volume Controller CIMOM Version 1.1.0.1, 1.2.0.4, 1.2.0.5 (1.3.2 patch available),
1.2.0.5, 1.2.1.x
ESS ICAT 1.1.0.2, 1.2.0.15, 1.2.0.29, 1.2.x, 1.2.1.40 and later for ESS


File system support
Data Manager supports the monitoring and reporting of the following file systems:
FAT
FAT32
NTFS4, NTFS5
EXT2, EXT3
AIX_JFS
HP_HFS
VXFS
UFS
TMPFS
AIX_OLD
NW_FAT
NW_NSS
NF
WAFL
FAKE
AIX_JFS2
SANFS
REISERFS

Network File System support
Data Manager currently supports the monitoring and reporting of the following Network File
Systems (NFS):
IBM TotalStorage SAN File System 1.0 (Version 1 Release 1), from AIX V5.1 (32-bit) and
Windows 2000 Server/Advanced Server clients
IBM TotalStorage SAN File System 2.1, 2.2 from AIX V5.1 (32-bit), Windows 2000
Server/Advanced Server, Red Hat Enterprise Linux 3.0 Advanced Server, and SUN
Solaris 9 clients
General Parallel File System (GPFS) v2.1, v2.2

RDBMS support
Data Manager currently supports the monitoring of the following relational database
management systems (RDBMS):
Microsoft SQL Server 7.0, 2000
Oracle 8i, 9i, 9i V2, 10G
Sybase SQL Server 11.0.9 and higher
DB2 Universal Database™ (UDB) 7.1, 7.2, 8.1, 8.2 (64-bit UDB DB2 instances are
supported)

3.12.3 Security considerations
This section describes the security issues that you must consider when installing Data
Manager.


User levels
There are two levels of users within IBM TotalStorage Productivity Center for Data:
non-administrator users and administrator users. The level of users determine how they use
IBM TotalStorage Productivity Center for Data.
Non-administrator users
– View the data collected by IBM TotalStorage Productivity Center for Data.
– Create, generate, and save reports.
IBM TotalStorage Productivity Center for Data administrators. These users can:
– Create, modify, and schedule Pings, Probes, and Scans
– Create, generate, and save reports
– Perform administrative tasks and customize the IBM TotalStorage Productivity Center
for Data environment
– Create Groups, Profiles, Quotas, and Constraints
– Set alerts

Important: Security is set up by using the certificates. You can use the demonstration
certificates or you can generate new certificates. It is recommended that you generate new
certificates when you install the Agent Manager.

Certificates
If you generated new certificates, follow this procedure:
1. Copy the CD image to your computer.
2. Copy the agentTrust.jks file from the Agent Manager directory AgentManager/certs to the
CommonAgentcerts directory of the manager CD image. This overwrites the existing
agentTrust.jks file.

You can write a new CD image with the new file or keep this image on your computer and
point the Suite Installer to the directory when requested.

Important: Before installing IBM TotalStorage Productivity Center for Data, define the
group within your environment that will have administrator rights within Data Manager. This
group must exist on the same machine where you are installing the Server component.
During the installation, you are prompted to enter the name of this group.


3.12.4 Creating the DB2 database
Before you install the component, create the IBM TotalStorage Productivity Center for Data
database.
1. From the start menu, select Start →Programs →IBM DB2 →General Administration
Tools →Control Center.
2. This launches the DB2 Control Center. Create a database that is used for IBM
TotalStorage Productivity Center for Data as shown in Figure 3-8. Select All Databases,
right-click and select Create Databases →Standard.

Figure 3-8 DB2 database creation


3. In the window that opens (Figure 3-9), complete the required database name information.
We used the database name of TPCDATA. Click Finish to complete the database
creation.

Figure 3-9 DB2 database information for creation


4

Chapter 4. Installing the IBM TotalStorage
Productivity Center suite
Installation of the TotalStorage Productivity Center suite of products is done using the install
wizards. The first, the Prerequisite Software Installer, installs all the products needed before
one can install the TotalStorage Productivity Center suite. The second, the Suite Installer,
installs the individual components or the entire suite of products.

This chapter documents the use of the Prerequisite Software Installer and the Suite Installer.
It also includes hints and tips based on our experience.


4.1 Installing the IBM TotalStorage Productivity Center
IBM TotalStorage Productivity Center provides a Prerequisite Software Installer and Suite
Installer that helps guide you through the installation process. You can also use the Suite
Installer to install stand-alone components.

The Prerequisite Software Installer installs the following products in this order:
1. DB2, which is required by all the managers
2. WebSphere Application Server, which is required by all the managers except for
TotalStorage Productivity Center for Data
3. Tivoli Agent Manager, which is required by Fabric Manager and Data Manager

The Suite Installer installs the following products or components in this order:
1. IBM Director, which is required by TotalStorage Productivity Center for Disk and
TotalStorage Productivity Center for Replication
2. Productivity Center for Disk and Replication Base, which is required by TotalStorage
Productivity Center for Disk and TotalStorage Productivity Center for Replication
3. TotalStorage Productivity Center for Disk
4. TotalStorage Productivity Center for Replication
5. TotalStorage Productivity Center for Fabric - Manager
6. TotalStorage Productivity Center for Data - Manager

In addition to the manager installations, the Suite Installer guides you through the installation
of other IBM TotalStorage Productivity Center components. You can select more than one
installation option at a time. This redbook separates the types of installations into several
sections to help explain them. The additional types of installation tasks are:
IBM TotalStorage Productivity Center Agent installations
IBM TotalStorage Productivity Center GUI/Client installations
Language Pack installations
IBM TotalStorage Productivity Center product uninstallations

4.1.1 Considerations
You may want to use IBM TotalStorage Productivity Center for Disk to manage the IBM
TotalStorage Enterprise Storage Server (ESS), DS8000, DS6000, Storage Area Network
(SAN) Volume Controller (SVC), IBM TotalStorage Fibre Array Storage Technology (FAStT),
or DS4000 storage subsystems. In this case, you must install the prerequisite input/output
(I/O) Subsystem Licensed Internal Code (SLIC) and Common Information Model (CIM) Agent
for the devices. See Chapter 6, “Configuring IBM TotalStorage Productivity Center for Disk”
on page 247, for more information.

If you are installing the CIM Agent for the ESS, or the DS8000 or DS6000 you must install it
on a separate machine.

TotalStorage Productivity Center 2.3 does not support Linux on zSeries or on S/390®. Nor
does IBM TotalStorage Productivity Center support Windows domains.


4.2 Prerequisite Software Installation
This section guides you step by step through the install process of the prerequisite software
components.

4.2.1 Best practices
Before you begin installing the prerequisite software components, we recommend that you
complete the following tasks:
1. Grant privileges to the user ID used to install the IBM TotalStorage Productivity Center
components, including the IBM TotalStorage Productivity Center for Disk and Replication
Base, IBM TotalStorage Productivity Center for Disk, IBM TotalStorage Productivity
Center for Replication, IBM TotalStorage Productivity Center for Data and IBM
TotalStorage Productivity Center for Fabric. For details refer to “Granting privileges” on
page 66.
2. Make sure Internet Information Services (IIS) is not installed on the server. If it is installed,
uninstall it using the procedure in 3.9, “Uninstalling Internet Information Services” on
page 73.
3. Install and configure Simple Network Management Protocol (SNMP) described in 3.10,
“Installing SNMP” on page 73.
4. Stop and disable Windows Management Instrumentation (Figure 3.7 on page 70) and
World Wide Web Publishing (3.8, “World Wide Web Publishing” on page 73) services.
5. Create a database for Agent Manager installation. To create the database, see 3.12.4,
“Creating the DB2 database” on page 81. The default database name for Agent Manager
is IBMCDB.

4.2.2 Installing prerequisite software
Follow these steps to install the prerequisite software components:
1. Insert the IBM TotalStorage Productivity Center Prerequisite Software Installer CD into the
CD-ROM drive.
If Windows autorun is enabled, the installation program should start automatically. If it
does not, open Windows Explorer and go to the IBM TotalStorage Productivity Center
CD-ROM drive. Double-click setup.exe.

Note: It may take a few moments for the installer program to initialize. Be patient.
Eventually, you see the language selection panel (Figure 4-1).

2. The installer language window (Figure 4-1) opens. From the list, select a language. This is
the language that is used to install this product. Click OK.

Figure 4-1 Prerequisite Software Installer language

Chapter 4. Installing the IBM TotalStorage Productivity Center suite 85

3. The Prerequisite Software Installer wizard welcome pane in Figure 4-2 opens. Click Next.
The Software License Agreement panel is then displayed. Read the terms of the license
agreement. If you agree with the terms of the license agreement select the I accept the
terms in the license agreement radio button and click Next to continue.

Figure 4-2 Prerequisite Software Installer wizard

4. The prerequisite operating system check panel in Figure 4-3 on page 87 opens. When it
completes successfully click Next.


Figure 4-3 Prerequisite Operating System check

5. The Tivoli Common Directory location panel (Figure 4-4) opens and prompts for a location
for the log files. Accept the default location or enter a different location. Click Next to
continue.

Figure 4-4 Tivoli Common Directory location


6. The product selection panel (Figure 4-5) opens. To install the entire TotalStorage
Productivity Center suite, check the boxes next to DB2, WebSphere, and Agent Manager.

Figure 4-5 Product selection

7. The DB2 Universal Database panel (Figure 4-6) opens. Select Enterprise Server Edition
and click Next to continue.

Figure 4-6 DB2 Universal Database


Note: After clicking Next (Figure 4-6), if you see the panel in Figure 4-7, you must first stop
and disable the Windows Management Instrumentation service before continuing with the
installation. See Figure 3.7 on page 70 for detailed instructions.

Figure 4-7 Windows Management Instrumentation service warning


8. The DB2 user name and password panel (Figure 4-8) opens. If the DB2 user name exists
on the system, the correct password must be entered or the DB2 installation will fail. If the
DB2 user name does not exist it will be created by the DB2 install. In our installation we
accepted the default user name and entered a unique password. Click Next to continue.

Figure 4-8 DB2 user configuration


9. The Target Directory Confirmation panel (Figure 4-9) opens. Accept the default target
directories for DB2 installation or enter a different location. Click Next.

Figure 4-9 Target Directory Confirmation

10.The select the languages panel (Figure 4-10) opens. This installs the languages selected
for DB2. Select your desired language(s). Click Next.

Figure 4-10 Language selection


11.The Preview Prerequisite Software Information panel (Figure 4-11) opens. Review the
information and click Next.

Figure 4-11 Preview Prerequisite Software Information

12.The WebSphere Application Server system prerequisites check panel (Figure 4-12)
opens. When the check completes successfully click Next.

Figure 4-12 WebSphere Application Server system prerequisites check


13.The installation options panel (Figure 4-13) opens. Select the type of installation you wish
to perform. The rest of this section guides you through Unattended Installation.
Unattended Installation guides you through copying all installation images to a central
location called the installation image depot. Once the copies are completed, the
component installations proceed with no further intervention needed. Attended Installation
prompts you to enter the location of each install image as needed.
Click Next to continue.

Figure 4-13 Installation options


14.The install image depot location panel opens (see Figure 4-14). Enter the location where
all installation images are to be copied. Click Next.

Figure 4-14 Install image depot location

15.You are first prompted for the location of the DB2 installation image (see Figure 4-15).
Browse to the installation image and select the path to the installation files or insert the
install CD and click Copy.

Figure 4-15 DB2 installation source


16.After the DB2 installation image is copied to the install image depot, you are prompted for
the location of the WebSphere installation image (see Figure 4-16). Browse to the
installation image and select the path to the installation files or insert the install CD and
click Copy.

Figure 4-16 WebSphere installation source

17.After the WebSphere installation image is copied, you are prompted for the location of the
WebSphere Cumulative fix 3 installation image (see Figure 4-17). Browse to the
installation image and select the path to the installation files or insert the install CD and
click Copy.

Figure 4-17 WebSphere fix 3 installation source


18.When an install image has been successfully copied to the Install Image Depot, a green
check mark appears to the right of the prerequisite. After all the prerequisite software
images are successfully copied to the install image depot (Figure 4-18), click Next.

Figure 4-18 Installation images copied successfully


19.The installation of DB2, WebSphere, and the WebSphere Fix Pack begins. When a
prerequisite is successfully installed, a green check mark appears to its left. If the
installation of a prerequisite fails, a red X appears to the left. If a prerequisite installation
fails, exit the installer, check the logs to determine and correct the problem, and restart the
installer.
When the installation completes successfully (see Figure 4-19), click Next.

Figure 4-19 DB2 and WebSphere installation complete


20.The Agent Manager Registry Information panel opens. Select the type of database,
specify the database name, and choose a local or remote database. The default DB2
database name is IBMCDB. For a local database connection, the DB2 database will be
created if it does not exist. We recommend that you take the default database name for a
local database. Click Next to continue (see Figure 4-20).

Attention: For a remote database connection, the database specified below must exist.
Refer to 3.12.4, “Creating the DB2 database” on page 81 for information on how to create a
database in DB2.

Figure 4-20 Agent Manager Registry Information


21.The Database Connection Information panel in Figure 4-21 opens. Specify the location of
the database software directory (for DB2, the default install location is C:Program
FilesIBMSQLLIB), the database user name and password. You must specify the
database host name and port if you are using a remote database. Click Next to continue.

Figure 4-21 Agent Manager database connection Information


Note: For a remote database connection the database specified in Figure 4-20 on page 98
must exist. If the database does not exist, you will see the error message shown in
Figure 4-22. Refer to 3.12.4, “Creating the DB2 database” on page 81 for information on
how to create a database in DB2.

Figure 4-22 DB2 database error


22.A panel opens prompting for a location to install Tivoli Agent Manager (see Figure 4-23).
Accept the default location or enter a different location. Click Next to continue.

Figure 4-23 Tivoli Agent Manager installation directory


23.The WebSphere Application Server Information panel (Figure 4-24) opens. This panel lets
you specify the host name or IP address, and the cell and node names on which to install
the Agent Manager. If you specify a host name, use the fully qualified host name. For
example, specify HELIUM.almaden.ibm.com. If you use the IP address, use a static IP
address. This value is used in the URLs for all Agent Manager services. We recommend
that you use the fully qualified host name, not the IP address of the Agent Manager server.
Typically the cell and node name are both the same as the host name of the computer. If
WebSphere was installed before you started the Agent Manager installation wizard, you
can look up the cell and node name values in the %WebSphere Application
Server_INSTALL_ROOT%binSetupCmdLine.bat file.
You can also specify the ports used by the Agent Manager. We recommend that you
accept the defaults.
– Registration Port: The default is 9511 for the server-side Secure Sockets Layer (SSL).
– Secure Port: The default is 9512 for client authentication, two-way SSL.
– Public Port: The default is 9513.
If you are using WebSphere network deployment or a customized deployment, make sure
that the cell and node names are correct. For more information about WebSphere
deployment, see your WebSphere documentation.
After filling in the required information in the WebSphere Application Server Information
panel, click Next.

Figure 4-24 WebSphere Application Server Information


Note: If an IP address is entered in the WebSphere Application Server Information panel
shown in Figure 4-24, the next panel (see Figure 4-25) explains why a host name is
recommended. Click Back to use a host name or click Next to use the IP address.

Figure 4-25 Agent Manager IP address warning


24.The Security Certificates panel (Figure 4-26) opens. Specify whether to create new
certificates or to use the demonstration certificates.
In a typical production environment, you would create new certificates. The ability to use
demonstration certificates is provided as a convenience for testing and demonstration
purposes. Make a selection and click Next.

Figure 4-26 Security Certificates


25.The Security Certificate Settings panel (see Figure 4-27) opens. Specify the certificate
authority name, security domain, and agent registration password.
The agent registration password is used to register the agents. You must provide this
password when you install the agents. This password also sets the Agent Manager key
store and trust store files. Record this password, it will be used again in the installation
process.
The domain name is used in the right-hand portion of the distinguished name (DN) of
every certificate issued by the Agent Manager. It is the name of the security domain
defined by the Agent Manager. Typically, this value is the registered domain name or
contains the registered domain name. For example, for the computer system
myserver.ibm.com, the domain name is ibm.com. This value must be unique in your
environment. If you have multiple Agent Managers installed, this value must be different on
each Agent Manager.
The default agent registration password is changeMe and it is case sensitive.

Figure 4-27 Security Certificate Settings


26.The User input summary panel for Agent Manager (see Figure 4-28) opens. Review the

Figure 4-28 User input summary


27.The summary information for Agent Manager panel (see Figure 4-29) opens. Click Next.

Figure 4-29 Agent Manager installation summary


28.You will see a panel indicating the status of the Agent Manager install process. the
IBMCDB database will be created and tables are added to the database. Once the
installation of agent manager completes the Summary of Installation and Configuration
Results panel (see Figure 4-30) opens. Click Next to continue.

Figure 4-30 Summary of Installation and Configuration Results


29.The next panel (Figure 4-31) informs you when the Agent Manager service started
successfully. Click Finish.

Figure 4-31 Agent Manager service started


30.The next panel (Figure 4-32) indicates the installation of prerequisite software is complete.
Click Finish to exit the prerequisite installer.

Figure 4-32 Prerequisite software installation complete

4.3 Suite installation
This section guides you through the step by step process to install the TotalStorage
Productivity Center components you select.

The Suite Installer launches the installation wizard for each manager you chose to install.

4.3.1 Best practices
Before you begin installing the suite of products complete the following tasks.
1. If you are running the Fabric Manager installation under Windows 2000, the Fabric
Manager installation requires the user ID to have the following user rights:
Act as part of the operating system
Log on as a service user rights
see Granting privileges under 3.5.1, “User IDs” on page 65
2. Enable Windows Management Instrumentation (see Figure 3.7 on page 70)
3. Install SNMP (see 3.10, “Installing SNMP” on page 73)
4. Create the database for the TotalStorage Productivity Center for Data installation (see
3.12.4, “Creating the DB2 database” on page 81).

4.3.2 Installing the TotalStorage Productivity Center suite
Follow these steps for successful installation:


1. Insert the IBM TotalStorage Productivity Center Suite Installer CD into the CD-ROM drive.
If Windows autorun is enabled, the installation program should start automatically. If it
does not, open Windows Explorer and go to the IBM TotalStorage Productivity Center
CD-ROM drive. Double-click setup.exe.

Note: It may take a few moments for the installer program to initialize. Be patient.
Eventually, you see the language selection panel (Figure 4-33).

2. The Installer language window (see Figure 4-33) opens. From the list, select a language.
This is the language used to install this product. Click OK.

Figure 4-33 Installer Wizard

3. You see the Welcome to the InstallShield Wizard for The IBM TotalStorage Productivity
Center panel (see Figure 4-34). Click Next.

Figure 4-34 Welcome to IBM TotalStorage Productivity Center panel

4. The Software License Agreement panel (Figure 4-35 on page 112) opens. Read the terms
of the license agreement. If you agree with the terms of the license agreement, select the
I accept the terms of the license agreement radio button. Then click Next.
If you do not accept the terms of the license agreement, the installation program ends
without installing IBM TotalStorage Productivity Center components.


Figure 4-35 License agreement

5. The next panel enables you to select the type of installation (Figure 4-36). Select Manager
installations of Data, Disk, Fabric, and Replication and then click Next.

Figure 4-36 IBM TotalStorage Productivity Center options panel


6. In the next panel (see Figure 4-37), select the components that you want to install. Click
Next to continue.

Figure 4-37 IBM TotalStorage Productivity Center components


7. The suite installer installs the IBM Director first (see Figure 4-38). Click Next.

Figure 4-38 IBM Director prerequisite install

8. The IBM Director installation is now ready to begin (see Figure 4-39). Click Next.

Figure 4-39 Begin IBM Director installation


9. The package location for IBM Director panel (see Figure 4-40) opens. Enter the
appropriate information and click Next.

Note: Make sure the Windows Management Instrumentation service is disabled (see
Figure 3.7 on page 70 for detailed instructions). If it is enabled, a window appears
prompting you to disable the service after you click Next to continue.

Figure 4-40 IBM Director package location

10.The next panel (see Figure 4-41) provides information about the IBM Director post
installation reboot option. When prompted, choose the option to reboot later. Click Next.

Figure 4-41 IBM Director information


11.The IBM Director Server - InstallShield Wizard panel (Figure 4-42) opens. It indicates that
the IBM Director installation wizard will launch. Click Next.

Figure 4-42 IBM Director InstallShield Wizard

12.The License Agreement window opens (Figure 4-43). Read the license agreement. Click I
accept the terms in the license agreement radio button and then click Next.

Figure 4-43 IBM Director license agreement


13.The next window (Figure 4-44) displays an advertisement for Enhance IBM Director with
the new Server Plus Pack window. Click Next.

Figure 4-44 IBM Director new Server Plus Pack window

14.The Feature and installation directory window (Figure 4-45) opens. Accept the default
settings and click Next.

Figure 4-45 IBM Director feature and installation directory window


15.The IBM Director service account information window (see Figure 4-46) opens.
a. Type the domain for the IBM Director system administrator. Alternatively, if there is no
domain, then type the local host name (the recommended setup).
b. Type a user name and password for IBM Director. The IBM Director will run under this
user name and you will log on to the IBM Director console using this user name. In our
installation we used the user ID we created to install the TotalStorage Productivity
Center. This user must be part of the Administrator group.
c. Click Next to continue.

Figure 4-46 Account information

16.The Encryption settings window (Figure 4-47) opens. Accept the default settings in the
Encryption settings window. Click Next.

Figure 4-47 Encryption settings


17.In the Software Distribution settings window (Figure 4-48), accept the default values and
click Next.

Note: The TotalStorage Productivity Center components do not use the
software-distribution packages function of IBM Director.

Figure 4-48 Installation target directory

18.The Ready to Install the Program window (Figure 4-49) opens. Click Install.

Figure 4-49 Installation ready


19.The Installing IBM Director server window (Figure 4-50) reports the status of the
installation.

Figure 4-50 Installation progress

20.The Network driver configuration window (Figure 4-51) opens. Accept the default settings
and click OK.

Figure 4-51 Network driver configuration

The secondary window closes and the installation wizard performs additional actions
which are tracked in the status window.


21.The Select the database to be configured window (Figure 4-52) opens. Select IBM DB2
Universal Database and click Next.

Figure 4-52 Database selection

22.The IBM Director DB2 Universal Database configuration window (Figure 4-53) opens. It
may be behind the status window. You must click this window to bring it to the foreground.
a. In the Database name field, type a new database name for the IBM Director database
table or type an existing database name.
b. In the User ID and Password fields, type the DB2 user ID and password that you
created during the DB2 installation.
c. Click Next to continue.

Figure 4-53 Database selection configuration


23.In the IBM Director DB2 Universal Database configuration secondary window
(Figure 4-54), accept the default DB2 node name LOCAL - DB2. Click OK.

Figure 4-54 Database node name selection

24.The Database configuration in progress window is displayed at the bottom of the IBM
Director DB2 Universal Database configuration window. Wait for the configuration to
complete and the secondary window to close.
25.When the InstallShield Wizard Completed window (Figure 4-55) opens, click Finish.

Figure 4-55 Completed installation

Important: Do not reboot the machine at the end of the IBM Director installation. The
Suite Installer reboots the machine.


26.When you see IBM Director Server Installer Information window (Figure 4-56), click No.

Figure 4-56 IBM Director reboot option

Important: Are you installing IBM TotalStorage Productivity Center for Data? If so,
have you created the database for IBM TotalStorage Productivity Center for Data or are
you using a existing database?

If you are installing Tivoli Disk manager, you must have created the administrative
superuser ID and group and set the privileges.

27.The Install Status panel (see Figure 4-57) opens after a successful installation. Click Next.

Figure 4-57 IBM Director Install Status successful


28.In the machine reboot window (see Figure 4-58), click Next to reboot the machine.

Important: If the server does not reboot at this point, cancel the installer and reboot the
server.

Figure 4-58 Install wizard completion


4.3.3 IBM TotalStorage Productivity Center for Disk and Replication Base
There are three separate installations to perform:
Install the IBM TotalStorage Productivity Center for Disk and Replication Base code.
Install the IBM TotalStorage Productivity Center for Disk.
Install the IBM TotalStorage Productivity Center for Replication.

IBM TotalStorage Productivity Center for Disk and Replication Base must be installed by a
user who is logged on as a local administrator (for example, as the administrator user) on the
system where the IBM TotalStorage Productivity Center for Disk and Replication Base will be
installed. If you intend to install IBM TotalStorage Productivity Center for Disk and Replication
Base as a server, you need the following required system privileges, called user rights, to
successfully complete the installation as described in 3.5.1, “User IDs” on page 65.
Act as part of the operating system
Create a token object
Increase quotas
Replace a process level token
Debug programs

After rebooting the machine the installer initializes to continue the suite install. A window
opens prompting you to select the installation language to be used for this wizard
(Figure 4-59). Select the language and click OK.

Figure 4-59 Selecting the language for the IBM TotalStorage Productivity Center installation wizard


1. The next panel enables you to select the type of installation (Figure 4-60). Select Manager
installations of Data, Disk, Fabric, and Replication and click Next.

Figure 4-60 IBM TotalStorage Productivity Center options panel

2. The next window (Figure 4-61) opens allowing you to select which components to install.
Select the components you wish to install (all components in this case) and click Next.

Figure 4-61 TotalStorage Productivity Center components


3. The installer checks that all prerequisite software is installed on your system (see
Figure 4-62). Click Next.

Figure 4-62 Prerequisite software check

4. Figure 4-63 shows the Installer window about to begin installation of Productivity Center
for Disk and Replication Base. The window also displays the products that are yet to be
installed. Click Next to begin the installation.

Figure 4-63 IBM TotalStorage Productivity Center installation information


5. The Package Location for Disk and Replication Manager window (Figure 4-64) opens.
Enter the appropriate information and click Next.

Figure 4-64 Package location for Productivity Center Disk and Replication

6. The Information for Disk and Replication Base Manager panel (see Figure 4-65) opens.
Click Next.

Figure 4-65 Installer information


7. The Welcome panel (see Figure 4-66) opens. It indicates that the Disk and Replication
Base Manager installation wizard will be launched. Click Next.

Figure 4-66 IBM TotalStorage Productivity Center for Disk and Replication Base welcome information


8. In the Destination Directory panel (Figure 4-67), you confirm the target directories. Enter
the directory path or accept the default directory and click Next.

Figure 4-67 IBM TotalStorage Productivity Center for Disk and Replication Base Installation directory

9. In the IBM WebSphere Instance Selection panel (see Figure 4-68), click Next.

Figure 4-68 WebSphere Application Server information


10.If the installation user ID privileges were not set, you see an information panel stating that
you need to set the privileges (see Figure 4-69). Click Yes.

Figure 4-69 Verifying the effective privileges

11.The required user privileges are set and an informational window opens (see Figure 4-70).
Click OK.

Figure 4-70 Message indicating the enablement of the required privileges

12.At this point, the installation terminates. You must close the installer. Log off of Windows,
log back on again, and then restart the installer.


13.In the Installation Type panel (Figure 4-71), select Typical and click Next.

Figure 4-71 IBM TotalStorage Productivity Center for Disk and Replication Base type of installation

14.If the IBM Director Support Program and IBM Director Server service is still running, the
Servers Check panel (see Figure 4-72) opens and prompts you to stop the services. Click
Next to stop the services.

Figure 4-72 Server checks


15.In the User Name Input 1 of 2 panel (Figure 4-73), enter the name and password for the
IBM TotalStorage Productivity Center for Disk and Replication Base super user ID. This
user name must be defined to the operating system. In our environment we used
tpccimom as our super user. After entering the required information click Next to continue.

Figure 4-73 IBM TotalStorage Productivity Center for Disk and Replication Base superuser information

16.If the specified super user ID is not defined to the operating system a window asking if you
would like to create it appears (see Figure 4-74). Click Yes to continue.

Figure 4-74 Create new local user account


17.In the User Name Input 2 of 2 panel (Figure 4-75), enter the user name and password for
the IBM DB2 Universal Database Server. This is the user ID that was specified when DB2
was installed (see Figure 4-8 on page 90). Click Next to continue.

Figure 4-75 IBM TotalStorage Productivity Center for Disk and Replication Base DB2 user information


18.The SSL Configuration panel (Figure 4-76) opens. If you selected IBM TotalStorage
Productivity Center for Disk and Replication Base Server, then you must enter the fully
qualified name of the two server key files that were generated previously or that must be
generated during or after the IBM TotalStorage Productivity Center for Disk and
Replication Base installation. The information that you enter will be used later.
a. Choose either of the following options:
• Generate a self-signed certificate: Select this option if you want the installer to
automatically generate these certificate files. We generate the certificates in our
installation.
• Defer the generation of the certificate as a manual post-installation task:
Select this option if you want to manually generate these certificate files after the
installation, using WebSphere Application Server ikeyman utility.
b. Enter the Key file and Trust file passwords. The passwords must be a minimum of six
characters in length and cannot contain spaces. You should record the passwords in
the worksheets provided in Appendix A, “Worksheets” on page 991.
c. Click Next.

Figure 4-76 Key and Trust file options


The Generate Self-Signed Certificate window opens (see Figure 4-77). Complete all the
required fields and click Next to continue.

Figure 4-77 IBM TotalStorage Productivity Center for Disk and Replication Base Certificate information


19.Next you see the Create Local Database window (Figure 4-78). Accept the default
database name of DMCOSERV, or optionally enter the database name. Click Next to
continue.

Note: The database name must be unique to IBM TotalStorage Productivity Center for
Disk and Replication Base. You cannot share the IBM TotalStorage Productivity Center
for Disk and Replication Base database with any other applications.

Figure 4-78 IBM TotalStorage Productivity Center for Disk and Replication Base database name


20.The Preview window (Figure 4-79) displays a summary of all of the choices that you made
during the customizing phase of the installation. Click Install to complete the installation.

Figure 4-79 IBM TotalStorage Productivity Center for Disk and Replication Base Installer information


21.The DB2 database is created, the keys are generated, and the Productivity Center for Disk
and Replication base is installed. The Finish window opens. You can view the log file for
any possible error messages. The log file is located in installeddirectorylogsdmlog.txt.
The dmlog.txt file contains a trace of the installation actions. Click Finish to complete the
installation.

Figure 4-80 Productivity Center for Disk and Replication Base Installer - Finish

Notepad opens and displays the post-installation tasks information. Read the information and
complete any required tasks.
22.The Install Status window (Figure 4-81) opens after the successful Productivity Center for
Disk and Replication Base installation. Click Next.

Figure 4-81 Install Status for Productivity Center for Disk and Replication Base successful


4.3.4 IBM TotalStorage Productivity Center for Disk
The next product to install is the Productivity Center for Disk as indicated in Figure 4-82. Click
Next to begin the installation.

Figure 4-82 IBM TotalStorage Productivity Center installer information

1. A window (Figure 4-83) opens that prompts you for the package location for CD-ROM
labeled IBM TotalStorage Productivity Center for Disk. Enter the appropriate information
and click Next.

Figure 4-83 Productivity Center for Disk installation package location


2. The next window that opens indicates that the IBM TotalStorage Productivity Center for
Disk installer wizard will be launched (see Figure 4-84). Click Next.

Figure 4-84 IBM TotalStorage Productivity Center for Disk installer

3. The Productivity Center for Disk Installer - Welcome panel (see Figure 4-85) opens. Click
Next.

Figure 4-85 IBM TotalStorage Productivity Center for Disk Installer Welcome


4. The Destination Directory panel (Figure 4-86) opens. Enter the directory path or accept
the default directory and click Next.

Figure 4-86 Productivity Center for Disk Installer - Destination Directory

5. The Installation Type panel (Figure 4-87) opens. Select Typical and click Next.

Figure 4-87 Productivity Center for Disk - Installation Type


6. The Create Local Database panel (Figure 4-88) opens. Accept the default database name
of PMDATA or re-enter a new database name. Then click Next.

Figure 4-88 IBM TotalStorage Productivity Center for Disk - Create Local Database


7. Review the information on the IBM TotalStorage Productivity Center for Disk – Preview
panel (Figure 4-89) and click Install.

Figure 4-89 IBM TotalStorage Productivity Center for Disk Installer - Preview

8. The installer creates the required database (see Figure 4-90) and installs the product. You
see a progress bar for the Productivity Center for Disk installation status.

Figure 4-90 Productivity Center for Disk DB2 database creation


9. When the installation is complete, you see the Finish panel (Figure 4-91). Review the post
installation tasks. Click Finish.

Figure 4-91 Productivity Center for Disk Installer - Finish

10.The Install Status window (Figure 4-92) opens after the successful Productivity Center for
Disk installation. Click Next.

Figure 4-92 Install Status for Productivity Center for Disk successful


4.3.5 IBM TotalStorage Productivity Center for Replication
A panel opens that indicates that the installation for IBM TotalStorage Productivity Center for
Replication is about to begin (see Figure 4-93). Click Next to begin the installation.

Figure 4-93 IBM TotalStorage Productivity Center installation overview

1. The Package Location for Replication Manager panel (Figure 4-94) opens. Enter the
appropriate information and click Next.

Figure 4-94 Productivity Center for Replication install package location


Replication installer wizard will be launched (see Figure 4-95). Click Next.

Figure 4-95 Productivity Center for Replication installer

3. The Welcome window (Figure 4-96) opens. It suggests documentation that you can review
prior to the installation. Click Next to continue or click Cancel to exit the installation.

Figure 4-96 IBM TotalStorage Productivity Center for Replication Installer – Welcome


4. The Destination Directory panel (Figure 4-97) opens. Enter the directory path or accept
the default directory. Click Next to continue.

Figure 4-97 IBM TotalStorage Productivity Center for Replication Installer – Destination Directory


5. The next panel (see Figure 4-98) asks you to select the installation type. Select the
Typical radio button and click Next.

Figure 4-98 Productivity Center for Replication Installer – Installation Type


6. In the Create Local Database for ‘Hardware’ Subcomponent window (see Figure 4-99), in
the Database name field, enter a value for the new Hardware subcomponent database or
accept the default. We recommend that you accept the default. Click Next.

Note: The database name must be unique to the Replication Manager subcomponent.
You cannot share the Replication Manager subcomponent database with any other
applications or with other Replication Manager subcomponents.

Figure 4-99 IBM TotalStorage Productivity Center for Replication: Hardware subcomponent


7. In the Create Local Database for ‘ElementCatalog’ Subcomponent window (see
Figure 4-100), in the Database name field, enter for the new Element Catalog
subcomponent database or accept the default. Click Next.


Figure 4-100 IBM TotalStorage Productivity Center for Replication: Element Catalog subcomponent


8. In the Create Local Database for ‘ReplicationManager’ Subcomponent window (see
Figure 4-101), in the Database name field, enter the new Replication Manager
subcomponent database or accept the default. Click Next.


Figure 4-101 TotalStorage Productivity Center for Replication: Replication Manager subcomponent


9. In the Create Local Database for ‘ReplicationManager’ Subcomponent window (see
Figure 4-102), in the Database name field, enter the new SVC hardware subcomponent
database or accept the default. Click Next.


Figure 4-102 IBM TotalStorage Productivity Center for Replication: SVC Hardware subcomponent


10.The Setting Tuning Cycle Parameter window (Figure 4-103) opens. Accept the default
value of tuning every 24 hours or change the value. You can change this value later in the
ElementCatalog.properties file. Click Next.

Figure 4-103 IBM TotalStorage Productivity Center for Replication: Database tuning cycle


11.Review the information in the TotalStorage Productivity Center for Replication Installer –
Preview panel (Figure 4-104). Click Install.

Figure 4-104 IBM TotalStorage Productivity Center for Replication Installer – Preview


12.You see the Productivity Center for Replication Installer - Finish panel (see Figure 4-105)
upon successful installation. Read the post installation tasks. Click Finish to complete the
installation.

Figure 4-105 Productivity Center for Replication Installer – Finish

13.The Install Status window (Figure 4-106) opens after the successful Productivity Center
for Disk and Replication Base installation. Click Next.

Figure 4-106 Install Status for Productivity Center for Replication successful


4.3.6 IBM TotalStorage Productivity Center for Fabric
Prior to installing IBM TotalStorage Productivity Center for Fabric, you must complete several
prerequisite tasks. These tasks are described in detail in 3.11, “IBM TotalStorage Productivity
Center for Fabric” on page 75. Specifically, complete the tasks in the following sections:
3.10, “Installing SNMP” on page 73
3.11.1, “The computer name” on page 75
Figure 3.11.2 on page 75
3.11.3, “Windows Terminal Services” on page 75
“User IDs and password considerations” on page 76
3.11.4, “Tivoli NetView” on page 76
3.11.5, “Personal firewall” on page 77
“Security considerations” on page 77

Installing the manager
After successful installation of the Productivity Center for Replication, the Suite Installer
begins the installation of Productivity Center for Fabric (see Figure 4-107). Click Next.



1. The panel that opens prompts you to specify the location of the install package for
Productivity Center for Fabric Manager (see Figure 4-108). Enter the appropriate path and
click Next.

Important: If you used the demonstration certificates, point to the CD-ROM drive. If
you generated new certificates, point to the manager CD image with the new
agentTrust.jks file.

Figure 4-108 Productivity Center for Fabric install package location


Fabric installer wizard will be launched (see Figure 4-109). Click Next.

Figure 4-109 Productivity Center for Fabric installer

3. A window opens in which you select the language to use for the wizard (see Figure 4-110).
Select the required language and click OK.

Figure 4-110 IBM TotalStorage Productivity Center for Fabric installer: Selecting the language


4. A panel opens asking you to select the type of installation you wish to perform
(Figure 4-111).
In this case, we install the IBM TotalStorage Productivity Center for Fabric code. You can
also use the Suite Installer to perform remote deployment of the Fabric Agent. You can
perform this operation only if you installed the common agent previously on machines. For
example, you may have installed the Data Agent on the machines and want to add the
Fabric Agent to the same machines. You must have the Fabric Manager installed before
you can deploy the Fabric Agent. You cannot select both Fabric Manager Installation and
Remote Fabric Agent Deployment at the same time. You can only select one option.
Click Next.

Figure 4-111 Fabric Manager installation


5. The Welcome panel (Figure 4-112) opens. Click Next.

Figure 4-112 IBM TotalStorage Productivity Center for Fabric: Welcome information

6. The next window that opens prompts you to confirm the target directory (see
Figure 4-113). Enter the directory path or accept the default directory. Click Next.

Figure 4-113 IBM TotalStorage Productivity Center for Fabric installation directory


7. In the next panel (see Figure 4-114), you specify the port number. This is a range of 25
port numbers for use by IBM TotalStorage Productivity Center for Fabric. The first port
number that you specify is considered the primary port number. You only need to enter
the primary port number. The primary port number and the next 24 numbers are reserved
for use by IBM TotalStorage Productivity Center for Fabric. For example, if you specify port
number 9550, IBM TotalStorage Productivity Center for Fabric uses port numbers 9550
through 9574.
Ensure that the port numbers you use are not used by other applications at the same time.
To determine which port numbers are in use on a particular computer, type either of the
following commands from a command prompt.
netstat -a
netstat -an
We recommend that you use the first of these two commands.
The port numbers in use on the system are listed in the Local Address column of the
output. This field has the format host:port. Enter the primary port number and click Next.

Figure 4-114 IBM TotalStorage Productivity Center for Fabric port number


8. As shown in Figure 4-115, select the database repository, either DB2 or Cloudscape. If
you select DB2, you must have previously installed DB2 on the server. DB2 is the
recommended installation option. Click Next.

Figure 4-115 IBM TotalStorage Productivity Center for Fabric database selection type

9. In the next panel (see Figure 4-117 on page 164), select the WebSphere Application
Server to use in the installation. WebSphere Application Server was installed as part of
the prerequisite software so we chose the Non Embedded (Full) WebSphere
Application Server option. If the Fabric manager is to be installed standalone on a server
choose the Embedded WebSphere Application Server - Express option. Click Next.

Figure 4-116 Productivity Center for Fabric WebSphere Application Server type selection


10.The Single/Multiple User ID/Password Choice panel (see Figure 4-117), using DB2,
opens. If you select DB2 as your database, you see this panel. This panel allows you to
use the DB2 administrative user ID and password for the DB2, WebSphere, Host
Authentication, and NetView.
If you select all the boxes, you are only prompted for the DB2 user ID and password which
is used for all instances. In our install we only selected DB2 and NetView. A different user
ID and password will be used for WebSphere and Host Authentication.

Note: If you selected IBM Cloudscape as your database, this panel is not displayed.

Click Next.

Figure 4-117 IBM TotalStorage Productivity Center for Fabric user and password options


11.The DB2 Administrator user ID and password panel (Figure 4-118), using DB2, opens. If
you selected DB2 as your database, you see this panel. This panel allows you to use the
DB2 administrative user ID and password for the DB2. The user ID and password
specified during the DB2 installation in Figure 4-8 on page 90 was used in this example.
Enter the required user ID and password. Click Next. The installer will verify that the user
ID entered exists.

Figure 4-118 IBM TotalStorage Productivity Center for Fabric database user information


12.In the next window (see Figure 4-119) that opens, type the name of the new database in
the Type database name: field or accept the default. In our install we accepted the default
database name. Click Next.

Note: The database name must be unique. You cannot share the IBM TotalStorage
Productivity Center for Fabric database with any other applications.

Figure 4-119 IBM TotalStorage Productivity Center for Fabric database name

13.Since we did not check the box for WebSphere in Figure 4-117 on page 164, the panel in
Figure 4-120 on page 167 opens prompting for a WebSphere user ID and password. We
used the tpcadmin user ID, which is what we used for the IBM Director service account
(refer to Figure 4-46 on page 118). Enter the required information and click Next.


Figure 4-120 WebSphere Application Server user ID and password

14.Since we also did not check the box for Host Authentication (Figure 4-117 on page 164),
the following panel (Figure 4-121) opens. Enter the password for Host Authentication. This
password is used by the Fabric agents. Click Next.

Figure 4-121 Host Authentication password

15.In the window (Figure 4-122 on page 168) that opens, enter the parameters for the Tivoli
NetView drive name. Click Next.


Figure 4-122 IBM TotalStorage Productivity Center for Fabric database drive information

16.The Agent Manager Information panel (Figure 4-123 on page 169) opens. You must
complete the following fields:
– Agent manager name or IP address: This is the host name or IP address of your
Agent Manager.
– Agent manager registration port: This is the port number of your Agent Manager.
The default value is 9511.
– Agent Manager public port: This is a public port. The default value is 9513.
– Agent registration password (twice): This is the password used to register the
common agent with the Agent Manager as shown in Figure 4-27 on page 105. If the
password is not set and the default is accepted, the password is changeMe. This
password is case sensitive.
The agent registration password resides in the AgentManager.properties file where the
Agent Manager is installed. It is located in the following directory:
%WSAS_INSTALL_ROOT%InstalledApps<cell>AgentManager.earAgentManag
er.warWEB-INFclassesresource
– Resource manager registration user ID: This is the user ID used to register the
resource manager with the Agent Manager. The default is manager.
The Resource Manager registration user ID and password reside in the
Authorization.xml file where the Agent Manager is installed. It is located in the following
directory:
<Agent_Manager_install_dir>config
– Resource manager registration password (twice): This is the password used to
register the resource manager with the Agent Manager. The default is password.
Fill in the information and click Next.


Figure 4-123 IBM TotalStorage Productivity Center for Fabric Agent Manager information

17.The next panel (Figure 4-124) that opens provides information about the location and size
of IBM TotalStorage Productivity Center for Fabric - Manager. Click Next.

Figure 4-124 IBM TotalStorage Productivity Center for Fabric installation information

18.You see the Status panel. The installation can take about 15 to 20 minutes to complete.
19.When the installation has completed, you see a panel indicating that the wizard
successfully installed the Fabric Manager (see Figure 4-125 on page 170). Click Next.


Figure 4-125 IBM TotalStorage Productivity Center for Fabric installation status

20.In the next panel (see Figure 4-126), you are prompted to restart your computer. Select
No, I will restart my computer later because you do not want to restart your computer
now. Click Finish to complete the installation.

Figure 4-126 IBM TotalStorage Productivity Center for Fabric restart options

21.The Install Status panel (see Figure 4-127 on page 171) opens. It indicates that the
Productivity Center for Fabric installation was successful. Click Next.



4.3.7 IBM TotalStorage Productivity Center for Data
Prior to installing IBM TotalStorage Productivity Center for Data, you need to complete
several prerequisite tasks. These tasks are described in detail in 3.12, “IBM TotalStorage
Productivity Center for Data” on page 78. Specifically you must complete the tasks in the
following sections:
3.12.1, “Server recommendations” on page 78
3.12.2, “Supported subsystems and databases” on page 78
3.12.3, “Security considerations” on page 79
3.12.4, “Creating the DB2 database” on page 81
The IBM TotalStorage Productivity Center for Data database needs to be created before
you begin the installation.

This section provides an overview of the steps you need to perform when installing IBM
TotalStorage Productivity Center for Data.

Important: Make sure that the Tivoli Agent Manager service is started before you begin
the installation.

You see the panel indicating that the installation of Productivity Center for Data - Manager is
about to begin (see Figure 4-128 on page 172). Click Next to begin the installation.


Figure 4-128 IBM TotalStorage Productivity Center for Data installation information

1. In the window that opens, you are prompted to enter the install package location for IBM
TotalStorage Productivity Center for Data (see Figure 4-129). Enter the appropriate

Figure 4-129 Productivity Center for Data install package location

Data installer wizard will be launched (see Figure 4-130 on page 173). Click Next.


Figure 4-130 Productivity Center for Data installer

3. In the next panel (see Figure 4-131), select Install Productivity Center for Data and click
Next.

Figure 4-131 IBM TotalStorage Productivity Center for Data install window


4. Read the License Agreement shown in Figure 4-132. Indicate your acceptance of the
agreement by selecting the I have read and AGREE to abide by the license agreement
above check box. Then click Next.

Figure 4-132 IBM TotalStorage Productivity Center for Data license agreement

5. The next panel asks you to confirm that you read the license agreement (see
Figure 4-133). Click Yes to indicate that you have read and accepted the license
agreement.

Figure 4-133 Confirmation the Productivity Center for Data license agreement has been read


6. The next window shown in Figure 4-134 allows you to choose the type of installation that
you are performing. Select The Productivity Center for Data Server and an Agent on
this machine.
This installs the server, agent, and user interface components on the machine where the
installation program is running. You must install the server on at least one machine within
your environment.
Click Next.

Figure 4-134 IBM TotalStorage Productivity Center for Data selection options


7. Review and enter the license key for the appropriate functions if required. See
Figure 4-135. Click Next.

Figure 4-135 IBM TotalStorage Productivity Center for Data license key information


8. The installation program validates the license key and you are asked to select the
relational database management system (RDBMS) that you want to host the Data
Manager repository. See Figure 4-136.
The repository is a set of relational database tables where Data Manager builds a
database of statistics to keep track of your environment. For our installation, we select IBM
DB2 UDB. Click Next.

Figure 4-136 IBM TotalStorage Productivity Center for Data database selection

9. The Create Service Account panel opens to create the TSRMsrv1 local account. Click
Yes.

Figure 4-137 Create Service Account


10.In the next window (see Figure 4-138), complete these tasks:
a. Select the database that was created as a prerequisite. Refer to 3.12.4, “Creating the
DB2 database” on page 81.
b. Fill in the required user ID and password. This is the DB2 user ID and password
defined previously.
c. Click Next.

Figure 4-138 IBM TotalStorage Productivity Center for Data database selection option

11.The Repository Creation Parameters panel (see Figure 4-139 on page 179) for UDB
opens. On this panel you can specify the database schema and tablespace name.
If you are using DB2 as the repository, you can also choose how you will manage the
database space:
– System Managed (SMS): This option indicates that the space is managed by the OS.
In this case you specify the Container Directory, which is then managed by the system,
and can grow as large as the free space on the file system.

Tip: If you do not have in house database skills, the System Managed approach is
recommended.

– Database Managed (DMS): This option means that the space is managed by the
database. In this case you need to specify the Container Directory, Container File, and
Size fields. The Container File specifies a filename for the repository, and Size is the
predefined space for that file. You can later change this by using the ALTER
TABLESPACE command. We accepted the defaults.


Tip: We recommend that you use meaningful names for Container Directory and
Container File at installation. This can help you in case you need to find the
Container File.

Enter the necessary information and click Next.

Figure 4-139 IBM TotalStorage Productivity Center for Data repository information


12.The Productivity Center for Data Parameters panel (Figure 4-140) opens. Use the Agent
Manager Parameters window (Figure 4-141 on page 182) to provide information about the
Agent Manager installed in your environment. Click Next.

Figure 4-140 IBM TotalStorage Productivity Center for Data installation parameters


13.The Agent Manager Parameters panel (Figure 4-141 on page 182) provides information
about the Agent Manager installed in your environment. Table 4-1 provides a description
of the fields in the panel.

Table 4-1 Agent Manager Parameters descriptions
Field Description

Hostname Enter the fully qualified network name or IP address of the Agent
Manager server as seen by the agents.

Registration Port Enter the port number of the Agent Manager. The default is 9511.

Public Port Enter the public port for Agent Manager. The default is 9513.

Resource Manager Username Enter the Agent Manager user ID. This is the user ID used to
register the common agent with the Agent Manager. The default is
manager.

Resource Manager Password Enter the password used to register the common agent with the
Agent Manager.

Agent Registration password This is the password that was set during the Tivoli Agent Manager
installation Figure 4-27 on page 105. The default password is
changeMe, the password is stored in the AgentManager.properties
file, in the %install dir%AgentManagerimage directory.

Click Next.

Note: If an error is displayed during this part of the installation, verify that the
Agenttrust.jks file was copied across and verify the Agent Registration password.


Figure 4-141 IBM TotalStorage Productivity Center for Data Agent Manager install information


14.Use the NAS Discovery Parameters panel in Figure 4-142 to configure Data Manager for
use with any network-attached storage (NAS) devices in your environment. Click Next.
You can leave the fields blank if you do not have any NAS devices.

Figure 4-142 IBM TotalStorage Productivity Center for Data NAS options


15.The Space Requirements panel for the Productivity Center for Data Server (Figure 4-143)
opens. Enter the directory path or accept the default directory.
If the current disk or device does not have enough space for the installation, then you can
enter a different location for the installation in the Choose the installation directory field. Or
you can click Browse to browse your system for an available and appropriate space. The
default installation directory is C:Program FilesIBMTPCData.
Click Next.

Figure 4-143 IBM TotalStorage Productivity Center for Data installation destination options

16.Confirm the path for installing the Productivity Center for Data Server as shown in
Figure 4-144. At this point, the installation process has gathered all of the information that
is needed to perform the installation. Click OK.

Figure 4-144 IBM TotalStorage Productivity Center for Data Server destination path confirmation


17.Review and change the Productivity Center for Data Agent Parameters (see Figure 4-145)
as required. We recommend that you accept the defaults. Click Next.

Figure 4-145 IBM TotalStorage Productivity Center for Data agent parameters


18.The Windows Service Account panel shown in Figure 4-146 opens. Choose Create a local
account for the agent to run under and click Next.

Figure 4-146 Windows Service Account


19.The Space Requirements panel (see Figure 4-147) opens for the Productivity Center for
Data Agent. Enter the directory path or accept the default directory.
If the current disk or device does not have enough space for the installation, then you can
enter a different location for the installation in the Choose the Common Agent installation
directory field. Or you can click Browse to browse your system for an available and
appropriate space. The default installation directory is C:Program FilesTivoliep.
Click Next.

Figure 4-147 IBM TotalStorage Productivity Center for Datacommon agent installation information

20.When you see a message similar to the one in Figure 4-148, confirm the path where
Productivity Center for Data Agent is to be installed. At this point, the installation process
has gathered all of the information necessary to perform the installation. Click OK.

Figure 4-148 IBM TotalStorage Productivity Center for Data Agent destination path confirmation


21.When you see a window similar to the example in Figure 4-149, review the choices that
you have made. Then click Next.

Figure 4-149 IBM TotalStorage Productivity Center for Data preview options


22.A window opens that tracks the progress of the installation (see Figure 4-150).

Figure 4-150 IBM TotalStorage Productivity Center for Data installation information


23.When the installation is done, the progress window shows a message indicating that the
installation completed successfully (see Figure 4-151). Review this panel and click Done.

Figure 4-151 IBM TotalStorage Productivity Center for Data success information

24.The Install Status panel opens showing the message The Productivity Center for Data
installation was successful. Click Next to complete the installation.

Figure 4-152 Install Status for Productivity Center for Data successful


5

Chapter 5. CIMOM install and configuration
This chapter provides a step-by-step guide to configure the Common Information Model
Object Manager (CIMOM), LSI Provider, and Service Location Protocol (SLP) that are
required to use the IBM TotalStorage Productivity Center.


5.1 Introduction
After you have completed the installation of TotalStorage Productivity Center for Disk,
TotalStorage Productivity Center for Replication, TotalStorage Productivity Center for Fabric,
or TotalStorage Productivity Center for Data, you will need to install and configure the
Common Information Model Object Manager (CIMOM) and Service Location Protocol (SLP)
agents.

Note: For the remainder of this chapter, we refer to the TotalStorage Productivity Center
for Disk, TotalStorage Productivity Center for Replication, TotalStorage Productivity Center
for Fabric, and TotalStorage Productivity Center for Data simply as the TotalStorage
Productivity Center.

The TotalStorage Productivity Center uses SLP as the method for CIM clients to locate
managed objects. The CIM clients may have built in or external CIM agents. When a CIM
agent implementation is available for a supported device, the device may be accessed and
configured by management applications using industry-standard XML-over-HTTP
transactions.

In this chapter we describe the steps for:
Planning considerations for Service Location Protocol (SLP)
SLP configuration recommendation
General performance guidelines
Planning considerations for CIMOM
Installing and configuring CIM agent for Enterprise Storage Server and DS6000/DS8000
Verifying connection to ESS
Verify connection to DS6000/DS8000
Setting up Service Location Protocol Directory Agent (SLP DA)
Installing and configuring CIM agent for DS 4000 Family
Configuring CIM agent for SAN Volume Controller

5.2 Planning considerations for Service Location Protocol
The Service Location Protocol (SLP) has three major components, Service Agent (SA) and
User Agent (UA) and a Directory Agent (DA). The SA and UA are required components and
DA is an optional component.

You may have to make a decision whether to use SLP DA in your environment based on
considerations as described below.

5.2.1 Considerations for using SLP DAs
You may consider to use a DA to reduce the amount of multicast traffic involved in service
discovery. In a large network with many UAs and SAs, the amount of multicast traffic involved
in service discovery can become so large that network performance degrades.


By deploying one or more DAs, UAs must unicast to DAs for service and SAs must register
with DAs using unicast. The only SLP-registered multicast in a network with DAs is for active
and passive DA discovery.

SAs register automatically with any DAs they discover within a set of common scopes.
Consequently, DAs within the UAs scopes reduce multicast. By eliminating multicast for
normal UA request, delays and time-outs are eliminated.

DAs act as a focal point for SA and UA activity. Deploying one or several DAs for a collection
of scopes provides a centralized point for monitoring SLP activity.

You may consider to use DAs in your enterprise if any of the following conditions are true:
Multicast SLP traffic exceeds 1% of the bandwidth on your network, as measured by
snoop.
UA clients experience long delays or time-outs during multicast service request.
You would like to centralize monitoring of SLP service advertisements for particular
scopes on one or several hosts. You can deploy any number of DAs for a particular scope
or scopes, depending on the need to balance the load.
Your network does not have multicast enabled and consists of multiple subnets that must
share services.
The configuration of an SLP DA is particularly recommended when there are more than 60
SAs that need to respond to any given multicast service request.

5.2.2 SLP configuration recommendation
Some configuration recommendations are provided for enabling TotalStorage Productivity
Center to discover a larger set of storage devices. These recommendations cover some of
the more common SLP configuration problems.

This topic discusses router configuration and SLP directory agent configuration.

Router configuration
Configure the routers in the network to enable general multicasting or to allow multicasting for
the SLP multicast address and port, 239.255.255.253, port 427. The routers of interest are
those that are associated with subnets that contain one or more storage devices that are to
be discovered and managed by TotalStorage Productivity Center. To configure your router
hardware and software, refer to your router reference and configuration documentation.

Attention: Routers are sometimes configured to prevent passing of multicast packets
between subnets. Routers configured this way prevent discovery of systems between
subnets using multicasting. Routers can also be configured to restrict the minimum
multicast TTL (time-to-live) for packets it passes between subnets, which can result in the
need to set the Multicast TTL higher to discover systems on the other subnets of the router.

The Multicast TTL controls the time-to-live for the multicast discovery packets. This value
typically corresponds to the number of times a packet is forwarded between subnets,
allowing control of the scope of subnets discovered.

Multicast discovery does not discover Director V1.x systems or systems using TCP/IP
protocol stacks that do not support multicasting (for example, some older Windows 3.x and
Novell 3.x TCP/IP implementations).

Chapter 5. CIMOM install and configuration 193

SLP directory agent configuration
Configure the SLP directory agents (DAs) to circumvent the multicast limitations. With
statically configured DAs, all service requests are unicast by the user agent. Therefore, it is
possible to configure one DA for each subnet that contains storage devices that are to be
discovered byTotalStorage Productivity Center. One DA is sufficient for each subnet. Each of
these DAs can discover all services within its own subnet, but no other services outside its
own subnet. To allow TotalStorage Productivity Center to discover all of the devices, it needs
to be statically configured with the addresses of each of these DAs. This can be
accomplished using the TotalStorage Productivity Center This setup is described in
“Configuring TotalStorage Productivity Center for SLP discovery” on page 223.

5.3 General performance guidelines
Here are some general performance considerations for configuring the TotalStorage
Productivity Center for Disk and TotalStorage Productivity Center for Replication
environment.
Do not overpopulate the SLP discovery panel with SLP agent hosts. Remember that
TotalStorage Productivity Center for Disk includes a built-in SLP User Agent (UA) that will
receive information about SLP Service Agents and Directory Agents (DA) that reside in
the same subnet as the TotalStorage Productivity Center for Disk installation.
You should have not more than one DA per subnet.
Misconfiguring the IBM Director discovery preferences may impact performance on auto
discovery or on device presence checking. It may also result in application time-outs, as
attempts are made to resolve and communicate with hosts that are not available.
It should be considered mandatory to run the ESS CLI and ESS CIM agent or DS CIM
agent, and LSI Provider software on another host of comparable size to the main
TotalStorage Productivity Center server. Attempting to run a full TotalStorage Productivity
Center implementation (Disk Manager, Data Manager, Fabric Manager, Replication
Manager, DB2, IBM Director and the WebSphere Application server) on the same host as
the CIM agent, will result in dramatically increased wait times for data retrieval. You may
also experience resource contention and port conflicts.

5.4 Planning considerations for CIMOM
The CIM agent includes a CIM Object Manager (CIMOM) which adapts various devices using
a plug-in called a provider. The CIM agent can work as a proxy or can be imbedded in
storage devices. When the CIM agent is installed as a proxy, the IBM CIM agent can be
installed on the same server that supports the device user interface. Figure 5-1 on page 195
shows overview of CIM agent.


Figure 5-1 CIM Agent overview

You may plan to install CIM agent code on the same server which also has device
management interface or you may install it on a separate server.

Attention: At this time only few devices come with an integrated CIM Agent, most devices
need a external CIMOM for CIM enable management applications (CIM Clients) to be able
to communicate with device.

For the ease of the installation IBM provides an ICAT (short for Integrated Configuration
Agent Technology) which is a bundle that mainly includes the CIMOM, the device provider
and an SLP SA.

5.4.1 CIMOM configuration recommendations
Following recommendations are based on our experience in ITSO Lab environment:
The CIMOM agent code which you are planning to use, must be supported by the installed
version of TotalStorage Productivity Center. You may refer to the link below for the latest
updates:
http://www-1.ibm.com/servers/storage/support/software/tpc/
You must have the CIMOM supported firmware level on the storage devices. It you have
an incorrect version of firmware, you may not be able to discover and manage any the
storage devices.
The data traffic between CIMOM agent and device can be very high, especially during
performance data collection. Hence it is recommended to have a dedicated server for the
CIMOM agent. Although, you may configure the same CIMOM agent for multiple devices
of same type.
You may also plan to locate this server within same data center where storage devices are
located. This is in consideration to firewall port requirements. Typically, it is best practice
to minimize firewall port openings between data center and external network. If you
consolidate the CIMOM servers within the data center then you may be able to minimize
the need to open the firewall ports only for TotalStorage Productivity Center
communication with CIMOM.
Co-location of CIM agent instances of the differing type on the same server is not
recommended because of resource contention.
It is strongly recommended to have a separate and dedicated servers for CIMOM agents
and TotalStorage Productivity Center. This is due to resource contention, TCP/IP port
requirements and system services co-existence.


5.5 Installing CIM agent for ESS
Before starting Multiple Device Manager discovery, you must first configure the Common
Information Model Object Manager (CIMOM) for ESS.

The ESS CIM Agent package is made up of the following parts (see Figure 5-2).

Figure 5-2 ESS CIM Agent Package

This section provides an overview of the installation and configuration of the ESS CIM Agent
on a Windows 2000 Advanced Server operating system.

5.5.1 ESS CLI Install
The following list of installation and configuration tasks are in the order in which they should
be performed:
Before you install the DS CIM Agent you must install the IBM TotalStorage Enterprise
Storage System Command Line Interface (ESS CLI) if you plan to manage 2105-F20s or
2105-800s with this CIM agent. The DS CIM Agent installation program checks your
system for the existence of the ESS CLI and provides the warning shown in Figure 5-16
on page 205 if no valid ESS CLI is found.

Attention: If you are upgrading from a previous version of the ESS CIM Agent, you must
uninstall the ESS CLI software that was required by the previous CIM Agent and reinstall
the latest ESS CLI software, you must have a minimum ESS CLI level of 2.4.0.236.

Perform the following steps to install the ESS CLI for Windows:
1. Insert the CD for the ESS CLI in the CD-ROM drive, run the setup and follow the
instructions as shown in Figure 5-3 on page 197 through Figure 5-11 on page 201.

Note: The ESS CLI installation wizard detects if you have an earlier level of the ESS CLI
software installed on your system and uninstalls the earlier level. After you uninstall the
previous version, you must restart the ESS CLI installation program to install the current
level of the ESS CLI.


Figure 5-3 ESS CLI InstallShield Wizard I

2. Select I accept the terms of the license agreement and click Next.

Figure 5-4 ESS CLI License agreement

3. Click Next.


Figure 5-5 ESS CLI choose target system panel

4. Click Next.

Figure 5-6 ESS CLI Setup Status panel

5. Click Next.


Figure 5-7 ESS CLI selected options summary

Figure 5-8 ESS CLI Installation Progress

6. Click Next.


Figure 5-9 ESS CLI installation complete panel

7. Read the information and click Next.

Figure 5-10 ESS CLI Readme

8. Reboot your system before proceeding with the ESS CIM Agent installation. You must do
this because the ESS CLI is dependent on environmental variable settings which will not
be in effect for the ESS CIM Agent. This is because the CIM Agent runs as a service
unless you reboot your system.


Figure 5-11 ESS CLI Reboot panel

9. Verify that the ESS CLI is installed:
– Click Start → Settings → Control Panel.
– Double-click the Add/Remove Programs icon.
– Verify that there is an IBM ESS CLI entry.
10.Verify that the ESS CLI is operational and can connect to the ESS. For example, from a
command prompt window, issue the following command:
esscli -u userid -p password -s 9.1.11.111 list server
Where:
– 9.1.11.111 represents the IP address of the Enterprise Storage Server
– usedid represents the Enterprise Storage Server Specialist user name
– password represents the Enterprise Storage Server Specialist password for the user
name

Figure 5-12 shows the response from the esscli command.

Figure 5-12 ESS CLI verification


5.5.2 DS CIM Agent install
To install the DS CIM Agent in your Windows system, perform the following steps:
1. Log on to your system as the local administrator.
2. Insert the CIM Agent for DS CD into the CD-ROM drive. The Install Wizard launchpad
should start automatically, if you have autorun mode set on your system. You should see
a launchpad window similar to Figure 5-13.
3. You may review the Readme file from the launchpad menu. Subsequently, you can click
Installation Wizard. The Installation Wizard starts executing the setup.exe program and
shows the Welcome panel in Figure 5-14 on page 203.

Note: The DS CIM Agent program should start within 15 - 30 seconds if you have autorun
mode set on your system. If the installer window does not open, perform the following
steps:

– Use a Command Prompt or Windows Explorer to change to the Windows directory on
the CD.
– If you are using a Command Prompt window, run launchpad.bat.
– If you are using Windows Explorer, double-click on the launchpad.bat file.

Note: If you using CIMOM code from IBM download Web site and not from the distribution
CD, then you must ensure to use a shorter windows directory pathname. Executing
Launchpad.bat from the longer pathname may fail. An example of a short pathname is
C:CIMOMsetup.exe.

Figure 5-13 DSCIM Agent launchpad


4. The Welcome window opens suggesting what documentation you should review prior to
installation. Click Next to continue (see Figure 5-14).

Figure 5-14 DS CIM Agent welcome window


5. The License Agreement window opens. Read the license agreement information. Select
I accept the terms of the license agreement, then click Next to accept the license
agreement (see Figure 5-15).

Figure 5-15 DS CIM Agent license agreement


The window shown in Figure 5-16 only appears if no valid ESS CLI installed. If you do not
plan to manage an ESS from this CIM agent, then click Next.

Important: If you plan to manage an ESS from this CIM agent, then click Cancel. Install
the ESS CLI following the instructions in 5.5.1, “ESS CLI Install” on page 196.

Figure 5-16 DS CIM Agent ESS CLI warning


6. The Destination Directory window opens. Accept the default directory and click Next (see
Figure 5-17).

Figure 5-17 DS CIM Agent destination directory panel


7. The Updating CIMOM Port window opens (see Figure 5-18). You Click Next to accept the
default port if it available and free in your environment. For our ITSO setup we used default
port 5989.

Note: If the default port is the same as another port already in use, modify the default
port and click Next. Use the following command to check which ports are in use:
netstat -a

Figure 5-18 DS CIM Agent port window


8. The Installation Confirmation window opens (see Figure 5-19). Click Install to confirm the
installation location and file size.

Figure 5-19 DS CIM Agent installation confirmation


9. The Installation Progress window opens (see Figure 5-20) indicating how much of the
installation has completed.

Figure 5-20 DS CIM Agent installation progress

10.When the Installation Progress window closes, the Finish window opens (see Figure 5-21
on page 210). Check the View post installation tasks check box if you want to view the
post installation tasks readme when the wizard closes. We recommend you review the
post installation tasks. Click Finish to exit the installation wizard (Figure 5-21 on
page 210).

Note: Before proceeding, you might want to review the log file for any error messages. The
log file is located in xxxlogsinstall.log, where xxx is the destination directory where the DS
CIM Agent for Windows is installed.


Figure 5-21 DS CIM Agent install successful

11.If you checked the view post installation tasks box, then the window shown in Figure 5-22
appears. Close the window when you have finished reviewing the post installation tasks.

Figure 5-22 DS CIM Agent post install readme

The launchpad window (Figure 5-13 on page 202) appears. Click Exit.


5.5.3 Post Installation tasks
Continue with the following post installation tasks for the ESS CIM Agent.

Verify the installation of the SLP
Proceed as follows:
Verify that the Service Location Protocol is started. Select Start → Settings → Control
Panel. Double-click the Administrative Tools icon. Double-click the Services icon.
Find Service Location Protocol in the Services window list. For this component, the Status
column should be marked Started as shown in Figure 5-23.

Figure 5-23 Verify Service Location Protocol started

If SLP is not started, right-click on the SLP and select Start from the pop-up menu. Wait for
the Status column to be changed to Started.

Verify the installation of the DS CIM Agent
Proceed as follows:
Verify that the CIMOM service is started. If you closed the Services window, select Start
→ Settings → Control Panel. Double-click the Administrative Tools icon. Double-click
the Services icon.
Find the IBM CIM Object Manager - ESS in the Services window list. For this component,
the Status column should be marked Started and the Startup Type column should be
marked Automatic, as shown in Figure 5-24 on page 212.


Figure 5-24 DS CIM Object Manager started confirmation

If the IBM CIM Object Manager is not started, right-click on the IBM CIM Object Manager -
ESS and select Start from the pop-up menu. Wait for the Status column to change to
Started.

If you are able to perform all of the verification tasks successfully, the ESS CIM Agent has
been successfully installed on your Windows system. Next, perform the configuration tasks.

5.6 Configuring the DS CIM Agent for Windows
This task configures the DS CIM Agent after it has been successfully installed.

Perform the following steps to configure the DS CIM Agent:
Configure the ESS CIM Agent with the information for each Enterprise Storage Server the
ESS CIM Agent is to access.
– Start → Programs → IBM TotalStorage CIM Agent for ESS → CIM agent for the
IBM TotalStorage DS Open API → Enable DS Communications as shown in
Figure 5-25.

Figure 5-25 Configuring the ESS CIM Agent

5.6.1 Registering DS Devices
Type the following command for each DS server that is configured:
addessserver <ip> <user> <password>

Where:
– <ip> represents the IP address of the Enterprise Storage Server
– <user> represents the DS Storage Server user name
– <password> represents the DS Storage Server password for the user name


Repeat the previous step for each additional DS device that you want to configure.

Note: CIMOM collects and caches the information from the defined DS servers at startup
time; the starting of the CIMOM might take a longer period of time the next time you start it.

Attention: If the username and password entered is incorrect or the DS CIM agent does
not connect to the DS this will cause a error and the DS CIM Agent will not start and stop
correctly, use following command to remove the ESS entry that is causing the problem and
reboot the server.
– rmessserver <ip>

Whenever you add or remove DS from CIMOM registration, you must re-start the CIMOM
to pick up the updated DS device list.

5.6.2 Registering ESS Devices
Proceed as follows:
Type the command addess <ip> <user> <password> command for each ESS (as shown
in Figure –):
– Where: <ip> represents the IP address of the cluster of Enterprise Storage Server
– <user> represents the Enterprise Storage Server Specialist user name
– <password> represents the Enterprise Storage Server Specialist password for the user
name. The addess command example is shown in Figure 5-26 on page 214.

Important:
DS CIM agent relies on ESS CLI connectivity from DS CIMOM server to ESS devices.
Make sure that the ESS devices you are registering are reachable and available at this
point. It is recommended to verify this by launching ESS specialist browser from the
ESS CIMOM server. You may logon to both ESS clusters for each ESS and make sure
you are authenticated with correct ESS passwords and IP addresses.
If the ESS are on the different subnet than the ESS CIMOM server and behind a
firewall, then you must authenticate through firewall first before registering the ESS with
CIMOM.
If you have a bi-directional firewall between ESS devices and CIMOM server then you
must verify the connection using rsTestConnection command of ESS CLI code. If the
ESS CLI connection is not successful, you must authenticate through the firewall in
both directions i.e from ESS to CIMOM server and also from CIMOM server to ESS.
Once you are satisfied that you are able to authenticate and receive ESS CLI heartbeat
with all the ESS successfully, you may proceed for entering ESS IP addresses.

If CIMOM agent fails to authenticate with ESSs, then it will not start-up properly and may
be very slow, since it re-tries the authentication.


Figure 5-26 The addess command example

5.6.3 Register ESS server for Copy services
Type the following command for each ESS server that is configured for Copy Services:
addesserver <ip> <user> <password>
– Where <ip> represents the IP address of the Enterprise Storage Server
– <user> represents the Enterprise Storage Server Specialist user name
– <password> represents the Enterprise Storage Server Specialist password for the user
name
Repeat the previous step for each additional ESS device that you want to configure.
Close the setdevice interactive session by typing exit.
Once you have defined all the ESS servers, you must stop and restart the CIMOM to
make the CIMOM initialize the information for the ESS servers.

Note: CIMOM collects and caches the information from the defined ESS servers at startup
time, the starting of the CIMOM might take a longer period of time the next time you start it.

Attention: If the username and password entered is incorrect or the ESS CIM agent does
not connect to the ESS this will cause a error and the ESS CIM Agent will not start and
stop correctly, use following command to remove the ESS entry that is causing the problem
and reboot the server.
– rmessserver <ip>

Whenever you add or remove an ESS from CIMOM registration, you must re-start the
CIMOM to pick up updated ESS device list.


5.6.4 Restart the CIMOM
Perform these steps to use the Windows Start Menu facility to stop and restart the CIMOM.
This is required so that CIMOM can register new devices or un-register deleted devices:
Stop the CIMOM by selecting Start → Programs → CIM Agent for the IBM
TotalStorage DS Open API → Stop CIMOM service. A Command Prompt window
opens to track the stoppage of the CIMOM (as shown in Figure 5-27). If the CIMOM has
stopped successfully, the following message is displayed:

Figure 5-27 Stop ESS CIM Agent

Restart the CIMOM by selecting Start → Programs → CIM Agent for the IBM
TotalStorage DS Open API → Start CIMOM service. A Command Prompt window
opens to track the progress of the starting of the CIMOM. If the CIMOM has started
successfully, the message shown in Figure 5-28 is displayed.

Figure 5-28 Restart ESS CIM Agent

Note: The restarting of the CIMOM may take a while because it is connecting to the
defined ESS servers and is caching that information for future use.

5.6.5 CIMOM user authentication
Use the setuser interactive tool to configure the CIMOM for the users who will have the
authority to use the CIMOM. The user is the TotalStorage Productivity Center for Disk and
Replication superuser.

Important: A TotalStorage Productivity Center for Disk and Replication superuserid and
password must be create. This userid is initially used to by TotalStorage Productivity
Center to connect to the CIM Agent. It is easiest if this superuserid is used for all CIM
Agents. It can be set individually for each CIM Agent if necessary. This user ID should be
less than or equal to eight characters.

Upon installation of the CIM Agent for ESS, the provided default user name is “superuser”
with a default password of “passw0rd”. The first time you use the setuser tool, you must use
this user name and password combination. Once you have defined other user names, you
can start the setuser command by specifying other defined CIMOM user names.


Note: The users which you configure to have authority to use the CIMOM are uniquely
defined to the CIMOM software and have no required relationship to operating system user
names, the ESS Specialist user names, or the ESS Copy Services user names.

Here is the procedure:
– Open a Command Prompt window and change directory to the ESS CIM Agent
directory, for example “C:Program FilesIBMcimagent”.
– Type the command setuser -u superuser -p passw0rd at the command prompt to
start the setuser interactive session to identify users to the CIMOM.
– Type the command adduser cimuser cimpass in the setuser interactive session to
define new users.
• where cimuser represents the new user name to access the ESS CIM Agent
CIMOM
• cimpass represents the password for the new user name to access the ESS CIM
Agent CIMOM
Close the setdevice interactive session by typing exit.

For our ITSO Lab setup we used TPCCIMOM as superuser and TPCCIMOM as the
password.

5.7 Verifying connection to the ESS
During this task the ESS CIM Agent software connectivity to the Enterprise Storage Server
(ESS) is verified.

The connection to the ESS is through the ESS CLI software.

If the network connectivity fails or if the user name and password that you set in the
configuration task is incorrect, the ESS CIM Agent cannot connect successfully to the ESS.

The installation, verification, and configuration of the ESS CIM Agent must be completed
before you verify the connection to the ESS.
Verify that you have network connectivity to the ESS from the system where the ESS CIM
Agent is installed. Issue a ping command to the ESS and check that you can see reply
statistics from the ESS IP address.
Verify that the SLP is active by selecting Start → Settings → Control Panel.
Double-click the Administrative Tools icon. Double-Click the Services icon. You should
see similar to Figure 5-23 on page 211. Ensure that Status is Started.
Verify that the CIMOM is active by selecting Start → Settings → Control Panel →
Administrative Tools → Services. Launch Services panel and select IBM CIM Object
Manager service. Verify the Status is shown as Started, as shown in Figure 5-29 on
page 217.


Figure 5-29 Verify ESS CIMOM has started

Verify that SLP has dependency on CIMOM, this is automatically configured when you
installed the CIM agent software. Verify this by selecting Start → Settings → Control
Panel. Double-click the Administrative Tools icon. Double-Click the Services icon and
subsequently select properties on Service Location Protocol as shown in Figure
Figure 5-30.

Figure 5-30 SLP properties panel

Click Properties and select the Dependencies tab as shown in Figure 5-31 on page 218.
You must ensure that IBM CIM Object Manager has a dependency on Service Location
Protocol (this should be the case by default).


Figure 5-31 SLP dependency on CIMOM

Verify CIMOM registration with SLP by selecting Start →Programs →CIM Agent for the
IBM TotalStorage DS Open API → Check CIMOM Registration. A window opens
displaying the WBEM services as shown in Figure 5-32. These services have either
registered themselves with SLP or you explicitly registered them with SLP using slptool.
If you changed the default ports for a CIMOM during installation, the port number should
be correctly listed here. It may take some time for a CIM Agent to register with SLP.

Figure 5-32 Verify CIM Agent registration with SLP

Note: If the verification of the CIMOM registration is not successful, stop and restart the
SLP and CIMOM services. Note that the ESS CIMOM will attempt to contact each ESS
registered to it. Therefore, the startup may take some time, especially if it is not able to
connect and authenticate to any of the registered ESSs.

Use the verifyconfig -u superuser -p passw0rd command, where superuser is the user
name and passw0rd is the password for the user name that you configured to manage the
CIMOM, to locate all WBEM services in the local network. You need to define the
TotalStorage Productivity Center for Disk superuser name and passw0rd in order for
TotalStorage Productivity Center for Disk to have the authority to manage the CIMOM. The
verifyconfig command checks the registration for the ESS CIM Agent and checks that it can


connect to the ESSs. At ITSO Lab we had configured two ESSs (as shown in Figure 5-33 on
page 219).

Figure 5-33 The verifyconfig command

5.7.1 Problem determination
You might run into the some errors. If that is the case, you may verify with the cimom.log file.

This file is located in C:Program FilesIBMcimagent directory. You may verify that you
have the entries with your current install timestamp as shown in Figure 5-34. The entries of
specific interest are:
CMMOM050OI Registered service service:wbem:https://x.x.x.x:5989 with SLP SA
CMMOM0409I Server waiting for connections

This first entry indicates that the CIMOM has successfully registered with SLP using the port
number specified at ESS CIM agent install time, and the second entry indicates that it has
started successfully and waiting for connections.

Figure 5-34 CIMOM Log Success


If you still have problems, Refer to the DS Open Application Programming Interface
Reference for an explanation and resolution of the error messages. You can find this
Guide in the doc directory at the root of the CIM Agent CD.

5.7.2 Confirming the ESS CIMOM is available
Before you proceed, you need to be sure that the DS CIMOM is listening for incoming
connections. To do this run a telnet command from the server where TotalStorage
Productivity Center for Disk resides. A successful telnet on the configured port (as indicated
by a black screen with cursor on the top left) will tell you that the DS CIMOM is active. You
selected this port during DS CIMOM code installation. If the telnet connection fails, you will
have a panel like the one shown in Figure 5-35. In such case, you have to investigate the
problem until you get a blank screen for telnet port.

Figure 5-35 Example of telnet fail connection

Another method to verify that DS CIMOM is up and running is to use the CIM Browser
interface. For Windows machines change the working directory to c:Program
Filesibmcimagent and run startcimbrowser. The WBEM browser in Figure 5-36 will
appear. The default user name is superuser and the default password is passw0rd. If you
have already changed it, using the setuser command, the new userid and password must be
provided. This should be set to the TotalStorage Productivity Center for Disk userid and
password.

Figure 5-36 WBEM Browser


When login is successful, you should see a panel like the one in Figure 5-37.

Figure 5-37 CIMOM Browser window

5.7.3 Setting up the Service Location Protocol Directory Agent
You can use the following procedure to set up the Service Location Protocol (SLP) Directory
Agent (DA) so that TotalStorage Productivity Center for Disk can discover devices that reside
in subnets other than the one in which TotalStorage Productivity Center for Disk resides.

Perform the following steps to set up the SLP DAs:
1. Identify the various subnets that contain devices that you want TotalStorage Productivity
Center for Disk to discover.
2. Each device is associated with a CIM Agent. There might be multiple CIM Agents for each
of the identified subnets. Pick one of the CIM Agents for each of the identified subnets. (It
is possible to pick more than one CIM Agent per subnet, but it is not necessary for
discovery purposes.)
3. Each of the identified CIM Agents contains an SLP service agent (SA), which runs as a
daemon process. Each of these SAs is configured using a configuration file named
slp.conf. Perform the following steps to edit the file:
– For example, if you have DS CIM agent installed in the default install directory path,
then go to the C:Program FilesIBMcimagentslp directory.
– Look for file named slp.conf.
– Make a backup copy of this file and name it slp.conf.bak.


– Open the slp.conf file and scroll down until you find (or search for) the line
;net.slp.isDA = true
Remove the semi-colon (;) at the beginning of the line. Ensure that this property is set
to true (= true) rather than false. Save the file.
– Copy this file (or replace it if the file already exists) to the main windows subdirectory
for Windows machines (for example c:winnt), or in the /etc directory for UNIX
machines.
4. It is recommended to reboot the SLP server at this stage. Otherwise, alternatively, you
may choose to restart the SLP and CIMOM services. You can do this from your Windows
desktop → Start Menu → Settings → Control Panel → Administrative tools → Services.
Launch the Services GUI → Locate the Service Location Protocol, right click and select
stop. It will pop-up another panel which will request to stop IBM CIM Object Manager
service. You may click Yes.
You may start the SLP daemon again after it has stopped successfully.

Alternatively, you may choose to re-start the CIMOM using command line as described in
“Restart the CIMOM” on page 215

Creating slp.reg file

Important: To avoid to register manually the CIMOM outside the subnet every time that the
Service Location Protocol (SLP) is restarted, create a file named slp.reg.

The default location for the registration is “C:winnt”.

slpd reads the slp.reg file on startup and re-reads it when ever the SIGHUP signal is
received.

slp.reg file example
Example 5-1 is a slp.reg file sample.

Example 5-1 slp.reg file
#############################################################################
#
# OpenSLP static registration file
#
# Format and contents conform to specification in IETF RFC 2614, see also
# http://www.openslp.org/doc/html/UsersGuide/SlpReg.html
#
#############################################################################

#----------------------------------------------------------------------------
# Register Service - SVC CIMOMS
#----------------------------------------------------------------------------

service:wbem:https://9.43.226.237:5989,en,65535
# use default scopes: scopes=test1,test2
description=SVC CIMOM Open Systems Lab, Cottle Road
authors=Aliprandi,Andrews,Cooper,Eggli,Lovelace,Zerbini
creation_date=04/02/20



description=SVC CIMOM Tucson L2 Lab

#service:wbem:https://9.42.164.175:5989,en,65535
#description=SVC CIMOM Raleigh SAN Central
#authors=Aliprandi,Andrews,Cooper,Eggli,Lovelace,Zerbini
#creation_date=04/02/20

#----------------------------------------------------------------------------
# Register Service - SANFS CIMOMS
#----------------------------------------------------------------------------

#Additional parameters for setting the appropriate namespace values
#CIM_InteropSchemaNamespace=root/cimv2
#Namespace=root/cimv2
#description=SANFS CIMOM Gaithersburg ATS Lab

#description=SANFS CIMOM Tucson L2 Lab

#----------------------------------------------------------------------------
# Register Service - FAStT CIMOM
#----------------------------------------------------------------------------

#CIM_InteropSchemaNamespace=root/lsissi
#ProtocolVersion=0
#Namespace=root/lsissi
#description=FAStT700 CIMOM ITSO Lab, Almaden

5.7.4 Configuring TotalStorage Productivity Center for SLP discovery
You can use this panel to enter a list of DA addresses. TotalStorage Productivity Center for
Disk sends unicast service requests to each of these statically configured DAs, and sends
multicast service requests on the local subnet on which TotalStorage Productivity Center for
Disk is installed. Configure an SLP DA by changing the configuration of the SLP service
agent (SA) that is included as part of an existing CIM Agent installation. This causes the
program that normally runs as an SLP SA to run as an SLP DA.


You have now converted the SLP SA of the CIM Agent to run as an SLP DA. The CIMOM is
not affected and will register itself with the DA instead of the SA. However, the DA will
automatically discover all other services registered with other SLP SAs in that subnet.

Attention: You will need to register the IP address of the server running the SLP DA
daemon with the IBM Director to facilitate MDM SLP discovery. You can do this using the
IBM Director console interface of TotalStorage Productivity Center for Disk. The procedure
to register the IP address is described in 6.2, “SLP DA definition” on page 248.

5.7.5 Registering the DS CIM Agent to SLP
You need to manually register the DS CIM agent to the SLP DA only when the following
conditions are both true:
There is no DS CIM Agent in the TotalStorage Productivity Center for Disk server subnet
(TotalStorage Productivity Center for Disk).
The SLP DA used by Multiple Device Manager is also not running an DS CIM Agent.

Tip: If either of the preceding conditions are false, you do not need to perform the following
steps.

To register the DS CIM Agent issue the following command on the SLP DA server:
C:>CD C:Program FilesIBMcimagentslp
slptool register service:wbem:https://ipaddress:port

Where ipaddress is the ESS CIM Agent ip address. For our ITSO setup, we used IP address
of our ESS CIMOM server as 9.1.38.48 and default port number 5989. Issue a verifyconfig
command as shown in Figure 5-33 on page 219 to confirm that SLP is aware of the
registration.

Attention: Whenever you update SLP configuration as shown above, you may have to
stop and start the slpd daemon. This will enable SLP to register and listen on newly
configured ports.

Also, whenever you re-start SLP daemon, ensure that IBM DS CIMOM agent has also
re-started. Otherwise you may issue startcimom.bat command, as shown in previous
steps. Another alternative is to reboot the CIMOM server. Please note the for DS CIMOM
startup takes longer time.

5.7.6 Verifying and managing CIMOM’s availability
You may now verify that TotalStorage Productivity Center for Disk can authenticate and
discover the CIMOM agent services which are registered to the SLP DA. See “Verifying and
managing CIMOMs availability” on page 256.


5.8 Installing CIM agent for IBM DS4000 family
The latest code for the IBM DS4000 family is available at the IBM support Web site. You need
to download the correct and supported level of CIMOM code for TotalStorage Productivity
Center for Disk Version 2.3.

You can navigate from the following IBM support Web site for TotalStorage Productivity
Center for Disk to acquire the correct CIMOM code:
http://www-1.ibm.com/servers/storage/support/software/tpcdisk/

You may to have traverse through multiple links to get to the download files. At the time of
writing this book, we go to the Web page shown in Figure 5-38.

Figure 5-38 IBM support matrix Web page


By scrolling down the same Web page, we go to the following link for DS 4000 CIMOM code
as shown in Figure 5-39. This link leads to the Engenio Provider site. The current supported
code level is 1.0.59, as indicated in the Web page.

Figure 5-39 Web download link for DS Family CIMOM code

From the Web site, select the operating system used for the server on which the IBM DS
family CIM Agent will be installed. You will download a setup.exe file. Save it to a directory on
the server on which you will be installing the DS 4000 CIM Agent (see Figure 5-40 on
page 227).


Figure 5-40 DS CIMOM Install

Launch the setup.exe file to begin the DS 4000 family CIM agent installation. The
InstallShield Wizard for LSI SMI-S Provider window opens (see Figure 5-41). Click Next to
continue.

Figure 5-41 LSI SMI-SProvider window


The LSI License Agreement window opens next. If you agree with the terms of the license
agreement, click Yes to accept the terms and continue the installation (see Figure 5-42).

Figure 5-42 LSI License Agreement

The LSI System Info window opens. The minimum requirements are listed along with the
install system disk free space and memory attributes as shown in Figure 5-43. If the install
system fails the minimum requirements evaluation, then a notification window will appear and
the installation will fail. Click Next to continue.

Figure 5-43 System Info window


The Choose Destination Location window appears. Click Browse to choose another location
or click Next to begin the installation of the FAStT CIM agent (see Figure 5-44).

Figure 5-44 Choose a destination

The InstallShield Wizard will now prepare and copy the files into the destination directory.
See Figure 5-45.

Figure 5-45 Install Preparation window


The README appears after the files have been installed. Read through it to become familiar
with the most current information (see Figure 5-46). Click Next when ready to continue.

Figure 5-46 README file

In the Enter IPs and/or Hostnames window, enter the IP addresses and hostnames of the
FAStT devices that this FAStT CIM agent will manage as shown in Figure 5-47.

Figure 5-47 FAStT device list


Use the Add New Entry button to add the IP addresses or hostnames of the FAStT devices
that this FAStT CIM agent will communicate with. Enter one IP address or hostname at a time
until all the FAStT devices have been entered and click Next (see Figure 5-48).

Figure 5-48 Enter hostname or IP address

Do not enter the IP address of a FAStT device in multiple FAStT CIM Agents within the same
subnet. This may cause unpredictable results on the TotalStorage Productivity Center for
Disk server and could cause a loss of communication with the FAStT devices.

If the list of hostnames or IP addresses has been previously written to a file, use the Add File
Contents button, which will open the Windows Explorer. Locate and select the file and then
click Open to import the file contents.

When all the FAStT device hostnames and IP addresses have been entered, click Next to
start the SMI-S Provider Service (see Figure 5-49).

Figure 5-49 Provider Service starting


When the Service has started, the installation of the FAStT CIM agent is complete (see
Figure 5-50).

Figure 5-50 Installation complete

Arrayhosts file
The installer will create a file called
– %installroot%SMI-SProviderwbemservicescimombinarrayhosts.txt

The arrayhosts file is shown in Figure 5-51. In this file the IP addresses of installed DS 4000
units can be reviewed, added, or edited.

Figure 5-51 Arrayhosts file

Verifying LSI Provider Service availability
You can verify from Windows Services Panel that the LSI Provider service has started as
shown in Figure 5-52 on page 233. If you change the contents of the arrayhost file for adding
and deleting DS 4000 devices, then you will need to restart the LSI Provider service using the
Windows Services Panel.


Figure 5-52 LSI Provider Service

Registering DS4000 CIM agent
The DS4000 CIM Agent needs to be registered with an SLP DA if the FAStT CIM Agent is in
a different subnet then that of IBM TotalStorage Productivity Center for Disk and Replication
Base environment. The registration is not currently provided automatically by the CIM Agent.
You register the DS 4000 CIM Agent with SLP DA from a command prompt using the slptool
command.

An example of the slptool command is shown below. You must change the IP address to
reflect the IP address of the workstation or server where you installed the DS 4000 family DS
4000 CIM Agent. The IP address of our FAStT CIM Agent is 9.1.38.79 and port 5988. You
need to execute this command on your SLP DA server. It our ITSO lab, we used SLP DA on
the ESS CIMOM server. You need to go to directory C:Program FilesIBMcimagentslp
and run:
slptool register service:wbem:http:9.1.38.79:5988

Important: You cannot have the FAStT management password set if you are using IBM

At this point you may run following command on the SLP DA server to verify that DS 4000
family FAStT CIM agent is registered with SLP DA.
slptool findsrvs wbem

The response from this command will show the available services which you may verify.

5.8.1 Verifying and Managing CIMOM availability
discover the CIMOM agent services which are registered by SLP DA. You can proceed to
your TotalStorage Productivity Center for Disk server. See “Verifying and managing CIMOMs
availability” on page 256.


5.9 Configuring CIMOM for SAN Volume Controller
The CIM Agent for SAN Volume Controller is part of the SAN Volume Controller Console and
provides the TotalStorage Productivity Center for Disk with access to SAN Volume Controller
clusters. You must customize the CIM Agents in your enterprise to accept the TotalStorage
Productivity Center for Disk user name and password. Figure 5-53 explains the
communication between the TotalStorage Productivity Center for Disk and SAN Volume
Controller Environment.

Figure 5-53 TotalStorage Productivity Center for Disk and SVC communication

For additional details on how to configure the SAN Volume Controller Console, refer to the
redbook IBM TotalStorage Introducing the SAN Volume Controller and SAN Integration
Server, SG24-6423.

To discover and manage the SAN Volume Controller, we need to ensure that our TotalStorage
Productivity Center for Disk superuser name and password (the account we specify in the
TotalStorage Productivity Center for Disk configuration panel, as shown in 5.9.1, “Adding the
SVC TotalStorage Productivity Center for Disk user account” on page 235) matches an
account defined on the SAN Volume Controller console. In our case we implemented
username TPCSUID and password ITSOSJ.

You may want to adapt a similar nomenclature and set up the username and password on
each SAN Volume Controller CIMOM to be monitored with TotalStorage Productivity Center
for Disk.


5.9.1 Adding the SVC TotalStorage Productivity Center for Disk user account
As stated previously, you should implement a unique userid to manage the SAN Volume
Controller devices in TotalStorage Productivity Center for Disk. This can be achieved at the
SAN Volume Controller console using the following steps:
1. Login to the SAN Volume Controller console with a superuser account
2. Click Users under My Work on the left side of the panel (see Figure 5-54).

Figure 5-54 SAN Volume Controller console


3. Select Add a user in the drop down under Users panel and click Go (see Figure 5-55).

Figure 5-55 SAN Volume Controller console Add a user


4. An introduction screen is opened, click Next (see Figure 5-56).

Figure 5-56 SAN Volume Controller Add a user wizard


5. Enter the User Name and Password and click Next (see Figure 5-57).

Figure 5-57 SAN Volume Controller Console Define users panel

6. Select your candidate cluster and move it to the right under Administrator Clusters (see
Figure 5-58). Click Next to continue.

Figure 5-58 SAN Volume Controller console Assign administrator roles


7. Click Next after you Assign service roles (see Figure 5-59).

Figure 5-59 SAN Volume Controller Console Assign user roles


8. Click Finish after you verify user roles (see Figure 5-60).

Figure 5-60 SAN Volume Controller Console Verify user roles

9. After you click Finish, the Viewing users panel opens (see Figure 5-61).

Figure 5-61 SAN Volume Controller Console Viewing Users


Confirming that the SAN Volume Controller CIMOM is available
Before you proceed, you need to be sure that the CIMOM on the SAN Volume Controller is
listening for incoming connections. To do this, issue a telnet command from the server
where TotalStorage Productivity Center for Disk resides. A successful telnet on port 5989 (as
indicated by a black screen with cursor on the top left) will tell you that the CIMOM SAN
Volume Controller console is active. If the telnet connection fails, you will have a panel like
the one in Figure 5-62.

Figure 5-62 Example of telnet fail connection

5.9.2 Registering the SAN Volume Controller host in SLP
The next step to detecting an SAN Volume Controller is to manually register the SAN Volume
Controller console to the SLP DA.

Tip: If your SAN Volume Controller console resides in the same subnet as the
TotalStorage Productivity Center server, SLP registration will be automatic so you do not
need to perform the following step.

To register the SAN Volume Controller Console perform the following command on the SLP
DA server:
slptool register service:wbem:https://ipaddress:5989
Where ipaddress is the SAN Volume Controller console ip address.

Run a verifyconfig command to confirm that SLP ia aware of the SAN Volume Controller
console registration.

5.10 Configuring CIMOM for TotalStorage Productivity Center
for Disk summary
The TotalStorage Productivity Center discovers both IBM storage devices that comply with
the Storage Management Initiative Specification (SMI-S) and SAN devices such as switches,
ports, and hosts. SMIS-compliant storage devices are discovered using the Service Location
Protocol (SLP).


The TotalStorage Productivity Center server software performs SLP discovery on the
network. The User Agent looks for all registered services with a service type of service:wbem.
The TotalStorage Productivity Center performs the following discovery tasks:
Locates individual storage devices
Retrieves vital characteristics for those storage devices
Populates The TotalStorage Productivity Center internal databases with the discovered
information

The TotalStorage Productivity Center can also access storage devices through the CIM
Agent software. Each CIM Agent can control one or more storage devices. After the CIMOM
services have been discovered through SLP, the TotalStorage Productivity Center contacts
each of the CIMOMs directly to retrieve the list of storage devices controlled by each CIMOM.
TotalStorage Productivity Center gathers the vital characteristics of each of these devices.

For the TotalStorage Productivity Center to successfully communicate with the CIMOMs, the
following conditions must be met:
A common user name and password must be configured for all the CIM Agent instances
that are associated with storage devices that are discoverable by TotalStorage
Productivity Center (use adduser as described in 5.6.5, “CIMOM user authentication” on
page 215). That same user name and password must also be configured for TotalStorage
Productivity Center using the Configure MDM task in the TotalStorage Productivity Center
interface. If a CIMOM is not configured with the matching user name and password, it will
be impossible to determine which devices the CIMOM supports. As a result, no devices
for that CIMOM will appear in the IBM Director Group Content pane.
The CIMOM service must be accessible through the IP network. The TCP/IP network
configuration on the host where TotalStorage Productivity Center is installed must include
in its list of domain names all the domains that contain storage devices that are
discoverable by the TotalStorage Productivity Center.

It is important to verify that CIMOM is up and running. To do that, use the following command
from TotalStorage Productivity Center server:
telnet CIMip port

Where: CIMip is the ip address where CIM Agent run and port is the port value used for the
communication (5989 for secure connection, 5988 for unsecure connection).

5.10.1 SLP registration and slptool
TotalStorage Productivity Center for Disk uses Service Location Protocol (SLP) discovery,
which requires that all of the CIMOMs that TotalStorage Productivity Center for Disk
discovers are registered using the Service Location Protocol (SLP).

SLP can only discover CIMOMs that are registered in its IP subnet. For CIMOMs outside of
the IP subnet, you need to use an SLP DA and register the CIMOM using slptool. Ensure
that the CIM_InteropSchemaNamespace and Namespace attributes are specified.

For example, type the following command:
slptool register service:wbem:https://myhost.com:port

Where: myhost.com is the name of the server hosting the CIMOM, and port is the port
number of the service, such as 5989.


5.10.2 Persistency of SLP registration
Although it is acceptable to register services manually into SLP, it is possible for SLP users to
to statically register legacy services (applications that were not compiled to use the SLP
library) using a configuration file that SLP reads at startup, called slp.reg.

All of the registrations are maintained by slpd and will remain registered as long as slpd is
alive. The Service Location Protocol (SLP) registration is lost if the server where SLP resides
is rebooted or when the Service Location Protocol (SLP) service is stopped.

A Service Location Protocol (SLP) manual registration is needed for all the CIMOMs outside
the subnet where SLP DA resides.

Important: To avoid to register manually the CIMOM outside the subnet every time that
the Service Location Protocol (SLP) is restarted, create a file named slp.reg.

The default location for the registration is for Windows machines “c:winnt”, or “/etc”
directory for UNIX machines.

slpd reads the slp.reg file on startup and re-reads it when ever the SIGHUP signal is
received

5.10.3 Configuring slp.reg file
Example 5-2 shows a typical slp.reg file:

Example 5-2 An slp.reg file
#############################################################################
#
# OpenSLP static registration file
#
# Format and contents conform to specification in IETF RFC 2614, see also
# http://www.openslp.org/doc/html/UsersGuide/SlpReg.html
#
#############################################################################

#----------------------------------------------------------------------------
# Register Service - SVC CIMOMS
#----------------------------------------------------------------------------

description=SVC CIMOM Open Systems Lab, Cottle Road

description=SVC CIMOM Tucson L2 Lab

#description=SVC CIMOM Raleigh SAN Central


#----------------------------------------------------------------------------
# Register Service - SANFS CIMOMS
#----------------------------------------------------------------------------

#description=SANFS CIMOM Gaithersburg ATS Lab

#description=SANFS CIMOM Tucson L2 Lab

#----------------------------------------------------------------------------
# Register Service - FAStT CIMOM
#----------------------------------------------------------------------------

#CIM_InteropSchemaNamespace=root/lsissi
#ProtocolVersion=0
#Namespace=root/lsissi
#description=FAStT700 CIMOM ITSO Lab, Almaden


Part 3

Part 3 Configuring the
IBM TotalStorage
Productivity Center
In this part of the book we provide information about customizing the components of the
IBM TotalStorage Productivity Center product suite, for the following components:

We also include a chapter on how to set up the individual (sub) agents on a managed host.


6

Chapter 6. Configuring IBM TotalStorage
This chapter provides information about the basic tasks that you need to complete after you
install IBM TotalStorage Productivity Center for Disk:
Define SLP DA servers to IBM TotalStorage Productivity Center for Disk
Discover CIM Agents
Configure CIM Agents to IBM TotalStorage Productivity Center for Disk
Discover Storage devices
Install the remote GUI


6.1 Productivity Center for Disk Discovery summary
Productivity Center for Disk discovers both IBM storage devices that comply with the SMI-S
and SAN devices such as switches, ports, and hosts. SMIS-compliant storage devices are
discovered using the SLP. The Productivity Center for Disk server software performs SLP
discovery in the network. The User Agent looks for all registered services with a service type
of service:wbem. Productivity Center for Disk performs the following discovery tasks:
Locates individual CIM Agents
Locates individual storage devices
Retrieves vital characteristics for those storage devices
Populates the internal Productivity Center for Disk databases with the discovered
information

Productivity Center for Disk can also access storage devices through the CIM Agent
software. Each CIM Agent can control one or more storage devices. After the CIMOM
services are discovered through SLP, Productivity Center for Disk contacts each of the
CIMOMs directly to retrieve the list of storage devices controlled by each CIMOM.
Productivity Center for Disk gathers the vital characteristics of each of these devices.

For Productivity Center for Disk to successfully communicate with the CIMOMs, you must
meet the following conditions:
A common user name (superuser) and password must be set during installation of the
IBM TotalStorage Productivity Center for Disk base. This user name and password can be
changed using the Configure MDM task in the Productivity Center for Disk interface. If a
CIMOM is not configured with the matching user name and password, then you must
configure each CIMOM with the correct userid and password using the panel shown in
Figure 6-16 on page 258. We recommend that the common user name and password be
used for each CIMOM.
The CIMOM service must be accessible through the IP network. The TCP/IP network
configuration on the host, where Productivity Center for Disk is installed, must include in
its list of domain names all the domains that contain storage devices that are discoverable
by Productivity Center for Disk.

It is important to verify that CIMOM is up and running. To do that, use the following command:
telnet CIMip port

Here, CIMip is the IP address where the CIM Agent runs, and port is the port value used for
the communication (5989 for a secure connection; 5988 for an unsecure connection).

6.2 SLP DA definition
Productivity Center for Disk can discover CIM Agents on the same subnet through SLP
without any additional configuration. SLP DA should be set up on each subnet as described in
5.7.3, “Setting up the Service Location Protocol Directory Agent” on page 221. The SLP DA
can then be defined to Productivity Center for Disk using the panel located at Options →
Discovery Preferences → MDM SLP Configuration as shown in Figure 6-1 on page 249.
Enter the IP address of the server with the SLP DA into the SLP directory agent host box and
click Add.


We are assuming that you have followed the steps outlined in Chapter 5, “CIMOM install and
configuration” on page 191. You should complete the following tasks in order to discover
devices defined to our Productivity Center common base host. Make sure that:
All CIM agents are running and are registered with the SLP server.
The SLP agent host is defined in the IBM Director options (Figure 6-1) if it resides in a
different subnet from that of the TotalStorage Productivity Center server (Options →
Discovery Preferences → MDM SLP Configuration tab).

Note: If the Productivity Center common base host server resides in the same subnet as
the CIMOM, then it is not a requirement that the SLP DA host IP address be specified in
the Discovery Preferences panel as shown in Figure 6-2. Refer to Chapter 2, “Key
concepts” on page 27 for details on SLP discovery.

Here we provide a step-by-step procedure:
1. Discovery will happen automatically based on preferences that are defined in the Options
→ Discovery Preferences → MDM SLP Configuration tab. The default values for Auto
discovery interval and Presence check interval is set to 0 (see Figure 6-1). These
values should be set to a more suitable value, for example, to 1 hour for Auto discovery
interval and 15 minutes for Presence check interval. The values you specify will have a
performance impact on the CIMOMs and Productivity Center common base server, so do
not set these values too low.

Figure 6-1 Setting discovery preferences

Chapter 6. Configuring IBM TotalStorage Productivity Center for Disk 249

Continue entering IP addresses for all SLP DA servers. Click OK when finished (see
Figure 6-2).

Figure 6-2 Discovery preference set

2. Turn off automatic inventory on discovery.

Important: Because of the time and CIMOM resources needed to perform inventory on
storage devices, it is undesirable and unnecessary to perform this operation each time
Productivity Center common base performs a device discovery.


Turn off automatic inventory by selecting Options → Server Preferences as shown in
Figure 6-3.

Figure 6-3 Selecting Server Preferences

Now uncheck the Collect On Discovery check box as shown in Figure 6-4, all other options
can remain unchanged. Select OK when done.

Figure 6-4 Server Preferences


3. You can click Discover all Systems in the top left corner of the IBM Director Console to
initiate an immediate discovery task (see Figure 6-5).

Figure 6-5 Discover All Systems icon


4. You can also use the IBM Director Scheduler to create a scheduled job for new device
discovery.
– Either click the scheduler icon in the IBM Director tool bar or use the menu, Tasks →
Scheduler (see Figure 6-6).

Figure 6-6 Tasks Scheduler option for Discovery

– In the Scheduler, click File → New Job (see Figure 6-7).

Figure 6-7 Task Scheduler Discovery job


– Establish parameters for the new job. Under the Date/Time tab, include date and time
to perform the job, and whether the job is to be repeated (see Figure 6-8).

Figure 6-8 Discover job parameters

– From the Task tab (see Figure 6-9), select Discover MDM storage devices/SAN
Elements, then click Select.

Figure 6-9 Discover job selection task


– Click File → Save as, or use the Save as icon.
– Provide a descriptive job name in the Save Job panel (see Figure 6-10) and click OK.

Figure 6-10 Discover task job name

Now run the discovery process by selecting Tasks → Discover Systems → All Systems
and Devices (Figure 6-11).

Figure 6-11 Perform discovery


Double-click the Manage CIMOM task to see the status of the discovery (Figure 6-12).

Figure 6-12 Configure CIMOMs

The CIMOMs will appear in the list as they are discovered.

6.2.1 Verifying and managing CIMOMs availability
discover the CIMOM agent services which are registered to the SLP DA.

Launch the IBM Director Console and select TotalStorage Productivity Center for Disk →
Manage CIMOMs in the tasks panel as shown in Figure 6-13. The panel shows the status of
connection to the respective CIMOM servers. Our ITSO DS CIMOM server connection status
is indicated in first line, with IP address 9.1.38.48, port 5996, and status as Success.

Figure 6-13 Manage CIMOM panel


It should not be necessary to change any information if you followed the recommendation to
use the same superuser id and password for all CIMOMs. Select the CIMOM to be configured
and click Properties to configure a CIMOM (Figure 6-14).

Figure 6-14 Select a CIMOM to configure

1. In order to verify and re-confirm the connection, you may select the respective connection
status and click Properties. Figure 6-16 on page 258 shows the properties panel. You
may verify the information and update if necessary. The namespace, username and
password are picked up automatically, hence they are not normally required to be entered
manually. This username is used by CIMOM to logon to TotalStorage Productivity Center
for Disk.
If you have difficulty getting a successful connection, then you may manually enter
namespace, username, and password. Update the properties panel and test the
connection to the CIMOM:
a. Enter the Namespace value. It is rootibm for the ESS, DS6000 and DS8000 It is
interop for the DS4000.
b. Select the protocol. It is typically https for ESS, DS6000 and DS8000. It is http for
DS4000.
c. Enter the User name and password. The default is the superuser password entered
earlier. If you entered a different user name and password with the setuser command
for the CIM agent, then enter that user name and password here.
d. Click Test Connection to verify correct configuration.
e. You should see the panel in Figure 6-15. Click Close on the panel.

Figure 6-15 Successful test of the connection to a CIMOM


f. Click OK on the panel shown in Figure 6-16 to save the properties.

Figure 6-16 CIMOM Properties

2. After the connection to the CIMOM is successful, then perform discovery again as shown
before in Figure 6-11 on page 255. This will discover the storage devices connected to
each CIMOM (Figure 6-17).

Figure 6-17 DS4000 CIMOM Properties Panel

3. Click the Test Connection button to see a panel similar to Figure 6-15 on page 257,
showing that the connection is successful.

Tip: If you move or delete CIMOMs in your environment, the old CIMOM entries are not
automatically updated, and entries with a Failure status will be seen as in Figure 6-13 on
page 256. These invalid entries can slow down discovery performance, as TotalStorage
Productivity Center tries to contact them each time it performs a discovery.

You cannot delete CIMOM entries directly from the Productivity Center common base
interface. Delete them using the DB2 control center tool as described in 16.6, “Manually
removing old CIMOM entries” on page 911.


6.3 Disk and Replication Manager remote GUI
It is possible to install a TotalStorage Productivity Center for Disk console on a server other
than the one on which the TotalStorage Productivity Center for Disk code is installed. This
allows you to manage TotalStorage Productivity Center for Disk from a secondary location.
Having a secondary TotalStorage Productivity Center for Disk console will offload workload
from the TotalStorage Productivity Center for Disk server.

Note: You are only installing the IBM Director and TotalStorage Productivity Center for
Disk console code. You do not need to install any other code for the remote console.

In our lab we installed the remote console on a dedicated Windows 2000 server with 2 GB
RAM. You must install all the consoles and clients on the same server. Here are the steps:
1. Install the IBM Director console.
2. Install the TotalStorage Productivity Center for Disk console.
3. Install the Performance Manager client if the Performance Manager component is
installed.

Installing the IBM Director console
Follow these steps:
1. Start the setup.exe of IBM Director.
2. The main IBM Director window (Figure 6-18) opens. Click INSTALL IBM DIRECTOR.

Figure 6-18 IBM Director installer


3. In the IBM Director Installation panel (Figure 6-19), select IBM Director Console
installation.

Figure 6-19 Installation options for IBM Director

4. After a moment, the InstallShield Wizard for IBM Director Console panel (Figure 6-20)
opens. Click Next.

Figure 6-20 Welcome panel


5. In the License Agreement panel (Figure 6-21), select I accept the terms in the license
agreement. Then click Next.

Figure 6-21 License Agreement

6. The next panel (Figure 6-22) contains information about enhancing IBM Director.

Figure 6-22 Enhance IBM Director


7. The Feature and installation directory selection panel (Figure 6-23) allows you to change
how a program feature is installed. Click Next.

Figure 6-23 Selecting the program features to install

8. In the Ready to Install the Program window (Figure 6-24), accept the default selection.
Then click Install to start the installation.

Figure 6-24 Ready to Install the Program panel


9. The installation takes a few minutes. When it is finished, you see the InstallShield Wizard
Completed window (Figure 6-25). Click Finish to complete the installation.

Figure 6-25 Installation finished

The remote console of IBM Director is now installed.


Installing the remote console for Productivity Center for Disk
To install the remote console for Productivity Center for Disk follow these steps:
1. Insert the installation media for Productivity Center for Disk and Replication Base.
2. Change to the W2K directory. Figure 6-26 shows the files in that directory.

Figure 6-26 Files in the W2K directory

3. Start the LaunchPad.bat batch file. Coincidently this file has the same name as the
TotalStorage Productivity Center Launchpad, although it has nothing to do with it.
4. Click Installation wizard to begin the installation (Figure 6-27 on page 265).


Figure 6-27 Multiple Device Manager LaunchPad

5. For a brief moment, you see a DOS box with the installer being unpacked. When this is
done, you see the Welcome window shown in Figure 6-28. Click Next.

Figure 6-28 Welcome window


6. The License Agreement window (Figure 6-29) is displayed. Select I accept the term in
the license agreement and click Next.



7. The Destination Directory window (Figure 6-30) opens. Accept the default path or enter
the target directory for the installation. Click Next.

Figure 6-30 Installation directory


8. In the Select Product Type window (Figure 6-31), select Productivity Center for Disk
and Replication Base Console for the product type. Click Next.



9. The Preview window (Figure 6-32) contains the installation information. Review it and click
Install to start the console install.

Figure 6-32 Summary

10.When you reach the Finish window, click Finish to exit the add-on installer (Figure 6-33).

Figure 6-33 Installation finished


11.You return to the IBM TotalStorage Productivity Center for Disk and Replication Base
installer window shown in Figure 6-27 on page 265. Click Exit to end the installation.

The IBM Director remote console is now installed. The add-ons for IBM TotalStorage
Productivity Center for Disk and Replication Base have been added.

If the TotalStorage Productivity Center Launchpad is installed, it detects that the IBM Director
remote console is available the next time the LaunchPad is started. Also the Launchpad can
now be used to start IBM Director.

6.3.1 Installing Remote Console for Performance Manager function
After installing IBM Director Console and TotalStorage Productivity Center for Disk base
console, you will need to install remote console for Performance Manager function. For this,
insert the CD-ROM which contains the code for TotalStorage Productivity Center for Disk and
click setup.exe. In our example, we used the downloaded code as shown in the screenshot
in Figure 6-34.

Figure 6-34 Screenshot of our lab download directory location


Next, you will see Welcome panel shown in Figure 6-35.Click Next.

Figure 6-35 Welcome panel from TotalStorage Productivity Center for Disk installer

The License Agreement panel shown in Figure 6-36 on page 272 appears. Select I accept
the terms in the license agreement and click Next to continue.


Figure 6-36 Accept the terms of license agreement.

Choose the default destination directory as shown in Figure 6-37 and click Next.

Figure 6-37 Choose default destination directory


In the next panel, choose to install Productivity Center for Disk Client and click Next as
shown in Figure 6-38.

Figure 6-38 Select Product Type


In the next panel, select both check boxes for products, if you would like to install the console
and command line client for the Performance Manager function (see Figure 6-39). Click Next.

Figure 6-39 TotalStorage Productivity Center for Disk features selection


The Productivity Center for Disk Installer - CoServer Parameters panel opens (see
Figure 6-40). Enter the TPC user ID and password and the IP that the remote console will use
to validate with the TPC server. This is the IP of the TPC server and IBM Director logon.

Figure 6-40 Productivity Center for Disk Installer - CoServer parameters

The Productivity Center for Disk Installer - Preview panel appears (see Figure 6-41 on
page 276). Review the information and click Install to start the process of installing the remote
console.


Figure 6-41 Productivity Center for Disk Installer - Preview

When the install is complete you will see the Productivity Center for Disk Installer - Finish
panel as shown in Figure 6-42. Click Finish to complete the install process.

Figure 6-42 TotalStorage Productivity Center for Disk finish panel


6.3.2 Launching Remote Console for TotalStorage Productivity Center
You can launch the remote console from the TotalStorage Productivity Center desktop icon
from the remote console server. You will see the window in Figure 6-43.

Figure 6-43 TotalStorage Productivity Center launch window

You may click Manage Disk Performance and Replication as highlighted in the figure. This
will launch IBM director remote console.

You may logon to director server and start using remote console functions except for
Replication Manager.

Note: At this point, you have installed the remote console for Performance Manager
function only and not for replication manager. You can install remote console for replication
manager if you wish.


7

Chapter 7. Configuring TotalStorage
Productivity Center for
Replication
This chapter provides information to help you customize the TotalStorage Productivity Center
for Replication component of the TotalStorage Productivity Center. In particular, we describe
how to set up a remote GUI and CLI.


7.1 Installing a remote GUI and CLI
A replication session can be managed remotely using the graphical user interface (GUI) and
command line interface (CLI). To install, follow the procedure below.
1. Copy the suite install and Replication manager code to the computer you wish to use.
2. In the suite install folder, double click on the setup.exe file to launch the installer wizard.
3. At the language panel (Figure 7-1), choose the language you wish to use during the
install.

Figure 7-1 Select a language

4. At the welcome screen (Figure 7-2), click Next.

Figure 7-2 Welcome screen


5. The software license agreement panel appears (Figure 7-3). Click the radio button next to
I accept the terms of the license agreement and click Next to continue.


6. In the TotalStorage Productivity Center install options panel (Figure 7-4), click the radio
button next to User interface installations of Data, Disk, Fabric, and Replication and
click Next.

Figure 7-4 TotalStorage Productivity Center install options

Chapter 7. Configuring TotalStorage Productivity Center for Replication 281

7. In the Remote GUI/Command LIne Client component window (Figure 7-5), check the box
by The Productivity Center for Replication - Command Line Client and click Next.

Figure 7-5 Select Remote GUI/Command Line Client

8. A window opens (Figure 7-6) to begin the replication command line client install.

Figure 7-6 Replication command client install


9. In the next window, enter the location of the Replication Manager install package
(Figure 7-7).

Figure 7-7 Install package location for replication

10.A window opens prompting you to interact with the Replication Manager install wizard
(Figure 7-8).

Figure 7-8 Launch Replication Manager installer


11.The window in Figure 7-9 appears until the install wizard is launched.

Figure 7-9 Launching installer

12.The Productivity Center for Replication Installer - Welcome wizard window (Figure 7-10)
opens. Click Next.

Figure 7-10 Replication remote CLI install wizard


13.Specify the directory path of the Replication Manager installation files in the window
shown in Figure 7-11. Click Next.

Figure 7-11 Replication remote CLI Installer - destination directory

14.In the CoServer Parameters window shown in Figure 7-12, enter the following information:
– Host Name: Host name or IP address of the Replication Manager server
– Host Port: Port number of the Replication Manager server (default value is 9443)
– User Name: User name of the CIM Agent managing the storage device(s)
– User Password: User password of the CIM Agent managing the storage device(s)


Figure 7-12 Replication remote CLI Installer - coserver parameters

15.Review the information in the Preview window shown in Figure 7-13 and click Install.

Figure 7-13 Replication remote CLI Installer - preview


16.After successfully installing the remote CLI, the window in Figure 7-14 appears. Click
Finish.

Figure 7-14 Replication remote CLI Installer - finish

17.After clicking Finish, the postinstall.txt file opens.You may read the file or close and view it
at a later time.
18.A window opens informing you of a successful installation (see Figure 7-15). Click Next to
finish.

Figure 7-15 Remote CLI installation successful


8

Productivity Center for Data
This chapter describes the necessary tasks to start using IBM TotalStorage Productivity
Center for Data in your environment. After you install Productivity Center for Data, there are a
few remaining steps to perform, but you can start to use it without performing these steps at
first. Most people use Productivity Center for Data to look at the environment and see how the
storage capacity is distributed. This chapter focuses on what is necessary to fulfill this task.

The following procedures are covered in this chapter:
Configuring a discovered IBM TotalStorage Enterprise Storage Server (ESS) Common
Information Model (CIM) Agent
Configuring a discovered Fibre Array Storage Technology (FAStT) CIM Agent
Adding a CIM Agent that is located in a remote network
Setting up the IBM TotalStorage Productivity Center for Data Web interface
Setting up a remote console

We also recommend that you perform the following actions, although we do not describe
them here:
Setting up the alerting dispositions: Simple Network Management Protocol (SNMP),
Tivoli Enterprise Console (TEC), and mail
Setting up retention of log files and other information


8.1 Configuring the CIM Agents
Configuration of the CIM Agents for IBM TotalStorage Productivity Center for Data is different
than the configuration you have to perform within Productivity Center for Disk. This section
explains how to set up the CIM Agents in two ways: if it was discovered by the Data Manager,
or if the CIM Agent is located in a different subnet and multicasts are not enabled.

Here is an overview of the procedure to work with CIM Agents:
1. Perform discovery of a new CIM Agent (using Service Location Protocols (SLP)).
2. Configure the discovered CIM Agent properties or definition of a new CIM Agent.
3. Discovery collects the device.
4. After the characteristics are available, set up the device for monitoring.
5. A probe on the device gathers information about the disks and logical unit numbers
(LUNs).

8.1.1 CIM and SLP interfaces within Data Manager
The CIM interface within Data Manager is used only to gather information about the disks, the
LUNs, and some asset information. The data is correlated with the data that the manager
receives from the agents.

Since there is no way to install the agent of Data Manager directly on a storage subsystem,
Data Manager obtains the information from storage subsystems by using the Storage
Management Initiative - Specification (SMI-S) standard. This standard uses another standard,
CIM.

Data Manager uses this interface to access a storage subsystem. A CIM Agent (also called
CIM Object Manager (OM)) that ideally runs within the subsystem, but can also run on a
separate host, announces its existence by using the SLP. You can learn more about this
protocol in 2.3, “Service Location Protocol (SLP) overview” on page 38.

Within Data Manager, an SLP User Agent (UA) is integrated, and that agents performs a
discovery of devices. This discovery is limited to the local subnet of the Data Manager, and is
expanded only if multicasts are enabled on the network routers. See “Multicast” on page 43
for details.

Unlike Productivity Center for Disk, the User Agent that is integrated within the Data Manager
cannot talk to an SLP Directory Agent (DA). This restriction requires you to manually
configure every storage subsystem that was not automatically discovered.

8.1.2 Configuring CIM Agents
The procedure to configure a CIM Agent is simple. If a CIM Agent was discovered, you simply
enter the security information. We use the term CIM Agent instead of CIMOM because this is
a more generic term.

Figure 8-1 on page 291 shows the panel where the CIM Agents are configured. In our
example, the first two entries show CIM Agents that were discovered but are not yet
configured. The last two entries show an ESS and a FAStT CIM Agent that have already been
configured.


If you want to configure a CIM Agent that cannot be discovered because of the restriction
explained in 8.1.1, “CIM and SLP interfaces within Data Manager” on page 290, then you also
need to enter the IP address and select the right protocol.

Figure 8-1 Selecting CIM Agent Logins

If you completed the worksheets (see Appendix A, “Worksheets” on page 991), have them
available for the next steps.

Configuring discovered CIM Agents
For discovered CIM Agents that are not configured, complete these steps:
1. In the CIM/OM Login Administration panel (Figure 8-1), highlight the discovered CIM
Agent. Click Edit.
2. The Edit CIM/OM Login Properties window (Figure 8-2 on page 292) opens. Proceed as
follows:
a. Verify the IP address, port, and protocol.

Note: Not all CIM Agents provide a secure communication via https. For example,
FAStT does not provide https, so you have to select http.

b. Enter the name and password for the user which was configured in the CIM Agent of
that device.

Note: At the time of this writing, a FAStT CIM Agent does not use a special user to
secure the access. Data Manager still requires an input in the user and password
field, so type anything you want.

c. If you selected https as the protocol to use, enter the complete path and file name of
the certificate file that is used to secure the communication between the CIM Agent
and the Data Manager.

Note: The Truststore file of the ESS CIM Agent is located in the C:Program
Filesibmcimagent directory on the CIM Agent Host.

Chapter 8. Configuring IBM TotalStorage Productivity Center for Data 291

d. Click Save to finish the configuration.

Figure 8-2 CIM Agent login properties

Configuring new CIM Agents
If you have to enter a CIM Agent manually, click New in the CIM/OM Login Administration
panel (Figure 8-1 on page 291). The New CIM/OM Login Properties window (Figure 8-3)
opens. You perform the same steps as described in “Configuring discovered CIM Agents” on
page 291. For a new CIM Agent, you must also specify the IP address and protocol to use.
The port is set depending on the protocol.


Figure 8-3 New CIM Agent login properties

Next steps
After you configure the CIM Agent properties, run discovery on the storage subsystems.
During this process, the Data Manager talks to the CIM Agent to gather information about the
devices.

When this is completed, you see an entry for the subsystem in the Storage Subsystem
Administration panel (Figure 8-5 on page 294).

8.1.3 Setting up a disk alias
Optionally, you can change the name of a disk subsystem to a more meaningful name:
1. In the Data Manager GUI, in the Navigation Tree, expand Administrative Services →
Configuration →Data Manager subtree as shown in Figure 8-4. Select Storage
Subsystem Administration.

Figure 8-4 Navigation Tree


2. The panel shown in Figure 8-5 on page 294 opens.
a. Highlight the subsystem.
b. Place a check mark in the Monitored column.

Note: Select the Monitored column if you want Data Manager to probe the
subsystem whenever a probe job is run against it. If you deselect the Monitored
check box for a storage subsystem, the following actions occur:
All the data gathered by the server for the storage subsystem is removed from
the enterprise repository.
You can no longer run Monitoring, Alerting, or Policy Management jobs against
the storage subsystem.

c. Click Set disk alias.

Figure 8-5 Storage Subsystem Administration

3. The Set Disk Alias window (Figure 8-6) opens.
a. Enter the Alias/Name.
b. Click OK to finish.

Figure 8-6 Set Disk Alias

4. You may need to refresh the GUI for the changes to become effective. Right-click an old
entry in the Navigation Tree, and select Refresh.

Next steps
Now that you have set up the CIM Agent properties and specified to monitor the subsystems,
run a probe against it to collect data about the disks and LUNs. After you do this, you can look
at the results in different reports.


8.2 Setting up the Web GUI
The Web GUI is basically the same as the remote GUI that you can install on any machine.
You simply use a Web browser to download a Java™ application that is then launched.

We show only the basic setup of the Web server, which may not be very secure. The objective
here is to gain access to the Data Manager from a machine that does not have the remote
GUI installed.

Attention: We had the Tivoli Agent Manager running on the same machine. The Agent
Manager comes with an application (the Agent Recovery Service) that uses port 80, so we
had to find an unused port on the same machine.

In addition, you must be careful if you use the Internet Information Server (IIS). IIS uses
several ports by default which may interfere with the installed WebSphere Application
Server. Therefore we recommend that you use the IBM HTTP Server.

8.2.1 Using IBM HTTP Server
This section explains how to set up the IBM HTTP Server to make the remote GUI available
via the Web. When you install WebSphere Application Server on a machine, the IBM HTTP
Server is installed on the same machine.

The IBM HTTP server does not come with a GUI for the administration. Instead you use
configuration files to modify any settings.

The HTTP server in installed in C:Program FilesWebSphereAppServerHTTPServer. This
directory contains the conf subdirectory, which contains the httpd.conf file, which is used to
configure the server.
1. In C:Program FilesWebSphereAppServerHTTPServerconf directory, open the
httpd.conf file.
2. Locate the line where the port is defined. See Example 8-1. Change the port number. In
our example, we used 2077.

Example 8-1 Abstracts of the httpd.conf file
ServerName GALLIUM
# This is the main server configuration file. See URL http://www.apache.org/
# for instructions.

# Do NOT simply read the instructions in here without understanding
# what they do, if you are unsure consult the online docs. You have been
# warned.

# Originally by Rob McCool

# Note: Where filenames are specified, you must use forward slashes
# instead of backslashes. e.g. "c:/apache" instead of "c:apache". If
# the drive letter is omitted, the drive where Apache.exe is located
# will be assumed

....

# Port: The port the standalone listens to.
#Port 80


Port 2077

3. Locate the line AfpaEnable. Comment out the three Afpa... lines as shown in Example 8-2.

Example 8-2 Afpa
#AfpaEnable
#AfpaCache on
#AfpaLogFile "C:Program FilesWebSphereAppServerHTTPServer/logs/afpalog" V-ECLF


4. Locate the line that starts with <Directory. Modify the line to point the directory to
C:Program FilesIBMTPCDatagui as shown in Example 8-3.

Example 8-3 Directory setting
# ---------------------------------------------------------------------------
# This section defines server settings which affect which types of services
# are allowed, and in what circumstances.

# Each directory to which Apache has access, can be configured with respect
# to which services and features are allowed and/or disabled in that
# directory (and its subdirectories).

# will be assumed

# First, we configure the "default" to be a very restrictive set of
# permissions.

# Note that from this point forward you must specifically allow
# particular features to be enabled - so if something's not working as
# you might expect, make sure that you have specifically enabled it
# below.

# This should be changed to whatever you set DocumentRoot to.

#<Directory "C:Program FilesWebSphereAppServerHTTPServer/htdocs/en_US">
<Directory "C:Program FilesIBMTPCDatagui">

5. Locate the line that starts with DocumentRoot. Modify the line to point the directory to
C:Program FilesIBMTPCDatagui as shown in Example 8-4.

Example 8-4 DocumentRoot
# --------------------------------------------------------------------------------
# In the following section, you define the name space that users see of your
# http server. This also defines server settings which affect how requests are
# serviced, and how results should be formatted.

# See the tutorials at http://www.apache.org/ for
# more information.

# will be assumed.

# DocumentRoot: The directory out of which you will serve your
# documents. By default, all requests are taken from this directory, but
# symbolic links and aliases may be used to point to other locations.

#DocumentRoot "C:Program FilesWebSphereAppServerHTTPServer/htdocs/en_US"
DocumentRoot "C:Program FilesIBMTPCDatagui"


6. Locate the line that starts with DirectoryIndex. Modify the line to use TPCD.html as the
index document as shown in Example 8-4 on page 297.

Example 8-5 Directory index
# DirectoryIndex: Name of the file or files to use as a pre-written HTML
# directory index. Separate multiple entries with spaces.

#DirectoryIndex index.html
DirectoryIndex tpcd.html

7. Save the file.
8. Start the HTTP server.
9. Open a command prompt.
a. Change to the directory C:Program FilesWebSphereAppServerHTTPServer.
b. Type apache, and press Enter.
10.This starts the HTTP server as a foreground application. Now when you use a Web
browser, simply enter:
http://servername:portumber
In our environment, we entered:
http://gallium:2077

You see a Web page, and a Java application is then loaded. (Java is installed if necessary.)

Note: Do not omit the http://. Since we do not use the default, you have to tell the
browser which protocol to use.


8.2.2 Using Internet Information Server
If you have IIS installed on the server running Data Manager, use these steps to enable the
access to the remote GUI via a Web site.

Attention: If you have WebSphere Application Server running on the same server, be
careful not to create port conflicts, especially since port 80 is in use by both applications.

1. Start the Internet Information Services administration GUI.
2. A window opens as shown in Figure 8-7. In the left panel, right-click the entry with your
host name and select New → Web Site to launch the Web Site Creation Wizard.

Figure 8-7 Internet Information Server administration GUI


3. The Web Site Creation Wizard opens, displaying the Welcome panel (see Figure 8-8).
Click Next.

Figure 8-8 Web Site Creation Wizard

4. The Web Site Description panel (Figure 8-9) opens. Enter a description in the panel and
click Next.

Figure 8-9 Web Site Description panel


5. The IP Address and Port Settings panel (Figure 8-10) opens. Enter an unused port
number and click Next.

Figure 8-10 IP Address and Port Setting panel

6. In the Web Site Home Directory panel (Figure 8-11), enter the home directory of the Web
server. This is the directory where the files for the remote Web GUI are stored. The default
is C:Program FilesIBMTPCDatagui.
Click Next.

Figure 8-11 Web Site Home Directory panel


7. The Web Site Access Permissions panel (Figure 8-12) opens. Accept the default access
permissions, and click Next.

Figure 8-12 Web Site Access Permissions panel

8. When you see the window indicating that you have successfully completed the Web Site
Creation Wizard (Figure 8-13), click Finish.

Figure 8-13 Setup finished

9. In the Internet Information Services window (Figure 8-7 on page 299), right-click the new
Web server entry, and select Properties.


10.The Data Manager Properties window (Figure 8-14) opens.
a. Select the Documents tab.

Figure 8-14 Adding a default document

b. Click Add.
c. In the window that opens, enter tpcd.html. Click OK.
d. Click OK to close the Properties window.
11.This starts the HTTP Server as a foreground application. Now, when you use a Web
browser, simply enter:
http://servername:portumber
In our installation, we entered:
http://gallium:2077

You see a Web page and a Java application is loaded. (Java is installed if necessary.)

Note: Do not omit the http://. Since we don’t use the default, you have to tell the browser
which protocol to use.

8.2.3 Configuring the URL in Fabric Manager
The user properties file of the Fabric Manager contains settings that control polling, SNMP
traps destination, and the fully qualified host name of Data Manager. As an administrator, you
can use srmcp manager service commands to display and set the values in the user
properties file.

The srmcp ConfigService set command sets the value of the specified property to a new
value in the user properties file (user.properties). This command can be run only on the
manager computer.


Issuing a command on Windows
Use these steps to enter a command using Windows.
1. Open a command prompt window.
2. Change the directory to installation directorymanagerbinw32-ix86. The default
installation directory is C:Program FilesIBMTPCFabricmanagerbinw32-ix86.
3. Enter the following command:
setenv
4. Enter the following command:
srmcp -u Administrator -p password ConfigService set SRMURL
http://data.itso.ibm.com:2077

The change is picked up immediately. There is no need to restart Fabric Manager.

8.3 Installing the Data Manager remote console
To install the remote console for Productivity Center for Data, use the procedure explained in
this section. You can also start the installation using the Suite Installer. However, when the
Data Manager installer is launched, you begin with the first step of the procedure that follows.
1. Select language. We selected English (Figure 8-16 on page 305).

Figure 8-15 Welcome panel

2. The next panel (see Figure 8-16 on page 305) is the Software License Agreement. Click I
accept the terms in the license agreement and click Next to continue.



3. The next panel allows you to select the components to be installed. For the remote
console installation, select the User interface installations of Data, Disk, Fabric, and
Replication (see Figure 8-17). Click Next to continue.

Figure 8-17 Product selection


4. The next panel allows you to select which Remote GUI will be installed. Select the
Productivity Center for Data (see Figure 8-18) and click Next to continue.

Figure 8-18 Remote GUI selection panel

5. The next panel is informational (see Figure 8-19) and verifies that the Productivity Center
for Data GUI will be installed.

Figure 8-19 Verification panel


6. The install package location panel is displayed. Specify the required information (see
Figure 8-20) and click Next to continue.

Figure 8-20 Install package location

7. Another information panel is displayed (see Figure 8-21) indicating that the product
installer will be launched. Click Next to continue.

Figure 8-21 The installer will be launched


8. .In the window that opens, like the one in Figure 8-22, select Install Productivity Center
for Data and click Next.

Figure 8-22 Installation action

9. The License Agreement panel (Figure 8-23) opens. Select I have read and AGREE to
abide by the license agreement above and click Next.



10.A License Agreement Confirmation window (Figure 8-24) opens. Click Yes to confirm.

Figure 8-24 License Agreement Confirmation

11.The next window that opens prompts you to specify what you want to install (see
Figure 8-25). In this example, we already had the agent installed on our machine.
Therefore all options are still available. Select The GUI for reporting and click Next.



12.In the Productivity Center for Data Parameters panel (Figure 8-26), enter the Data
Manager connection details and a Data Manager server name. Change the port if
necessary and click Next.

Figure 8-26 Data Manager connection details


13.In the Space Requirements panel (Figure 8-27), you can change the installation directory
or leave the default. Click Next.

Figure 8-27 Installation directory

14.If the directory does not exist, you see the message shown in Figure 8-28. Click OK to
continue or Cancel to change the directory.

Figure 8-28 Directory does not exist


15.You see the window shown in Figure 8-29 indicating that Productivity Center for Data has
verified your entries and is ready to start the installation. Click Next to start the installation.

Figure 8-29 Ready to start the installation


16.During the installation, you see a progress indicator. When the installation is finished, you
see the Install Progress panel (Figure 8-30). Click Done to exit the installer.

Figure 8-30 Installation completed

The IBM TotalStorage Productivity Center for Data remote console is now installed.

If the TotalStorage Productivity Center Launchpad is installed, it detects that Productivity
Center for Data remote console is available the next time the LaunchPad is started. The
LaunchPad can now be used to start Productivity Center for Data.

8.4 Configuring Data Manager for Databases
Complete the following steps before attempting to monitor your databases with Data
Manager.
1. Go to Administrative Services → Configuration → General → License Keys and
double-click IBM TPC for Data - Databases (Figure 8-31 on page 314).


Figure 8-31 TPC for Data - Databases License Keys

2. From the list of agents, select those you wish to monitor by checking the box under
Licensed (Figure 8-32). After checking the desired boxes, click the RDBMS Logins tab.

Figure 8-32 TPC for Data - Databases Licensing tab


3. To successfully scan a database, you must provide a login name and password for each
instance. Click Add New... (Figure 8-33).

Figure 8-33 RDBMS Logins

4. In the RDBMS Login Editor window, enter the required information:
– Database - the database type you wish to monitor
– Agent Host - the host you wish to monitor
– Instance - the name of the instance
– User - login ID for the instance
– Password - password for the instance
– Port - port where database is listening

Figure 8-34 RDBMS Login Editor


5. After the database is successfully registered, click OK (Figure 8-35).

Figure 8-35 RDBMS successfully registered

8.5 Alert Disposition
This section describes the available alerting options one can configure. This option defines
how the Alerts are generated when a corresponding event is discovered. This panel is shown
in Figure 8-36 by going to Administrative Services → Configuration → General → Alert
Disposition.

Figure 8-36 Alert Disposition panel


You can specify these parameters:
SNMP
– Community - The name of the SNMP community for sending traps
– Host - The system (event manager) which will receive the traps
– Port - The port on which traps will be sent (the standard port is 162)
TEC (Tivoli Enterprise Console)
– TEC Server - for sending traps to; the system (TEC) that will receive the traps
– TEC Port - to which traps will be sent (the standard port is 5529)
E-mail
– Mail Server - The mail server which will be used for sending the e-mail.
– Mail Port - The port used for sending the mail to the mail server.
– Default Domain - Default domain to be used for sending the e-mail.
– Return To - The return address for undeliverable e-mail.
– Reply To - The address to use when will replying to an Alert-triggered e-mail.
Alert Log Disposition
– Delete Alert Log Records - older than how long the Alert Log files will be kept.


9

Productivity Center for Fabric
This chapter explains the steps that you must follow, after you install IBM TotalStorage
Productivity Center for Fabric from the CD, to configure the environment.

Refer to 4.3.6, “IBM TotalStorage Productivity Center for Fabric” on page 157, which shows
the installation procedure for installing IBM TotalStorage Productivity Center for Fabric using
the Suite Installer.

IBM TotalStorage Productivity Center for Fabric is a rebranding of IBM Tivoli Storage Area
Network Manager. Since the configuration process has not changed, the information provided
is still applicable.

This IBM Redbook complements the IBM Redbook IBM Tivoli Storage Area Network
Manager: A Practical Introduction, SG24-6848. You may also want to refer to that redbook to
learn about design or deployment considerations, which are not covered in this redbook.


9.1 TotalStorage Productivity Center component interaction
This section discusses the interaction between IBM TotalStorage Productivity Center for
Fabric and the other IBM TotalStorage Productivity Center components. IBM TotalStorage
Productivity Center interaction includes external products and devices. IBM TotalStorage
Productivity Center for Fabric uses standard calls to devices to provide and gather
information to enable it to provide information about your environment.

9.1.1 IBM TotalStorage Productivity Center for Disk and Replication Base
When a supported storage area network (SAN) Manager is installed and configured, IBM
TotalStorage Productivity Center for Disk and IBM TotalStorage Productivity Center for
Replication leverages the SAN Manager to provide enhanced function. Along with basic
device configuration functions, such as logical unit number (LUN) creation, allocation,
assignment, and deletion for single and multiple devices, basic SAN management functions
such as LUN discovery, allocation, and zoning are provided in one step. In Version 2.1 of
TotalStorage Productivity Center, IBM TotalStorage Productivity Center for Fabric is the
supported SAN Manager.

The set of SAN Manager functions that are exploited are:
The ability to retrieve SAN topology information, including switches, hosts, ports, and
storage devices.
The ability to retrieve and to modify the zoning configuration on the SAN.
The ability to register for event notification — this ensures that IBM TotalStorage
Productivity Center for Disk is aware when the topology or zoning changes as new
devices are discovered by the SAN Manager, and when host LUN configurations change.

9.1.2 SNMP
IBM TotalStorage Productivity Center for Fabric acts as a Simple Network Management
Protocol (SNMP) manager to receive traps from managed devices in the event of status
changes or updates. These traps are used to manage all the devices that the Productivity
Center for Fabric is monitoring to provide the status window shown by NetView. These traps
should then be passed onto a product, such as Tivoli Event Console (TEC), for central
monitoring and management of multiple devices and products within your environment. When
using the IBM TotalStorage Productivity Center Suite Installer, the SNMP configuration is
performed for you. If you install IBM TotalStorage Productivity Center for Fabric manually,
then you need to configure SNMP.

The NetView code that is provided when you install IBM TotalStorage Productivity Center for
Fabric is to be used only for this product. If you configure this NetView as your SNMP
listening device for non-IBM TotalStorage Productivity Center for Fabric purposes, then you
need to purchase the relevant NetView license.

9.1.3 Tivoli Provisioning Manager
Tivoli Provisioning Manager uses IBM TotalStorage Productivity Center for Fabric when it
performs its data resource provisioning. Provisioning is the use of workflows to provide
resources (data or server) whenever workloads exceed specified thresholds and dictate that
a resource change is necessary to continue to satisfy service-level agreements or business
objectives. If the new resources are data resources which are part of the SAN fabric, then
IBM TotalStorage Productivity Center for Fabric is invoked to provide LUN allocation, path
definition, or zoning changes as necessary.


Refer to the IBM Redbook Exploring Storage Management Efficiencies and Provisioning:
Understanding IBM TotalStorage Productivity Center and IBM TotalStorage Productivity
Center with Advanced Provisioning, SG24-6373, which presents an overview of the product
components and functions. It explains the architecture and shows the use of storage
provisioning workflows.

9.2 Post-installation procedures
This section discusses the next steps that we performed after the initial product installation
from the CD, to take advantage of the function IBM TotalStorage Productivity Center for
Fabric provides. After you install the Fabric Manager, you need to decide on which machines
you will install the Agent and on which machines you will install the Remote Console. The
following sections show how to install these components.

9.2.1 Installing Productivity Center for Fabric – Agent
This section explains how to install the Productivity Center for Fabric – Agent. The installation
must be performed by someone who has a user ID with administrator rights (Windows) or root
authority (UNIX).

We used the Suite Installer to install the Agent. You can also install directly from the
appropriate subdirectory on the CD. Because the installation is Java based, it looks the same
on all platforms.
1. In the window that opens, select the language for installation. We chose English. Click
Next. You will see the Welcome screen (Figure 9-1). Click Next to continue.


Chapter 9. Configuring IBM TotalStorage Productivity Center for Fabric 321

2. The next screen (see Figure 9-2) is the Software License Agreement. Click I accept the
terms in the license agreement and click Next to continue.


3. In the Suite Installer panel (Figure 9-3), select the Agent installations of Data, Fabric,
and CIM Agent option. Then click Next.

Figure 9-3 Suite Installer panel for selecting Agent installations


4. In the next window, select one or more Agent components (see Figure 9-4). In this
example, we chose The Productivity Center for Fabric - Agent option. Click Next.

Figure 9-4 Agent type selection panel

5. In the next panel, confirm the components to install. See Figure 9-5. Click Next.

Figure 9-5 Productivity Center for Fabric - Agent confirmation


6. As shown in Figure 9-6, enter the install package location. In our case, we installed from
the I: drive. You will most likely use the product CD. Click Next.

Figure 9-6 Input location selection panel

7. A panel (Figure 9-7) opens indicating that the Productivity Center for Fabric installer will be
launched. At this point, the Suite Installer is invoking the Installer process for the individual
agent install. If you install the Agent directly from the CD, without using the Suite Installer,
you commence the process after this point. The Suite Installer masks a few displays from
you when it calls the product installer.
Click Next.

Figure 9-7 Product Installer will be launched


8. In the window that opens, select the language for installation. We chose English. Click
Next.


9. As In Figure 9-9 on page 326, specify the Fabric Manager Name and Fabric Manager Port
Number. Type the Fabric Manager Name with the machine name where the Productivity
Center for Fabric - Manager is installed. If it is a different domain, you must fully qualify the
server name. In our case, colorado is the machine name of the server where Productivity
Center for Fabric - Manager is installed.
The port number is automatically inserted, but you can change it if you used a different
port when you installed the Productivity Center for Fabric - Manager.
Click Next.


Figure 9-9 Fabric Manager name and port option

10.The Host Authentication password is entered in the panel shown in Figure 9-10. This was
specified during the Agent Manager install and is used for agent installs.

Figure 9-10 Host Authentication password


11.In the next panel (Figure 9-11), you have the option to change the default installation
directory. We clicked Next to accept the default.

Figure 9-11 Selecting the installation directory

12.The Agent Information panel (Figure 9-12) asks you to specify a label which is applied to
the Agent on this machine. We used the name of the machine. The port number is the port
through which this Agent communicates. Click Next.

Figure 9-12 Agent label and port


13.In the panel in Figure 9-13, specify the account that the Fabric Agent is to run under. We
used the Administrator account.

Figure 9-13 Fabric agent account

14.In the Agent Management Information panel (Figure 9-14 on page 329), enter the location
of the Tivoli Agent Manager.
In our configuration, colorado is the machine name in our Domain Name Server (DNS)
where Tivoli Agent Manager is installed. The Registration Password is the password that
you used when you installed Tivoli Agent Manager.
Click Next.


Figure 9-14 Agent Manager information

15.Finally you see a confirmation panel (Figure 9-15) that shows the installation summary.
Review the information and click Next.

Figure 9-15 Installation summary panel


16.You see the installation status bar. Then you see a panel indicating a successful
installation (Figure 9-16). Click Finish.

Figure 9-16 Successful installation panel

17.The panel in Figure 9-17 indicates the successful install of the Fabric agent.

Figure 9-17 Successful install of fabric agent panel.

18.You return to the Suite Installer window where you have the option to install other Agents.
Click Cancel to finish.


Upon successful installation, you notice that nothing is added to the Start menu. The only
evidence that the Agent is installed and running is that a Service is automatically started.
Figure 9-18 shows the started Services in our Windows environment.

Figure 9-18 Common Agent Service indicator

If you look in the Control Panel, under Add/Remove Programs, there is now an entry for IBM
TotalStorage Productivity Center for Fabric - Agent. To remove the Agent, you click this entry.

9.2.2 Installing Productivity Center for Fabric – Remote Console
This section explains how to install the Productivity Center for Fabric – Remote Console.

Pre-installation tasks
Before you begin the installation, make sure that you have met the requirements that are
discussed in the following sections.

SNMP service installed
Make sure that you have installed the SNMP service and have an SNMP community name of
Public defined. For more information, see 3.10, “Installing SNMP” on page 73.

Existing Tivoli NetView installation
If you have an existing Tivoli NetView 7.1.4 installation, you can use it with Productivity
Center for Fabric installation. If you have any other version installed, you must uninstall it
before you install Productivity Center for Fabric – Remote Console.


Installing the console
The Productivity Center for Fabric Console remotely displays information about the monitored
SAN. A user who has Administration rights must perform the installation. At this time of writing
this redbook, this installation was supported on the Windows 2000 and Windows XP
platforms. The following steps show a successful installation. We used the Suite Installer to
install the Console, and the following windows reflect that process.
1. Select the language. We selected English. The next panel (see Figure 9-19) is the
installer Welcome panel.

Figure 9-19 Installer Welcome panel.


2. The next screen (see Figure 9-20) is the Software License Agreement. Click I accept the
terms in the license agreement and click Next to continue.

Figure 9-20 License agreement panel

3. The first Suite Installer window (Figure 9-21) opens. Select the User interface
Installations of Data, Fabric, and Replication option and then click Next.

Figure 9-21 Suite Installer for selecting Console


4. In the next panel (Figure 9-22), select one or more remote GUI or command line
components. To install the console, select The Productivity Center for Fabric - Remote
GUI Client. Click Next.

Figure 9-22 Selecting the Remote Console

5. In the installation confirmation panel (Figure 9-23), click Next.

Figure 9-23 Remote GUI Client installation confirmation


6. As shown in Figure 9-24, enter the location of the source code for the installation. In most
cases, this the product CD drive. In our case, we installed the code from the E: drive. Click
Next.

Figure 9-24 Source code location panel

7. The next panel (Figure 9-25) indicates that the Fabric Installer will be launched. If you
install the Agent directly from the CD, without using the Suite Installer, you begin the
process after this point. The Suite Installer masks a few displays from you when it calls the
product installer. Click Next.

Figure 9-25 Installer will be launched


8. Select the language. We selected English. The Suite Installer launches the Fabric installer
(see Figure 9-26).

Figure 9-26 Productivity Center for Fabric installer launched

9. The InstallShield Wizard opens for IBM TotalStorage Productivity Center for Fabric -
Console (see Figure 9-27, “InstallShield Wizard for Console” on page 336). Click Next.

Figure 9-27 InstallShield Wizard for Console


10.In the next panel, you can specify the location of the directory into which the product will
be installed. Figure 9-28 shows the default location. Click Next.

Figure 9-28 Default installation directory

11.Specify the name and port number of the host where the Productivity Center for Fabric
Manager is installed. See Figure 9-29. Click Next.

Figure 9-29 Productivity Center for Fabric Manager details


12.In the next panel (Figure 9-30), specify a starting port number from which the installer will
allocate a series of ports for communication. We used the default. Click Next.

Figure 9-30 Starting port number

13.Type the password that you will use for all remote consoles or that the managed hosts will
use for authentication with the manager (see Figure 9-31). This password must be the
same as the one you entered in the Fabric Manager Installation. Click Next.

Figure 9-31 Host Authentication panel


14.Specify the drive where NetView is to be installed or accept the default (see Figure 9-32).
Click Next.

Figure 9-32 Selecting the NetView installation drive

15.As shown in Figure 9-33, specify a password which will be used to run the NetView
Service. Then click Next.

Figure 9-33 NetView Service password


16.A panel opens that displays a summary of the installation (Figure 9-34). Click Next to
begin the installation.

Figure 9-34 Summary panel

17.The installation completes successfully as indicated by the message in the panel shown in
Figure 9-35. Click Next.

Figure 9-35 Installation successful message


18.You are prompted to restart your machine (Figure 9-36). You may elect to restart
immediately or at another time. We chose Yes, restart my computer. Click Finish.

Figure 9-36 restart Computer request

After rebooting your system, you see a new Service is automatically started, as shown in
Figure 9-37.

Figure 9-37 NetView Service

To start the Remote Console, click Start →Programs →Tivoli NetView →NetView
Console.


9.3 Configuring IBM TotalStorage Productivity Center for
Fabric
This section explains how to configure IBM TotalStorage Productivity Center for Fabric.

9.3.1 Configuring SNMP
When using the IBM TotalStorage Productivity Center Suite Installer, the SNMP configuration
is performed for you. If you install IBM TotalStorage Productivity Center for Fabric manually,
then you need to configure the Productivity Center for Fabric.

There are several ways to configure Productivity Center for Fabric for SNMP traps.

Method 1: Forward traps to the local Tivoli NetView console
In this scenario, you set up the devices to send SNMP traps to the NetView console, which is
installed on the Productivity Center for Fabric Server. Figure 9-38 shows an example of this
setup.

Managed Host
(Agent)
Disk array
Managed Host
(Agent)

Disk array
Managed Host
(Agent)

Disk array
SAN
Switch
SNMP

Disk array
Disk array

Productivity
Centre for
Fabric Manager

Figure 9-38 SNMP traps to local NetView console


NetView listens for SNMP traps on port 162, and the default community is public. When the
trap arrives to the Tivoli NetView console, it is logged in the NetView Event browser and then
forwarded to Productivity Center for Fabric as shown in Figure 9-39. Tivoli NetView is
configured during installation of the Productivity Center for Fabric Server for trap forwarding to
the Productivity Center for Fabric Server.

Productivity Center for
Fabric Server
SNMP Trap
TCP Tivoli NetView SAN Manager
162

fibre channel switch
trapfrwd .conf
(trap forwarding
to TCP /IP port
9556 )

Figure 9-39 SNMP trap reception

NetView forwards SNMP traps to the defined TCP/IP port, which is the sixth port derived from
the base port defined during installation. We used the base port 9550, so the trap forwarding
port is 9556.

With this setup, the SNMP trap information appears in the NetView Event browser.
Productivity Center for Fabric uses this information for changing the topology map.

Note: If the traps are not forwarded to Productivity Center for Fabric, the topology map is
updated based on the information coming from Agents at regular polling intervals. The
default Productivity Center for Fabric Server installation (including the NetView installation)
sets up the trap forwarding correctly.

Existing NetView installation
If you installed Productivity Center for Fabric with an existing NetView, you need to set up
trap forwarding:
1. Configure the Tivoli NetView trapfrwd daemon. Edit the trapfrwd.conf file in the
usrovconf directory. This file has two sections: Hosts and Traps.
a. Modify the Hosts section to specify the host name and port to forward traps to (in our
case, port 9556 on host COLORADO.ALMADEN.IBM.COM).
b. Modify the Traps section to specify which traps Tivoli NetView should forward. The
traps to forward for Productivity Center for Fabric are:
1.3.6.1.2 *(Includes MIB-2 traps (and McDATA’s FC Management MIB traps)
1.3.6.1.3 *(Includes FE MIB and FC Management MIB traps)
1.3.6.1.4 *(Includes proprietary MIB traps (and QLogic’s FC Management MIB traps))


Example 9-1 shows a sample trapfrwd.conf file.

Example 9-1 trapfrwd.conf file
[Hosts]
#host1.tivoli.com 0
#localhost 1662
colorado.almaden.ibm.com 9556
[End Hosts]
[Traps]
#1.3.6.1.4.1.2.6.3 *
#mgmt
1.3.6.1.2 *
#experimental
1.3.6.1.3 *
#Andiamo
1.3.6.1.4.1.9524 *
#Brocade
1.3.6.1.4.1.1588 *
#Cisco
1.3.6.1.4.1.9 *
#Gadzoox
1.3.6.1.4.1.1754 *
#Inrange
1.3.6.1.4.1.5808 *
#McData
1.3.6.1.4.1.289 *
#Nishan
1.3.6.1.4.1.4369 *
#QLogic
1.3.6.1.4.1.1663 *
[End Traps]

2. The trapfrwd daemon must be running before traps are forwarded. Tivoli NetView does
not start this daemon by default. To configure Tivoli NetView to start the trapfrwd
daemon, enter these commands at a command prompt:
ovaddobj usrovlrftrapfrwd.lrf
ovstart trapfrwd
3. To verify that trapfrwd is running, in NetView, select Options →Server Setup. In the
Server Setup – Tivoli NetView window (Figure 9-40 on page 345), you see that trapfrwd
is running.


Figure 9-40 Trapfwd daemon

After trap forwarding is enabled, configure the SAN components, such as switches, to send
their SNMP traps to the NetView console.

Note: This type of setup gives you the best results, especially for devices where you
cannot change the number of SNMP recipients and the destination ports.

Method 2: Forward traps directly to Productivity Center for Fabric
In this example, you configure the SAN devices to send SNMP traps directly to the
Productivity Center for Fabric Server. The receiving port number is the primary port number
plus six ports. In this case, traps are only used to reflect the topology changes and they are
not shown in the NetView Event browser.

Note: Some of the devices do not allow you to change the SNMP port. They only send
traps to port 162. In such cases, this scenario is not useful.


Method 3: Traps to the Productivity Center for Fabric and SNMP console
In this example, you set up the SAN devices to send SNMP traps to both the Productivity
Center for Fabric Server and to a separate SNMP console, which you installed in your
organization. See Figure 9-41.
p

Managed Host
(Agent)
Disk array
Managed Host
(Agent)

Disk array
Managed Host
(Agent)

Disk array
SAN
Switch

Disk array
Disk array

TotalStorage
Productivity Center
SNMP For Fabric Server
Console
port 162

Figure 9-41 SNMP traps for two destinations

The receiving port number for the Productivity Center for Fabric Server is the primary port
number plus six ports. The receiving port number for the SNMP console is 162.

In this case traps are used to reflect the topology changes and they will display in the SNMP
console events. The SNMP console, in this case, can be another Tivoli NetView installation or
any other SNMP management application.

For such a setup, the devices have to support setting multiple traps receivers and changing
the trap destination port. Since this functionality is not supported in all devices, we do not
recommend this scenario.

9.3.2 Configuring the outband agents
Productivity Center for Fabric Server uses agents to discover the storage environment and to
monitor the status. These agents are setup in the Agent Configuration panel.
1. From the NetView console, select SAN →Configuration →Configure Manager.


2. The SAN Configuration window (Figure 9-42) opens.
a. Select the Switches and Other SNMP Agents tab on the left side.

Figure 9-42 Selecting switches and other SNMP agents

b. You see the outband agents in the right panel. Define all the switches in the SAN that
you want to monitor. To define such an Agent, click Add.
c. The Enter IP Address window (Figure 9-43) opens. Enter the host name or IP address
of the switch and click OK.

Figure 9-43 Outband agent definition

d. The agent appears in the agent list as shown in Figure 9-42. The state of the agent
must be Contacted if you want Productivity Center for Fabric to get data from it.
e. To remove an already defined agent, select it and click Remove.

Defining a logon ID for zone information
Productivity Center for Fabric can retrieve the zone information from IBM Fibre Channel
Switches and from Brocade Silkworm Fibre Channel Switches. To accomplish this,
Productivity Center for Fabric uses application programming interface (API) calls to retrieve
zoning information. To use this API, Productivity Center for Fabric must login into the switch
with administrative rights. If you want to see zoning information, you need to specify the login
ID for the Agents you define.


Here is the procedure:
1. In the SAN Configuration window (Figure 9-42 on page 347), select the defined Agent and
click Advanced.
2. In the SNMP Agent Configuration window (Figure 9-44), enter the user name and
password for the switch login. Click OK to save this information.

Figure 9-44 Logon ID definition

You can now see zone information for your switches.

Tip: You must enter user ID and password information only for one switch in each SAN to
retrieve the zoning information. We recommend that you enter this information for at least
two switches for redundancy. Enabling more switches than necessary for API zone
discovery may slow performance.

9.3.3 Checking inband agents
After you install agents on the managed systems, as explained in 9.2.1, “Installing
Productivity Center for Fabric – Agent” on page 321, the Agents should appear in the Agent
Configuration window with an Agent state of Contacted (see Figure 9-42 on page 347). If the
Agent does not appear in the panel, check the Agent log file for the cause. You can only
remove Agents which are no longer responding to the server. Such Agents display a status of
Not responding, as shown in Figure 9-45.

Figure 9-45 Not responding inband Agent


9.3.4 Performing an initial poll and setting up the poll interval
After you set up the Agents and devices for use with the Productivity Center for Fabric Server,
you perform the initial poll. You can manually poll using the SAN Configuration panel
(Figure 9-46):
1. In NetView, select SAN →Configure.
2. In the SAN Configuration window, click Poll Now to perform a manual poll.

Note: Polling takes time, and depends on the size of the SAN.

3. If you did not configure trap forwarding for the SAN devices, (as described in 9.3.1,
“Configuring SNMP” on page 342), you must define the polling interval. In this case, the
topology change will not be event driven from the devices, but will be updated regularly at
the polling interval.
You set the poll interval in the SAN Configuration panel (Figure 9-46). You can specify the
polling interval in:
– Minutes
– Hours
– Days: You can specify the time of the day for polling.
– Weeks: You can specify the day of the week and time of the day for polling.
After you set the poll interval, click OK to save the changes.

Figure 9-46 SAN Configuration

Tip: You do not need to configure the polling interval if all your devices are set to send
SNMP traps to the local NetView console or the Productivity Center for Fabric Server.


10

Chapter 10. Deployment of agents
Chapter 9, “Configuring IBM TotalStorage Productivity Center for Fabric” on page 319, covers
the installation of the managers that are the central part of IBM TotalStorage Productivity
Center. During that installation, the Resource Managers of Productivity Center for Data and
Productivity Center for Fabric were installed and registered to a Tivoli Agent Manager, either
an existing one, or one that was installed as a prerequisite in the first phase of the installation.

This chapter explains how to set up the individual agents (subagents) on a managed host.
The agents of Data Manager and Fabric Manager are called subagents, because they reside
within the scope of the common agent.


10.1 Installing the agents
There are two ways to set up a new subagent on a host, depending on the state of the target
machine:
The common agent is not installed. In this case, install the software using an installer.
The common agent is installed. In this case, deploy the agent from the data or fabric
manager, or install it using the installer.

To install the agent, follow these steps:
1. In the Suite Installer panel (Figure 10-1), select Agent installations of Data, Fabric, and
CIM Agent. Click Next.

Figure 10-1 Suite installer installation action


2. In the next panel (Figure 10-2), select one or more agents to install. The options include
the IBM TotalStorage Enterprise Storage Server (ESS) Common Information Model (CIM)
Agent. However, this agent does not use any functions of Tivoli Agent Manager.

Figure 10-2 Agent selection panel

The next window asks you to enter the location of the installation code. Then the panel that
follows tells you that the individual product installer is launched and you are asked to interact
with it.

Chapter 10. Deployment of agents 353

10.2 Data Agent installation using the installer
After the product installer for IBM TotalStorage Productivity Center for Data starts, you can
choose to install, uninstall, or apply maintenance to this component. Since no component of
the Productivity Center for Data is installed on the server, only one option is available to
install.
1. In the Install window (Figure 10-3), select Install Productivity Center for Data and click
Next.

Figure 10-3 IBM TotalStorage Productivity Center for Data installation action


2. A window opens showing the license agreement (see Figure 10-4). Select the I have read
and AGREE to abide by the license agreement above check box and click Next.


3. The License Agreement Confirmation window (Figure 10-5) opens. It asks you to confirm
that you have read the license agreement. Click Yes.

Figure 10-5 License Agreement Confirmation


4. In the next panel (Figure 10-6), choose the option of Productivity Center for Data that you
want to install. To install the agent locally on the same machine on which the installer is
currently running, select An agent on this machine and click Next.



5. In the Productivity Center for Data Parameters panel (Figure 10-7), enter the server name
and port of your Productivity Center for Data Manager. In our environment, the name of
the server was gallium, and the port was the default, which is 2078. We did not need to
use the fully qualified host name, but this may be different in your environment. Click Next.

Figure 10-7 Data Manager server details

The installer tries to contact the Data Manager server. If this is successful, you see a
message like Server gallium:2078 connection successful - server parameters
verified in the Progress Log section of the installation window Figure 10-7.
6. The installer checks whether a common agent is already installed on the machine.
Because in our environment no common agent was installed on the machine, the installer
issues the message No compatible Common Agents were discovered so one will be
installed. See the Progress log in Figure 10-8 on page 358.


7. As shown in Figure 10-8, enter the parameters of the agent:
a. Use the suggested default port 9510.
b. Deselect Agent should perform a SCAN when first brought up because this may
take a long time and you want to schedule this during the night.
c. Leave Agent may run scripts sent by server as selected.
d. The Agent Registration Information is the password that you specified during the
installation of the Tivoli Agent Manager.

Note: Do not change the common agent port, because this may prevent the
deployment of agents later.

e. Click Next to continue the installation.

Figure 10-8 Parameter for the common agent and Data Agent


8. In the Space Requirements panel (Figure 10-9), accept the default directory for the
common agent installation. Click Next to proceed with the installation.

Figure 10-9 Common Agent installation directory

9. If the directory that you specify does not exist, you see the message shown in
Figure 10-10. Click OK to acknowledge this message and continue the installation.

Figure 10-10 Creating the directory


10.Figure 10-11 shows the last panel before the installation starts. Review the progress log. If
you want to review the parameters, click Prev to go to the previous panels. Then click
Next.

Figure 10-11 Review settings


11.The installation starts and displays the progress in the Install Progress window
(Figure 10-12). The progress bar is not shown in the picture, but you can see the
messages in the progress log. When the installation is complete, click Done to end the
installation.

Figure 10-12 Installation progress

12.The installer closes now, and the Suite Installer is active again. It also reports the
successful installation. Click Next to return to the panel shown in Figure 10-1 on page 352
to install another agent (for example a Fabric Agent) or click Cancel to exit the installation.

10.3 Deploying the agent
The deployment of an agent is a convenient way to install a subagent onto multiple machines
at the same time. You can also use this method if you do not want to install agents directly on
each machine where an agent should be installed.

The most important prerequisite software to install on the target machines is the common
agent. If the common agent is not already installed on the target machine, the deployment will
not work. For example, if you installed one of the two Productivity Center agents, on the
targets, you can deploy the other agent using the methods described here.

At the time of this writing, Suite Installer does not have the option to deploy agents, so you
have to use the native installer setup.exe program for Fabric Manager. The packaging of Data
Manager is different and you can use the Suite Installer to install it.


Note: For agent deployment, you do not need to have the certificate files available,
because the target machines already have the necessary certificates installed during the
common agent installation.

Data Agent
You can perform this installation from any machine. It does not have to be the Data Manager
server itself.

When you use the Suite Installer, there is no option to deploy agents. However, you can
choose to install an agent, to launch the Data Manager installer, and later deploy an agent
instead of installing it (see Figure 10-6 on page 356). We did not use the Suite Installer for the
agent deployment.
1. Start the installer by running setup.exe from the Data Manager installation CD.
2. After a few seconds, you see the panel shown in Figure 10-13. If you have the Data
Manager or the agent already installed on that machine where you started the installer,
select Uninstall Productivity Center for Data or Apply maintenance to Productivity
Center for Data. Click Next.

Figure 10-13 Productivity Center for Data Installation action

3. A window opens displaying the license agreement (see Figure 10-4 on page 355). Follow
the steps as explained in steps 2 on page 355 and 3 on page 355.


4. In the next window that opens (Figure 10-14), select Agents on other machines. Then
click Next.

Figure 10-14 Productivity Center for Data Install agents options


5. In the Productivity Center for Data Parameters panel (Figure 10-15), enter the Productivity
Center for Data server name and the port number. Then click Next.

Figure 10-15 Data Manager server details


6. The installer tries to verify your input by connecting to the Data Manager server. The
message Server gallium:2078 connection successful - server parameters verified
is displayed in the progress log (see Figure 10-16) if it is successful. Click Next.
7. In our environment, we did not have a Windows domain, so we entered the details of the
target machines manually. Click Manually Enter Agents.

Figure 10-16 Select the Remote Agents to install: Manually entering Agents


8. In the Manually Enter Agent window (Figure 10-17), enter the IP address or host name of
the target computer and a user ID and password of valid Windows users on that machine.
You can only enter more than one machine here, if all the machines can be managed with
the same user ID and password. Click OK after you enter all computers that can be
managed with the same user ID.

Figure 10-17 Manually Enter Agents panel


9. The list with the computers that installs the subagent is updated and now appears as
shown in Figure 10-18. If you want to install the subagent onto a second computer, but the
computer uses a different user ID than the previous one, click Manually Enter Agents
again to enter the information for that second computer. Repeat this step for every
computer that uses a different user ID and password.
After you enter all target computers, click Next.

Figure 10-18 Selecting the Remote Agents to install: Computers targeted for a remote agent install

10.At this time, the installer tries to contact the common agent on target computers to get
information about them. This may take a while, so at first you cannot select anything in the
window that is presented next (see Figure 10-19 on page 368).
Look at the progress log in the lower section of the window to determine what is currently
happening. If the installer cannot contact the target computer, verify that the common
agent is running. You can do that by looking at the status of the Windows services of the
target machine. Another way is to open a telnet connection from a Command Prompt to
that machine on port 9510.
c:>telnet 9.1.38.104 9510
If the common agent is running, it listens for requests on that port and opens a connection.
You simply see a blank screen. If the common agent is not running, you see the message
Connecting To 9.1.38.104...Could not open a connection to host on port 9510 :
Connect failed.
When the installer is done with this step, you see the message Productivity Center for
Data subagent an 9.1.38.104 will be installed at C:Program
FilestivoliepTPCData.
Deselect Agent should perform a SCAN when first brought up, because this may take
a long time and you want to schedule this during the night. Click Install to start the
deployment.


Figure 10-19 Common agent status

11.When the deployment is finished, you see the message shown in Figure 10-20. Review
the progress log. Click OK to end the installation.

Figure 10-20 Agent deployment installation completed


Fabric Agent
There are differences between the Data Agent deployment and the Fabric Agent deployment.
To remotely deploy one or many fabric manager subagents, you must be logged on to the
fabric manager server. This is different than the data subagent deployment where you can
start the installation from any machine.
At this time, there is no way to use the Suite Installer, so you have to use the native fabric
manager installer.
The Fabric Manager comes with a separate package for the Fabric Agent. Data Manager
comes with only one installation program for all the possible install options (server, agent,
remote agent or GUI).

To start the deployment, you start the Fabric Manager installer. You do not start the installer
for the Fabric Agent.
1. Launch setup.exe from the fabric manager installation media.
2. After a Java Virtual Machine is prepared and you select the language of the installer, a
window opens that prompts you to select the type of installation to perform. See
Figure 10-21. Select Remote Fabric Agent Deployment and click Next.

Figure 10-21 Installation action

3. A Welcome window opens. Click Next.


4. The IBM License Agreement Panel (Figure 10-22) opens. Select I accept the terms in
the license agreement and click Next.


5. The installer connects to the Tivoli Agent Manager and presents a list of hosts. Select the
hosts to deploy the agents. See Figure 10-23. Click Next to start the deployment.

Figure 10-23 Remote host selection


6. The next panel (Figure 10-24) displays the selected hosts. Verify the information. You can
click Back to change your selection or click Next to start the installation.

Figure 10-24 Remote host confirmation

7. When the installation is completed, you see a summary window similar to the example in
Figure 10-25. Click Finish.

Figure 10-25 Agent Deployment summary

Your agent should now be installed on the remote hosts.


Part 4

Part 4 Using the
IBM TotalStorage
Productivity Center
In this part of the book we provide information about using the components of the
IBM TotalStorage Productivity Center product suite. We include a chapter filled with hints and
tips about setting up the IBM TotalStorage Productivity Center environment and problem
determination basics, as well as a chapter on maintaining the DB2 database.


11

Chapter 11. Using TotalStorage Productivity
Center for Disk
This chapter provides information about the functions of the Productivity Center common
base. Components of the Productivity Center common base include these topics:
Launching and logging on to TotalStorage Productivity Center
Launching device managers
Performing device inventory collection
Working with the ESS, DS6000, and DS8000 families
Working with SAN Volume Controller
Working with the IBM DS4000 family (formerly FAStT)
Event management


11.1 Productivity Center common base: Introduction
Before using Productivity Center common base features, you need to perform some
configuration steps. This will permit you to detect storage devices to be managed. Version 2.3
of Productivity Center common base permits you to discover and manage:
ESS 2105-F20, 2105-800, 2105-750
DS6000 and DS8000 family
SAN Volume Controller (SVC)
DS4000 family (formally FAStT product range)

Provided that you have discovered a supported IBM storage device, Productivity Center
common base storage management functions will be available for drag-and-drop operations.
Alternatively, right-clicking the discovered device will display a drop-down menu with all
available functions specific to it.

We review the available operations that can be performed in the sections that follow.

Note: Not all functions of TotalStorage Productivity Center are applicable to all device
types. For example, you cannot display the virtual disks on a DS4000 because the virtual
disks concept is only applicable to the SAN Volume Controller. The sections that follow
cover the functions available for each of the supported device types.

11.2 Launching TotalStorage Productivity Center
Productivity Center common base along with TotalStorage Productivity Center for Disk and
TotalStorage Productivity Center for Replication are accessed via the TotalStorage
Productivity Center Launchpad (Figure 11-1) icon on your desktop. Select Manage Disk
Performance and Replication to start the IBM Director console interface.

Figure 11-1 TotalStorage Productivity Center launchpad


Alternatively access IBM Director from Windows Start → Programs → IBM Director →
IBM Director Console

Log on to IBM Director using the superuser id and password defined at installation. Please
note that passwords are case sensitive. Login values are:-
IBM Director Server: Hostname of the machine where IBM Director is installed
User ID: The username to logon with. This is the superuser ID. Enter it in the form
<hostname><username>
Password: The case sensitive superuser ID password

Figure 11-2 shows the IBM Director Login panel you will see after launching IBM Director.

Figure 11-2 IBM Director Log on

11.3 Exploiting Productivity Center common base
The Productivity Center common base module adds the Multiple Device Manager submenu
task on the right-hand Tasks pane of the IBM Director Console as shown in Figure 11-3 on
page 378.

Note: The Multiple Device Manager product has been rebranded to TotalStorage
Productivity Center for Disk and TotalStorage Productivity Center for Replication. You will
still see the name Multiple Device Manager in some panels and messages.

Productivity Center common base will install the following sub-components into the Multiple
Device Manager menu:
Launch Device Manager
Launch Tivoli SAN Manager (now called TotalStorage Productivity Center for Fabric)
Manage CIMOMs
Manage Storage Units (menu)
– Inventory Status
– Managed Disks
– Virtual Disks
– Volumes

Chapter 11. Using TotalStorage Productivity Center for Disk 377

Note: The Manage Performance and Manage Replication tasks that you see in
Figure 11-3 become visible when TotalStorage Productivity Center for Disk or TotalStorage
Productivity Center for Replication are installed. Although this chapter covers Productivity
Center common base, you would have installed either TotalStorage Productivity Center for
Disk, TotalStorage Productivity Center for Replication, or both.

Figure 11-3 IBM Director Console with Productivity Center common base

11.3.1 Launch Device Manager
The Launch Device Manager task may be dragged onto an available storage device.

For ESS, this will open the ESS Specialist window for a chosen device. For SAN Volume
Controller, it will launch a browser session to that device. For DS4000 or FAStT devices, the
function is not available.

11.4 Performing volume inventory
This function is used to collect the detailed volume information from a discovered device and
place it into the Productivity Center common base databases. You need to do this at least
once before Productivity Center common base can start to work with a device.


When the Productivity Center common base functions are subsequently used to
create/remove LUN’s the volume inventory is automatically kept up to date and it is therefore
not necessary to repeatedly run inventory collection from the storage devices.

Version 2.3 of Productivity Center common base does not currently contain the full feature set
of all functions for the supported storage devices. This will make it necessary to use the
storage devices own management tools for some tasks. For instance you can create new
VDisks with Productivity Center common base on a SAN Volume Controller but you can not
delete them. You will need to use the SAN Volume Controller’s own management tools to do
this. For these types of changes to be reflected in Productivity Center common base an
inventory collection will be necessary to re-synchronize the storage device and Productivity
Center common base inventory.

Attention: The use of volume inventory is common to ALL supported storage devices and
must be performed before disk management functions are available.

To start inventory collection, right-click the chosen device and select Perform Inventory
Collection as shown in Figure 11-4.

Figure 11-4 Launch Perform Inventory Collection

A new panel will appear (Figure 11-5 on page 380) as a progress indication that the inventory
process is running. At this stage Productivity Center common base is talking to the relevant
CIMOM to collect volume information from the storage device. After a short while the
information panel will indicated that the collection has been successful. You can now close
this window.


Figure 11-5 Inventory collection in progress

Attention: When the panel in Figure 11-5 indicates that the collection has been done
successfully, it does not necessarily mean that the volume information has been fully
processed by Productivity Center common base at this point. To track the detailed
processing status, launch the Inventory Status task as seen in Figure 11-6.

Figure 11-6 Launch Inventory Status


To see the processing status of an inventory collection, launch the Inventory Status task as

Figure 11-7 Inventory Status

The example Inventory Status panel seen in Figure 11-7 shows the progress of the
processing for a SAN Volume Controller. Use the Refresh button in the bottom left of the
panel to update it with the latest progress. You can also launch the Inventory Status panel
before starting an inventory collection to watch the process end to end.

In our test lab the inventory process time for an SVC took around 2 minutes, end to end.


11.5 Changing the display name of a storage device
You can change the display name of a discovered storage device to something more
meaningful to your organization. Right-click the chosen storage device (Figure 11-8) and
select the Rename option.

Figure 11-8 Changing the display name of a storage device

Enter a more meaningful device name as in Figure 11-9 and click OK.

Figure 11-9 Entering a user defined storage device name


11.6 Working with ESS
This section covers the Productivity Center common base functions that are available when
managing ESS devices.

There are two ways to access Productivity Center functions for a given device, and these can
be seen in Figure 11-10:
Tasks access: You will see in the right-hand task panel that there are a number of
available tasks under the Manage Storage Units section. These management functions
can be invoked by dragging them onto the chosen device. However, not all functions are
applicable to all supported devices.
Right-click access: To access all functions available for a specific device, simply
right-click it to see a drop-down menu of options for that device. Figure 11-10 shows the
drop-down menu for an ESS.

Figure 11-10 also shows the functions of TotalStorage Productivity Center for Disk and
TotalStorage Productivity Center for Replication. Although this chapter only covers the
Productivity Center common base functions, you would always have installed either
TotalStorage Productivity Center for Disk, TotalStorage Productivity Center for Replication, or
both.

Figure 11-10 Accessing Productivity Center common base functions


11.6.1 ESS Volume inventory
To view the status of the volumes available within a given ESS device, perform one of the
following actions:
Right-click the ESS device and select Volumes as in Figure 11-11.
On the right-hand side under the Tasks column, drag Managed Storage Units →
Volumes onto the storage device you want to query.

Tip: Before volumes can be displayed, as with other storage devices managed by
Productivity Center common base, an initial inventory must be completed.

If you try to view volumes for an ESS that has not been inventoried, you will receive a
notification that this needs to be done. To perform an inventory collection, see 11.4,
“Performing volume inventory” on page 378.

Figure 11-11 Working with ESS volumes


In either case, in the bottom left corner, the status will change from Ready to Starting Task,
and it will remain this way until the volume inventory appears. Figure 11-12 shows the
Volumes panel for the select ESS device that will appear.

Figure 11-12 ESS volume inventory panel

11.6.2 Assigning and unassigning ESS Volumes
From the ESS volume inventory panel (Figure 11-12), you can modify existing volume
assignments by either assigning a volume to a new host port(s) or by unassigning a host from
an existing volume to host port(s) mapping.

To assign a volume to a host port, select the volume then click the Assign host button on the
right side of the volume inventory panel (Figure 11-12). You will be presented with a panel like
the one shown below in Figure 11-13 on page 386.

Select from the list of available host port world wide port names (WWPNs), and select either a
single host port WWPN, or select more than one by holding down the control <Ctrl> key and
selecting multiple host ports. When the desired host ports have been selected for Volume
assignment, click OK.


Figure 11-13 Assigning ESS LUN’s

When you click OK, TotalStorage Productivity Center for Fabric will be called to assist with
zoning this volume to the host. If TotalStorage Productivity Center for Fabric is not installed,
you will see a message panel as shown in Figure 11-14.

When the volume has been successfully assigned to the selected host port, the Assign host
ports panel will disappear and the ESS Volumes panel will be displayed once again, reflecting
now the additional host port mapping number in the far right side of the panel, in the Number
of host ports column.

Note: If TotalStorage Productivity Center for Fabric is installed, refer to Chapter 14, “Using
TotalStorage Productivity Center for Fabric” on page 703, for complete details of its
operation.

Also note that TotalStorage Productivity Center for Fabric is only invoked for zoning when
assigning hosts to ports. It is not invoked to remove zones when hosts are unassigned.

Figure 11-14 Tivoli SAN Manager warning


11.6.3 Creating new ESS volumes
To create new ESS volumes select the Create button from the Volumes panel as seen in
Figure 11-12 on page 385. The Create volume panel will appear (Figure 11-15).

Figure 11-15 ESS create volume

Use the drop-down fields to select the Storage type and choose from Available arrays on
the ESS. Then enter the number of volumes you want to create and the Volume quantity
along with the Requested size. Finally select the host ports you want to have access to the
new volumes from the Defined host ports scrolling list. You can select multiple hosts by
holding down the control key <Ctrl> while clicking hosts.

On clicking OK TotalStorage Productivity Center for Fabric will be called to assist with zoning
the new volumes to the host(s). If TotalStorage Productivity Center for Fabric (formally known
as TSANM) is not installed you will see a message panel as seen in Figure 11-16.

If TotalStorage Productivity Center for Fabric is installed, refer to Chapter 14, “Using
TotalStorage Productivity Center for Fabric” on page 703 for complete details of its operation.



Figure 11-17 Remove a host path from a volume

Figure 11-18 Display ESS volume properties

11.6.4 Launch device manager for an ESS device
This option allows you to link directly to the ESS Specialist of the chosen device:
Right-click the ESS storage resource, and select Launch Device Manager.
Launch Device Managers onto the storage device you want to query.


Figure 11-19 ESS specialist launched by Productivity Center common base

11.7 Working with DS8000
managing DS8000 devices.

be seen in Figure 11-20 on page 390.
available tasks under the Manage Storage Units section. These management function
applicable to all supported devices.
Right-click access: To access all functions available for a specific device, simply
right-click it to see a drop-down menu of options for that device. Figure 11-10 on page 383
shows the drop-down menu for an ESS.

Figure 11-20 on page 390 also shows the functions of TotalStorage Productivity Center for
Disk and TotalStorage Productivity Center for Replication. Although this chapter only covers
the Productivity Center common base functions, you would always have installed either
TotalStorage Productivity Center for Disk, TotalStorage Productivity Center for Replication, or
both.


Figure 11-20 Accessing Productivity Center common base functions

11.7.1 DS8000 Volume inventory
To view the status of the volumes available within a given DS8000 device, perform one of the
following actions:
Right-click the ESS device and select Volumes as in Figure 11-21 on page 391.

Tip: Before volumes can be displayed, as with other storage devices managed by
Productivity Center common base, an initial inventory must be completed.

If you try to view volumes for an DS8000 that has not been inventoried, you will receive a
notification that this needs to be done. To perform an inventory collection, see 11.4,
“Performing volume inventory” on page 378.


Figure 11-21 Working with DS8000 volumes

In either case, in the bottom left corner, the status will change from Ready to Starting Task,
and it will remain this way until the volume inventory appears. Figure 11-22 shows the
Volumes panel for the select DS8000 device that will appear.

Figure 11-22 DS8000 volume inventory panel


11.7.2 Assigning and unassigning DS8000 Volumes
From the DS8000 volume inventory panel (Figure 11-12 on page 385) you can modify
existing volume assignments by either assigning a volume to a new host port(s) or by
unassigning a host from an existing volume to host port(s) mapping.

To assign a volume to a host port, you can click the Assign host button on the right side of
the volume inventory panel. You will be presented with a panel like the one in Figure 11-23.
Select from the list of available host port world wide port names (WWPNs), and select either a
single host port WWPN, or select more than one, by holding down the control <Ctrl> key and
selecting multiple host ports. When the desired host ports have been selected for Volume

.

Figure 11-23 Assigning DS8000 LUN’s

When you click OK, TotalStorage Productivity Center for Fabric will be called to assist with
zoning this volume to the host. If TotalStorage Productivity Center for Fabric is not installed,
you will see a message panel as seen in Figure 11-24 on page 393.

When the volume has been successfully assigned to the selected host port, the Assign host
ports panel will disappear and the ESS Volumes panel will be displayed once again, reflecting
now the additional host port mapping number in the far right side of the panel, in the Number
of host ports column.

Note: If TotalStorage Productivity Center for Fabric is installed, refer to Chapter 14, “Using
TotalStorage Productivity Center for Fabric” on page 703 for complete details of its
operation.

Also note that TotalStorage Productivity Center for Fabric is only invoked for zoning when
assigning hosts to ports. It is not invoked to remove zones when hosts are unassigned.



11.7.3 Creating new DS8000 volumes
To create new ESS volumes, select the Create button from the Volumes panel as seen in
Figure 11-12 on page 385. The Create volume panel will appear (Figure 11-25).

Figure 11-25 DS8000 create volume

Use the drop-down fields to select the Storage type and choose from Available arrays on
the DS8000. Then enter the number of volumes you want to create and the Volume quantity
along with the Requested size. Finally select the host ports you want to have access to the
new volumes from the Defined host ports scrolling list. You can select multiple hosts by
holding down the control key <Ctrl> while clicking hosts.

On clicking OK, TotalStorage Productivity Center for Fabric will be called to assist with zoning
the new volumes to the host(s). If TotalStorage Productivity Center for Fabric is not installed
you will see a message panel as shown in Figure 11-26 on page 394.


I


11.7.4 Launch device manager for an DS8000 device
This option allows you to link directly to the DS8000 device manager of the chosen device:
Right-click the DS8000 storage resource, and select Launch Device Manager.
Launch Device Managers onto the storage device you want to query.
We received a message that TotalStorage Productivity Center for Disk could not
automatically logon (Figure 11-27). Click OK to get the DS8000 storage manager screen
as shown in Figure 11-28 on page 395.

Figure 11-27 DS8000 storage manager launch warning


Figure 11-28 shows the DS8000 device manager launched by Productivity Center common
base.

Figure 11-28 DS8000 device manager launched by Productivity Center common base


11.8 Working with SAN Volume Controller
managing SAN Volume Controller subsystems.

be seen in Figure 11-29 on page 397:
available tasks under the Manage Storage Units section. These management functions
appropriate to all supported devices.
Right-click access: To access all functions available for a specific device, right-click it to
see a drop-down menu of options for that device. Figure 11-29 on page 397 shows the
drop-down menu for a SAN Volume Controller.

Note: Overall, the SAN Volume Controller functionality offered in Productivity Center
common base compared to that of the native SAN Volume Controller Web based GUI is
fairly limited in version 2.1. There is the ability to add existing unmanaged LUNs to existing
MDisk groups, but there are no tools to remove MDisks from a group or create/delete
MDisk groups. The functions available for VDisks are similar too. Productivity Center
common base can create new VDisks in a given MDisk group, but there is little other
control over the placement of these volumes. It is not possible to remove VDisks or
reassign them to other hosts using Productivity Center common base.

11.8.1 Working with SAN Volume Controller MDisks
To view the properties of SAN Volume Controller managed disks (MDisk) as shown in
Figure 11-30 on page 398, perform one of the following actions:
Right-click the SVC storage resource, and select Managed Disks (Figure 11-29 on
page 397).
Managed Disks onto the storage device you want to query.

Tip: Before SAN Volume Controller managed disk properties (MDisks) can be displayed,
as with other storage devices managed by Productivity Center common base, an initial
inventory must be completed.

If you try to use the Managed Disk function on a SAN Volume Controller that has not been
inventoried, you will receive a notification that this needs to be done. Refer to 11.4,
“Performing volume inventory” on page 378 for details on performing this operation.


Here is the panel for selecting managed disks (Figure 11-29).

Figure 11-29 Select managed disk


Next, you should see the panel shown in Figure 11-30.

Figure 11-30 The MDisk properties panel for SAN Volume Controller

Figure 11-30 shows candidate or unmanaged MDisks, which are available for inclusion into
an existing MDisk group.

To add one or more unmanaged disks to an existing MDisk group:
Select the MDisk group from the pull-down menu.
Select one MDisk from the list of candidate MDisks, or use the <Ctrl> key to select
multiple disks.
Click the OK button at the bottom of the screen and the selected MDisk(s) will be added to
the MDisk group (Figure 11-31 on page 399).


Figure 11-31 Add MDisk to a managed disk group

11.8.2 Creating new MDisks on supported storage devices

Attention: The Create button, as seen in Figure 11-30 on page 398, is not for creating
new MDisk groups. It is for creating new MDisks on storage devices serving the SAN
Volume controller. It is not possible to create new MDisk groups using Version 2.3 of
Productivity Center common base.

Select the MDisk group from the pull-down menu (Figure 11-30 on page 398).
Select the Create button.
A new panel opens to create the storage volume (Figure 11-32 on page 400).
Select a device accessible to the SVC (devices not marked by an asterisk).
Devices marked with an asterisk are not acting as storage to the selected SAN Volume
Controller. Figure 11-32 on page 400 shows an ESS with an asterisk next to it. This is
because of the setup on the test environment. Make sure the device you select does not
have an asterisk next to it.
Specify the number of MDisks in the Volume quantity and size in the Requested volume
size.
Select the Defined SVC ports that should be assigned to these new MDisks.


Note: If TotalStorage Productivity Center for Fabric is installed and configured, extra
panels will appear to create appropriate zoning for this operation. See Chapter 14, “Using
TotalStorage Productivity Center for Fabric” on page 703 for details.

Click OK to start a process that will create a new volume on the selected storage device
and then add it to the SAN Volume Controllers MDisk group.

Figure 11-32 Create volumes to be added as MDisks

Productivity Center common base will now request the specified storage amount from the
specified back-end storage device (see Figure 11-33).

Figure 11-33 Volume creation results


The next step is to add the MDisks to an MDisk group (see Figure 11-34).

Figure 11-34 Assign MDisk to an MDisk group


Figure 11-35 shows the result of adding the mdisk4 to the selected MDisk group.

Figure 11-35 Result of adding the mdisk4 to the MDisk group

11.8.3 Create and view SAN Volume Controller VDisks
To create or view the properties of SAN Volume Controller virtual disks (VDisk) as shown in
Figure 11-36 on page 403, perform one of the following actions:
Right-click the SVC storage resource, and select Virtual Disks.
On the right-hand side under the Tasks column, drag Managed Storage Units → Virtual
Disks onto the storage device you want to query.

In version 2.3 of Productivity Center common base, it is not possible to delete VDisks. It is
also not possible to assign or reassign VDisks to a host after the creation process. Keep this
in mind when working with storage using Productivity Center common base on a SAN
Volume Controller. These tasks can still be performed using the native SAN Volume
Controller Web based GUI.


Tip: Before SAN Volume Controller virtual disk properties (VDisks) can be displayed, as
with other storage devices managed by Productivity Center common base, an initial
inventory must be completed.

If you try to use the Virtual Disk function on a SAN Volume Controller that has not been
inventoried, you will receive a notification that this needs to be done. To perform an
inventory collection, see 11.4, “Performing volume inventory” on page 378.

Figure 11-36 Launch Virtual Disks


Viewing VDisks
Figure 11-37 shows the VDisk inventory and volume attributes for the selected SAN Volume
controller.

Figure 11-37 The VDisk properties panel


Creating a VDisk
To create a new VDisk, use the Create button as shown in Figure 11-37 on page 404. You
need to provide a suitable VDisk name and select the MDisk group from which you want to
create the VDisk. Specify the number of VDisks to be created and the size in megabytes or
gigabytes that each VDisk should be. Figure 11-38 shows some example input in these fields.

Figure 11-38 SAN Volume Controller VDisk creation

The Host ports section of the VDisk properties panel allows you to use TotalStorage
Productivity Center for Fabric functionality to perform zoning actions to provide VDisk access
to specific host WWPNS. If TSANM is not installed, you will receive a warning

TotalStorage Productivity Center for Fabric” on page 703 for details on how to configure and
use it.


Figure 11-39 shows that the creation of the VDisk was successful.

Figure 11-39 Volume creation results

11.9 Working with DS4000 family or FAStT storage
managing DS4000 and FAStT type subsystems.

be seen in Figure 11-40 on page 407:
available tasks under the Manage Storage Units section. These management function
appropriate to all supported devices.
Right-click access: To access all functions available for the selected device, right-click it
to see a drop-down menu of options for it.

Figure 11-40 on page 407 shows the functions of TotalStorage Productivity Center for Disk
and TotalStorage Productivity Center for Replication. Although this chapter only covers the
Productivity Center common base functions you would always have either or both
TotalStorage Productivity Center for Disk, TotalStorage Productivity Center for Replication
installed.


11.9.1 Working with DS4000 or FAStT volumes
To view the status of the volumes available within a selected DS4000 or FAStT device,
perform one of the following actions:
Right-click the DS4000 or FAStT storage resource, and select Volumes (Figure 11-40).

In either case, in the bottom left corner, the status will change from Ready to Starting Task
and it will remain this way until the volume inventory is completed (see Figure 11-41 on
page 408).

Note: Before DS4000 or FAStT volume properties can be displayed, as with other storage
devices managed by Productivity Center common base, an initial inventory must be
completed.

Refer to 11.4, “Performing volume inventory” on page 378 for details.

Figure 11-40 Working with DS4000 and FAStT volumes


Figure 11-41 DS4000 and FAStT volumes panel

Figure 11-41 shows the volume inventory for the selected device. From this panel you can
Create and Delete volumes or assign and unassign volumes to hosts.


11.9.2 Creating DS4000 or FAStT volumes
To create new storage volumes on a DS4000 or FAStT, select the Create button from the right
side of the Volumes panel (Figure 11-41 on page 408). You will be presented with the Create
volume panel as in Figure 11-42.

Figure 11-42 DS4000 or FAStT create volumes

Select the desired Storage Type and array from Available arrays using the drop-down
menus. Then enter the Volume quantity and Requested volume size of the new volumes.
Finally select the host posts you wish to assign to the new volumes from the Defined host
ports scroll box, holding the <Crtl> key to select multiple ports.

The Defined host ports section of the panel allows you to use TotalStorage Productivity
Center for Fabric (formally TSANM) functionality to perform zoning actions to provide volume
access to specific host WWPNS. If TSANM is not installed, you will receive the warning

If TotalStorage Productivity Center for Fabric is installed refer to Chapter 14, “Using
use it.


If TotalStorage Productivity Center for Fabric is not installed, click OK to continue.

You should then see the panels shown below (Figure 11-43 through Figure 11-48 on
page 412).


Figure 11-44 Volume creation results (1)



11.9.3 Assigning hosts to DS4000 and FAStT Volumes
Use this feature to assign hosts to an existing DS4000 or FAStT volume.

To assign a DS4000 or FAStT volume to a host port, first select a volume by clicking it from
the volumes panel (Figure 11-41 on page 408). Now click the Assign host button from the
right side of the Volumes panel. You will be presented with a panel as shown in Figure 11-49.

Select from the list of available host ports world wide port names (WWPNs), and select
either a single host port WWPN, or more than one by holding down the control <Ctrl> key and
selecting multiple host ports. When the desired host ports have been selected for host

Figure 11-49 Assign host ports to DS4000 or FAStT

The Defined host ports section of the panel allows you to use TotalStorage Productivity
Center for Fabric (formally TSANM) functionality to perform zoning actions to provide volume
access to specific host WWPNS. If TSANM is not installed, you will receive the warning

use it.



If TotalStorage Productivity Center for Fabric is not installed, click OK to continue
(Figure 11-51).

Figure 11-51 DS4000 volumes successfully assigned to a host

11.9.4 Unassigning hosts from DS4000 or FAStT volumes
To unassign a DS4000 or FAStT volume from a host port, first select a volume by clicking it
from the volumes panel (Figure 11-41 on page 408). Now click the Unassign host button
from the right side of the Volumes panel. You will be presented with a panel as in
Figure 11-52. Select from the list of available host port world wide port names (WWPNs), and
select either a single host port WWPN, or more than one by holding down the control <Ctrl>
key and selecting multiple host ports. When the desired host ports have been selected for
host assignment, click OK.

Note: If the Unassign host button is grayed out when you have selected a volume, this
means that there are no current hosts assignments for that volume. If you believe this is
incorrect, it could be that the Productivity Center common base inventory is out of step with
this device’s configuration. This situation can arise when an administrator makes changes
to the device outside of the Productivity Center common base interface.

To correct this problem, perform an inventory for the DS4000 or FAStT and repeat. Refer
to 11.4, “Performing volume inventory” on page 378

Figure 11-52 Unassign host ports from DS4000 or FAStT


TotalStorage Productivity Center for Fabric is not called to perform zoning cleanup in version
2.1. This functionality is planned in a future release.

Figure 11-53 Volume unassignment results

11.9.5 Volume properties

Figure 11-54 DS4000 or FAStT volume properties


11.10 Event Action Plan Builder
The IBM Director includes sophisticated event-handling support. Event Action Plans can be
set up that specify what steps, if any, should be taken when particular events occur in the
environment.

Understanding Event Action Plans
An Event Action Plan associates one or more event filters with one or more actions.

For example, an Event Action Plan can be created to send a page to the network
administrator's pager if an event with a severity level of critical or fatal is received by the IBM
Director Server. You can include as many event filter and action pairs as needed in a single
Event Action Plan.

An Event Action Plan is activated only when you apply it to a managed system or group. If an
event targets a system to which the plan is applied and that event meets the filtering criteria
defined in the plan, the associated actions are performed. Multiple event filters can be
associated with the same action, and a single event filter can be associated with multiple
actions.

The list of action templates you can use to define actions are listed in the Actions pane of the
Event Action Plan Builder window (see Figure 11-55).

Figure 11-55 Action templates


Creating an Event Action Plan
Event Action Plans are created in the Event Action Plan Builder window. To open this window
from the Director Console, click the Event Action Plan Builder icon on the toolbar. The
Event Action Plan Builder window is displayed (see Figure 11-56).

Figure 11-56 Event Action Plan Builder

Here are the tasks to create an Event Action Plan.
1. To begin, do one of the following actions:
– Right-click Event Actions Plan in the Event Action Plans pane to access the context
menu, and then select New.
– Select File → New → Event Action Plan from the menu bar.
– Double-click the Event Action Plan folder in the Event Action Plans pane (see
Figure 11-57).

Figure 11-57 Create Event Action Plan


2. Enter the name you want to assign to the plan and click OK to save the new plan. The new
plan entry with the name you assigned is displayed in the Event Action Plans pane. The
plan is also added to the Event Action Plans task as a child entry in the Director Console
(see Figure 11-58). Now that you have defined an event action plan, you can assign one or
more filters and actions to the plan.

Figure 11-58 New Event Action Plan

Notes:
You can create a plan without having defined any filters or actions.
The order in which you build a filter, action, and Event Action Plan does not matter.

3. Assign at least one filter to the Event Action Plan using one of the following methods:
– Drag the event filter from the Event Filters pane to the Event Action Plan in the Event
Action Plans pane.
– Highlight the Event Action Plan, then right-click the event filter to display the context
menu and select Add to Event Action Plan.
– Highlight the event filter, then right-click the Event Action Plan to display the context
menu and select Add Event Filter (see Figure 11-59 on page 419).


Figure 11-59 Add events to the action plan

The filter is now displayed as a child entry under the plan (see Figure 11-60).

Figure 11-60 Events added to action plan


4. Assign at least one action to at least one filter in the Event Action Plan using one of the
following methods:
– Drag the action from the Actions pane to the target event filter under the desired Event
Action Plan in the Event Action Plans pane.
– Highlight the target filter, then right-click the desired action to display the context menu
and select Add to Event Action Plan.
– Highlight the desired action, then right-click the target filter to display the context menu
and select Add Action.
The action is now displayed as a child entry under the filter (see Figure 11-61).

Figure 11-61 Action as child of Display Events Action Plan

5. Repeat the previous two steps for as many filter and action pairings as you want to add to
the plan. You can assign multiple actions to a single filter and multiple filters to a single
plan.

Note: The plan you have just created is not active because it has not been applied to a
managed system or a group. In the next section we explain how to apply an Event
Action Plan to a managed system or group.


11.10.1 Applying an Event Action Plan to a managed system or group
An Event Action Plan is activated only when it is applied to a managed system or group.
To activate a plan:
Drag the plan from the Tasks pane of the Director Console to a managed system in the
Group Contents pane or to a group in the Groups pane.
Drag the system or group to the plan.
Select the plan, right-click the system or group, and select Add Event Action Plan (see
Figure 11-62).

Figure 11-62 Notification of Event Action Plan added to group/system(s)

Repeat this step for all associations you want to make. You can activate the same Event
Action Plan for multiple systems (see Figure 11-63).

Figure 11-63 Director with Event Action Plan - Display Events

Once applied, the plan is activated and displayed as a child entry of the managed system or
group to which it is applied when the Associations - Event Action Plans item is checked.


Message Browser
When an event occurs, the Message Browser (see Figure 11-64) pops up on the server
console.

Figure 11-64 Message Browser

If the message has not yet been viewed, then that Status for that message will be blank.
When viewed, a checked envelope icon will appear under the Status column next to the
message.

To see greater detail on a particular message, select the message in the left pain and click the
Event Details button (see Figure 11-65).

Figure 11-65 Event Details window


11.10.2 Exporting and importing Event Action Plans
With the Event Action Plan Builder, you can import and export action plans to files. This
enables you to move action plans quickly from one IBM Director Server to another or to
import action plans that others have provided.

Export
Event Action Plans can be exported to three types of files:
Archive: Backs up the selected action plan to a file that can be imported into any IBM
Director Server.
HTML: Creates a detailed listing of the selected action plans, including its filters and
actions, in an HTML file format.
XML: Creates a detailed listing of the selected action plans, including its filters and
actions, in an XML file format.

To export an Event Action Plan, do the following steps:
1. Open the Event Action Plan Builder.
2. Select an Event Action Plan from those available under the Event Action Plan folder.
3. Select File → Export, then click the type of file you want to export to (see Figure 11-66).
If this Event Action Plan will be imported by an IBM Director Server, then select Archive.

Figure 11-66 Archiving an Event Action Plan


4. Name the archive and set a location to save in the Select Archive File for Export window
as shown in Figure 11-67.

Figure 11-67 Select destination and file name

Tip: When you export an action plan, regardless of the type, the file is created on a local
drive on the IBM Director Server. If an IBM Director Console is used to access the IBM
Director Server, then the file could be saved to either the Server or the Console by
selecting Server or Local from the Destinations pull-down. It cannot be saved to a network
drive. Use the File Transfer task if you want to copy the file elsewhere.


Import
Event Action Plans can be imported from a file, which must be an Archive export of an action
plan from another IBM Director Server. Follow these steps to import an Event Action Plan:
1. Transfer the archive file to be imported to a drive on the IBM Director Server.
2. Open the Event Action Plan Builder from the main Console window.
3. Click File → Import → Archive (see Figure 11-68).

Figure 11-68 Importing an Event Action Plan

4. From the Select File for Import window (see Figure 11-69), select the archive file and
location. The file must be located on the IBM Director Server. If using the Console, you
must transfer the file to the IBM Director Server before it can be imported.

Figure 11-69 Select file for import


5. Click OK to begin the import process.
The Import Action Plan window opens, displaying the action plan to import (see
Figure 11-70). If the action plan had been assigned previously to systems or groups, you
will be given the option to preserve associations during the import. Select Import to
complete the import process.

Figure 11-70 Verifying import of Event Action Plan


12

Center Performance Manager
This chapter provides a step-by-step guide to help you configure and use the Performance
Manager functions provided by the TotalStorage Productivity Center for Disk.


12.1 Exploiting Performance Manager
You can use the Performance Manager component of TotalStorage Productivity Center for
Disk to manage and monitor the performance of the storage devices that TotalStorage
Productivity Center for Disk supports.

Performance Manager provides the following functions:
Collecting data from devices
Performance Manager collects data from the IBM TotalStorage Enterprise Storage Server
(ESS) and IBM TotalStorage SAN Volume Controller in the first release.
Configuring performance thresholds
You can use the Performance Manager to set performance thresholds for each device
type. Setting thresholds for certain criteria allows Performance Manager to notify you
when a certain threshold has been crossed, thus enabling you to take action before a
critical event occurs.
Viewing performance data
You can view performance data from the Performance Manager database using the gauge
application programming interfaces (APIs). These gauges present performance data in
graphical and tabular forms.
Using Volume Performance Advisor (VPA)
The Volume performance advisor is an automated tool that helps you select the best
possible placement of a new LUN from a performance perspective. This function is
integrated with Device Manager so that, when the VPA has recommended locations for
requested LUNs, the LUNs can ne allocated and assigned to the appropriate host without
going back to Device Manager.
Managing workload profiles
You can use Performance Manager to select a predefined workload profile or to create a
new workload profile that is based on historical performance data or on an existing
workload profile. Performance Manager uses these profiles to create a performance
recommendation for volume allocation on an IBM storage server.

The installation of the Performance Manager component onto an existing TotalStorage
Productivity Center for Disk server provides a new ‘Manage Performance’ task tree
(Figure 12-1) on the right-hand side of the TotalStorage Productivity Center for Disk host. This
task tree includes the various elements shown.

Figure 12-1 New Performance Manager tasks


12.1.1 Performance Manager GUI
The Performance Manager Graphical User Interface can be launched from the IBM Director
console Interface. After logging on to IBM Director, you will see a screen as in Figure 12-2.
On the rightmost Tasks pane, you will see Manage Performance launch menu. It is
highlighted and expanded in the figure shown.

Figure 12-2 IBM Director Console with Performance Manager

12.1.2 Performance Manager data collection
To collect performance data for the Enterprise Storage Server (ESS), Performance Manager
invokes the ESS Specialist server, setting a particular performance data collection frequency
and duration of collection. Specialist collects the performance statistics from an ESS,
establishes a connection, and sends the collected performance data to Performance
Manager. Performance Manager then processes the performance data and saves it in
Performance Manager database tables.

From this section you can create data collection tasks for the supported, discovered IBM
storage devices. There are two ways to use the Data Collection task to begin gathering device
performance data.
1. Drag and drop the data collection task option from the right-hand side of the Multiple
Device Manager application, onto the Storage Device for which you want to create the new
task.

Chapter 12. Using TotalStorage Productivity Center Performance Manager 429

2. Or, right-click a storage device in the center column, and select the Performance Data
Collection Panel menu option as shown in Figure 12-3.

Figure 12-3 ESS tasks panel

Either operation results in a new window named Create Performance Data Collection Task
(Figure 12-4).

In this window you will specify:
A task name
A brief description of the task
The sample frequency in minutes
The duration of data collection task (in hours)

Figure 12-4 Create Performance Data Collection Task for ESS


In our example, we are setting up a data collection task on an ESS with Device ID
2105.16603. After we have created a task name Cottle _ESS with sample frequency of 5
minutes and duration is 1 hour.

It is possible to add more ESSs to the same data collection task, by clicking the Add button
on the right-hand side. You can click individual devices, or select multiples by making use of
the Ctrl key. See Figure 12-5 for an example of this panel. In our example, we created a task
for the ESS with device ID 2105.22513.

Figure 12-5 Adding multiple devices to a single task


Once we have established the scope of our data collection task and have clicked the OK
button, we see our new data collection task available in the right-hand task column (see
Figure 12-6). We have created task Cottle_ESS in the example.

Tip: When providing a description for a new data collection task, you may elect to provide
information about the duration and frequency of the task.

Figure 12-6 A new data collection task


In order to schedule it, right-click the selected task (see Figure 12-7).

Figure 12-7 Scheduling new data collection task

You will see another window as shown in Figure 12-8.

Figure 12-8 Scheduling task

You have the option to use the job scheduling facility of TotalStorage Productivity Center for
Disk, or to execute the task immediately.


If you select Execute Now, you will see a panel similar to the one in Figure 12-9, providing
you with some information about task name and task status, including the time the task was
initialized.

Figure 12-9 Task progress panel

If you would rather schedule the task to occur at a future time, or to specify additional
parameters for the job schedule, you can walk through the panel in Figure 12-10. You may
provide a scheduled job description for the scheduled job. In our example, we created a job,
24March Cottle ESS.

Figure 12-10 New scheduled job panel


12.1.3 Using IBM Director Scheduler function
You may specify additional scheduled job parameters by using the Advanced button.
You will see the panel in Figure 12-11. You can also launch this panel from IBM Director
Console → Tasks → Scheduler → File → New Job. You can also set up the repeat
frequency of the task.

Figure 12-11 New scheduled job, advanced tab

Once you are finished customizing the job options, you may save it using File → Save as
menu.

Or, you may do this by clicking the diskette icon in the top left corner of the advanced
panel.


When you save with advanced job options, you may provide a descriptive name for the job as

Figure 12-12 Save job panel with advanced options

You should receive a confirmation that your job has been saved as shown in Figure 12-13.

Figure 12-13 scheduled job is saved


12.1.4 Reviewing data collection task status
You can review the task status using Task Status under the rightmost column Tasks. See
Figure 12-14.

Figure 12-14 Task Status


Upon double-clicking Task Status, it launches the following panel as shown in Figure 12-15.

Figure 12-15 Task Status Panel

To review the task status, you can click the task shown under the Task name column.
For example, we selected the task FCA18P, which was aborted, as shown in Figure 12-16 on
page 439. Subsequently, it will show the details with Device ID, Device status, and Error
Message ID in the Device status box. You can click the entry in the device status box. It will
further show the Error message in the Error message box.


Figure 12-16 Task status details

12.1.5 Managing Performance Manager Database
The collected performance data is stored in a back-end DB2 database. This database needs
to be maintained in order to keep only relevant data in the database. Your may decide on a
frequency for purging old data based on your organization’s requirements.


The performance database panel can be launched on clicking Performance Database as
shown in Figure 12-17. It will display the Performance Database Properties panel as shown in
Figure 12-18 on page 441.

Figure 12-17 Launch Performance Manager database


You can use the performance database panel to specify properties for a performance
database purge task. The sizing function on this panel shows used space and free space in
the database. You can choose to purge performance data based on age of the data, the type
of the data, and the storage devices associated with the data (Figure 12-18).

Figure 12-18 Properties of Performance database

The Performance database properties panel shows the following data:
Database name
The name of the database
Database location
The file system on which the database resides.
Total file system capacity
The total capacity available to the file system, in gigabytes.
Space currently used on file system
Space is shown in gigabytes and also by percentage.
Performance manager database full
The amount of space used by Performance Manager. The percentage shown is the
percentage of available space (total space - currently used space) used by the
Performance Manager database.


The following formula is used to derive the percentage of disk space full in the
Performance Manager database:
a = the total capacity of the file system
b = the total allocated space for Performance Manager database on the file system
c = the portion of the allocated space that is used by the Performance Manager database
For any decimal amount over a particular number, the percentage is rounded up to the
next largest integer. For example, 5.1% is rounded to and displayed as 6%.
Space status advisor
The Space status advisor monitors the amount of space used by the Performance
Manager database and advises you as to whether you should purge data. The advisor
levels are:
Low: You do not need to purge data now
High: You should purge data soon
Critical: You need to purge data now
Disk space thresholds for status categories:
low if utilization <0.8, high if 0.8 <= utilization <0.9 and critical otherwise.
The delimiters between low/high/critical are 80% and 90% full. Purge database options
Groups the database purge information.
Name Type
A name for the performance database purge task. The maximum length for a name can be
from 1 to 250 characters.
Description (optional)
Type a description for the performance database purge task. The maximum length for a
description can be from 1 to 250 characters.
Device type
Select one or more storage device types for the performance database purge. Options are
SVC, ESS, or All. (Default is All.)
Purge performance data older than
Select the maximum age for data to be retained when the purge task is run. You can
specify this value in days (1-365) or years (1-10). For example, if you select the Days
button and a value of 10, the database purge task will purge all data older than 10 days
when it is run. Therefore, if it has been more than 10 days since the task was run, all
performance data would be purged. Defaults are 365 days or 10 years.
Purge data containing threshold exception information
Deselecting this option will preserve performance data that contains information about
threshold exceptions. This information is required to display exception gauges. This option
is selected by default.
Save as task button
When you click Save as task, the information you specified is saved and the panel closes.
The newly created task is saved to the IBM Director Task pane under the Performance
Manager Database. Once it is saved, the task can be scheduled using the IBM Director
scheduler function.


12.1.6 Performance Manager gauges
Once data collection is complete, you may use the Gauges task to retrieve information about
a variety of storage device metrics.

Gauges are used to tunnel down to the level of detail necessary to isolate performance issues
on the storage device. To view information collected by the Performance Manager, a gauge
must be created or a custom script written to access the DB2 tables/fields directly.

Creating a gauge
Open the IBM Director and do one of the following tasks:
Right-click the storage device in the center pane and select Gauges (see Figure 12-19).

Figure 12-19 Right-click gauge opening

You can click Gauges on the panel shown and it will produce the Job Status window as
shown in Figure 12-21 on page 444. It is also possible to launch Gauge creation by
expanding Multiple Device Manager - Manage Performance in the rightmost column. Drag
the Gauges item to the storage device desired and drop to open the gauges for that device
(see Figure 12-20 on page 444).


Figure 12-20 Drag-n-drop gauge opening

This will produce the Job status window (see Figure 12-21) while the Performance gauges
window opens. You will see the Job status window while other selected windows are opening.

Figure 12-21 Opening Performance gauges job status

The Performance gauges window will be empty until a gauge is created for use. We have
created three gauges.(see Figure 12-22).

Figure 12-22 Performance gauges

Clicking the Create button to the left brings up the Job status window while the Create
performance gauge window opens.


The Create performance gauge window changes values depending on whether the cluster,
array, or volume items are selected in the left pane. Clicking the cluster item in the left pane
produces a window as seen in Figure 12-23.

Figure 12-23 Create performance gauge - Performance

Under the Type pull-down menu, select Performance or Exception.

Performance
Cluster Performance gauges provide details on the average cache holding time in seconds as
well as the percent of I/O requests that were delayed due to NVS memory shortages. Two
Cluster Performance gauges are required per ESS to view the available historical data for
each cluster. Additional gauges can be created to view live performance data.
Device: Select the storage device and time period from which to build the performance
gauge. The time period can be changed for this device within the gauge window, thus
allowing an overall or detailed view of the data.
Name: Enter a name that is both descriptive of the type of gauge as well as the detail
provided by the gauge. The name must not contain white space, special characters, or
exceed 100 characters in length. Also, the name must be unique on the TotalStorage
Productivity Center for Disk Performance Manager Server.
If “test” were used as a gauge name, then it could not be used for another gauge -— even
if another storage device were selected — as it would not be unique in the database.
Example names: 28019P_C1H would represent the ESS serial number (28019), the
performance gauge type (P), the cluster (C1), and historical (H), while 28019E would
represent the exception (E) gauge for the same ESS. Gauges for the clusters and arrays
would build on that nomenclature to group the gauges by ESS on the Gauges window.


Description: Use this space to enter a detailed description of the gauge that will appear
on the gauge and in the Gauges window.
Metric(s): Click the metric(s) that will be display by default when the gauge is opened for
viewing. Those metrics with the same value under the Units column in the Metrics table
can be selected together using either Shift mouse-click or Ctrl mouse-click.
The metrics in this field can be changed on the historical gauge after the gauge has been
opened for viewing. In other words, a historical gauge for each metric or group of metrics
is not necessary. However, these metrics cannot be changed for live gauges. A new
gauge is required for each metric or group of metrics desired.
Component: Select a single device from the Component table. This field cannot be
changed when the gauge is opened for viewing.
Data points: Selecting this radio button enables the gauge to display most recent data
being obtained from currently running performance collectors against the storage device.
One most recent performance data gauge is required per cluster and per metric to view
live collection data.
The Device pull-down menu displays text informing the user whether or not a performance
collection task is running against this Device.
You can select number of datapoints for your requirements to display the last “x” data
points from the date of the last collection. The data collection could be currently running or
the most recent one.
Date Range: Selecting this radio button presents data over a range of dates/times. Enter
the range of dates this gauge will use as a default for the gauge.
The date and time values may be adjusted within the gauge to any value before or after
the default values and the gauge will display any relevant data for the updated time period.
Display gauge: Checking this box will display the newly created gauge after clicking the
OK button. Otherwise, if left blank, the gauge will be saved without displaying.


Click the OK button when ready to save the performance gauge (see Figure 12-24). In this
example, we have created a gauge with the name 22513C1H and the description, average
cache holding time. We selected a starting and ending date as 11-March-2005. This
corresponds with our data collection task schedule.

Figure 12-24 Ready to save performance gauge

The gauge appears after clicking the OK button with the Display gauge box checked or when
the Display button is clicked after selecting the appropriate gauge on the Performance
gauges window (see Figure 12-26 on page 448). If you decide not to display gauge and save
only, then you will see a panel as shown here in Figure 12-25.

Figure 12-25 Saved performance gauges


Figure 12-26 Cluster performance gauge - upper

The top of the gauge contains the following labels:
Graph Name The name of the gauge
Description The description of the gauge
Device The storage device selected for the gauge
Component level Cluster, array, volume
Component ID The ID # of the component (cluster, array, volume)
Threshold The thresholds that were applied to the metrics
Time of last data collection Date and time of the last data collection

The center of the gauge contains the only fields that may be altered in the Display Properties
section. The Metrics may be selected either individually or in groups as long as the data types
are the same (for example, seconds with seconds, milliseconds with milliseconds, or percent
with percent). Click the Apply button to force a Performance Gauge section update with the
new y-axis data.

The Start Date:, End Date:, Start Time:, and End Time: fields may be varied to either expand
the scope of the gauge or narrow it for a more granular view of the data. Click the Apply
button to force a Performance Gauge section update with the new x-axis data. For example,
we applied Total I/O Rate metric to the saved gauge, and the resultant graph is shown in
Figure 12-27 on page 449. Here, the Performance Gauge section of the gauge displays
graphically, the information over time selected by the gauge and the options in the Display
Properties section.


Figure 12-27 Cluster performance gauge with applied I/O rate metric

Click the Refresh button in the Performance Gauge section to update the graph with the
original metrics and date/time criteria. The date and time of the last refresh appear to the right
of the Refresh button. The date and time displayed are updated first, followed by the contents
of the graph, which can take up to several minutes to update.

Finally, the data used to generate the graph is displayed at the bottom of the window (see
Figure 12-28 on page 450). Each of the columns in the data section can be sorted up or down
by clicking the column heading (see Figure 12-32 on page 453). The sort reads the data from
left to right, so the results may not be as expected.

The gauges for the array and volume components function in the same manner as the cluster
gauge created above.


Figure 12-28 Create Performance Gauge- Lower

Exception
Exception gauges display data only for those active thresholds that were crossed during the
reporting period. One Exception gauge displays threshold exceptions for the entire storage
device based on the thresholds active at the time of collection.


To create an exception gauge, select Exception from the Type pull-down menu (see
Figure 12-29).

Figure 12-29 Create performance gauge - Exception

By default, the Cluster will be highlighted in the left pane and the metrics and component
sections will not be available.
Device: Select the storage device and time period from which to build the performance
gauge. The time period can be changed for this device within the gauge window thus
allowing an overall or detailed view of the data.
Name: Enter a name that is both descriptive of the type of gauge as well as the detail
provided by the gauge. The name must not contain white space, special characters, or
exceed 100 characters in length. Also, the name must be unique on the TotalStorage
Productivity Center for Disk Performance Manager Server.
Description: Use this space to enter a detailed description of the gauge that will appear
on the gauge and in the Gauges window
Date Range: Selecting this radio button presents data over a range of dates/times. Enter
the range of dates this gauge will use as a default for the gauge.
The date and time values may be adjusted within the gauge to any value before or after
the default values and the gauge will display any relevant data for the updated time period.
Display gauge: Checking this box will display the newly created gauge after clicking the
OK button. Otherwise, if left blank, the gauge will be saved without displaying.


Click the OK button when ready to save the performance gauge. We created an exception
gauge as shown in Figure 12-30.

Figure 12-30 Ready to save exception gauge

The top of the gauge contains the following labels:
Graph Name The name of the gauge
Description The description of the gauge
Device The storage device selected for the gauge
Threshold The thresholds that were applied to the metrics
Time of last data collection Date and time of the last data collection

The center of the gauge contains the only fields that may be altered in the Display Properties
section. The Start Date: and End Date: fields may be varied to either expand the scope of the
gauge or narrow it for a more granular view of the data. Click the Apply button to force an
Exceptions Gauge section update with the new x-axis data.

The Exceptions Gauge section of the gauge displays graphically, the information over time
selected by the gauge, and the options in the Display Properties section (see Figure 12-31 on
page 453).


Figure 12-31 Exceptions gauge - upper

Click the Refresh button in the Exceptions Gauge section to update the graph with the
original date criteria. The date and time of the last refresh appear to the right of the Refresh
button. The date and time displayed are updated first, followed by the contents of the graph,
which can take up to several minutes to update. Finally, the data used to generate the graph
are displayed at the bottom of the window.

Each of the columns in the data section can be sorted up or down by clicking the column
heading (see Figure 12-32).

Figure 12-32 Data sort options


Display Gauges
To display previously created gauges, either right-click the storage device and select gauges
(see Figure 12-19 on page 443) or drag and drop the Gauges item on the storage device
(see Figure 12-20 on page 444) to open the Performance gauges window, shown here in
Figure 12-33).

Figure 12-33 Performance gauges window

Select one of the gauges and then click Display.

Gauge Properties
The Properties button allows the following fields or choices to be modified.

Performance
These are the performance related possibilities:
Description
Metrics
Component
Data points
Date range — date and time ranges


You can change the data displayed in the gauge from Data points with an active data
collection to Date range (see Figure 12-34). Selecting Date range allows you to choose the
Start date and End Date using the performance data stored in the DB2 database.

Figure 12-34 Performance gauge properties


Exception
You can change the Type property of the gauge definition from Performance to Exception. For
a gauge type of Exception, you can only choose to view data for a Date range (see
Figure 12-35).

Figure 12-35 Exception gauge properties

Delete a gauge
To delete a previously created gauge, either right-click the storage device and select gauges
(see Figure 12-19 on page 443) or drag and drop the Gauges item on the storage device (see
Figure 12-20 on page 444) to open the Performance gauges window shown in Figure 12-33
on page 454.

Select the gauge to remove and click Delete. A pop-up window will prompt for confirmation to
remove the gauge (see Figure 12-36).

Figure 12-36 Confirm gauge removal


To confirm, click Yes and the gauge will be removed. The gauge name may now be reused, if
desired.

12.1.7 ESS thresholds
Thresholds are used to determine watermarks for warning and error indicators for an
assortment of storage metrics, including:
Disk Utilization
Cache Holding Time
NVS Cache Full
Total I/O Requests

Thresholds are used either by:
1. Right-clicking a storage device in the center panel of TotalStorage Productivity Center for
Disk, and selecting the thresholds menu option (Figure 12-37)
2. Or, by dragging and dropping the thresholds task from the right tasks panel in Multiple
Device Manager, onto the desired storage device, to display or modify the thresholds for
that device

Figure 12-37 Opening the thresholds panel


Upon opening the thresholds submenu, you will see the following display, which shows the
default thresholds in place for ESS, as shown in Figure 12-38.

Figure 12-38 Performance Thresholds main panel

On the right-hand side, there are buttons for Enable, Disable, Copy Threshold Properties,
Filters, and Properties.

If the selected task is already enabled, then the Enable button will appear greyed out, as in
our case. If we attempt to disable a threshold that is currently enabled, by clicking the Disable
button, a message will be displayed as shown in Figure 12-39.

Figure 12-39 Disabling threshold warning panel

You may elect to continue, and disable the selected threshold, or to cancel the operation by
clicking Don’t disable threshold.


The copy threshold properties button will allow you to copy existing thresholds to other
devices of similar type (ESS, in our case). The window in Figure 12-40 is displayed.

Figure 12-40 Copying thresholds panel

Note: As shown in Figure 12-40, the copying threshold panel is aware that we have
registered on our ESS CIM agent host both clusters of our model 800 ESS, as indicated by
the semicolon delimited IP address field for the device ID “2105.22219”.

The Filters window is another available thresholds option. From this panel, you can enable,
disable, and modify existing filter values against selected thresholds as shown in
Figure 12-41.

Figure 12-41 Threshold filters panel


Finally, you can open the properties panel for a selected threshold, and are shown the panel
in Figure 12-42. You have options to acknowledge the values at their current settings, or
modify the warning or error levels, or select the alert level (none, warning only, and warning or
error are the available options).

Figure 12-42 Threshold properties panel

12.1.8 Data collection for SAN Volume Controller
Performance Manager uses an integrated configuration assistant tool (ICAT) interface of a
SAN Volume Controller (SVC) to start and stop performance statistics collection on an SAN
Volume Controller device.

The process for performing data collection on SAN Volume Controller is similar to that of ESS.

You will need to setup a new Performance Data Collection Task for the SAN Volume
Controller device. Figure 12-43 on page 461 is an example of the panel you should see when
you drag the Data Collection task onto the SAN Volume Controller device, or right-click the
device and left-click Data Collection.

As with the ESS data collection task:
Define a task name and description
Select sample frequency and duration of the task and click OK

Note: The SAN Volume Controller can perform data collection at a minimum 15 minute
interval.

You may use the Add button to include additional SAN Volume Controller devices in the same
data collection task, or use the Remove button to exclude SAN Volume Controllers from an
existing task. In our case we are performing data collection against a single SAN Volume
Controller.


Figure 12-43 The SVC Performance Data Collection Task

As long as at least one data collection task has been completed, you are able to proceed with
the steps to create a gauge to view your performance data.

12.1.9 SAN Volume Controller thresholds
To view the available Performance Manager Thresholds, you can right-click the SAN Volume
Controller device and click Thresholds, or drag the Threshold task from the right-hand panel
onto the SAN Volume Controller device you want to query.

A panel like the one in Figure 12-44 appears.

Figure 12-44 The SVC performance thresholds panel

SVC has following thresholds with their default properties:
VDisk I/Os rate
Total number of virtual disk I/Os for each I/O group. SAN Volume Controller defaults:
– Status: Disabled
– Warning: None
– Error: None
VDisk bytes per second
Virtual disk bytes per second for each I/O group. SAN Volume Controller defaults:
– Warning: None
– Error: None


MDisk I/O rate
Total number of managed disk I/Os for each managed disk group. SAN Volume Controller
defaults:
– Warning: None
– Error: None
MDisk bytes per second
Managed disk bytes per second for each managed disk group. SAN Volume Controller
defaults:
– Warning: None
– Error: None

You may only enable a particular threshold once minimum values for warning and error levels
have been defined. If you attempt to select a threshold and enable it without first modifying
these values, you will see a notification like the one in Figure 12-45.

Figure 12-45 SAN Volume Controller threshold enable warning

Tip: In TotalStorage Productivity Center for Disk, default threshold warning or error values
of -1.0 are indicators that there is no recommended minimum value for the threshold and
are therefore entirely user defined. You may elect to provide any reasonable value for these
thresholds, keeping in mind the workload in your environment.

To modify the warning and error values for a given threshold, you may select the threshold,
and click the Properties button. The panel in Figure 12-46 will be shown. You can modify the
threshold as appropriate, and accept the new values by selecting the OK button.

Figure 12-46 Modifying threshold warning and error values


12.1.10 Data collection for the DS6000 and DS8000
.The process for performing data collection on DS6000/DS8000 is similar to that of ESS.

You will need to set up a new Performance Data Collection Task for the DS6000/DS8000
device. Figure 12-47 is an example of the panel you should see when you drag the Data
Collection task onto the SAN Volume Controller device, or right-click the device and left-click
Data Collection. Figure 12-48 shows user validation.

As with the ESS data collection task:
Define a task name and description
Select sample frequency and duration of the task and click OK.

Figure 12-47 DS6000/DS8000 user name and password

Figure 12-48 The DS6000/DS8000 Data Collection Task

As long as at least one data collection task has been completed, you are able to proceed with
the steps to create a gauge to view your performance data (Figure 12-49 on page 464
through Figure 12-51 on page 466).


Figure 12-49 DS6000/DS8000 Cluster level gauge values


Figure 12-50 DS6000/DS8000 Rank Group level gauges


Figure 12-51 DS6000/DS8000 Volume level gauges

12.1.11 DS6000 and DS8000 thresholds
To view the available Performance Manager Thresholds, you can right-click the
DS6000/Ds8000 device and click Thresholds, or drag the Threshold task from the right-hand
panel onto the DS6000/DS8000 device you want to query.

A panel like the one in Figure 12-52 appears.

Figure 12-52 The DS6000/DS8000 performance thresholds panel


You may only enable a particular threshold once minimum values for warning and error levels
have been defined. If you attempt to select a threshold and enable it without first modifying
these values, you will see a notification like the one in Figure 12-45 on page 462.

Figure 12-53 DS6000/DS8000 threshold enable warning

Tip: In TotalStorage Productivity Center for Disk, default threshold warning or error values
of -1.0 are indicators that there is no recommended minimum value for the threshold and
are therefore entirely user defined. You may elect to provide any reasonable value for these
thresholds, keeping in mind the workload in your environment.

To modify the warning and error values for a given threshold, you may select the threshold,
and click the Properties button. The panel in Figure 12-46 on page 462 will be shown. You
can modify the threshold as appropriate, and accept the new values by clicking the OK
button.

Figure 12-54 Modifying DS6000/DS8000 threshold warning and error values

12.2 Exploiting gauges
Gauges are a very useful tool and help in identifying performance bottlenecks. In this section
we show the drill down capabilities of gauges. The purpose of this section is not to cover
performance analysis in detail for a specific product, but to highlight capabilities of the tool.
You may adopt and use a similar approach for the performance analysis.


12.2.1 Before you begin
Before you begin with customizing gauges, ensure that enough correct samples of data are
collected in the performance database. This is true for any performance analysis.

The data samples you collect must cover an appropriate time period that corresponds with
high/low instances of the I/O workload. Also, the samples should cover sufficient iterations of
the peak activity to perform analysis over a period of time. This is true for analyzing a pattern.
You may use the advanced scheduler function of IBM Director to configure a repetitive task.

If you plan to perform analysis for one specific instance of activity, then you can ensure that
the performance data collection task covers the specific time period.

12.2.2 Creating gauges: an example
In this example, we will cover creation and customization of gauges for ESS.

First of all, we scheduled an ESS performance data collection task at every 3-hour interval
using the IBM Director scheduler function for 8 days. For details on using the IBM Director
scheduler, refer to 12.1.3, “Using IBM Director Scheduler function” on page 435.

For creating the gauge, we launched the Performance gauges panel as shown in
Figure 12-55, by right-clicking the ESS device.

Figure 12-55 Gauges panel


Click the Create button to create a new gauge. You will see a panel similar to Figure 12-56.

Figure 12-56 Create performance gauge

We selected Cluster in the top left corner, Total I/O Rate metric in the metrics box, and
Cluster 1 in the component box. Also, we entered the following parameters:
Name: 22219P_drilldown_analysis
Description: Eiderdown analysis for 22219 ESS


For the Date range, we selected our historical data collection sampling period and clicked
Display gauge. Upon clicking OK, we got the next panel as shown in Figure 12-57.

Figure 12-57 Gauge for ESS 22219 Cluster performance


12.2.3 Zooming in on the specific time period
The previous chart shows some peaks of high cluster I/O rate between the period from April
6th to 8th. We decided to zoom into the peak activity and hence selected a more narrow time
period as shown in Figure 12-58 and clicked the Apply button.

Figure 12-58 Zooming on specific time period for Total IO rate metric

12.2.4 Modify gauge to view array level metrics
For the next chart, we decided to have an array level metric for the same time period as
before. Hence, we selected the gauge that we created earlier and clicked Properties as

Figure 12-59 Properties for a defined gauge


The subsequent panel is shown in Figure 12-60. We selected Array level metric for Cluster 1,
Device Adapter 1, Loop A, and Disk Group 2 for Avg. Response time as circled in
Figure 12-60.

Figure 12-60 Customizing gauge for array level metric


The resultant chart is shown in Figure 12-61.

Figure 12-61 Modified gauge with Avg. response time chart


12.2.5 Modify gauge to review multiple metrics in same chart
Next, we decided to review Total I/O, read/sec and writes/sec in the same chart for
comparison purpose. We selected these three metrics in the Gauge properties panel and
clicked the Apply button. The resultant chart is shown in Figure 12-62.

Tip: For selecting multiple metrics in the same chart, click the first metric, hold the shift
key, and click the last metric. If the metrics you plan to choose are not in the continuous
list, but are separated, then hold the control key instead of the shift key.

Figure 12-62 Viewing multiple metrics in the same chart

The chart Writes and Total I/O are shown as overlapping and Reads are shown as zero.

Tip: If you select multiple metrics that do not have the same units for the y-axis, then the
error is displayed as shown in Figure 12-63.

Figure 12-63 Error displayed if there are no common units


12.3 Performance Manager command line interface
The Performance Manager module includes a command line interface known as ‘perfcli’,
located in the directory c:Program FilesIBMmdmpmpmcli.

In its present release, the perfcli utility includes support for ESS and SAN Volume Controller
data collection task creation and management (starting and stopping data collection tasks).
There are also executables that support viewing and management of task filters, alert
thresholds, and gauges.

There is detailed help available at the command line, with information about syntax and
specific examples of usage.

12.3.1 Performance Manager CLI commands
The Performance Manager Command Line Interface (perfcli) includes the following
commands shown in Figure 12-64.

Figure 12-64 Directory listing of the perfcli commands

startesscollection/startsvccollection: These commands are used to build and run
data collection against the ESS or SAN Volume Controller, respectively.
lscollection: This command is used to list the running, aborted, or finished data
collection tasks on the Performance Management server.
stopcollection: This command may be used to stop data collection against a specified
task name.
lsgauge: You can use the lsgauge command to display a list of existing gauge names,
types, device types, device IDs, modified dates, and description information.
rmgauge: Use this command to remove existing gauges.
showgauge: This command is used to display performance data output using an existing
defined gauge.
setessthresh/setsvcthresh: These two commands are respectively used to set ESS and
SAN Volume Controller performance thresholds.
cpthresh: You can use the cpthresh command to copy threshold properties from one
selected device to one or more other devices.
setfilter: You can use setfilter to set or change the existing threshold filters.
lsfilter: This command may be used to display the threshold filter settings for all
devices specified.


setoutput: This command may be used to view or modify the existing data collection
output formats, including settings for paging, row printing, format (default, XML, or
character delimited), header printing, and output verbosity.
lsdev: This command can be used list the storage devices that are used by TotalStorage
Productivity Center for Disk.
lslun: This command can be used list the LUNs or Performance Manager volumes
associated with storage devices.
lsthreshold: This command can be used to list the threshold status associated with
storage devices.
lsgauge: This command can be used list the existing gauge names, gauge type, device
name, device ID, date modified, and optionally device information.
showgauge: Use this command to display performance output by triggering an existing
gauge.
showcapacity: This command displays managed capacity, the sum of managed capacity
by device type, and the total of all ESS and SAN Volume Controller managed storage.
showdbinfo: This command displays percent full, used space, and free space of the
Performance Manager database.
lsprofile: Use this command to display Volume Performance Advisor profiles.
cpprofile: Use this command to copy Volume Performance Advisor profiles.
mkprofile: Use this command to create a workload profile that you can use later with
mkrecom command to create a performance recommendation for ESS volume allocation.
mkreom: Use this command to generate and, optionally, apply a performance LUN advisor
recommendation for ESS volumes.
lsdbpurge: This command can be used to display the status of database purge tasks
running in TotalStorage Productivity Center for Disk.
tracklun: This command can be used to obtain historical performance statistics used to
create a profile.
startdbpurge: Use this command to start a database purge task.
showdev: Use this command to display device properties.
setoutput: This command sets output format for the administrative command line
interface.
cpthresh: This command can be used to copy threshold properties from one device to
other devices that are of the same type.
rmprofile: Use this command to remove delete performance LUN advisor profiles.


12.3.2 Sample command outputs
We show some sample commands in Figure 12-65. This sample shows how to invoke
perfcli commands from the Windows command line interface.

Figure 12-65 Sample perfcli command from Windows command line interface

Figure 12-66 and Figure 12-67 show perfcli sample commands within the perfcli tool.

Figure 12-66 perfcli sample command within perfcli tool

Figure 12-67 perfcli lslun sample command within perfcli tool


12.4 Volume Performance Advisor (VPA)
The Volume Performance Advisor (VPA) is designed to be an expert advisor that
recommends allocations for storage space based on considerations of the size of the request,
an estimate of the performance requirement and type of workload, as well as the existing load
on an ESS that might compete with the new request. The Volume Performance Advisor will
then make a recommendation as to the number and size of Logical Unit Numbers (logical
volumes or LUNs) to allocate, and a location within the ESS which is a good placement with
respect to the defined performance considerations. The user is given the option of
implementing the recommendation (allocating the storage), or obtaining subsequent
recommendations.

12.4.1 VPA introduction
Data placement within a large, complex storage subsystem has long been recognized as a
storage and performance management issue. Performance may suffer if done casually or
carelessly. It can also be costly to discover and correct those performance problems, adding
to the total cost of ownership.

Performance Manager is designed to contain an automated approach for storage allocation
through the functions of a storage performance advisor. It is called the Volume Performance
Advisor (VPA). The advisor is designed to automate decisions that could be achieved by an
expert storage analyst given the time and sufficient information. The goal is to give very good
advice by allowing VPA to consider the same factors that an administrator would in deciding
where to best allocate storage.

Note: At this point in time, the VPA tool is available for IBM ESS only.

12.4.2 The provisioning challenge
You want to allocate a specific amount of storage to run a particular workload. You could be a
storage administrator interacting through a user interface, or the user could be another
system component (such as a SAN management product, file system, DataBase
Management System (DBMS), or logical volume manager) interacting with the VPA
Application Programming Interface (API).

A storage request is satisfied by selecting some number of logical volumes (Logical Unit
Numbers (LUNs). For example, if you ask for 400GB of storage, then a low I/O rate,
cache-friendly workload could be handled on a single 400GB logical disk residing on a single
disk array; whereas a cache-unfriendly, high-bandwidth application might need several logical
volumes allocated across multiple disk arrays, using LVM, file system, or database striping to
achieve the required performance.

The performance of those logical disks depends on their placement on physical storage, and
what other applications might be sharing the arrays. The job of the Volume Performance
Advisor (VPA) is to select an appropriate set (number and placement) of logical disks that:
Consider the performance requirements of the new workload
Balance the workload across the physical resources
Consider the effects of the other workloads competing for the resources


Storage administrators and application developers need tools that pull together all the
components of the decision process used for provisioning storage. They need tools to
characterize and manage workload profiles. They need tools to monitor existing performance,
and tools to help them understand the impact of future workloads on current performance.
What they need is a tool that automates this entire process, which is what VPA for ESS does.

12.4.3 Workload characterization and workload profiles
Intelligent data placement requires a rudimentary understanding of the application workload,
and the demand likely to be placed on the storage system. For example, cache-unfriendly
workloads with high I/O intensity require a larger number of physical disks than cache-friendly
or lightweight workloads. To account for this, the VPA requires specific workload descriptions
to drive its decision-making process. These workload descriptions are precise, indicating I/O
intensity rates; percentages of read, write, random, and sequential content; cache
information; and transfer sizes. This workload-based approach is designed to allow the VPA
to correctly match performance attributes of the storage with the workload attributes with a
high degree of accuracy. For example, high random-write content workloads might best be
pointed to RAID10 storage. High cache hit ratio environments can probably be satisfied with
fewer numbers of logical disks.

Most users have little experience or capability for specifying detailed workload characteristics.
The VPA is designed to deal with this problem in three ways:
Predefined workload definitions based on characterizations of environments across
various industries and applications. They include standard OLTP type workloads, such as
“OLTP High”, and “Batch Sequential”.
Capturing existing workloads by observing storage access patterns in the environment.
The VPA allows the user to point to a grouping of volumes and a particular window of time,
and create a workload profile based on the observed behavior of those volumes.
Creation of hypothetical workloads that are similar to existing profiles, but differ in some
specific metrics.

The VPA has tools to manage a library of predefined and custom workloads, to create new
workload profiles, and to modify profiles for specific purposes.

12.4.4 Workload profile values
It is possible to change many specific values in the workload profile. For example, the access
density may be high because a test workload used small files. It can be adjusted to a more
accurate number. Average transfer size always defaults to 8KB, and should be modified if
other information is available for the actual transfer size.

The peak activity information should also be adjusted. It defaults to the time when the profile
workload was measured. In an existing environment it should specify the time period for
contention analysis between existing workloads and the new workload. Figure 12-68 on
page 480 shows a user defined VPA workload profile.


Figure 12-68 User defined workload profile details example

12.4.5 How the Volume Performance Advisor makes decisions
As mentioned previously, the VPA is designed to take several factors into account when
recommending volume allocation:
Total amount of space required
Minimum and maximum number of volumes, and sizes of volumes
Workload requirements
Contention from other workloads
VPA tries to allocate volumes on the least busy resources, at the same time balancing
workload across available resources. It uses the workload profile to estimate how busy
internal ESS resources will become if that workload is allocated on those resources. So it
estimates how busy the raid arrays, disk adapters, and controllers will become. The workload
profile is very important in making that decision.
For example, cache hit ratios affect the activity on the disk adapters and raid arrays. When
creating a workload profile from existing data, it's important for you to pick a representative
time sample to analyze. Also you should examine the IO/sec per GB. Many applications have
access density in the range of 0.1 to 3.0. If it is significantly outside this range, then this might
not be an appropriate sample.


The VPA will tend to utilize resources that can best accommodate a particular type of
workload. For example, high write content will make Raid 5 arrays busier than RAID 10 and
VPA will therefore bias to RAID 10. Faster devices will be less busy, so VPA biases allocations
to the faster devices. VPA also analyzes the historical data to determine how busy the internal
ESS components (arrays, disk adapters, clusters) are due to other workloads. In this way,
VPA tries to avoid allocating on already busy ESS components.

If VPA has a choice among several places to allocate volumes, and they appear to be about
equal, it is designed to apply a randomizing factor. This keeps the advisor from always giving
the same advice, which might cause certain resources to be overloaded if everyone followed
that advice. This also means that several usages of VPA by the same user may not
necessarily get the same advice, even if the workload profiles are identical.

Note: VPA tries to allocate the fewest possible volumes, as long as it can allocate on low
utilization components. If the components look too busy, it will allocate more (smaller)
volumes as a way of spreading the workload.It will not recommend more volumes than the
maximum specified by the user.

VPA may however be required to recommend allocation on very busy components. A
utilization indicator in the user panels will indicate whether allocations would cause
components to become heavily utilized. The I/O demand specified in the workload profile for
the new storage being allocated is not a Service Level Agreement (SLA). In other words,
there is no guarantee that the new storage, once allocated, will perform at or above the
specified access density. The VPA will make recommendations unless the available space on
the target devices is exhausted.

An invocation of VPA can be used for multiple recommendations. To handle a situation when
multiple sets of volumes are to be allocated with different workload profiles, it is important that
the same VPA wizard be used for all sets of recommendations. Select Make additional
recommendations on the View Recommendations page, as opposed to starting a
completely new sequence for each separate set of volumes to be allocated. VPA is designed
to remember each additional (hypothetical) workload when making additional
recommendations.

There are, of course, limitations to the use of an expert advisor such as VPA. There may well
be other constraints (like source and target Flashcopy requirements), which must be
considered. Sometimes these constraints can be accommodated with careful use of the tool,
and sometimes they may be so severe that the tool must be used very carefully. That is why
VPA is designed as an advisor.

In summary, the Volume Performance Advisor (VPA) provides you a tool to help automate
complex decisions involved in data placement and provisioning. It short, it represents a future
direction of storage management software! Computers should monitor their resources and
make autonomic adjustments based on the information. The VPA is an expert advisor which
provides you a step in that direction.

12.4.6 Enabling the Trace Logging for Director GUI Interface
Enabling GUI logging can be a useful for troubleshooting GUI problems, however unlikely they
may occur, which you may encounter while using VPA. Since this function requires a server
reboot where TotalStorage Productivity Center for Disk is running, you may consider doing
this prior to engaging in use of the VPA.


On the Windows platform, follow these steps:
1. Start → Run regedit.exe
2. Open the HKEY_LOCAL_MACHINE →SOFTWARE →Tivoli →Director →CurrentVersion
file.
3. Modify the LogOutput. Set the value to be equal to 1.
4. Reboot the server

The output log location from the instructions above is X:/program files/ibm/director/log
(where X is the drive where the Director application was installed). The log file for the Director
is com.tivoli.console.ConsoleLauncher.stderr.

On the Linux platform, TWGRas.properties sets the output logging on. You need to remove
the comment from the last line in the file (twg.sysout=1) and ensure that you have set
TWG_DEBUG_CONSOLE as an environment variable.
For example in bash: $export TWG_DEBUG_CONSOLE=true

12.4.7 Getting started
In this section, we provide detailed steps of using VPA with pre-defined performance
parameters (workload profile) you can utilize for advice in optimal volume placement in your
environment.

For detailed steps on creating customized workload profiles, you may refer to 12.4.8,
“Creating and managing workload profiles” on page 508.

To use VPA with customized workload profile, the major steps are:
Create a data collection task in Performance Manager
In order to utilize the VPA, you must first have a useful amount of performance data
collected from the device you want to examine. Refer to “Performance Manager data
collection” on page 429 for more detailed instructions regarding use of the Performance
data collection feature of the Performance Manager.
Schedule and run a successful performance data collection task
It is important to have an adequate amount of historical to provide you a statistically
relevant sampling population.
Create or use a user-defined workload profile
Use the Volume Performance Advisor to:
– Add Devices
– Specify Settings
– Select workload profile (predefined or user defined)
– View Profile Details
– Choose Candidate Location
– Verify Settings
– Approve Recommendations or restart VPA process with different parameters)

Workload profiles
The basic VPA concept, and the storage administrator’s goal, is to balance the workload
across all device components. This requires detailed ESS configuration information including
all components (clusters, device adapters, logical subsystems, ranks, and volumes)


To express the workload represented by the new volumes, they are assigned a workload
profile. A workload profile contains various performance attributes:
I/O demand, in I/O operations per second per GB of volume size
Average transfer size, in KBs per second
Percentage mix of I/O - sequential or random, and read or write
Cache utilization - percent of: cache hits for random reads, cache misses for random
writes
Peak activity time - the time period when the workload is most active

You can create your own workload profile definitions in two ways
By copying existing profiles, and editing their attributes
By performing an analysis of existing volumes in the environment
This second option is known as a Workload Analysis. You may select one or more
existing volumes, and the historical performance data for these volumes retrieved, to
determine their (average) performance behavior over time.

Using VPA with pre-defined workload profile
This section describes a VPA example using a default workload profile. The purpose of this
section to help you familiarize for using VPA tool.

Although, it is recommended to generate and use your customized workload profile after
gathering performance data. The customized profile will be realistic in terms of your
application performance requirements.

The VPA provides five predefined (canned) Workload Profile definitions. They are:
1. OLTP Standard: for general Online Transaction Processing Environment (OLTP)
2. OLTP High: for higher demand OLTP applications
3. Data Warehouse: for data warehousing applications
4. Batch Sequential: for batch applications accessing data sequentially
5. Document Archival: for archival applications, write-once, read-infrequently

Note: Online Transaction Processing (OLTP) is a type of program that facilitates and
manages transaction-oriented applications. OLTP is frequently used for data entry and
retrieval transactions in a number of industries, including banking, airlines, mail order,
supermarkets, and manufacturers. Probably the most widely installed OLTP product is
IBM's Customer Information Control System (CICS®).

Launching VPA tool
The steps to utilize a default workload profile to have the Volume Performance Advisor
examine and advise you on volume placement are:
1. In the IBM Director Task pane, click Multiple Device Manager.
2. Click Manage Performance
3. Click Volume Performance Advisor


4. You can choose two methods to launch VPA:
a. “Drag and Drop” the VPA icon to the storage device to be examined (see Figure 12-69).

Figure 12-69 “Drag and Drop” the VPA icon to the storage device


b. Select storage device → right-click the device → select Volume Performance Advisor
(see Figure 12-70).

Figure 12-70 Select ESS and right-click for VPA

If a storage device is selected for the drag and drop step, that is not in the scope of the VPA,
the following message will open (see Figure 12-71). Devices such as a CIMOM or an SNMP
device will generate this error. Only ESS is supported at this time.

Figure 12-71 Error launching VPA example

ESS User Validation
If this is the first time your are using VPA tool for the selected ESS device, then the ESS User
Validation panel will display as shown in Figure 12-72 on page 486. Otherwise, if you have
already validated the ESS user for VPA usage, then it will skip this panel and it will launch the
VPA setting default panel as shown in Figure 12-77 on page 488.


Figure 12-72 ESS User validation screen example

In the ESS User Validation panel, specify the user name, password, and port for each of
the IBM TotalStorage Enterprise Storage Servers (ESSs) that you want to examine.
During the initial setup of the VPA, on the ESS User Validation window, you need to first
select the ESS (as shown in Figure 12-73 on page 487) and then input correct username,
correct password and password verification.
You must click Set after you have input the correct username, password, and password
verification in the appropriate fields (see highlighted portion with circle in Figure 12-74 on
page 487). When you click Set, the application will populate the data you input (masked)
into the correct Device Information fields in the Device Information box (see Figure 12-75
on page 487).
If you do not click Set, before selecting OK, the following error(s) will appear depending on
what data needs to be entered.
BWN005921E (ESS Specialist username has not been entered correctly or applied)
BWN005922E (ESS Specialist password has not been entered correctly or applied)
If you encounter these errors, ensure you have correctly input the values in the input fields
in the lower part of the ESS user validation window and then retry by clicking OK.
The ESS user validation window contains the following fields:
– Devices table - Select an ESS from this table. It includes device IDs and device IP
addresses of the ESS devices on which this task was dropped.
– ESS Specialist username - Type a valid ESS Specialist user name and password for
the selected ESS. Subsequent displays of the same information for this ESS show the
user name and password that was entered. You can change the user name by
entering a new user name in this field.
– ESS Specialist password - Type a valid ESS Specialist password for the selected
ESS. Any existing password entries are removed when you change the ESS user
name.
– Confirm password - Type the valid ESS Specialist password again exactly as you
typed it in the password field.
– ESS Specialist port - Type a valid ESS port number. The default is 80.
– Set button - Click to set names, passwords, and ports without closing the panel.
– Remove button - Click to remove the selected information.
– Add button - Click to invoke the Add devices panel.


– OK button - Click to save the changes and close the panel.

Figure 12-73 ESS User validation - select ESS

Figure 12-74 Apply ESS Specialist user defined input

Figure 12-75 Applied ESS Specialist user defined input


Click the OK button to save the changes and close the panel. The application will attempt
to access the ESS storage device.
The error message in Figure 12-76 can be indicative of use of an incorrect username or
password for authentication. Additionally, If you have a firewall and are not adequately
authenticating to the storage device, the error may appear. If this does occur, check to
ensure you are using the correct username and password for the authentication and have
firewall access and are properly authenticating to establish storage device connectivity.

Figure 12-76 Authentication error example

Configuring VPA settings for the ESS diskspace request
After you have successfully completed the User Validation step, the VPA Settings window
will open (see Figure 12-77).

Figure 12-77 VPA Settings default panel


You use the Volume performance advisor - Settings window to identify your requirements
for host attachment and the total amount of space that you need. You can also use this
panel to specify volume number and size constraints, if any. We will begin with our
example as shown in Figure 12-78.

Figure 12-78 VPA settings for example


Here we describe the fields in this window:
– Total space required (GB) - Type the total space required in gigabytes. The smallest
allowed value is 0.1 GB. We requested 3 GB for our example.

Note: You cannot exceed the volume space available for examination on the server(s) you
select. To show the error, in this example we selected host Zombie and Total required
space as 400 GB. We got the error shown in Figure 12-79.
Action: Retry with different values and look at the server log for details.
Solution(s):
– Select a smaller maximum Total (volume) Space required GB and retry this step.
– Select more hosts which will include adequate volume space for this task.
– You may want to select the box entitled Consider volumes that have already been
allocated but not assigned in the performance recommendation.

Director log file enabling will generate logs for troubleshooting Director GUI components,
including the PM coswearer. In this example, the file we reference is:
com.tivoli.console.ConsoleLauncher.stderr.
(com.tivoli.console.ConsoleLauncher.stdout is also useful)

The sample log is shown in Figure 12-80.

Figure 12-79 Error showing exceeded the space requested

Figure 12-80 Director GUI console errorlog

– Specify a volume size range button - Click the button to activate the field, then use
the Minimum size (GB) spinner and the Maximum size (GB) spinner to specify the
range. In this example, we selected 1 GB as minimum and 3 GB as maximum.


– Specify a volume quantity range button - Click the button to activate the field, then
use the Minimum number spinner and the Maximum number spinner to specify the
range.
– Consider volumes that have already been allocated but not assigned to hosts in
the performance recommendation. If you check this box, VPA will use these types of
volumes in the volume performance examination process.
When this box (Consider volumes...) is checked and you click Next, the VPA wizard
will open the following warning window (see Figure 12-81).

Figure 12-81 Consider volumes - warning window example

Note: The BWN005996W message is a warning (W).

You have selected to reuse unassigned existing volumes which could potentially cause
data loss. Go Back to the VPA Settings window by clicking OK if you do not want to
consider unassigned volumes. Click the Help button for more information.

Explanation: The Volume Performance Advisor will assume that all currently unassigned
volumes are not in use, and may recommend the reuse of these volumes. If any of these
unassigned volumes are in use — for example, as replication targets or other data
replication purposes — and these volumes are recommended for reuse, the result could
be potential data loss.

Action: Go back to the Settings window and unselect Consider volumes that have already
been allocated but not assigned to hosts in the performance recommendation if you do not
want to consider volumes that may potentially be used for other purposes. If you want to
continue to consider unassigned volumes in your recommendations, then continue.

– Host Attachments table - Select one or more hosts from this table. This table lists all
hosts (by device ID) known to the ESS that you selected for this task.
It is important to only choose hosts for volume consideration that are the same server
type. It is also important to note that the VPA takes into consideration the maximum
volume limitations of server type such as (Windows 256 volumes maximum) and AIX
(approximately 4000 volumes). If you select a volume range above the server limit,
VPA will display an error. In our example we used the host “Zombie”.
– Next button - Click to invoke the Choose workload profile window. You use this
window to select a workload profile from a list of existing profile templates.


5. Click Next, after inputting your preferred parameters, and the Choose workload profile
window will display (see Figure 12-82).

Figure 12-82 VPA Choose workload profile window example

Choosing a workload profile
You can use the Choose workload profile window to select a workload profile from a list
of existing profiles. The Volume performance advisor uses the workload profile and other
performance information to advise you about where volumes should be created. For our
example we have selected the OLTP Standard default profile type.
– Workload profiles table - Select a profile from this table to view or modify. The table
lists predefined or existing workload profile names and descriptions. Predefined
workload profiles are shipped with Performance Manager. Workload profiles that you
previously created, if any, are also listed.
– Manage profiles button - Click to invoke the Manage workload profile panel.
– Profile details button - Click to see details about the selected profile in the Profile
details panel as shown in Figure 12-83 on page 493. Details include the following types
of information:
• Total I/O per second per GB
• Random read cache hits
• Sequential and random reads and writes
• Start and end dates
• Duration (days)


Note: You cannot modify the properties of the workload profile from this panel. The panel
options are “greyed out” (inactive). You can make changes to a workload profile from
Manage Profile → Create like panel.

– Next button - Click to invoke the Choose candidate locations window. You can use
this panel to select volume locations for the VPA to consider.

Figure 12-83 Properties for OLTP Standard profile

6. After reviewing the properties for predefined workload profiles, you may select a workload
profile from the table which closely resemble your workload profile requirements. For our
scenario, we have selected the OLTP Standard workload name from the Choose
workload profile window. We are going to use this workload profile for the LUN
placement recommendations.
– Name - Shows the default profile name. The following restrictions apply to the profile
name.
• The workload profile name must be between 1 to 64 characters.
• Legal characters are A-Z, a-z, 0-9, “-”, “_”, “.”, and “:”
• The first character cannot be “-” or “_”.


• Spaces are not acceptable characters.
– Description - Shows the description of workload profile.
– Total I/O per second per GB - Shows the values for the selected workload profile
Total I/O per second rate.
– Average transfer size (KB) - Shows the values for the selected workload profile.
– Caching information box - Shows the cache hits and destage percentages:
• Random read cache hits
Range from 1 - 100%. The default is 40%.
• Random write destage
Range from 1 - 100%. The default is 33%.
– Read/Write information box - Shows the read and write values.
The percentages for the four fields must equal 100%
• Sequential reads - The default is 14%.
• Sequential writes - The default is 23%.
• Random reads - The default is 36%.
• Random writes - The default is 32%.
– Peak activity information box
Since currently we are only viewing properties of an existing profile, the parameters for
this box are not selectable. But as reference for subsequent usage, you may review
this box. After you review properties for this box, you may click the Close button.
While creating new profile, this box will allow you to input following parameters:
• Use all available performance data radio button. You can select this option if you
want to include all available performance data previously collected in consideration
for this workload profile.
• Use the specified peak activity period radio button. You can select this button as
an alternate option (instead of using the Use all available performance data
option) for consideration in this workload profile definition.
• Time setting drop down menu. Select from the following options for the time setting
you want to use for this workload profile.
- Device time
- Client time
- Server time
- GMT
• Past days to analyze spinner. Use this (or manually enter the number) to select
the number of days of historical information you want to consider for this workload
profile analysis
• Time Range drop down lists. Select the Start time and End time to consider using
the appropriate fields.
– Close button - Click to close the panel. You will be returned to the Choose workload
profile window.


Choosing candidate locations
Select the name of the profile you want to use from the VPA Choose workload profile
window and then the Choose Candidate Locations window will open (see Figure 12-84).
We chose our OTLP Standard workload profile for the VPA analysis.

Figure 12-84 Choose candidate locations window

You can use the Choose candidate locations page to select volume locations for the
performance advisor to consider. You can choose to either include or exclude the selected
locations for the advisor's consideration.
The VPA uses historical performance information to advise you about where volumes
should be created. The Choose candidate locations page is one of the panels the
performance advisor uses to collect and evaluate the information.
– Device list - Displays device IDs or names for each ESS on which the task was
activated (each ESS on which you dropped the Volume advisor icon).
– Component Type tree - When you select a device from the Device list, the selection
tree opens on the left side of the panel.
The ESS component levels are shown in the tree. The following objects might be
included:
• ESS
• cluster
• device adapter


• array
• disk group
The component level names are followed by information about the capacity and the
disk utilization of the component level. For example, we used System component level.
It shows Component ID - 2105-F20-16603,Type- System, Description -
2105-F20-16603-IBM, Available capacity - 311GB, Utilization - Low.(see
Figure 12-84 on page 495).

Tip: You can select the different ESS component types and the VPA will reconsider the
volume placement advise based on that particular select. To familiarize yourself with the
options, select each component in turn to determine which component type centric advise
you prefer before proceeding to the next step.

Select a component type from the tree to display a list of the available volumes for that
component in the Candidates table (see Figure 12-84 on page 495). We chose system
for this example. It represents entire ESS system in this case.
Click Add button to add the component selected in the Candidates table to the Selected
candidates table. See Figure 12-85. It shows Selected candidate as 2105-F20-16603.

Figure 12-85 VPA Chose candidate locations Component Type tree example (system)


Verify settings for VPA
Click the Next button to invoke the Verify Settings window (see Figure 12-86).

Figure 12-86 VPA Verify settings window example

You can use the Verify settings panel to verify the volume settings that you specified in
the previous panels of the VPA.


Approve recommendations
After you have successfully completed the Verify Settings step, click the Next button,
and the Approve Recommendations window opens (see Figure 12-87).

Figure 12-87 VPA Recommendations window example

You use the Recommendations window to first view the recommendations from the VPA
and then to create new volumes based on the recommendations.
In this example, VPA also recommends the location of volume as 16603:2:4:1:1700 in the
Component ID column. This means recommended volume location is at ESS with ID
16603, Cluster 2, Device Adapter 4, Array 1 and volume ID1700.
With this information, it is also possible to create volume manually via ESS specialist
Browser interface or use VPA to create the same.
In the Recommendations window of the wizard, you can choose whether the
recommendations are to be implemented, and whether to loop around for another set of
recommendations. At this time, you have two options (other than to cancel the operation).
Make your final selection to Finish or return to the VPA for further recommendations.
a. If you do not want to assign the volumes using the current VPA advice, or want the
VPA to make another recommendation, check only the Make Additional
Recommendations box.


b. If you want to use the current VPA recommendation and make additional volume
assistants at this time, select both the Implement Recommendations and Make
Additional Recommendations check boxes. If you choose both options, you must
first wait until the current set of volume recommendations are created, or created and
assigned, before continuing. If you make this type of selection, a secondary window
will appear which runs synchronously within the VPA.

Tip: Stay in the same VPA session if you are going to implement volumes and add new
volumes. This will enable VPA to provide advice for your current selections, checking for
previous assignments, and verifying that no other VPA is processing the same volumes.

VPA loopback after Implement Recommendations selected
In the following example, we show the results of a VPA session.
1. In this example, we decided to Implement recommendations and also Make additional
recommendations. Hence we selected both check boxes (see Figure 12-88).

Figure 12-88 VPA Recommendation selected check box


2. Click the Continue button to proceed with VPA advice (see Figure 12-88 on page 499).

Figure 12-89 VPA results - in progress panel

3. In Figure 12-89, we can see that the volumes are being created on the server we selected
previously. This process takes a little time, so be patient.
4. Figure 12-90 indicates that the volume creation and assigning to ESS has completed. Be
patient and momentarily, the VPA loopback sequence will continue.

Figure 12-90 VPA final results


5. After the volume creation step has successfully completed, the following Settings window
will again open so that you may add more volumes (see Figure 12-91).

Figure 12-91 VPA settings default


For the additional recommendations, we decided to use same server. But, we specified the
Volume quantity range instead of Volume size range for the requested space of 2GB. See
Figure 12-92.

Figure 12-92 VPA additional space request


After clicking Next, the Choose Profile Panels opens. We selected the same profile as before:
OLTP Standard. See Figure 12-93.

Figure 12-93 Choose Profile


After clicking Next, the Choose candidate locations panel opens. We selected Cluster from
the Component Type drop-down list. See Figure 12-94.

Figure 12-94 Choose candidate location

The Component Type Cluster shows Component ID as 2105-F20-16603:2, Types as
Cluster, Descriptor as 2, Available capacity as 308GB and Utilization as Low. This indicates
that VPA plans to provision additional capacity on this Cluster 2 of ESS.


After clicking the Add button, Cluster 2 is a selected candidate for new volume. See
Figure 12-95.

Figure 12-95 Choose candidate location - select cluster


Upon clicking Next, the Verify settings panel opens as shown in Figure 12-96.

Figure 12-96 Verify settings


After verifying settings and clicking Next, VPA recommendations window opens. See
Figure 12-97.

Figure 12-97 VPA recommendations


Since the purpose of this example is to show our readers the VPA looping only, we decided to
un-check both check boxes for Implement Recommendations and Make additional
recommendations. Clicking Finish completed the VPA example (Figure 12-98).

Figure 12-98 Finish VPA panel

12.4.8 Creating and managing workload profiles
The VPA makes decisions based on characteristics of the workload profile to decide volume
placement recommendations. VPA decisions will not be accurate if an improper workload
profile is chosen, and it may cause future performance issues for application.

It is a must to have a valid and appropriate workload profile created prior to using VPA for any
application. Therefore, creating and managing workload profile is an important task, which
involves regular upkeep of workload profiles for each application disk I/O served by ESS.
Figure 12-99 on page 509 shows a typical sequence of managing workload profiles.


D eterm in e I/O w o rk lo a d
typ e o f ta rg e t C re a te I/O p erfo rm a n c e
a p p lic a tio n da ta c o lle c tio n ta s k
M
A N o m atch w ith C lo s e m a tc h w ith
N p re-d efined p ro file pre d e fine d p ro file

A
G
C C o o s e e P re -dfin e d d
h h o s P re -d e e fin e In itia te I/O p e rfo rm a n c e d a ta
I o r r C retete lik e
o C re a a lik e c o lle c tio n c ove rin g p e a k lo ad
tim e s an d g a th e r s u fficie n t
N p ro file
p ro file s a m p le s
G

S p e c ify tim e p e rio d o f
P p ea k a c tiv ity

R C h o o s e C reate
p ro file
O
F
I
L V a lid ate an alysis res u lts If res u lts n o t a c ce p tab le ,
re-v a lid a te d a ta c o lle c tio n
p a ra m e te rs
E
S R es u lts ac c ep ted

S a ve P ro file

Figure 12-99 Typical sequence for managing workload profiles

Before using VPA for any additional disk space requirement for an application, you will need
to:
determine typical I/O workload type of that application and;
have performance data collected which covers peak load time periods

You will need to determine the broad category in selected I/O workload fits in, e.g. whether it
is OLTP high, OLTP standard, Data Warehouse, Batch sequential or Document archival. This
is shown as highlighted box in the diagram. The TotalStorage Productivity Center for Disk
provides pre-defined profiles for these workload types and it allows you to create additional
similar profiles by choosing Create like profiles.

If do not find any match with pre-defined profiles, then you may prefer to Create a new profile.
While choosing Create like or Create profiles, you will also need to specify historical
performance data samples covering peak load activity time period. Optionally, you may
specify additional I/O parameters.

Upon submitting the Create or Create like profile, the performance analysis will performed
and results will be displayed. Depending upon the outcome of the results, you may need to
re-validate the parameters for data collection task and ensure that peak load samples are
taken correctly. If the results are acceptable, you may Save the profile. This profile can be
referenced for future usage by VPA.

In “Choosing workload profiles” on page 510, we cover step-by-step tasks using an example.


Choosing workload profiles
You can use Performance Manager to select a predefined workload profile or to create a new
workload profile based on historical performance data or on an existing workload profile.

Performance Manager uses these profiles to create a performance recommendation for
volume allocation on an IBM storage server. You can also use a set of Performance Manager
panels to create and manage the workload profiles. There are three methods you can use to
choose a workload profile as shown in Figure 12-100.

Figure 12-100 Choosing workload profiles

Note: Using a predefined profile does not require pre-existing performance data, but the
other two methods require historical performance data from the target storage device.

You can launch the workload profiles management tool using the drag and drop method from
the IBM Director console GUI interface. Drag the Manage Workload Profile task to the target
storage device as shown in Figure 12-101.

Figure 12-101 Launch Manage Workload Profile


If you are using Manage Workload Profile or VPA tool for first time of the selected ESS
device, then you will need to authorize ESS user validation. This has been described in detail
in “ESS User Validation” on page 485. The ESS User Validation is the same for VPA and
Manage Workload Profile tools.

After the successful ESS User validation, the Manage Workload Profile panel will be opened

Figure 12-102 Manage workload profiles

You can create or manage a workload profile using the following three methods:
1. Selecting a predefined workload profile
Several predefined workloads are shipped with Performance Manager. You can use the
Choose workload profile panel to select the predefined workload profile that most closely
matches your storage allocation needs. The default profiles shipped with Performance
Manager are shown in Figure 12-103.

Figure 12-103 Default workload profiles

You can select the properties panel of the respective pre-defined profile to verify the
profile details. A sample profile for OLTP Standard is shown in Figure 12-83 on page 493.


2. Creating a workload profile similar to another profile
You can use the Create like panel to modify the details of a selected workload profile.You
can then save the changes and assign a new name to create a new workload profile from
the existing profile. To Create like a particular profile, these are the tasks involved:
a. Create a performance data collection task for target storage device: You may
need to include multiple storage devices based on your profile requirements for the
application.
b. Schedule data collection task: You may need to ensure that a data collection task
runs over a sufficient period of time, which truly represents a typical I/O load of the
respective application. The key is to have sufficient historical data.

Tip: Best practice is to schedule frequency of a performance data collection task in such a
way that it covers peak load periods of I/O activity and it has at least a few samples of peak
loads. The number of samples depends on I/O characteristics of the application.

c. Determine the closest workload profile match: Determine whether new workload
profile matches w.r.t existing or pre-defined profiles. Note that it may not be the exact
fit, but should be of somewhat similar type.

d. Create the new similar profile: Using the Manage Workload Profile task, create new
profile. You will need to select appropriate time period for historical data, which you
have collected earlier.
In our example, we created similar profile using Batch Sequential pre-defined profile.
First, we select Batch Sequential profile and click Create like button as shown in
Figure 12-104.

Figure 12-104 Manage workload profile - create like


The Properties panel for Batch Sequential is opened, as shown in Figure 12-105.

Figure 12-105 Properties for Batch sequential profile

We changed the following values for our new profile:
Name: ITSO_Batch_Daily
Description: For ITSO batch applications
Average transfer size: 20KB
Sequential reads: 65%
Random reads: 10%
Peak Activity information: We used time period as past 24 days from 12AM to 11PM.


We saved our new profile (see Figure 12-106).

Figure 12-106 New Profile


This new profile - ITSO_Daily_batch is now available in Manage workload profile panel as
shown in Figure 12-107. This profile can now be used for VPA analysis. This completes our
example.

Figure 12-107 Manage profile panel with new profile

3. Creating a new workload profile from historical data

You can use the Manage workload profile panel to create a workload profile based on
historical data about existing volumes. You can select one or more volumes as the base for
the new workload profile. You can then assign a name to the workload profile, optionally
provide a description, and finally create the new profile.

To create a new workload profile, click the Create button as shown in Figure 12-108.

Figure 12-108 Create a new workload profile


This will launch a new panel for Creating workload profile as shown in Figure 12-109. At this
stage, you will need to specify the volumes for performance data analysis. In our example, we
selected all volumes. For selecting multiple volumes but not all, click the first volume, hold
the Shift key and click the last volume in the list. After all the required volumes are selected
(shown as dark blue), click the Add button. See Figure 12-109.

Note: The ESS volumes you specify should be representative of I/O behavior of the
application, for which you are planning to allocate space using the VPA tool.

Figure 12-109 Create new profile and add volumes


Upon clicking the Add button, all the selected volumes will be moved to the selected volumes
box as shown in Figure 12-110.

Figure 12-110 Selected volumes and performance period for new workload profile

In the Peak activity information box, you will need to specify an activity sample period for
Volume performance analysis. You can select the option Use all available performance
data or select Use the specified peak activity period. Based on your application peak I/O
behavior, you may specify the sample period with Start date, Duration in days, and Start / End
time.

For the time setting, you can choose the drop-down box:
Device time, or
Client time, or
Server time, or
GMT


After you have entered all the fields, click Next. You will see the Create workload profile -
Review panel as shown in Figure 12-111.

Figure 12-111 Review new workload profile parameters

You can specify a Name for the new workload profile and a Description. You may put in
detailed description that covers:
The application name for which the profile is being created
What application I/O activity is represented by the peak activity sample
When it was created
Who created it (optional)
Any other relevant information your organization requires

In our example, we created profile named New_ITSO_app1_profile. At this point you may click
Finish.

At this point, the TotalStorage Productivity Center for Disk will begin Volume performance
analysis based on the parameters you have provided. This process may take some time
depending upon number of volumes and sampling time period. Hence, be patient. Finally, it
will show the outcome of the analysis.


In our example, we got the results notification message as shown in Figure 12-112. Analysis
yielded that results are not statistically significant, as shown message:

BWN005965E: Analysis results are not significant.

This may indicate that:
There is not enough I/O activity on selected volumes, OR
The time period chosen for sampling is not correct, OR
Correct volumes were not chosen

You have an option to Save or Discard the profile. We decided to save the profile
(Figure 12-113).

Figure 12-112 Results for Create Profile

Upon saving the profile, it is now listed in the Manage workload profile panel as shown in
Figure 12-113.

Figure 12-113 Manage workload profile with new saved profile

– The new profile can now be referenced by VPA for future usage.


13

Center for Data
This chapter introduces you to the TotalStorage Productivity Center for Data and discusses
the available functions.

The information in this chapter provides the information necessary to accomplish the
following tasks:
Discover and monitor storage assets enterprise-wide
Report on enterprise-wide assets, files and filesystems, databases, users, and
applications
Provide alerts (set by the user) on issues such as capacity problems, policy violations, etc.
Support chargebacks by usage or capacity


13.1 TotalStorage Productivity Center for Data overview
This section describes the business purpose of TotalStorage Productivity Center for Data
(Data Manager), its architecture, components, and supported platforms.

13.1.1 Business purpose of TotalStorage Productivity Center for Data
The primary business purpose of TotalStorage Productivity Center for Data is to help the
storage administrator keep data available to applications so the company can produce
revenue.

Through monitoring and reporting, TotalStorage Productivity Center for Data helps the
storage administrator prevent outages in the storage infrastructure. Armed with timely
information, the storage administrator can take action to keep storage and data available to
the application. TotalStorage Productivity Center for Data also helps to make the most
efficient use of storage budgets by allowing administrators to use their existing storage more
efficiently, and more accurately predict future storage growth.

13.1.2 Components of TotalStorage Productivity Center for Data
At a high level, the major components of TotalStorage Productivity Center for Data are:
Server, running on a managing server, with access to a database repository
Agents, running on one or more Managed Devices
Clients (using either a locally installed GUI, or a browser-based Web GUI) which users
and administrators use to perform storage monitoring tasks.

Data Manager Server
The Data Manager Server:
Controls the discovery, reporting, and Alert functions
Stores all data in the central repository
Issues commands to Agents for jobs (either scheduled or ad hoc)
Receives requests from the user interface clients for information, and retrieves the
requested information from the central data repository.
Extends filesystems automatically
Reports on the IBM TotalStorage Enterprise Storage Server (ESS) and can also provide
LUN provisioning

An RDBMS (either locally or remote) manages the repository of data collected from the
Agents, and the reporting and monitoring capabilities defined by the users.

WWW Server
The Web Server is optional, and handles communications to allow remote Web access to the
Server. The WWW Server can run on the same physical server as the Data Manager Server.

Data Agent (on a Managed System)
The Agent runs Probes and Scans, collects storage-related information from the managed
system, and forwards it to the Manager to be stored in the database repository, and acted on
if so defined. An Agent is required for every host system to be monitored, with the exception
of NetWare and NAS devices.


Novell NetWare and NAS devices do not currently support locally installed Agents - they are
managed through an Agent installed on a machine that uses (accesses) the NetWare or NAS
device. The Agent will discover information on the volumes or filesystems that are accessible
to the Agent’s host.

The Agents are quite lightweight. Agents listen for commands from the host, and then
perform a Probe (against the operating system), and/or a Scan (against selected
filesystems). Normal operations might see one scheduled Scan per day or week, plus various
ad hoc Scans. Scans and Probes are discussed later in this chapter.

Clients (direct-connected and Web connected)
Direct-connect Clients have the GUI to the Server installed locally. They communicate
directly to the Manager to perform administration, monitoring, and reporting. The Manager
retrieves information requested by the Clients from the database repository.

Web-connect clients use the WWW Server to access the user interface through a Web
browser. The Java administrative applet is downloaded to the Web Client machine and
presents the same user interface that Direct-connect Clients see.

13.1.3 Security considerations
TotalStorage Productivity Center for Data has two security levels: non-administrative users
and administrators:
Non-administrator users can:
– View the data collected by TotalStorage Productivity Center for Data
Administrators can:
– Create, modify, and schedule Pings, Probes, and Scans
– Perform administrative tasks and customize the TotalStorage Productivity Center for
Data environment
– Create Groups, Profiles, Quotas, and Constraints
– Set Alerts

13.2 Functions of TotalStorage Productivity Center for Data
An overview of the functions of TotalStorage Productivity Center for Data is provided in this
section and explored in detail in the rest of the chapter. TotalStorage Productivity Center for
Data is designed to be easy to use, quick to install, with flexible and powerful configuration.
The main functions of the product are:
Automatically discover and monitor disks, partitions, shared directories, and servers
Reporting to track asset usage and availability
– Physical inventory - disks, partitions, servers
– Logical inventory - filesystems and files, databases and tables
– Forecasting demand versus capacity
– Standardized and customized reports, on-demand and batched
– Various user-defined levels of grouping
– From summary level down to individual file for userID granularity
Alerts - execute scripts, e-mail, SNMP traps, event log
Quotas
Chargeback

Chapter 13. Using TotalStorage Productivity Center for Data 523

13.2.1 Basic menu displays
Figure 13-1 shows the main menu for TotalStorage Productivity Center for Data. You can see
that the Agents configured show under the Agents entry. This display thus shows a quick
summary of the state of each Agent. There are several icons to indicate the status of the
Agents.
Green circle - Agent is communicating with the Server
Red crossed circle - Agent is down.
Red triangle - Agent on that system is not reachable
Red crossed square - Agent was connected, but currently there is an update for
TotalStorage Productivity Center for Data agent running.

Figure 13-1 Agent summary


Figure 13-2 shows the TotalStorage Productivity Center for Data dashboard. This is the
default right-hand pane display when you start TotalStorage Productivity Center for Data and
shows a quick summary of the overall health of the storage environment. It can quickly show
you potential problem areas for further investigation.

Figure 13-2 TotalStorage Productivity Center for Data - dashboard

The dashboard contains four viewable areas, which cycle among seven pre-defined sets of
panels. To cycle, use the Cycle Panels button. Use the Refresh button to update the display.

Enterprise-wide summary
The Enterprise-wide Summary panel shows statistics accumulated from all the Agents. The
statistics are:
Total filesystem capacity available
Total filesystem capacity used
Total filesystem free capacity
Total allocated and unallocated disk space
Total disk space unallocated to filesystems
Total LUN capacity
Total usable LUN capacity
Total number of monitored servers
Total number of unmonitored servers
Total number of storage subsystems
Total number of users
Total number of disks
Total number of LUNs
Total number of filesystems
Total number of directories
Total number of files


Filesystem Used Space
This panel displays a pie chart showing the distribution of used and free space in all
filesystems. Different chart types can be selected here. This provides a quick snapshot of
your filesystem space utilization efficiency.

Users Consuming the Most Space
By default this panel displays a bar chart (different chart types can be selected) of the users
who are using the largest amount of filesystem space.

Monitored Server Summary
This panel shows a table of total disk filesystem capacity for the monitored servers sorted by
OS type.

Filesystems with Least Free Space Percentage
This panel shows a table of the most full filesystems, including the percent of space free, the
total filesystem capacity, and the filesystem mount point.

Users Consuming the Most Space Report
This panel shows the same information as the Users Consuming the Most Space panel, but
in a table format.

Alerts Pending
This panel shows active Alerts that have been triggered but are still pending.

13.2.2 Discover and monitor Agents, disks, filesystems, and databases
TotalStorage Productivity Center for Data uses three methods to discover information about
the assets in the storage environment: Pings, Probes, and Scans. These are typically set up
to run automatically as scheduled tasks. You can define different Ping, Probe, and Scan jobs
to run against different Agents or groups of Agents (for example, to run a regular Probe of all
Windows systems) according to your particular requirements.

Pings
A Ping is a standard ICMP Ping which checks registered Agents for availability. If an Agent
does not respond to a Ping (or a pre-defined number of Pings) you can set up an Alert to take
some action. The actions could be one, any, or all of:
SNMP trap
TEC Event
Notification at login
Entry in the Windows event log
Run a script
Send e-mail to a specified user(s)


Pings are used to generate Availability Reports, which list the percentage of times a computer
has responded to the Ping. An example of an Availability Report for Ping is shown in
Figure 13-3. Availability Reports are discussed in detail in 13.11.3, “Availability Reporting” on
page 604.

Figure 13-3 Availability Report - Ping

Probes
Probes are used to gather information about the assets and system resources of monitored
servers, such as processor count and speed, memory size, disk count and size, filesystems,
etc. The data collected by the Probe process is used in the Asset Reports described in
13.11.1, “Asset Reporting” on page 595.

Figure 13-4 shows an Asset report for detected disks.

Figure 13-4 Asset Report of discovered disks


Figure 13-5 shows an Asset Report for detected database tablespaces.

Figure 13-5 Asset Report of database tablespaces

Scans
The Scan process is used to gather statistics about usage and trends of the server storage.
Data collected by the Scan jobs are tailored by Profiles. Results of Scan jobs are stored in
the enterprise repository. This data supplies the data for the Capacity, Usage, Usage
Violations, and Backup Reporting functions. These reports can be scheduled to run regularly,
or they can be run ad hoc by the administrator.

Profiles limit the scanning according to the parameters specified in the Profile. Profiles are
used in Scan jobs to specify what file patterns will be scanned, what attributes will be
gathered, what summary view will be available in reports and the retention period for the
statistics. TotalStorage Productivity Center for Data supplies a number of default Profiles
which can be used, or additional Profiles can be defined. Table 13-1 on page 547 shows the
default Profiles provided. Some of these include:
Largest files - Gathers statistics on the largest files
Largest directories - Gathers statistics on the largest directories
Most at risk - Gathers statistics on the files that have been modified the longest time ago
and have not been backed up since modified (Windows Agents only)


Figure 13-6 shows a sample of a report produced from data collected in Scans.

Figure 13-6 Summary View - by filesystem, disk space used and disk space free

This report shows a list of the filesystems on each Agent, the amount of space used in each,
expressed in bytes and as a percentage, the amount of free space, and the total capacity
available in the filesystem.

13.2.3 Reporting
Reporting in TotalStorage Productivity Center for Data is very powerful, with over 300
pre-defined views, and the capability to customize those standard views, save the custom
report, and add it to your menu for scheduled or ad hoc reports. You can also create your own
individual reports according to particular needs and set them to run as needed, or in batch
(regularly). Reports can be produced in table format for a variety of charting (graph) views.
You can export reports to CSV or HTML formats for external usage.

Reports are generated against data already in the repository. A common practice is to
schedule Scans and Probes just before running reports.

Reporting can be done at almost any level in the system, from the enterprise down to a
specific entity and any level in between. Figure 13-6 shows a high-level summary report. Or,
you can drill down to something very specific. Figure 13-7 is an example of a lower-level
report, where the administrator has focused on a particular Agent, KANAGA, to look at a
particular disk on a particular controller.


Figure 13-7 Asset Report - KANAGA assets

Reports can be produced either system-wide or grouped into views, such as by computer, or
OS type.

Restriction: Currently, there is a maximum of 32,767 (216 -1) rows per report. Therefore,
you cannot produce a report to list all the .HTM files in a directory containing a million files.
However, you can (and it would be more productive to do so) produce a report of the 20
largest files in the directory, or the 20 oldest files, for example.

TotalStorage Productivity Center for Data allows you to group information about similar
entities (disk, filesystems, etc.) from different servers or business units into a summary report,
so that business and technology administrators can manage an enterprise infrastructure. Or,
you can summarize information from a specific server - the flexibility and choice of
configuration is entirely up to the administrator.

You can report as at a point in time, or produce a historical report, showing storage growth
trends over time. Reporting lets you track actual demand for disk over time, and then use this
information to forecast future demand for the next quarter, two quarters, year, etc. Figure 13-8
is an example of a historical report, showing a graph of the number of files on the C drive on
the Agent KANAGA.


Figure 13-8 Historical report of filesystem utilization

TotalStorage Productivity Center for Data has three basic types of reports:
Computers and filesystems
Databases
Chargeback

Reporting categories
Major reporting categories for filesystems and databases are:
Asset Reporting uses the data collected Probes to build a hardware inventory of the
storage assets. You can then navigate through a hierarchical view of the assets by drilling
down through computers, controllers, disks, filesystems, directories, and exports. For
database reporting, information on instances, databases, tables, and data files is
presented for reporting.
Storage Subsystems Reporting provides information showing storage capacity at a
computer, filesystem, storage subsystem, LUN, and disk level. These reports also enable
you to view the relationships among the components of a storage subsystem. For a list of
supported devices, see <table>.
Availability Reporting shows responses to Ping jobs, as well as computer uptime.
Capacity Reporting shows how much storage capacity is installed, how much of the
installed capacity is being used, and how much is available for future growth. Reporting is
done by disk and filesystem, and for databases, by database.
Usage Reporting shows the usage and growth of storage consumption, grouped by
filesystem, and computers, individual users, or enterprise-wide.
Usage Violation Reporting shows violations to the corporate storage usage policies, as
defined through TotalStorage Productivity Center for Data. Violations are either of Quota
(defining how much storage a user or group of users is allowed) or Constraint (defining
which file types, owners and file sizes are allowed on a computer or storage entity). You
can define what action should be taken when a violation is detected - for example, SNMP
trap, e-mail, or running a user-written script.
Backup Reporting identifies files which are at risk because they have not been backed up.


Reporting on the Web
It is easy to customize Tivoli Storage Resource Manager to set up a reports Web site, so that
anyone in the organization can view selected reports through their browser. 13.16, “Setting up
a reports Web site” on page 698 explains how to do this. Figure 13-9 <change> shows an
example of a simple Web site to view TotalStorage Productivity Center for Data reports.

Figure 13-9 TotalStorage Productivity Center for Data Reports on the Web

13.2.4 Alerts
An Alert defines an action to be performed if a particular event occurs or condition is found.
Alerts can be set on physical objects (computers and disks) or a logical objects (filesystems,
directories, users, databases, and OS user groups). Alerts can tell you, for instance, if a disk
has a lot of recent defects, or if a filesystem or database is approaching capacity.

Alerts on computers and disks come from the output of Probe jobs and are generated for
each object that meets the triggering condition. If you have specified a triggered action
(running a script, sending an e-mail, etc.) then that action will be performed if the condition is
met.

Alerts on filesystems, directories, users, and OS user groups come from the combined output
of a Probe and a Scan. Again, if you have specified an action, that action will be performed if
the condition is met.

An Alert will register in the Alert log, plus you can also define one, some, or all of the following
actions to be performed in addition:
Send an e-mail indicating the nature of the Alert.
Run a specific script with relevant parameters supplied from the content of the Alert.
Make an entry into the Windows event log.
Pop up next time the user logs in to TotalStorage Productivity Center for Data.
Send an SNMP trap.
Log a TEC event

Refer to 13.4, “OS Alerts” on page 555 for details on alerts.


13.2.5 Chargeback: Charging for storage usage
TotalStorage Productivity Center for Data provides the ability to produce Chargeback
information for storage usage. The following items can have charges allocated against them:
Operating system storage by user
Operating system disk capacity by computer
Storage usage by database user
Total size by database tablespace

TotalStorage Productivity Center for Data can directly produce an invoice or create a file in
CIMS format. CIMS is a set of resource accounting tools that allow you to track, manage,
allocate, and charge for IT resources and costs. For more information on CIMS see the Web
site:
http://www.cims.com

Chargeback is a very powerful tool for raising the awareness within the organization of the
cost of storage, and the need to have the appropriate tools and processes in place to manage
storage effectively and efficiently.

Refer to 13.17, “Charging for storage usage” on page 700 for more details on Chargebacks.

13.3 OS Monitoring
The Monitoring features of TotalStorage Productivity Center for Data enable you to run
regularly scheduled or ad hoc data collection jobs. These jobs gather statistics about the
storage assets and their availability and their usage within your enterprise, and make the
collected data available for reporting.

This section gives a quick overview of the monitoring jobs, and explains how they work
through practical examples.

Reporting on the collected data is explained in “Data Manager reporting capabilities” on
page 592.


13.3.1 Navigation tree
Figure 13-10 shows the complete navigation tree for OS Monitoring which includes Groups,
Discovery, Pings, Probes, Scans, and Profiles.

Figure 13-10 OS Monitoring tree

Except for Discovery, you can create multiple definitions for each of those monitoring features
of TotalStorage Productivity Center for Data. To create a new definition, right-click the feature
and select Create <feature>. Figure 13-11 shows how to create a new Scan job.

Figure 13-11 Create Scan job creation


Once saved, any definition within TotalStorage Productivity Center for Data can be updated
by clicking the object. This will put you in Edit mode. Save your changes by clicking the floppy
disk icon in the top menu bar.

Discovery, Pings, Probes, and Scan menus contain jobs that can run on a scheduled basis or
ad hoc. To execute a job immediately, right-click the job then select Run Now (see
Figure 13-12). Each execution of a job creates a time-stamped output that can be displayed
by expanding the tree under the job (you may need to right-click the job and select Refresh
Job List).

Figure 13-12 OS Monitoring - Jobs list

The color of the job output represents the job status:
Green - Successful run
Brown - Warnings occurred during the run
Red - Errors occurred during the run
Blue - Running jobs

To view the output of a job, double click the job.

Groups and Profiles are definitions that may be used by other jobs - they do not produce an
output in themselves.

As shown in Figure 13-12, all objects created within Data Manager are prefixed with the user
ID of the creator. Default definitions, created during product installation, are prefixed with
TPCUser.Default.

Groups, Discovery, Probes, Scans, and Profiles are explained in the following sections.

13.3.2 Groups
Before defining monitoring and management jobs, it may be useful to group your resources
so you can limit the scope of monitoring or data collection.


Computer Groups
Computer Groups allow you to target management jobs on specific computers based on your
own criteria. Some criteria you might consider for grouping computers are platform type,
application type, database type, and environment type (for example, test or production).

Our lab environment contains Windows 2000 servers. In order to target specific servers for
monitoring based on OS and/or database type, we will defined the following groups:
Windows Systems
Windows DB Systems

To create the first group, expand Data Manager → Monitoring → Groups → Computer,
right-click Computer and select Create Computer Group. Our first group will contain all
Windows systems as shown in Figure 13-13. To add or remove a host from the group,
highlight it in either the Available or Current Selections panel and use the arrow buttons. You
can also enter a meaningful description in the field.

Figure 13-13 Computer Group definition

To save the new Group, click the floppy disk icon in the menu bar, and enter the Group name
in the confirmation box shown in Figure 13-14.

Figure 13-14 Save a new Computer Group


We created the other group using the same process, and named it Windows DB Systems.

Important: To avoid redundant data collection, a computer can belong to only one Group
at a time. If you add a system which is already in a Group, to a second Group, it will
automatically be removed from the first Group.

Figure 13-15 shows the final Group configuration, with the members of the Windows Systems
group.

Figure 13-15 Final Computers Group definitions

Note: The default group TPCUser.Default Computer Group contains all servers that have
been discovered, but not yet assigned to a Group.

Filesystem Groups
Filesystem Groups are used to associate together filesystems from different computers that
have some commonality. You can then use this group definition to focus the Scan and the
Alert processes to those filesystems.

To create a Filesystem Group, you have to select explicitly each filesystem for each computer
you want to include in the group. There is no way to do a grouped selection, e.g. / (root)
filesystem for all UNIX servers or C: for all Windows platforms.

Note: As for computers, a filesystem can belong to only one Group.


Directory Groups
Use Directory Groups to group together directories to which you want to apply the same
storage management rules.

Figure 13-16 shows the Directory Group definition screen by going to Data Manager →
Monitoring → Groups → Directory and right-clicking Directory and selecting Create
Directory Group.

Figure 13-16 Directory group definition

The Directory Group definition has two views for directory selection:
Use directories by computer to specify several directories for one computer.
Use computers by directory to specify one directory for several computers.

The button on the bottom of the screen toggles between New computer and New directory
depending on the view you select.


We will define one Directory Group with a DB2 directory for a specific computer (Colorado).
To define the Group:
1. Select directories by computer.
2. Click New computer.
3. Select colorado from the pull-down Computer field.
4. Enter C:DB2NODE0000 in the Directories field and click Add (see Figure 13-17).

Figure 13-17 Directories for computer configuration

5. Click OK.
6. Save the group as DB2 Node.

Figure 13-18 shows our final Groups configuration and details of the OracleArchive Group.

Figure 13-18 Final Directories Group definition


User Groups
You can define Groups made up of selected user IDs. These groupings will enable you to
easily define and focus storage management rules such as scanning and Constraints on the
defined IDs.

Note: You can include in a User Group only user IDs defined on the discovered hosts,
which have files belonging to them.

Note: As with computers, a user can be defined in only one Group.

OS User Group Groups
You can define Groups consisting of operating system user groups such as Administrators
for Windows or adm for UNIX. To define a Group consisting of user groups, select OS User
Group from the Groups entry on the left hand panel.

Note: As for users, an OS User Group will be added to the list of available Groups only
when a Scan job finds at least one file owned by a user belonging to that Group.

Note: As with users, an OS User Group can belong to only one Group at a time.

13.3.3 Discovery
The Discovery process is used to discover new computers within your enterprise that have
not yet been monitored by Data Manager. The discovery process will:
Request a list of Windows systems from the Windows Domain Controller
Contact, through SNMP, all NAS filers and check if they are registered in the nas.config
file
Discover all NetWare servers in the NetWare trees reported by Agents
Search UNIX Agents’ mount tables, looking for remote filesystems and discover NAS filers

More details of NAS and NetWare discovery are given in the manual IBM Tivoli Storage
Resource Manager: A Practical Introduction, SG24-6886.

Use the path Data Manager → Monitoring → Discovery to change the settings of the
Discovery job. The following options are available.

When to run tab
The initial tab, When to Run (Figure 13-19) is used to modify the scheduling settings. You can
specify to execute the Discovery:
Now - Run once when the job is saved.
Once - at a specified time in the future
Repeatedly - Choose the frequency in minutes, hours, days, weeks, or months. You can
limit the run to specific days of the week.


Figure 13-19 Discovery When to Run options

Alert tab
The second tab, Alert, enables you to be notified when a new computer is discovered. See
13.4, “OS Alerts” on page 555 for more details on the Alerting process.

Options tab
The third tab, Options (Figure 13-20) sets the discovery runtime properties.

Figure 13-20 Discovery job options


Uncheck the Skip Workstations field if you want to discover the Windows workstations
reported by the Windows Domain Controller.

13.3.4 Pings
The Ping process will:
launch TCP/IP pings against monitored computers
generate statistics on computer Availability in the central repository
generate an Alert if the process fails because of an unavailable host
summarizes the Ping process.

Pings gather statistics about the availability of monitored servers. The scheduled job will Ping
your servers and consider them active if it gets an answer. This is purely ICMP-protocol
based - there is no measurement of individual application availability. When you create a new
Ping job, you can set the following options.

Computers tab
Figure 13-21 shows the Computers tab, which is used to limit the scope of the computers that
are to be Pinged.

Figure 13-21 Ping job configuration - Computers


When to Ping tab
The tab, When to PING, sets the frequency used for checking. We selected a frequency of 10
minutes as shown in Figure 13-22 on page 543.

Figure 13-22 Ping job configuration - When to Ping

Options tab
On the Options tab, you specify how often the Ping statistics are saved in the database
repository. By default, TotalStorage Productivity Center for Data keeps its Ping statistics in
memory for eight Pings before flushing them to the database and calculating an average
availability. You can change the flushing interval to another time amount, or a number of
Pings (for example, to calculate availability after every 10 Pings). The system availability is
calculated as:
(Count of successful pings) / (Count of pings)

A lower interval can increase database size, but gives you more accuracy on the availability
history.

We selected to save to the database at each Ping (), which means we will have an availability
of 100% or of 0%, but we have a more granular view of the availability of our servers
(Figure 13-23).


Figure 13-23 Ping job configuration - Options

Alert tab
The Alert tab (shown in Figure 13-24) is used to generate Alerts for each host that is
unavailable. Alert mechanisms are explained in more detail in 13.4, “OS Alerts” on page 555.

You can choose any Alert type from the following:
SNMP trap to send a trap to the Event manager defined in Administrative services →
Configuration →General → Alert Disposition
TEC Event to send an event to a Tivoli Enterprise Console
Login Notification to direct the Alert to the specified user in the Alert Log (see 13.4, “OS
Alerts” on page 555)
Windows Event Log to generate an event to the Windows event log
Run Script to run a script on the specified server
Email to send a mail to the specified user through the Mail server defined in
Administrative services → Configuration → General → Alert Disposition


Figure 13-24 Ping job configuration - Alert

We selected to run a script that will send popup messages to selected administrators. The
script is listed in Example 13-1. Optimally, you would send an event to a central console such
as the Tivoli Enterprise Console. Note that certain parameters are passed to the script - more
information is given in “Alerts tab” on page 560.

Example 13-1 Script PINGFAILED.BAT
net send /DOMAIN:Colorado Computer %1 did not respond to last %2 ping(s). Please check it

We then saved the Ping job as PingHosts, and tested it by right-clicking and selecting Run
now. As the hosts did not respond, we received notifications as shown in Figure 13-25.

Figure 13-25 Ping failed popup for GALLIUM

More details about the related reporting features of TotalStorage Productivity Center for
DataTotalStorage Productivity Center for Data are in 13.11.3, “Availability Reporting” on
page 604.

13.3.5 Probes
The Probe process will:
Gather Assets data on monitored computers
Store data in the central repository
Generate an Alert if the process fails


The Probe process gathers data about the assets and system resources of Agents such as:
Memory size
Processor count and speed
Hard disks
Filesystems

The data collected by the Probe process is used by the Asset Reports described in 13.11.1,
“Asset Reporting” on page 595.

Computers tab
Figure 13-26 shows that we included the TPCUser.Default Computer Group in the Probe so
that all computers, including those not yet assigned to an existing Group, will be Probed. We
saved the Probe as ProbeHosts.

Figure 13-26 New Probe configuration

Important: Only the filesystems that have been returned by a Probe job will be available
for further use by Scan, Alerts, and policy management within TotalStorage Productivity
Center for Data.

When to Probe tab
This tab has the same configuration as for the Ping process.

We set up a weekly Probe to run on Sunday for all computers. We recommend running the
Probe job at a time where all the production data you want to monitor is available to the
system.

Alert tab
As this is not a business-critical process, we asked to be alerted by mail for any failed Probe.
Figure 13-27 shows the default mail text configuration for a Probe failure.


Figure 13-27 Probe alert - mail configuration

13.3.6 Profiles
Profiles are used:
In Scan jobs
To limit files to be scanned
To specify files tabulates to be scanned
To select the summary view
Directories and filesystems
User ids
OS user groups
To set statistics retention period

TotalStorage Productivity Center for Disk provides default profiles that provide data for all the
default reports.

Profiles are used in Scan jobs to specify:
The pattern of files to be scanned
The attributes of files to be gathered
The summary view that will be available in reports
The statistics retention period

Specifying correct profiles avoids gathering unnecessary information that may lead to space
problems within the repository. However, you will not be able to report on or check Quotas on
files that are not used by the Profile.

Data Manager comes with several default profiles, (shown in Table 13-1) prefixed with
TPCUser, which can be reused in any Scan jobs you define.
Table 13-1 Default profile

Default profile name Description

BY_ACCESS Gathers statistics by length of time since last access of files

BY_CREATION Gathers statistics by length of time since creation of files


Default profile name Description

BY_MOD_NOT_BACKED_UP Gathers statistics by length of time since last modification (only for
files not backed up since modification). Windows only

BY_MODIFICATION Gathers statistics by length of time since last modification of files

FILE_SIZE_DISTRIBUTION Gathers file size distribution statistics

LARGEST_DIRECTORIES Gathers statistics on the n largest directories. (20 is the default
amount.)

LARGEST_FILES Gathers statistics on the n largest files. (20 is the default amount.)

LARGEST_ORPHANS Gathers statistics on the n largest orphan files. (20 is the default
amount.)

MOST_AT_RISK Gathers statistics on the n files that have been modified the longest
time ago and have not yet been backed up since they were
modified. Windows only. (20 is the default amount.)

OLDEST_ORPHANS Gathers statistics on the n oldest orphan files. (20 is the default
amount.)

MOST_OBSOLETE_FILES Gathers statistics on the n “most obsolete” files (i.e., files that have
not been accessed or modified for the longest period of time). (20
is the default amount.)

SUMMARY_BY_FILE_TYPE Summarizes space usage by file extension

SUMMARY_BY_FILESYSTEM Summarizes space usage by Filesystem or Directory
/DIRECTORY

SUMMARY_BY_GROUP Summarizes space usage by OS Group

SUMMARY_BY_OWNER Summarizes space usage by Owner

TEMPORARY_FILES Gathers statistics on network-wide space consumed by temporary
files

WASTED_SPACE Gathers statistics on non-OS files not accessed in the last year and
orphaned files

Those default profiles, when set in a Scan job, gather data needed for all the default Data
Manager reports.

As an example, we will define an additional Profile to limit a Scan job to the 500 largest
Postscript or PDF files unused in the last six months. We also want to keep weekly statistics
at a filesystem and directory level for two weeks.

Statistics tab
On the Statistics tab (shown in Figure 13-28), we specified:
Retain filesystem summary for two weeks
Gather data based on creation data
Select the 500 largest files

The Statistics tab is used to specify the type of data that is gathered, and has a direct
impact on the type of reports that will be available. In our specific case, the Scan associated
with this profile will not create data for reports based on user IDs and users groups. Neither
will it create data for reports on directory size.


Figure 13-28 New Profile - Statistics tab

The Summarize space usage by section of the Statistics tab specifies how the space usage
data must be summarized. If no summary level is checked, the data will not be summarized,
and therefore will not be available for reporting in the corresponding level of Usage Reporting
section of TotalStorage Productivity Center for Data.

In our particular case, because we select to summarize by filesystem and directory, we will
see space used by PDF and Postscript files at those levels, providing we set up the Scan
profile correctly. See 13.3.7, “Scans” on page 552 for information on this. We will not see
which users or groups have allocated those PDF and Postscript files.

Restriction: For Windows servers, users and groups statistics will not be created for FAT
filesystems.

The Accumulate history section sets the retention period of the collected data. In this case,
we will see a weekly summary for the last two weeks.

The Gather statistics by length of time since section sets the base date used to calculate the
file load. It determines if data will be gathered and summarized for the Data Manager →
Reporting → Usage → Files reporting view.

The Gather information on the section sets the amount of files to retrieve for each of the
report views available under Data Manager → Reporting → Usage → Access Load.

Files filter tab
The Files filter tab is used to limit the scope of files that are returned by the Scan job. To
create a selection, right-click the All files selected context-menu option as shown in
Figure 13-29.


Figure 13-29 New Profile - File filter

With the New Condition menu, you can create a single filter on the files while the New
Group enables you to combine several conditions with:
All of The file is selected if all conditions are met (AND)
Any of The file is selected if at least one condition is met (OR)
None of The file is NOT selected if at least one condition is met (NOT OR)
Not all of The file is selected if none of the conditions are met (NOT AND)

The Condition Group can contain individual conditions or other condition groups.

Each individual condition will filter files based on one of the listed items:
Name
Last access time
Last modified
Creation time
Owner user ID
Owner group
Windows files attributes
Size
Type
Length

We want to select files that meet our conditions: (name is *.ps or name is *.pdf) and
unused since six months. The AND between our two conditions will be translated to All of,
while the OR within our first condition will be translated to Any of.

On the screen shown in Figure 13-29, we selected New Group. From the popup screen,
Figure 13-30, we selected All of and clicked OK.


Figure 13-30 New Condition Group

Now, within our All of group we will create one dependant Any of group using the same
sequence. The result is shown in Figure 13-31.

Figure 13-31 New Profile - Conditions Groups

Now, we create individual conditions within each group by right-clicking New Condition on
the group where the conditions must be created. Figure 13-32 shows the creation of our first
condition for the Any of group. We enter in our file specifications (*.ps and *.pdf) here.

Figure 13-32 New Profile - New condition


We repeated the operation for the second condition (All of). The final result is shown in
Figure 13-33.

Figure 13-33 New Profile - Conditions

The bottom of the right pane shows the textual form of the created condition. You can see that
it corresponds to our initial condition. We saved the profile as PS_PDF_FILES (Figure 13-34).

Figure 13-34 Profile save

13.3.7 Scans
The Scan process is used to gather data about files and to summarize Usage statistics as
specified in the associated profiles. It is mandatory for Quotas and Constraints management.
The Scan process gathers statistics about the usage and trends of the server storage. Scan
job results are stored in the repository and supply the data necessary for the Capacity,
Usage, Usage Violations, and Backup Reporting facilities. To create a new Scan job, Data
Manager → Monitoring → Scans, right-click and select Create Scan. The scope of each
Scan job is set by five different tabs on the right pane.

Filesystems tab
You can specify a specific filesystem for one computer, a filesystem Group (see “Filesystem
Groups” on page 537) or all filesystems for a specific computer. Only the filesystems you
have selected will be scanned. Figure 13-35 shows how to configure the Scan to gather data
on all our servers.

Note: Only filesystems found by the Probe process will be available for Scan.


Figure 13-35 New Scan configuration - Filesystem tab

Directory Groups tab
Use this tab to extend the scope of the Scan and also summarize data for the selected
directories. Only directories in the previously selected filesystems will be scanned.

Profiles tab
As explained in 13.3.6, “Profiles” on page 547, the Profiles are used to select the files that are
scanned for information gathering. A Scan job scans and gathers data only for files that are
scoped by selected Profiles. You can specify Profiles at two levels:
Filesystems: All selected filesystems will be scanned and data summarized for each
filesystem.
Directory: All selected directories (if included in the filesystem) will be scanned and data
summarized for each directory.


Figure 13-36 shows how to configure a Scan to have data summarized at both the filesystem
and directory level.

.

Figure 13-36 New Scan configuration - Profiles tab

When to SCAN tab
As with the Probe and Ping jobs, the scheduling of the job is specified on the When to Scan
tab.

Alert tab
You can be alerted through mail, script, Windows Event Log, SNMP trap, TEC event, or Login
notification if the Scan job fails. The Scan job may fail if an Agent is unreachable.

Click the floppy icon to save your new Scan job, shown in Figure 13-37.

Figure 13-37 New Scan - Save


Putting it all together
Table 13-2 summarizes the reports views for filesystems and directories that will be available
depending on the settings of the Profiles and the Scan jobs. We assume the Profiles have
been defined with the Summarize space by Filesystem/Directory option. Note that in order
to get reports by filesystem or directory, you need to select either or both in the Scan Profile.

Table 13-2 Profiles/Scans versus Reports
Scan Jobs settings Available reports

Filesystem Directory Filesystem Directory What is scanned By Filesystem By Directory
/Computer profile profile Reports Reports

x - - - FS - -

x x - - FS - -
Dir if in specified FS

x x x - FS x -

x x x x FS x x

x x x FS x
Dir scanned if in
specified FS

x - x x FS x -

x - - x FS - -

13.4 OS Alerts
TotalStorage Productivity Center for Data enables you to define Alerts on computers,
filesystems, and directories. Once the Alerts are defined, it will monitor the results of the
Probe and Scan jobs, and will trigger an Alert when the threshold or the condition is met.

TotalStorage Productivity Center for Data provides a number options for Alert mechanisms
from which you can choose depending on the severity you assign to the Alert.

Depending on the severity of the triggered event or the functions available in your
environment, you may want to be alerted with:


An SNMP trap to an event manager. Figure 13-38 shows a Filesystem space low Alert as
displayed in our SNMP application, IBM Tivoli NetView. Defining the event manager is
explained in 8.5, “Alert Disposition” on page 316.

Figure 13-38 Alert - SNMP trap sample

A TEC (Tivoli Enterprise Console) event.
An entry in the Alert Log (see Figure 13-39). You can configure Data Manager, so that the
Alert Log will be automatically displayed when you log on to the GUI by using Preferences
→ Edit General (see Figure 13-40).

Figure 13-39 Alert - Logged alerts sample


Figure 13-40 Alert - Preferences

An entry in the Windows Event log, as shown in Figure 13-41. This is useful for lower
severity alerts or when you are monitoring your Windows event logs with an automated
tool such as IBM Tivoli Distributed Monitoring.

Figure 13-41 Alerts - Windows Event Viewer sample

Running a specified script - The script runs on the specified computer with the authority of
the Agent (root or Administrator). See 13.5.5, “Scheduled Actions” on page 582 for special
considerations with scripts execution.
An e-mail - TotalStorage Productivity Center for Data must be configured with a valid
SMTP server and port as explained in 8.5, “Alert Disposition” on page 316.


13.4.1 Alerting navigation tree
Figure 13-42 shows the complete navigation tree for OS Alerting which includes Computer
Alerts, Filesystem Alerts, Directory Alerts, and Alert Log.

Figure 13-42 OS Alerting tree


Except for the Alert Log, you can create multiple definitions for each of those Alert features of
TotalStorage Productivity Center for Data. To create a new definition, right-click the feature
and select Create <feature>. Figure 13-43 shows how to create a new Filesystem Alert.

Figure 13-43 Filesystem alert creation


13.4.2 Computer Alerts
Computer Alerts act on the output of Probe jobs (see 13.3.5, “Probes” on page 545) and
generate an Alert for each computer that meets the triggering condition. Figure 13-44 shows
the configuration screen for a Computer Alert.

Figure 13-44 Computer alerts - Alerts

Alerts tab
The Alerts tab contains two parts:
Triggering condition to specify the computer component you want to be monitored. You
can monitor a computer for:
– RAM increased
– RAM decreased
– Virtual Memory increased
– Virtual Memory decreased
– New disk detected
– Disk not found
– New disk defect found
– Total disk defects exceed. You will have to specify a threshold.
– Disk failure predicted
– New filesystem detected
Information about disk failures is gathered through commands against disks with the
following exceptions:
– IDE disks do support only Disk failure predicted queries
– AIX SCSI disks do not support failures and predicted failures queries
Triggered action where you specify the action that must be executed. If you choose to run
a script, it will receive several positional parameters that depends on the triggering
condition. The parameters display on the Specify Script panel, which is accessed by
checking Run Script and clicking the Define button.


Figure 13-45 shows the parameters passed to the script for a RAM decreased condition.

Figure 13-45 Computer alerts - RAM decreased script parameters

Figure 13-46 shows the parameters passed to the script for a Disk not found condition.

Figure 13-46 Computer alerts - Disk not found script parameters

Computers tab
This limits the Alert process to specific computers or computer Groups (Figure 13-47).

Figure 13-47 Computer alerts - Computers tab


13.4.3 Filesystem Alerts
Filesystem Alerts will act on the output of Probe and Scan jobs and generate an Alert for each
filesystem that meets the specified threshold. Figure 13-48 shows the configuration screen for
a Filesystem Alert.

Figure 13-48 Filesystem Alerts - Alert

Alerts tab
As for Computer Alerts, the Alerts tab contains two parts. In the Triggering condition section
you can specify to be alerted if a:
Filesystem is not found, which means the filesystem was not mounted during the most
recent Probe or Scan.
Filesystem is reconfigured.
Filesystem free space is less than a threshold specified in percent, KB, MB, or GB.
Free UNIX filesystem inode count is less than a threshold (either percent or inodes count).


You can choose to run a script (click the Define button next to Run Script), or you can also
change the content of the default generated mail by clicking Edit Email. You will see a popup
with the default mail skeleton which is editable. Figure 13-49 shows the default e-mail
message.

Figure 13-49 Filesystem alert - Freespace default mail

13.4.4 Directory Alerts
Directory Alerts will act on the output of Scan jobs.

Alerts tab
Directory Alerts configuration is similar to Filesystem alerts. The supported triggers are:
Directory not found
Directory consumes more than the specified threshold set in percent, KB, MB or GB.

Directories tab
Since Probe jobs do not report on directories and Scan jobs report only on directories. if a
directory Profile has been assigned (See “Putting it all together” on page 555) you can only
choose to be alerted for any directory that has already been included in a Scan and actually
scanned.


13.4.5 Alert logs
The Data Manager → Alerting → Alert log menu (Figure 13-50) lists all Alerts that have
been generated.

Figure 13-50 Alerts log

There are nine different views. Each of them will show only the Alerts related to the selected
view except:
All view - Shows all Alerts
Alerts Directed to <logged user> - Shows all Alerts where the current logged user has
been specified in the Login notification field

When you click the icon on the left of a listed Alert, you will see detailed information on the
selected Alert as shown in Figure 13-51.


Figure 13-51 Detailed Alert information

13.5 Policy management
The Policy Management functions of Data Manager enable you to:
Define space limits (Quotas) on storage resources used by user IDs and user groups.
These limits can be set at a network (whole environment), computer, and filesystem level.
To define space limits (Quotas) on NAS resources used by user IDs and user groups.
To perform checks (Constraints) on specific files owned by the users and perform any
action on those files.
Define a filesystem extension policy can be used to automatically increase filesystem
capacity for managed hosts when utilization reaches a specified level. The LUN
provisioning option can be enabled to extend filesystems within an ESS.
To schedule scripts against your storage resources.

13.5.1 Quotas
Quotas can be set at either a user or at an OS User Group level. For the OS User Group
level, this could be either an OS User Group, (see “OS User Group Groups” on page 540), or
a standard OS group (such as system on UNIX, or Administrators on Windows). The User
Quotas trigger an action when one of the monitored users has reached the limit while the OS
User group Quotas trigger the action when the sum of space used by all users of monitored
groups has reached the limit. The Quotas definition mechanism is the same for both except
for the following differences:
The menu tree to use:
– Data Manager → Policy Management → Quotas → User
– Data Manager → Policy Management → Quotas → OS User group


The monitored elements you can specify:
– User and user groups for User Quotas
– OS User Group and OS User Group Groups for OS User Group Quota

We will show how to configure User Quotas. User Group Quotas are configured similarly.

Note that the Quota enforcement is soft - that is, users are not automatically prevented from
exceeding their defined Quota, but the defined actions will trigger if that happens. There are
three sub-entries for Quotas: Network Quotas, Computer Quotas, and Filesystem Quotas

Network Quotas
A Network Quota defines the maximum cumulated space a user can occupy on all the
scanned servers. An Alert will be triggered for each user that exceeds the limit specified in the
Quota definition.

Use Data Manager → Policy Management → Quotas → User → Network, right-click
and select Create Quota to create a new Quota. The right pane displays the Quota
configuration screen with four tabs.

Users tab
Figure 13-52 shows the Users tab for Network Quotas.

Figure 13-52 User Network Quotas - Users tab

From the Available column, select any user ID or OS User Group you want to monitor for
space usage.

The Profile pull-down menu is used to specify the file types that will be subject to the Quota.
The list will display all Profiles that create summaries by user (by file owner). Select the
Profile you want to use from the pull-down. The default Profile Summary by Owner collects
information about all files and summarizes them on the user level. The ALLGIFFILES profile
collects information about GIF files and creates a summary at a user level as displayed in
Figure 13-53. This (non-default) profile was created using the process shown in 13.3.6,
“Profiles” on page 547.


Figure 13-53 Profile with user summary

Using this profile option, we can define general Quotas for all files and more restrictive
Quotas for some multimedia files such as GIF and MP3.

Filesystem tab
On the Filesystem tab shown Figure 13-54, select the filesystems or computers you want to
be included in the space usage for Quota management.

Figure 13-54 User Network Quotas - Filesystem tab

In this configuration, for each user, his cumulated space usage on all servers will be
calculated and checked against the Quota limit.


When to check
The Quota management is based on the output of the Scan jobs. Therefore, each Quota
definition must be scheduled to run after the Scan jobs that collect the adequate information.

The When to CHECK tab is standard, and allows you to define a one off or a recurring job.

Alert tab
On the Alert tab, specify the Quota limit in: KB, MB or GB, and the action to run when the
Quota is exceeded.

Figure 13-55 User Network Quotas - Alert tab

You can choose from the standard Alerts type available with TotalStorage Productivity Center
for Data. Each Alert will be fired once for each user exceeding their Quota. We have selected
to run a script that we wrote, QUOTAUSERNET.BAT, listed in Example 13-2.

Example 13-2 QUOTAUSERNET.BAT script
echo NETWORK quota exceeded - %1 %2 uses %3 - Limit set to %4 >>quotausernet.txt

Example 13-3 shows the output file created by QUOTAUSERNET.BAT.

Example 13-3 Content of quotausernet.txt
NETWORK quota exceeded - user root uses 11.16GB - Limit set to 5.0GB
NETWORK quota exceeded - user Administrators@BUILTIN uses 11.97GB - Limit set to 5GB

The Alert has fired for user root and Administrators. This clearly shows that administrative
users such as root and Administrators should not normally be included in standard Quotas
monitoring.


Computer Quotas
Computer Quotas enable you to fire Alerts when a user exceeds their space Quota on a
specific computer as shown in Figure 13-56. Multiple Alerts are generated if a user violates
the Quota on separate computers.

Figure 13-56 Computer Quota - Alerts log

Filesystem Quotas
A Filesystem Quota defines a space usage limit at the filesystem level. An Alert will be fired
for each filesystem where a user exceeds the limit specified in the Quota definition.

Use Data Manager → Policy Management → Quota → User → Filesystem, right-click,
and select Create Quota to create a new Quota. After setting up and running a Quota for
selected filesystems, we received the following entries in the Alert History, shown in
Figure 13-57.

Figure 13-57 Filesystem Quota - Alerts log


13.5.2 Network Appliance Quotas
Using Data Manager → Policy Management → Network Appliance Quotas →
Schedules, you can compare the space used by users against Quotas defined inside
Network Appliance filers, using the appropriate software, and raise an Alert whenever a user
is close to reaching the NetApp Quota.

When you run a Network Appliance Quota job, the NetApp Quota definitions will be imported
into TotalStorage Productivity Center for Data for read-only purposes.

Note: Network Appliance Quotas jobs must be scheduled after the Scan jobs, since they
use the statistics gathered by the latest Scan to trigger any NetApp Quota violation.

With Data Manager → Policy Management → Network Appliance Quotas → Imported
User Quotas and Imported OS User Group Quotas, you can view the definitions of the
Quotas defined on your NetApp filers.

13.5.3 Constraints
The main features of Constraints are listed in Figure 13-58.

Constraints
Reports and triggers actions based on specific files which use
too much space on monitored servers

Files can be selected based on
server and filesystem
name pattern (eg: *.mp3, *.avi)
owner
age
size
attributes

Actions triggered through standard Alerting mechanism when
total space used by files exceeds a threshold

ibm.com/redbooks

Figure 13-58 Constraints

Constraints are used to generate Alerts when files matching specified criteria are consuming
too much space on the monitored servers.


Constraints provide a deeper level of Data Management. Quotas will allow reporting on users
who have exceeded their space limitations. With Constraints, we can get more detailed
information to specify limits on particular file types or other attributes, such as owner, age,
and so on. The output of a Constraint when applied to a Scan will return a list of the files that
are consuming too much space.

Note: Unlike Quotas, Constraints are automatically checked during Scan jobs and do not
need to be scheduled. Also, the Scan does not need to be associated with Profiles that will
cause data to be stored for reporting.

Filesystems tab
The Filesystems tab helps you select the computers and filesystems you want to check for
the current Constraint. The selection method for computers and filesystems is the same as
for Scan jobs (see 13.3.7, “Scans” on page 552).

File Types tab
On the File Types tab, you can explicitly allow or disallow certain file patterns (Figure 13-59).

Figure 13-59 Constraint - File Types

Use the buttons on the top of the screen, to allow or forbid files depending on their name. The
left column shows some default file patterns, or you can use the bottom field to create your
own pattern. Click >> to add your pattern to the allowed/forbidden files.


Users tab
The Users tab (shown in Figure 13-60) is used to allow or restrict the selected users in the
Constraint.

Figure 13-60 Constraint - Users

Important: The file condition is logically ORed with the User condition. A file will be
selected for Constraint processing if it meets at least one of the conditions.

Options tab
The Options tab provides additional conditions for file selection, and limits the number of
selected files to store in the central repository.

Once again, the conditions added in the tab will be logically ORed with the previous set in the
File Types and Users tab.


The bottom part of the tab, shown in Figure 13-61, contains the textual form of the Condition,
taking into account all the entries made in the Filesystems, File Types, Users and Options
tabs.

Figure 13-61 Constraints - Options

You can change this condition or add additional conditions, by using the Edit Filter button. It
displays the file filter popup (Figure 13-62) to change, add, and remove conditions or
conditions groups as previously explained in 13.3.6, “Profiles” on page 547.

Figure 13-62 Constraints - File filter


We changed the file filter to a more appropriate one by changing the OR operator to AND
(Figure 13-63).

Figure 13-63 Constraints - File filter changed

Alert tab
After selecting the files, you may want to generate an Alert only if the total used space
meeting the Constraint conditions exceeds a predefined limit. Use the Alert tab to specify
the triggering condition and action (Figure 13-64).

Figure 13-64 Constraints - Alert


In our Constraint definition, a script is triggered for each filesystem where the selected files
exceed one Gigabyte. We select the script by checking the Run Script option and selecting
Change... as shown in Figure 13-65. The script will be passed several parameters including a
path to a file that contains the list of files meeting the Constraint. You can use this list to
execute any action including delete or archive commands.

Figure 13-65 Constraints - Script parameters

Our example uses a sample script (tsm_arch_del.vbs) which is shipped with TotalStorage
Productivity Center for Data, which archives all the files in the produced list to a Tivoli Storage
Manager server, and then deletes them from local storage. This script is installed with
TotalStorage Productivity Center for Data server, and stored in the scripts subdirectory of the
server installation. It can be edited or customized if required - we recommend that you save
the original files first.

Versions for Windows (tsm_arch_del.vbs) and UNIX (tsm_arch_del) are provided. If you will
run this Constraint on a UNIX agent, then PERL is required to be installed on the agent. A
Tivoli Storage Manager server must be available and configured for this script to work. For
more information on the sample scripts, see Appendix A of the IBM Tivoli Storage Resource
Manager User’s Guide, SC32-9069.


13.5.4 Filesystem extension and LUN provisioning
The main functions of Filesystem Extension are shown in Figure 13-66.

Filesystem Extension

Automates filesystem extension

Supported platforms
AIX using JFS
SUN using VxFS

Support for automatic LUN provisioning with IBM ESS Storage
Subsystem

Actions triggered through standard Alerting mechanism when
a filesystem is performed

ibm.com/redbooks

Figure 13-66 Filesystem Extension


We use filesystem extension policy to automatically extend filesystems when utilization
reaches a specified threshold. We can also enable LUN provisioning to extend filesystems
within an ESS.

To set up filesystem extension policy, select Data Manager → Policy Management →
Filesystem Extension. Right-click Filesystem Extension and select Create Filesystem
Extension Rules as seen in Figure 13-67.

Figure 13-67 Create Filesystem Extension Rules


In the Filesystems tab, select the filesystems which will use filesystem extension policy by
moving them to the Current Selections panel.

Note the Enabled checkbox - the default is to check it, meaning the rule will be active. If you
uncheck the box, it will toggle to Disabled - you can still save the rule, but the job will not run.

To specify the extension parameters, select the Extension tab (Figure 13-68).

Figure 13-68 Filesystem Extension - Extension

This tab specifies how a filesystem will be extended. An explanation of the fields is provided
below.

Amount to Extend
We have the following options:
Add - the amount of space used for extension in MB or GB, or as a percentage of
filesystem capacity.
Make Freespace - the amount of freespace that will be maintained in the filesystems by
this policy. If freespace falls below the amount that is specified, the difference will be
added. Freespace can be specified in MB or GB increments, or by a percentage of
filesystem capacity.
Make Capacity - the total capacity that will be maintained in the selected filesystems. If
the capacity falls below the amount specified, the difference will be added.


Limit Maximum Filesystem Capacity?
When this option is enabled, the Filesystem Maximum Capacity is used in conjunction with
the Add or Make Freespace under Amount to Extend. If you enter a maximum capacity for a
filesystem in the Filesystem Maximum Capacity field and the filesystem reaches the
specified size, the filesystem will be removed from the policy and an Alert will be triggered.

Condition for Filesystem Extension
The options are:
Extend filesystems regardless of remaining freespace - the filesystem will be
expanded regardless of the available free space.
Extend filesystems when freespace is less than - defines the threshold for the
freespace which will be used to trigger the filesystem expansion. If freespace falls below
this value, the policy will be executed. Freespace can be specified in MB or GB
increments, or by a percentage of filesystem capacity.

Note: If you select Make Capacity under Amount to Extend, the Extend filesystems
when freespace is less than option is not available.

Use LOG ONLY Mode
Enable Do Not Extend Filesystems - Log Only when you want the policy to log the
filesystem extension. The extension actions that would have taken place are written to the log
file, but no extension takes place.

In the Provisioning tab (Figure 13-69) we define LUN provision parameters. Note that LUN
provisioning is available at the time of writing for filesystems on an ESS only.

Figure 13-69 Filesystem Extension - Provisioning


LUN Provisioning is an optional feature for filesystem extension. When the Enable
Automatic LUN Provisioning is selected, LUN provisioning is enabled.

In the Create LUNs that are at least field, you can specify a minimum size for new LUNs. If
you select this option, LUNs of at least the size specified will be created. If no size is
specified, then the Amount to Extend option specified for the filesystem (in “Amount to
Extend” on page 578) will be used. For more information on LUN provisioning, see IBM Tivoli
Storage Resource Manager 1.2 User’s Guide.

The Model for New LUNs feature means that new LUNs will be created similar to existing
LUNs in your setup. At least one ESS LUN must be currently assigned to the TotalStorage
Productivity Center for Data Agent associated with the filesystem you want to extend. There
are two options for LUN modeling:
Model new LUNs on others in the volume group of the filesystem being extended -
provisioned LUNS are modeled on existing LUNs in the extended filesystem’s volume
group.
Model new LUNs on others on the same host as the filesystem being extended -
provisioned LUNS are modeled on existing LUNs in the extended filesystem’s volume
group. If the corresponding LUN model cannot satisfy the requirements. it will look for
other LUNs on the same host.

The LUN Source option defines the location of the new LUN in the ESS, and has two options:
Same Storage Pool - provisioned LUNs will be created using space in an existing
Storage Pool. In ESS terminology this is called the Logical Sub System or LSS.
Same Storage Subsystem - provisioned LUNs can be created in any Storage Pool or
ESS LSS.

The When to Enforce Policy tab (Figure 13-70) specifies when to apply the filesystem
extension policy to the selected filesystems.

Figure 13-70 When to Enforce Policy tab


The options are:

Enforce Policy after every Probe or Scan automatically enforces the policy after every
Probe or Scan job. The policy will stay in effect until you either change this setting or disable
the policy.

Enforce Policy Now enforces the policy immediately for a single instance.

Enforce Policy Once at enforces the policy once at the specified time, specifying the month,
day, year, hour, minute, and AM/PM

The Alert tab (Figure 13-71) can define an Alert that will be triggered by the filesystem
extension job.

Figure 13-71 Alert tab


Currently the only available condition is A filesystem extension action started
automatically.

Refer to “Alert tab” on page 544 for an explanation of the definitions.

Important: After making configuration changes to any of the above filesystem extension
options, you must save the policy, as shown in Figure 13-72. If you selected Enforce
Policy Now, the policy will be executed after saving.

Figure 13-72 Save filesystem changes

For more information on filesystem extension and LUN provisioning, see IBM Tivoli Storage
Resource Manager: A Practical Introduction.

13.5.5 Scheduled Actions
TotalStorage Productivity Center for Data comes with an integrated tool to schedule script
execution on any of the Agents. If a script fails due to an unreachable Agent, the standard
Alert processes can be used. To create a Scheduled action, select Data Manager → Policy
Management → Scheduled Actions → Scripts, right-click and select Create Script.

Computers tab
On the Computers tab, select the computers or computer groups to execute the script.

Script Options tab
From the pull-down field, select a script that exists on the server. You can also enter the
name of a script not yet existing on the server or that only resides on the Agents.


The Script options tab is shown in Figure 13-73.

Figure 13-73 Scheduled action - Script options

The Script Name pull-down field lists all files (including non-script files) in the servers’ script
directory.

Attention: For Windows Agents, the script must have an extension that has an associated
script engine on the computer running the script (for example: .BAT, .CMD, or .VBS).

For UNIX Agents:
The extension is removed from the specified script name
The path to the shell (for example, /bin/bsh, /bin/ksh) must be specified in the first line of
the script
If the script is located in a Windows TotalStorage Productivity Center for Data Server
scripts directory, the script must have been created on a UNIX platform, and then
transferred in binary mode to the Server or you can use UNIX OS tools such as
dos2unix to convert the scripts. This will ensure that the CR/LF characters are correctly
inserted for execution under UNIX.

When to Run tab
As for other Data Manager jobs, you can choose to run a script once or repeatedly at a
predefined interval.

Alerts tab
With the Alert tab you can choose to be notified when a script fails due to an unreachable
Agent or a script not found condition. The standard Alert Mechanism described in 13.4, “OS
Alerts” on page 555 is used.

13.6 Database monitoring
The Monitoring functions of Data Manager are extended to databases when the license key
is enabled (8.4, “Configuring Data Manager for Databases” on page 313). Currently, MS
SQL-Server, Oracle, DB2, and Sybase are supported.


We will now review the Groups, Probes, Scans, and Profiles definitions for Data Manager for
Databases, and show the main differences compared to the core Data Manager monitoring
functions.

Figure 13-74 shows the navigation tree for Data Manager for Databases.

Figure 13-74 Databases - Navigation Tree

13.6.1 Groups
To get targeted monitoring of your database assets, you can create Groups consisting of:
Computers
Databases-Tablespaces
Tables
Users

Computer Groups
All databases residing on the selected computers will be probed, scanned, and managed for
Quotas.

The groups you have created using TotalStorage Productivity Center for Data remain
available for TotalStorage Productivity Center for Data for Databases. If you create a new
Group, the computers you put in it will be removed from the Group they currently belong to.

To create a Computer Group, use Data Manager - Databases → Monitoring → Groups
→ Computer, right-click, and select Create Computer Group.

“Computer Groups” on page 536 gives more information on creating Computer Groups.

Databases-Tablespaces Groups
Creating Groups with specific databases and tablespaces may be useful for applying identical
management rules for databases with the same functional role within your enterprise.


An example could be to create a group with all the Oracle-Server system databases, as you
will probably apply the same rules for space and alerting on those databases. This is shown
in Figure 13-75.

Figure 13-75 Database group definition

Table Groups
You can use Table Groups to create Groups of the same set of tables for selected or all
database instances.

You can use two different views to create a table group:
Tables by instance selects several tables for one instance.
Instances by table selects several instances for one table.

You can combine both views as each entry you add will be added to the group.

User Groups
As for core TotalStorage Productivity Center for Data, you can put user IDs in groups. The
user groups you create will be available for the whole TotalStorage Productivity Center for
Data product set.

Tip: The Oracle and MS SQL-Server user IDs (SYSTEM, sa, ...) are also included in the
available users list after the first database Probe.

13.6.2 Probes
The Probe process is used to gather data about the files, instances, logs, and objects that
make up monitored databases. The results of Probe jobs are stored in the repository and are
used to supply the data necessary for Asset Reporting.

Use Data Manager - Databases → Monitoring → Probe, right-click, and select Create
Probe to define a new Probe job. In the Instance tab of the Probe configuration, you can
select specific instances, computers, and computer groups (Figure 13-76).


Figure 13-76 Database Probe definition

The Computers list contains only computers that have been defined for Data Manager for
Databases. The definition procedure is described in “Configuring Data Manager for
Databases” on page 313.

13.6.3 Profiles
As for TotalStorage Productivity Center for Data, Profiles in Data Manager for Databases are
used to determine the databases attributes that are to be scanned. They also determine the
summary level and retention time to keep in the repository.

Use Data Manager - Databases → Monitoring → Profiles, right-click, and select Create
Profile to define a new profile. Figure 13-77 shows the Profile definition screen.

Figure 13-77 Database profile definition


You can choose to gather data on tables size, database extents, or database free space and
summarize the results at the database or user level.

13.6.4 Scans
Scan jobs in Data Manager for Databases collect statistics about the storage usage and
trends within your databases. The gathered data is used as input to the usage reporting and
Quota analysis.

Defining a Scan job requires defining:
The database, computer, and instances to Scan
The tables to monitor for detailed information such as size, used space, indexes, rows
count
The profile that will determine the data that is gathered and the report views that will be
made available by the Scan
The job scheduling frequency
Oracle-only additional options to gather information about pages allocated to a segment
that has enough free space for additional rows
The alerting mechanism to use should the Scan fail

All this information is set through the Scan definition screen that contains one tab for each
previously listed item. To define a new Scan, select Data Manager - Databases →
Monitoring → Scans, right-click and select Create Scan as in Figure 13-78.

Figure 13-78 Database Scan definition

Note: If you request detailed scanning of tables, the tables will only be scanned if their
respective databases have also been selected for scanning.


13.7 Database Alerts
TotalStorage Productivity Center for Data for Databases enables you to define Alerts on
instances, databases, and tables. The Probe and Scan jobs output are processed and
compared to the defined alerts. If a threshold is reached, an Alert will be triggered.

Tivoli Storage Resource Manage for Databases uses the standard Alert mechanisms
described in 13.4, “OS Alerts” on page 555.

13.7.1 Instance Alerts
Data Manager - Databases → Alerting → Instance Alerts, right-click and select Create
Alert lets you define some alerts as shown in Table 13-3. Those Alerts are triggered during
the Probe process.

Table 13-3 Instance Alerts
Alert type Oracle Sybase MSSQL

New database discovered x x

New tablespace discovered x

Archive log contains more than X units x

New device discovered x

Device dropped x

Device free space greater than X units x

Device free space less than X units x

An interesting Alert is the Archive Log Directory Contains More Than for Oracle, since the
Oracle application can hang if there is no more space available for its archive log. This Alert
can be used to monitor the space used in this specific directory and trigger a script that will
archive the files to an external manager such as Tivoli Storage Manager once the predefined
threshold is reached. For a detailed example, refer to IBM Tivoli Storage Resource Manager:
A Practical Introduction, SG24-6886.

13.7.2 Database-Tablespace Alerts
To define a Database-Tablespace Alert, select Data Manager - Databases → Alerting →
Database-Tablespace Alerts, right-click, and select Create Alert. You can define various
monitoring options on your databases as shown in Table 13-4. Those Alerts are triggered
during the Probe process.

Table 13-4 Instance alerts

Database/Tablespace freespace lower than x x x

Database/Tablespace offline x x x

Database/Tablespace dropped x x x

Freespace fragmented in more than n extents x

Largest free extent lower than x



Database Log freespace lower than x x

Last dump time previous to n days x

13.7.3 Table Alerts
To define a new Table Alert, use Data Manager - Databases → Alerting → Table Alerts,
right-click, and select Create Alert. With this option you can set up monitoring on database
tables. The Alerts that can be triggered for a table as shown in Table 13-5 below. Those Alerts
are triggered during the Scan processes and only if the Scan includes a Table Group.

Table 13-5 Table alerts
Alert type Oracle Sybase MsSQL

Total Table Size Greater Than x x x

Table Dropped x x x

(Max Extents - Allocated) < x
Segment Has More Than x

Chained Row Count Greater Than x

Empty Used Segment Space Exceeds x

Forwarded Row Count Greater Than x

13.7.4 Alert log
The Data Manager - Databases → Alerting → Alert Log menu lists all Alerts that have
been fired by the Probe jobs, the Scan jobs, the defined Alerts, and the violated Quotas.

Tip: Please refer to 13.4.5, “Alert logs” on page 564 for more information about using the
Alert log tree.

13.8 Databases policy management
The Policy Management functions of Data Manager for Databases enable you to:
Define space limits (Quotas) on database space used by tables owners. Those limits can
be set at a network (whole environment), at an instance or at a database level.
Schedule scripts against your database resources.


13.8.1 Network Quotas
A Network Quota will define the maximum cumulated space a user can occupy on all the
scanned databases. An Alert will be fired for each user that exceeds the limit specified in the
Quota definition.

We used Data Manager - Databases → Policy Management → Quotas → Network,
right-click and select Create Quota to create a new Quota. The right pane will switch to a
Quota configuration screen with four tabs.

Users tab
On the Users tab, specify the database users you want to be monitored for Quotas. You can
also select a profile in the Profile pull-down field on the top right of the tab. In this field, you
can select any Profile that stores summary data on a user level. The Quota will only be fired
for databases that have been scanned using this Profile (Figure 13-79).

Figure 13-79 Database Quota - Users tab

Database-Tablespace tab
Use this tab to restrict Quota checking to certain databases. You can choose several
databases or computers. If you choose a computer, all the databases running on it will be
included for Quota management.

When to run tab
As with Data Manager, you can select the time to run from:
Immediate
Once at a schedule date and time
Repetitive at predefined intervals

Alert tab
On the Alert tab you can specify the space limit allowed for each user and the action to run. If
no action is selected, the Quota violation will only be logged in the Alert log.


13.8.2 Instance Quota
The Instance Quota mechanism is similar to the Network Quota, except that it is set at the
instance level. Whenever a user reaches the Quota on one instance, an Alert will be fired.

13.8.3 Database Quota
With Database Quota, the Quota is set at the database level. Each monitored user will be
reported back as soon as he reaches the limit on at least one of the monitored database.

13.9 Database administration samples
We now list some typical checks done regularly by Oracle database administrators and show
how they can be automated using Data Manager for Databases.

13.9.1 Database up
Data Manager for Databases can be used to test for database availability using Probe and
Scan jobs since they will fail and trigger an Alert if either the database or the listener is not
available. Since those jobs use system resources to execute, you may instead choose
scheduled scripts to test for database availability.

Due to limited scheduling options and the need for user-written scripts, we recommend using
dedicated monitoring products such as Tivoli Monitoring for Databases.

13.9.2 Database utilization
There are a number of different levels where system utilization can be monitored and
checked in a database environment.

Tablespace space usage
This is a standard Alert provided by Data Manager for Databases. This Alert will be triggered
by the Probe jobs.

Archive log directory space usage
This is a standard alert provided by Data Manager for Databases. This Alert will be triggered
by the Probe jobs as shown in 13.7.1, “Instance Alerts” on page 588.

Maximum extents used
Your application may become unavailable if a table reaches its maximum allowed number of
extents. This is an indicator that can be monitored using the (Max Events - Allocated Extents)
< Table Alert.

13.9.3 Need for reorganization
To ensure good application performance, it is important to be notified promptly if a database
reorganization is required.

Count of Used table extents
You can monitor for table reorganization need using the table Alert trigger Segment has more
than n extents.


Count of chained rows
Chained rows can have an impact on database access performance. This issue can be
monitored using the Chained Row Count Greater than table Alert trigger.

Count of Used table extents
You can monitor the need for table reorganization using the table Alert trigger Segment has
more than n extents.

Freelist count
You cannot monitor the count of freelists in an Oracle table using Data Manager for
Databases.

13.10 Data Manager reporting capabilities
The reporting capabilities of Data Manager are very rich, with over 300 predefined views. You
can see the data from a very high-level; for example, the total amount of free space available
over the enterprise; or from a low-level, for example, the amount of free space available on a
particular volume or a table in a database.

The data can be displayed in tabular or graphical format, or can be exported as HTML,
Comma Separated Variable (CSV), or formatted report files.

The reporting function uses the data stored in the Data Manager repository. Therefore, in
order for reporting to be accurate in terms of using current data, regular discovery, Ping,
Probe, and Scan jobs must be scheduled. These jobs are discussed in 13.3, “OS Monitoring”
on page 533.

Figure 13-80 shows the Data Manager main screen with the reporting options highlighted.

The Reporting sections are used for interactive reporting. They can be used to answer ad hoc
questions such as, “How much free space is available on my UNIX systems?” Typically, you
will start looking at data at a high-level and drill down to find specific detail. Much of the
information can also be displayed in graphical form as well as in the default table form.

The My Reports sections give you access to predefined reports. Some of these reports are
pre-defined by Data Manager, others can be created by individual users saving reporting
criteria in the Reporting options. You can also set up Batch Reports to create reports
automatically on a schedule.

My Reports will be covered in more detail in 13.14, “Creating customized reports” on
page 683, and 13.15, “Setting up a schedule for daily reports” on page 697.

The additional feature, TotalStorage Productivity Center for Data for Chargeback produces
storage usage Chargeback data, as described in 13.17, “Charging for storage usage” on
page 700.


Predefined reports provided by
TotalStorage Productivity
Center for Data
Reports customized and saved
by user tpcadmin

Schedule reports to run in
batch mode

Interactive reporting options

Database reporting options

Figure 13-80 TotalStorage Productivity Center for Data main screen showing reporting options

13.10.1 Major reporting categories
Data Manager collects data for reporting purposes in seven major categories. These will be
covered in the following sections. Within each major category there are a number of
sub-categories.

Most categories are available for both operating system level reporting and database
reporting. However, a few are for operating system reporting only. The description of each
category specifies which applies, and in the more detailed following sections for each
category, we present the capabilities separately for both Data Manager and Data Manager for
Databases as appropriate.

Asset Reporting
Asset data is collected by Probe processes and reports on physical components such as
systems, disk drives, and controllers. Currently, Asset Reporting down to the disk level is only
available for locally attached devices. Asset Reporting is available for both operating system
and database reporting.


Storage Subsystems Reporting
Storage Subsystem data is collected by Probe processes. It provides a mechanism for viewing
storage capacity at a computer, filesystem, storage subsystem, LUN, and disk level. These
reports also enable you to view the relationships among the components of a storage
subsystem. Storage Subsystem reporting is currently only available for IBM TotalStorage
Enterprise Storage Servers (ESS). Storage Subsystems Reporting is available for operating
system only.

Availability Reporting
Availability data is collected by Ping processes and allows you to report on the availability of
your storage resources and computer systems. Availability Reporting is provided for
operating system reporting only.

Capacity Reporting
Capacity Reporting shows how much storage you have and how much of it is being used.
You can report at anywhere from an entire network level down to an individual filesystem.
Capacity Reporting is provided for both operating system and database reporting.

Usage Reporting
Usage Reporting goes down a level from Capacity Reporting. It is concerned not so much
with how much space is in use, but rather with how the space is actually being used for. For
example, you can create a report that shows usage by user, or a wasted space report. You
define what wasted space means, but it could be for example files of a particular type or files
within a certain directory, which are more than 30 days old. Usage Reporting is provided for
both operating system and database reporting.

Usage Violation Reporting
Usage Violation Reporting allows you to set up rules for the type and/or amount of data that
can be stored, and then report on exceptions to those rules. For example, you could have a
rule that says that MP3 and AVI files are not allowed to be stored on file servers. You can also
set Quotas for how much space an individual user can consume. Note that usage violations
are only softly enforced - Data Manager will not enforce the rules in real time, but will
generate an exception report after the fact. Usage Violation Reporting is provided for both
operating system and database reporting.

Backup Reporting
Backup Reporting identifies files that have not been backed up. Backup Reporting is provided
for operating system reporting only.

13.11 Using the standard reporting functions
This section discusses Data Manager’s standard reporting capabilities. Customized reporting
is covered in 13.14, “Creating customized reports” on page 683.

This section is not intended to cover exhaustively all of the reporting options available, as
these are very numerous, and are covered in detail in the Reporting section of the manual
IBM Tivoli Storage Resource Manager V1.1 Reference Guide SC32-9069. Instead, this
section provides a basic overview of Data Manager reporting, with some examples of what
types of reports can be produced, and additional information on some of the less
straightforward reporting options.


To demonstrate the reporting capabilities of TotalStorage Productivity Center for Data, we
installed the Server code on a Windows 2000 system called Colorado, and deployed these
Windows Agents:
Gallium
Wisla
Lochness
Colorado is also an Agent as well as being the Server.

The host GALLIUM has both Microsoft SQL-Server and Oracle database installed to
demonstrate database reporting. The Agent on LOCHNESS also provides data for a NAS
device call NAS200. The Agent on VMWAREW2KSRV1 also provides data for a NetWare
server called ITSOSJNW6.

The lab setup is shown in Figure 13-81.

Tivoli Storage Resource Manager: Lab Environment

ITSRM
Scan
ITSRM
Database

A23BLTZM
WNT LOCHNESS
ITSRM W2K
Agent & ITSRM
GUI Server

Ethernet

NetWare VMWAREW2KSRV1 SOL-E GALLIUM CRETE BRAZIL IBM
EASTER
W2K (Vmware) Solaris W2K AIX AIX NAS200
HP-UX
ITSRM ITSRM ITSRM ITSRM ITSRM
ITSRM
Agent Agent Agent Agent Agent
Agent

ITSRM
Scan
ibm.com/redbooks
VMWAREW2KSRV1
W2K (Vmware)
ITSRM
A t
Figure 13-81 TotalStorage Productivity Center for Data Lab Environment

13.11.1 Asset Reporting
Asset Reporting provides configuration information for the TotalStorage Productivity Center
for Data Agents. The information available includes typical asset details such as disk system
name and disk capacities, but provides a large amount of additional detail.

Figure 13-82 shows the major subtypes within Asset Reporting. Note that unlike the other
reporting categories where most of the drill-down functions are chosen from the right-hand
panel, in Asset Reporting the drill-down functions are mostly available on the left-hand pane.


Figure 13-82 Reporting - Asset

By Cluster View
Click By Cluster to drill down into a virtual server or cluster node. You can drill down further
to a specific controller to see the disks under it and/or drill down on a disk to see the file
systems under it.

By Computer view
Click By Computer to see a list of all of the monitored systems (Figure 13-83.)

Figure 13-83 Reporting - Asset - By Computer


From there we can drill down on the assets associated with each system. We will take a look
at node GALLIUM. In Figure 13-84 we have shown most of the items for GALLIUM expanded,
with the details for Disk 2 displayed in the right-hand bottom pane.

You will see a detailed level of information, both in terms of the type of objects for which data
is collected (for example, Exports or Shares), and the specific detail for a given device.

Figure 13-84 Report - GALLIUM assets

By OS Type view
This view of the Asset data provides the same information as the By Computer view, with the
difference that the Agent systems are displayed sorted by operating system platform.

By Storage Subsystem view
Data Manager provides reporting for storage subsystems, any disk array subsystems whose
SMI-S Providers are CTP certified by SNIA for SMI-S 1.0.2, and IBM SAN Volume Controller
clusters.

For disk array subsystems, you can view information about:
– Disk groups (for IBM TotalStorage ESS subsystems)
– Array sites (for IBM TotalStorage DS6000/8000 only)
– Ranks (for IBM TotalStorage DS6000/8000 only)
– Storage pools (for disk array subsystems)
– Disks (for disk array subsystems)
– LUNs (for disk array subsystems)

For IBM SAN Volume Controllers, you can view information about
– Managed disk groups
– Managed disks
– Virtual disks


System-wide view
The System-wide view however does provide additional capability, as it can give a
System-wide view rather than a node-by-node view of some of the data. A graphical view of
some of the data is also available. Figure 13-85 shows most of the options available from the
System-wide view and in the main panel, the report of all exports or shares available.

Figure 13-85 Reporting - Assets - System-wide view

Each of the options available under the System-wide view are self explanatory with the
possible exception of Monitored Directories. Data Manager can monitor utilization at a
directory level as well as a device or filesystem level. However, by default, directory level
monitoring is disabled.

To enable directory monitoring, define a Directory Group by selecting Data Manager →
Monitoring → Groups → Directory, right-click Directory and choose Create Directory
Group. The process of setting up Directory Groups is discussed in more detail in 13.3.2,
“Groups” on page 535. Once the Directory Group is created it must be assigned to a Scan
job, and that job must be run on the systems where the directories to be monitored exist.

By setting up a monitored directory you will get additional information for that directory. Note
that the information collected includes any subdirectories. Information collected about the
directory tree includes the number of files, number of subdirectories, total space used, and
average file size. This can be graphed over time to determine space usage patterns.

IBM TotalStorage Productivity Center for Data for Databases
Asset Reporting for databases is similar to that for filesystems; however, filesystem entities
like controllers, disks, filesystems, and shares are replaced with database instances,
databases, tables, and data files.


Very specific information regarding an individual database is available as shown in
Figure 13-86 for the database DMCOSERV on node COLORADO.

Figure 13-86 DMCOSERV database asset details

Or you can see rollup information for all databases on a given system (using the System-wide
view) as shown in Figure 13-87.

Figure 13-87 System-wide view of database assets


All of the database Asset Reporting options are quite straightforward with the exception of
one. In order to receive table level asset information, one or more Table Groups needs to be
defined. This is a similar process to that for Directory Groups as described in “System-wide
view” on page 598.

You would not typically include all database tables within Table Groups, but perhaps either
critical or rapidly growing tables. We will set up a group for UDB.

To set up a Table Group, Data Manager - Databases → Monitoring → Groups → Table,
right-click Table and choose Create Table Group (Figure 13-88).

Figure 13-88 Create a new database table group

We have entered a description of Colorado Table Group. Now we click New Instance to
enter the details of the database and tables that we want to monitor. From the drop down box,
we select the database instance, in this case the UDB instance on Colorado. We then enter
three tables in turn. For each table, we entered the database name (DMCOSERV), the creator
name (db2admin) and a table name. After entering the values, click Add to enter more tables
or finish. We entered the table names of BASEENTITY, DMSTORAGEPOOL, and DMVOLUME, as
shown in Figure 13-89. Once all of the tables have been entered click OK.


Figure 13-89 Add UDB tables to table group

Now we return to the Create Table Group panel, and we see in Figure 13-90 the information
about the newly entered tables.

Figure 13-90 Tables added to table group

Now we Save by clicking the floppy disk icon and when prompted, we entered the Table
Group name of ColoradoTableGroup.

In order for the information for our tables to be collected, the Table Group needs to be
assigned to a Scan job. We will assign it to the default database scan job called
Tivoli.Default DB Scan by choosing Data Manager - Databases → Monitoring → Scans
→ TPCUser.Default Db Scan.


The definition for this scan job is shown in Figure 13-91 and in particular we see the Table
Groups tab. Our new Table Group is shown initially in the left hand pane. We moved it to the
right hand pane by selecting it and clicking >>. We then save the updates to the Scan job by
choosing File → Save (or with the floppy disk icon from the tool bar). Finally, we can execute
the Scan job by right-clicking it and choosing Run Now. Figure 13-91 shows the Scan job
definition after the Table Group had been assigned to it.

Figure 13-91 Table group added to scan job

Example 13-4 is an extract from the Scan job log showing that the table information is now
being collected. You can view the Scan job log through the TotalStorage Productivity Center
for Data GUI by first expanding the particular Scan job definition. A list of Scan execution
reports will be shown; select the one of interest. You may need to right-click the Scan job
definition and choose Refresh Job List. The list of Scan executions for the Tivoli.Default DB
Scan is shown in Figure 13-92.

Figure 13-92 Displaying Scan job list


Once you have the actual job chosen you can click the detail icon for the system that you are
interested in to display the job log. The actual file specification of the log file on the Agent
system will be displayed at the top of the output when viewed through the GUI. Example 13-4
shows the actual file output.

Example 13-4 Database scan job showing table monitoring
09-19 18:01:01 DBA0036I: The following databases-tablespaces will be scanned:
MS SQLServer gallium/gallium Databases:
master
model
msdb
Northwind
pubs
tempdb
Oracle itsrm Tablespaces:
ITSRM.DRSYS
ITSRM.INDX
ITSRM.RBS
ITSRM.SYSTEM
ITSRM.TEMP
ITSRM.TOOLS
ITSRM.USERS
09-19 18:01:01 DBA0041I: Monitored Tables:
.CTXSYS.DR$OBJECT
Northwind.dbo.Employees
Northwind.dbo.Customers
Northwind.dbo.Suppliers

Finally, we can produce table level asset reports by choosing for example, Data Manager -
Databases → Reporting → Asset → System-wide → All DBMSs → Tables → By
Total Size. This is shown in Figure 13-93.

Figure 13-93 Tables by total size asset report


13.11.2 Storage Subsystems Reporting
Storage Subsystems Reporting is covered in detail in 13.12, “TotalStorage Productivity
Center for Data ESS Reporting” on page 634.

13.11.3 Availability Reporting
Availability Reporting is quite simple. Two different sets of numbers are reported - Ping and
Computer Uptime. Ping is only concerned with whether or not the system is up and
responding to the ICMP requires - it does not care whether the Data Agent is running or not.
Ping results are collected by a Ping job, so this must be scheduled to run on a regular basis.
See 13.3.4, “Pings” on page 542.

Computer Uptime detects whether or not the Data Agent is running. Computer Uptime
statistics are gathered by a Probe job so this must be scheduled to run on a regular basis.
See 13.3.5, “Probes” on page 545.

Figure 13-94 shows the Ping report for our TotalStorage Productivity Center for Data
environment, and Figure 13-95 shows the Computer Uptime report. To generate these
reports, we had to select the computers of interest and select Generate Report.

Figure 13-94 Reports - Availability - Ping


Figure 13-95 Reports - Availability - Computer Uptime

13.11.4 Capacity Reporting
Capacity Reporting shows how much storage capacity is installed, and of that capacity, how
much is being used and how much is available for future growth.

There are four capacity report views within TotalStorage Productivity Center for Data:
Disk Capacity
Filesystem Capacity
Filesystem Used Space
Filesystem Free Space

However, in reality there are really only two views, or perhaps three. The Filesystem Capacity
and Filesystem Used Space views are nearly identical - the only differences being in the
order of the columns and the row sort order.

And there is relatively little difference between these two views and the Filesystem Free
Space view. The Filesystem Capacity and Filesystem Used Space views report on used
space, so include columns like percent used space whereas Filesystem Free Space includes
columns like percent free space. All other data is identical.

Therefore, there are really only two views: a Disk Capacity view and a Filesystem Capacity
view.

The Disk Capacity view provides information about physical or logical disk devices and what
proportion of them has been allocated. Figure 13-96 shows the Disk Capacity by Disk
selection window.


Figure 13-96 Disk capacity report selection window

Often there is a one-to-one relationship between devices and filesystems as seen in
Figure 13-97, particularly on Windows systems. However, if a single physical disk has two
partitions the detailed description will show two partitions at the bottom of the right-hand
pane.

Figure 13-97 Capacity report - Gallium Disk 0


IBM TotalStorage Productivity Center for Data for Databases
Capacity Reporting for databases is very straightforward. You can report on:
All databases of any type
All databases of a given type on a particular system or group of systems
On a specific database

Example 13-98 shows a Capacity Report by Computer Group. We actually have databases in
just one Computer Group, WindowsDBServers. We then drilled down to see all systems within
the WindowsDBServers group, then specifically to node GALLIUM, so that we could see all
databases on GALLIUM.

Figure 13-98 Database Capacity report by Computer Group

13.11.5 Usage Reporting
The reporting categories covered so far have been mostly concerned with reporting at the
system or device level. Usage Reporting goes down one more step to report at a level lower
than the filesystem. You can produce reports that answer questions such as:
How old is my data? When was it created, last accessed, or modified?
What are my largest files? What are my largest directories?
Do I have any orphan files?

Data Manager
With Usage Reporting, you will be able to:
Identify orphan files and either update their ownership or delete them to free up space
Identify the largest files and determine whether they are needed or whether parts of the
data could be archived
Identify obsolete files so that they can be either deleted or archived


There are a few restrictions on Usage Reporting:
In order to report by directory or by Directory Group you will need to set them up in Data
Manager → Monitoring → Groups → Directory.
UNIX systems do not record file create dates, so no reporting by creation time is available
for these systems.

Data Manager for Databases
Like database Asset Reporting, all of the database Usage Reporting options are quite
straightforward with the exception of table level reporting.

From a usage perspective there are two types of table report available:
Largest tables
Monitored tables

We can report on database largest tables by choosing for example, Data Manager -
Databases → Reporting → Usage → All DBMSs → Tables → Largest Tables → By
RDBMS Type. This report is shown in Figure 13-99.

Figure 13-99 Largest tables by RDBMS type


A Monitored Tables by RDBMS Type report is shown in Figure 13-100. In this case, only
tables which are part of a Table Group, which is included in a Scan job will be reported on.

Figure 13-100 Monitored tables by RDBMS type


13.11.6 Usage Violation Reporting
Usage Violation Reporting enforces Data Manager Constraints and Quotas. A Constraint is a
limit, by file name syntax, on the type of data that can be stored on a system. A Quota is a
storage usage limit placed on a user or operating system User Group, and can be defined at
the network, computer, or filesystem level. Constraints and Quotas were described in 13.5,
“Policy management” on page 565. It is important to remember that Quotas and Constraints
are not hard limits - users will not be stopped from working if a Quota or Constraint is violated,
but this event will trigger an exception, which will be reported.

Data Manager Constraint Violation Reporting
There are a number of predefined Constraints in Data Manager. Before we produce a
Constraint violation report, we need to set up a new Constraint called forbidden files. Setting
up Constraints was described in 13.5.3, “Constraints” on page 570.

First navigate Data Manager → Policy Management → Constraints. Existing Constraints
will be listed. Right-click Constraints and choose Create Constraint. On the Filesystems tab
we entered a description of forbidden files, chose Computer Groups, then selected
tpcadmin.Windows Systems and tpcadmin.Windows DB Systems and clicked >>. The
completed Filesystems tab is shown in Figure 13-101.

Figure 13-101 Create a Constraint - Filesystems tab


We then need to specify in the File Types tab, what a forbidden file is. You can define the
criteria as either inclusive or exclusive; that is, you can specify just those files types that will
violate the Constraint, or you can specify that all files will violate the Constraint except those
specified. There are a number of predefined file types included; you can also chose additional
files by entering appropriate values in the “Or enter a pattern field” at the bottom of the
form. We have chosen MP3 and AVI files. The completed File Types tab is shown in
Figure 13-102.

Figure 13-102 Create a Constraint - File Types tab

The Users tab is very similar to the File Types tab - you can specify which users should be
included or excluded from the selection criteria. We have taken the default, which is to include
all users.

In the Options tab, we nominate a maximum number of rows to be returned. We can also
apply some more specific selection criteria here such as only including files that are larger
than a defined size. Note, however that these criteria are added to the file list. For example, if
we specified here that we only wanted to include files greater than 1 MB, the search criteria
would be changed to ((NAME matches any of ('*.AVI', '*.mp3') AND TYPE <> DIRECTORY)
OR SIZE > 1 MB). So the returned list of files would be any file greater than 1 MB in size plus
any *.MP3 or *.AVI files.


If you wish to change the selection criteria so that instead you select any *.MP3 or *.AVI files
that are larger than 1 MB, you can enter 1 MB against the bigger than option, and then click
the Edit Filter button shown in Figure 13-105. You will then see the file filter as shown in
Figure 13-103. To add the size criteria to the file type criteria, click the Size > 1MB entry and
drag it up to the All of tag. The changed filter is shown in Figure 13-104. You can also see
the Boolean expression for the filter has changed to reflect this condition.

Figure 13-103 Edit a Constraint file filter - before change

Figure 13-104 Edit a Constraint file filter - after change


In this case we did not want to apply a size criteria, so we left the Option tab entries at their
defaults as shown in Figure 13-105.

Figure 13-105 Create a Constraint - Options tab

Finally, we can specify that we want an Alert generated if a triggering condition is met. The
only choice here is to specify a maximum amount of space consumed by the files that meet
our selection criteria. We left all of the Alert tab options at their defaults other than specifying
an upper limit of 100 MB for files that have met our selection criteria.


The Alert tab is shown in Figure 13-106. Alerting is covered in more detail in 13.4, “OS Alerts”
on page 555.

Figure 13-106 Create a Constraint - Alert tab

We then clicked the Save button and entered a name of Forbidden Files as shown in
Figure 13-107.

Figure 13-107 Create a Constraint - save

Before we can report against the Constraint, we need to ensure that a Scan job has been run
to collect the appropriate information.

Once the Scan has completed successfully, you can go ahead and produce Constraint
Violation Reports. Note that you cannot produce a report of violations of a particular
Constraint - the report will include entries for any Constraint violation. However, once the
report is generated, you can drill down into specific Constraint violations.

We produced the report by choosing Data Manager → Reporting → Usage Violations →
Constraint Violation → By Computer. You will see a screen like Figure 13-108 where you
can select a subset of the clients if appropriate - after selecting, click Generate Report.


Figure 13-108 Constraint violation report selection screen

You will then see a list of all of those instances of Constraint violations as shown in
Figure 13-109.

The report shows multiple types of Constraints. Some of these Constraints were predefined
(Orphaned File Constraint and Obsolete File Constraint) and others (ALLFILES and forbidden
files) we defined. An orphaned file is any file that does not have an owner. This allows you to
easily identify files that belonged to users who have left your organization or have had an
incorrect ownership set.

Figure 13-109 Constraint violations by computer


From there you can drill down on a specific Constraint, then filesystems within the Constraint,
and finally to a list of files that violated the Constraint on that filesystem by selecting the
magnifying glass icon next to the entry of interest. Or, as shown in Figure 13-110, by clicking
the pie chart icon next to the entry for forbidden files, you can produce a graph indicating what
proportion of capacity is being utilized by files violating the Constraint. Position the cursor
over any segment of the pie chart to show the percentage and number of bytes consumed by
that segment.

Figure 13-110 Graph of capacity used by Constraint violating files

Constraint violations are also written to the Data Manager Alert Log. Figure 13-111 shows the
same list of violations as if you had produced a Constraint Violations by computer report.

Figure 13-111 Alert log showing Constraint violations


Quota Violation Reporting
The process of producing a Quota violation report is very similar to producing a Constraint
violation report, but with some key differences.

One difference between Quotas and Constraints is the process of collecting data. For
Constraints, the data is collected as part of a standard Scan job in a similar way to adding an
additional Profile to a Scan. Quota data collections are performed in a separately scheduled
job. So, when you set up a Quota you need to specify scheduling parameters.

We set up a Quota rule called Big Windows Users by choosing Data Manager → Policy
Management → Quotas → User → Computer, right-clicking Computer and selecting
Create Quota. On the Users screen we entered a description of Big Windows Users and then
selected User Groups and then TPCUser.Default User Group as show in Figure 13-112.

Figure 13-112 Create Quota - Users tab


On the Computers tab we chose our Windows group: tpcadmin.Windows Systems
(Figure 13-113).

Figure 13-113 Create Quota - Computers tab

We then had to specify when and how often we wanted the Quota job to run. We chose to run
the job weekly under the When to CHECK tab as shown in Figure 13-114.

Figure 13-114 Create Quota - When to Check


On the Alert tab, shown in Figure 13-115, we accepted all of the defaults other than to specify
the limit under User Consumes More Than, in this case, 1 GB.

No Alerts will be generated other than to log any exceptions in the Data Manager Alert Log.

Figure 13-115 Create Quota - Alert

Finally, we save the Quota definition, calling it Big Windows Users as shown in Figure 13-116.

Figure 13-116 Create Quota - save


The new Quota now appears under Data Manager → Policy Management → Quotas →
User > Computer as tpcadmin.Big Windows Users (where tpcadmin is our Data Manager
username). We right-clicked the Quota and chose Run Now as in Figure 13-117.

Figure 13-117 Run new Quota job

This job will collect data related to the Quota, and add any Quota Violations to the Alert Log

Figure 13-118 Alert Log - Quota violations


We then drilled down on one of the Alerts to see the details (Figure 13-119).

Figure 13-119 Alert Log - Quota violation detail

And finally we can create a Quota Violation report by choosing Data Manager → Reporting
→ Usage Violations → Quota Violations → Computer Quotas → By Computer. The
high-level report is shown in Figure 13-120.

Figure 13-120 Quota violations by computer


We can then drill down further for additional detail or to produce a graphical representation of
the data behind the violation. The graph in Figure 13-121 shows a breakdown of the users’
data by file size.

Figure 13-121 Quota violation graphical breakdown by file size

Filesystem Usage Violation Reporting includes both Quota and Constraint violations.
However, for databases, only Quota violations are available.

You can place a Quota on users, user groups, or all users and you can limit the Quota by
computer, computer group, database instance, database tablespace group or tablespace.

We will set up an Instance Quota that limits any individual user to 100 MB of space per
instance for any database on any server in the tpcadmin.WindowsDBServers computer
group.

To do this, navigate to Data Manager - Databases → Policy Management → Quotas →
Instance. Right-click Instance and choose Create Quota. Figure 13-122 shows the Quota
definition screen. We entered a description of Big DB Users and selected the
TPCUser.Default User Group by expanding User Groups, clicking TPCUser.Default User
Group, and then clicking >>.


Figure 13-122 Create database Quota - Users tab

On the Instances tab, expand Computer Groups, select tpcadmin.Windows DB Systems
and then click >> to add it to the Current Selections as shown in Figure 13-123.

Figure 13-123 Create database Quota - Instances tab


On the When to Run tab shown in Figure 13-124, we chose to run the Quota job weekly and
chose a time of day for the job to run. Other values were left at the defaults.

Figure 13-124 Create a database Quota - When to Run tab

On the Alert tab (shown in Figure 13-125) we specified the actual Quota that we wanted
enforced, which was a 100 MB per user Quota. Other values were left as defaults.

Figure 13-125 Create a database Quota - Alert tab


We saved the new Quota definition with a name of Big DB Users as shown Figure 13-126.

Figure 13-126 Create a database Quota - Save

We now run the Quota by right-clicking it and choosing Run Now as seen in Figure 13-127.

Figure 13-127 Run the database Quota


To check if any user has violated the Quota, navigate Data Manager - Databases →
Alerting → Alert Log → All DBMSs → All. We see one violation as shown in
Figure 13-128.

Figure 13-128 DB Quota violation

We can also now run a database Quota violation report by choosing Data Manager -
Databases → Reporting → Usage Violations → Quota Violations → All Quotas → By
User Quota. This report can be seen in Figure 13-129.

Figure 13-129 Database Quota violation report


13.11.7 Backup Reporting
Backup Reporting is designed to do two things: It can alert you to situations where files have
been modified but not backed up, and it can provide data on the volume of data that will be
backed up. Figure 13-130 shows the options that are available for Backup Reporting.

Figure 13-130 Backup Reporting options

Most at Risk Files
Data Manager defines most at risk files as those that are least-recently modified, but have not
been backed up.

There are some points worth noting about this report:
Since the report relies on the archive bit being set to determine whether the file has
changed, this report will only work on Windows systems as UNIX systems have no
equivalent to the archive bit
When using most backup products, once a file has been backed up the archive bit is
cleared. Before Version 5.2, IBM Tivoli Storage Manager did not do this, therefore if this
level of Tivoli Storage Manager was used, this report would list files that actually may have
been backed up. IBM Tivoli Storage Manager Version 5.2 has the ability to reset the
Windows archive bit after a successful backup of a file.


By default, information on only 20 files will be returned. Figure 13-131 shows the selection
screen for the report. You will notice that the report uses the Profile TPCUser.Most at Risk. It
is in this Profile that the 20 file limit is set, although the value can be changed. You can
override the value on the selection screen, but you can only reduce the value here, not
increase it.

By updating the Profile you can also exclude files from the report. By default, any file in the
WINNTsystem* directory tree on any device will be excluded. You can add entries to the
exclusion list if appropriate. Ideally, the exclusion list should be the same as that in your
backup product.

Figure 13-131 Files most at risk report - selection

Modified Files Not Backed Up
The report provides an aging analysis of your data that has been modified but not backed up.
It will show what proportion of the data has been modified within the past 24 hours, between
one and seven days, between one week and one month, and so on. Figure 13-132 shows the
selection taken in our Windows environment. Like the Most at Risk Files report, this report
also relies on the archive bit, so check to see if your backup application uses this.


Figure 13-132 Modified Files not backed up selection

To view the report, click Generate report. We choose to view it as a graphic by then clicking
the pie icon and selecting Chart: Space Distribution for All. This is shown in Figure 13-133.
This chart tells you the amount of space consumed by files have not been backed up since
the last backup was run for this server.

Figure 13-133 Modified Files not backed up chart overall view


We can also select Chart: Count Distribution for All as shown in Figure 13-134 to show the
number of files in each category.

Figure 13-134 Files need backed up chart in detail view

The different charts can be viewed in different ways. To select another type of chart,
right-click in the chart area, select Customize this chart, and click the radio button next to
the desired chart type.

Backup Storage Requirements Reporting
This option allows you determine how much data would be backed up if you were to perform
either a full or an incremental backup. The Full Backup Size option can be used regardless of
the OS type and the backup application in use.


In Figure 13-135, the report is run against Windows systems by filesystem.

Figure 13-135 Backup storage requirements per filesystem

The selection can also run by computer, as shown in Figure 13-136.

Figure 13-136 Backup storage requirement per computer and per filesystem


The Incremental Backup Size option makes use of the archive bit, so it can only be used on
Windows systems, and if Tivoli Storage Manager is the backup application, the
resetarchiveattribute option must be used (for Version 5.2). A sample report is shown in
Figure 13-137.

Figure 13-137 Incremental reporting per Node and Filesystem based on files


The third report type here is Incremental Range Sizes Reporting. This does not rely on the
archive bit (instead, it uses the modification date) so is more generically applicable. It is
possible to show through the use of this report the actual difference between a traditional
weekly full/daily incremental backup process versus Tivoli Storage Manager’s progressive
incremental approach. To generate this report, select Data Manager → Reporting →
Backup → Backup Storage Requirements → Incremental Range Sizes → By
Computer as shown in Figure 13-138.

Figure 13-138 Incremental Range Size select By Computer


After you select the Computers of interest, click Generate Report. Figure 13-139 shows the
output from this report, with the amount of data changed for different time ranges. Note that
the values are cumulative, so for each time range; the values shown include the smaller time
periods.

Figure 13-139 Incremental Range Sizes Report

13.12 TotalStorage Productivity Center for Data ESS Reporting
The reporting capabilities in TotalStorage Productivity Center for Data are expanded in
Version 1.2 to include Enterprise Storage Subsystem (ESS) reporting. IBM Tivoli Storage
Resource Manager uses Probe jobs to collect information about the ESS. We can then use
the reporting facility to view that information. The new subsystem reports show the capacity,
controllers, disks, and LUNs of an ESS and their relationships to computers and filesystems
within a network.

13.12.1 ESS Reporting
For this section we discuss ESS asset and storage subsystem reporting, making references
to the ESS lab environment in Figure 13-140 below. Note that the host which accesses the
ESS had a TotalStorage Productivity Center for Data Agent installed. This provides the fullest
combination of reporting ability for the ESS. If an ESS-attached host does not have a
TotalStorage Productivity Center for Data Agent installed, items such as filesystem, logical
volume, and device logical names will not be displayed.


Win2k Srv sp3
CIM/OM server
w2kadvtsm
172.31.1.135

43p
AIX 5.1 ML 4 ESSF20
ITSRM Agent 172.31.1.1
tsmsrv43p
172.31.1.155 2109

Win2k Srv sp3
ITSRM Server
w2kadvtsrm
172.31.1.133

Intranet

Figure 13-140 ESS reporting lab

Prerequisites to ESS Reporting
Before doing ESS reporting with Data Manager, the following conditions are required:
CIM/OM server successfully installed.
Data Manager successfully logs into CIM/OM server.
Data Manager successfully runs a discovery and probes the ESS.

Important: Refer to Chapter 5, “CIMOM install and configuration” on page 191 and 8.1,
“Configuring the CIM Agents” on page 290 for additional details on confirming these
prerequisites.

Data Manager will run a discovery to locate the CIM/OM server in our environment, which in
turn discovers the ESSs. See 8.1.2, “Configuring CIM Agents” on page 290.

Creating the ESS Probe
IBM Tivoli Storage Resource Manager will then run a Probe to query the discovered ESS.
The Probe collects detailed statistics about the storage assets in our enterprise, such as
computers, storage subsystems, disk controllers, hard disks, and filesystems.


Next, we show how to create a Probe for an ESS-F20. Select Probes → Select new probe,
then under the Computers tab, choose Storage Subsystems. See Figure 13-141.

Figure 13-141 Creating ESS probe

On the When to PROBE tab, we selected PROBE Now because we need to populate the
backend repository. See Figure 13-142.

Figure 13-142 ESS - When to probe


Next is the Alert tab, shown in Figure 13-143. This defines the type of notification for a Probe.

Figure 13-143 ESS - Alert tab

After all parameters are defined, save the Probe definition. At this point the Probe is
submitted and will run immediately.

Note: For additional information on creating Probes, see 13.3.5, “Probes” on page 545.

There are several ways to check the status of the Probe job. First, we can check the color of
the Probe job entry in the navigation tree, then in the content panel. There are two colors that
represent job status. They are:
GREEN - Job successfully complete with no errors
RED - Job completed with errors


The status of the Probe job is displayed in text and in color, as shown in Figure 13-144, after
selecting the Probe job output in the navigation tree. The job at 1:55 pm is in green, indicating
success.

Figure 13-144 ESS - probe job status

We open the Probe job by selecting it and double-clicking the spy glass icon next to the job
in the content window. We see the contents of the job, including detailed information on the
status, as in Figure 13-145. Here, we have selected the successful Probe.

Figure 13-145 Probe job log


Asset Reports - By Storage Subsystem
With Asset reporting by storage subsystem, you can view the centralized asset repository that
Data Manager constructs during a Probe. The Probe itemizes the information about
computers, disks, controllers, and filesystems, and builds a hardware inventory of assets.
With the backend repository now populated with DS6000 asset information, we will show how
to view reports to display the storage resources.

We choose Data Manager → Reporting → Asset → By Storage Subsystem → Tucson
DS6000. This report provides specific resource information of the DS6000 and allows us to
view storage capacity by a computer, filesystem, storage subsystem, LUN, and disk level. We
can also view the relationships between the components of a storage subsystem. Notice that
the navigation tree is hierarchical. See Figure 13-146.

Figure 13-146 Asset by storage subsystem


We drill down to the Disk Groups. The disk group contains information related to the ESS, as
well as the volume spaces and disks associated with those Disk Groups. Expanding the Disk
Group node, a list of all Disk Groups on the ESS displays (Figure 13-147).

Figure 13-147 ESS disk group

Continuing, we expand the disk group DG1 to view the disks and volume spaces within it.
We open Volume Space VS3, which shows the disks and LUNs associated with it. The Disks
subsection shows the individual disks associated with the Volume Space (see Figure 13-148).


Figure 13-148 Disks in volume spaces

Notice the LUNs subsection for disk DD0105 (Figure 13-149). This shows the LUN to disk
relationship. The LUNs shown here are just a subset of all the LUNs. You can see that the
LUN is spread across all the displayed disks in the content window.

Figure 13-149 Disk and LUN association with volume space


Figure 13-150 shows the discovery of a disk with no LUN associations. This is known as a hot
spare. It can be used when one of the other seven disks in the disk group fails.

Figure 13-150 Hot spare LUN


We now show a high level view of all disks in ESSF20. There are 32 disks in the ESS, as
shown in Figure 13-146 on page 639 in the Number of Disks field. Figure 13-151 shows a
partial listing of the disks.

Figure 13-151 ESS all disks


We can also display a report of all the LUNs in the ESS. This report provides the physical disk
association with each LUN. We have a total of 56 LUNs in the ESSF20 as shown in
Figure 13-146 on page 639 (number of LUNS). A partial listing is shown in Figure 13-152.

Figure 13-152 ESS all LUNs


Storage Subsystem Reporting
We now open Reporting → Storage subsystems. Storage Subsystems Reporting allows
viewing storage capacity at a computer, filesystem, storage subsystem, LUN, and disk level.

By Computer
We drill down Computers Views → By Computer. The report displays the association of
filesystems to the storage subsystem, LUNS, and disks on ESSF20. These reports are useful
for relating computers and filesystems to different storage subsystem components. There are
three options available in the Relate Computers to: pull down, as shown in Figure 13-153.

Figure 13-153 By Computer - Relate Computer to

We select Storage Subsystems from the pull down, select the desired computer and click
Generate. Figure 13-154 shows that the generated report TSMSRV43P uses 9.24 GB in the
ESS.

Figure 13-154 By Computer - storage subsystem


Returning to the selection screen tab (Figure 13-153 on page 645) we select LUNs. We
choose the same host, and click Generate. Figure 13-155 shows the generated report; the
relationship between TSMSRV43P and its assigned LUNs. TSMSRV43P has one LUN
created on the ESS.

Figure 13-155 By Computer - LUNs

Finally, from the Selection tab (Figure 13-153 on page 645), we select Disks, our host
TSMSRV43P, and click Generate. Figure 13-156 shows the report: the ESS disks assigned to
the LUN on the host.

Figure 13-156 By Computer - disk


By Filesystem/Logical Volume
We will now drill to Computer Views → By Filesystem/Logical Volume. The report
displays the association of filesystems to the storage subsystem, LUNS, and disks on
ESSF20. These reports are useful for relating computers and filesystems to different storage
subsystem components. There are three options available in the Relate Filesystem/Logical
Volumes to pull down, shown in Figure 13-157.

Figure 13-157 By filesystem/logical volume

Select Storage Subsystem, the host (TSMSRV43P), and click Generate. Figure 13-158
shows the filesystems on the host, which are located on the ESS.

Figure 13-158 By filesystem/logical volumes - storage subsystem


From the Selection tab (Figure 13-157 on page 647) we now choose LUNs, the host
(TSMSRV43P), and click Generate. Figure 13-159 shows the LUN location of each
filesystem on the host.

Figure 13-159 By filesystem/logical volume - LUN

From the Selection tab (Figure 13-157 on page 647) we now choose Disks, the host
(TSMSRV43P), and click Generate. Figure 13-160 shows which disks are comprising each
filesystem and logical volume.

Figure 13-160 By filesystem/logical volume - Disk


By Storage Subsystem
We will now drill down Storage Subsystem Views → By Storage Subsystem. These
reports display the relationships of the ESS components (storage subsystems, LUNs, and
disks) to the computers and filesystems and logical volumes. There are two options available
in the Relate Storage Subsystems to: the pull down, shown in Figure 13-161.

Figure 13-161 By Storage Subsystems

Select Computers from the pull down, the subsystem ESSF20, and click Generate.
Figure 13-162 shows the space used by each host on the storage subsystem.

Figure 13-162 By Storage subsystem - Computer


Now, select Filesystem/logical Volumes from Figure 13-161, the ESSF20 subsystem, and
click Generate. Figure 13-163 shows each host’s filesystems and logical volumes, with their
capacity and free space.

Figure 13-163 By storage subsystem - filesystem/logical volume

By LUN
Continuing, we drill down Storage Subsystem Views → By LUNs (Figure 13-164).

Figure 13-164 By LUNs


Select Computer from the Relate LUNs to: pull down, select the subsystem (ESSF20) with
the associated disks (default is all), and click Generate Report. Figure 13-165 shows the
LUNs assigned to each host, with the host’s logical name for the LUN (/dev/hdisk1 in this
case).

Figure 13-165 By LUN - computer

Now select Filesystem/Logical Volumes from the Relate LUNS to pull down, the ESSF20
subsystem with associated logical disks (default is all), and click. Next, we clicked Generate
Report. Figure 13-166 shows the relationships between the LUNs, computers, and
filesystems/logical volumes, including free space and host device logical names.

Figure 13-166 By LUNS - filesystem/logical volumes


Disks
Now we drill to Storage Subsystem Views → Disks. There are two options available in the
Relate Disks to: pull down, shown in Figure 13-167.

Figure 13-167 Disks

Select Computer from the pull down, the ESSF20 subsystem with related disks (default is
all), and click Generate Report. Figure 13-168 shows the relationships of the disks to the
hosts.

Figure 13-168 Disks - computer


Now select Filesystem/Logical Volumes from the pull down (Figure 13-167 on page 652),
the ESSF20 subsystem with related disks (default is all), and click Generate Report.
Figure 13-169 shows the relationship between the ESS disks and the filesystems and logical
volumes.

Figure 13-169 Disks - filesystem/logical volumes

Note: For demonstration purposes, we have reduced some of the fields in the reports.

13.13 IBM Tivoli Storage Resource Manager top 10 reports
After analyzing typical customer scenarios, we have compiled the following list of “Top 10
reports” which we recommend running regularly for best practices:
ESS used and free storage
ESS attached hosts report
Computer Uptime
Growth in storage used and number of files
Incremental backup trends
Database reports against DBMS size
Database Instance storage report
Database reports size by instance and by computer
Locate the LUN on which a database is allocated
Finding important files on your systems

13.13.1 ESS used and free storage
This report shows the free and used storage on an ESS system. To generate this filesystem
logical view report, navigate Data Manager → Reporting → Storage Subsystem →
Computer Views → By Filesystem/Logical Volumes. Select the computers to report on,
and select Disks from the pull-down Relate Filesystems/Logical Volumes To as in
Figure 13-170.


Figure 13-170 ESS relation to computer selected by disk

Click Generate Report. The report is shown in Figure 13-171. Various columns are
displayed:
Storage Subsystem
Storage Subsystem Type
Manufacturer
Model
Serial Number
Computer
Filesystem/Logical Volume Path
Capacity
Free Space
Physical Allocation

Figure 13-171 Report for Filesystem/Logical Volumes Part 1


Figure 13-172 shows the right hand columns of the same report.

Figure 13-172 Report for Filesystem/Logical Volumes Part 2

This report provides quick answers to how much space on the ESS is allocated to each
filesystem.

Select LUNs this time from the pull-down in Figure 13-170 on page 654. The report in
Figure 13-173 shows the LUN to host mapping for the ESS, which filesystem is associated
with each LUN, and the free space.

Figure 13-173 Computer view to the filesystem with capacity and free space


13.13.2 ESS attached hosts report
This report shows which systems are using storage on an ESS. This is useful when ESS
maintenance is applied so that the administrators of affected systems can be informed.

To generate this report, select Data Manager → Reporting → Storage Subsystem →
Computer Views → By Computer tree. We have selected all computers as in
Figure 13-174.

Figure 13-174 ESS selection per computer

Click the Generate Report field - the report is shown in Figure 13-175.

Figure 13-175 ESS connections to computer report


Note that you can sort the report on a different column heading by clicking it. The current sort
field is indicated by the small pointer next to the field name. Clicking again in the same
column reverses the sort order.

13.13.3 Computer Uptime Reporting
Uptime is an important IT metric in the enterprise. To generate a Computer Uptime report,
select Data Manager → Reporting → Availability → Computer Uptime → By
Computer. Select the computers of interest by clicking the Selection... button and checking
the boxes next to the desired computers in the Computer Selection window (Figure 13-176)
and click OK.

Figure 13-176 Computer Uptime Report - computer selection

In the Selection window, specify a date range (optional), and click Generate Report, as

Figure 13-177 Computer Uptime report selection


For each computer, percent availability, number of reboots, total down time, and average
downtime is given, as in Figure 13-178 shows the selection. The default sort order is by
descending Total Down Time.

Figure 13-178 Computer Uptime report part 1

You can also display this information graphically, by selecting the pie chart icon at the top of
the report, as shown in Figure 13-179.

Figure 13-179 Computer Uptime report graphical combined (stacked bar)


Figure 13-180 shows an unstacked bar chart of the same information (right-click and select
Bar Chart).

Figure 13-180 Computer Uptime report graphical (bar chart)

13.13.4 Growth in storage used and number of files
The Backup Reporting features of Data Manager also give a convenient way to track the total
storage used by files in each computer, as well as the number of files stored. It can be
presented graphically, to show historical numbers and future trends. This information helps
you plan future storage requirements, be alerted to potential problems, and also (if using a
traditional full and incremental backup product), plan your backup server storage
requirements, since this report shows the size of a full backup on each computer.
Select Data Manager → Reporting → Backup → Backup Storage Requirements →
Full Backup Size → By Computer. We used the Profile: TPCUser.Summary By Filesystem/
Directory and selected all computers, as in Figure 13-181. Click Generate Report.

Figure 13-181 Generate Full Backup Size report


Figure 13-182 shows the total disk space used by all the files, and the number of files on each
computer. The top column shows the totals for all Agents.

Figure 13-182 Select History chart for File count

To drill down, select all the computers (using the Shift key) so they are highlighted, then click
the pie icon, and select History Chart: Space Usage for Selected. The generated report
(Figure 13-183), shows how the total full backup size has fluctuated, and is predicted to
change in the future (dotted lines - to disable this, click Hide Trends).

Figure 13-183 History Chart: Space Used


To display the file count graph, select History Chart: File count from the pie icon in
Figure 13-182. The output report is shown in Figure 13-184, which shows trends in the
number of files on each computer.

Figure 13-184 History chart: File Count

These reports will help you find potential problems (e.g. a computer system that shows an
unexpected sudden upward or downward spike) and also predicts disk and backup
requirements for the future.

13.13.5 Incremental backup trends
This report shows the rate of modification of files, which is very useful for incremental backup
planning.

Select Data Manager → Reporting → Backup → Backup Storage Requirements →
Incremental Range Size → By Filesystem. Select Profile: TPCUser.By Modification as


Figure 13-185 Incremental Range selection based on filespace

The generated report shows all the filesystems on the selected computers as in
Figure 13-186.

Figure 13-186 Summary of all filespace

The third column shows the total number and total size of files (for all the systems, then
broken down by filesystem). Then there are “Last Modified” columns for one day, one week,
one month, two months, three, six, nine, and one year selections. Each of these gives the
number and size of the modified files.


To generate charts, highlight all the systems, and click the pie icon. Select Chart: Count
Distribution for Selected, as shown in Figure 13-187.

Figure 13-187 Selection for Filesystem and computer to generate a graphic

The chart is shown in Figure 13-188. Note that when your cursor passes over a bar, a pop-up
shows the number of files associated with that bar.

Figure 13-188 Bar chart for Incremental Range Size by Filesystem


You can display other filesystems using the Next 2 and Prev 2 buttons. Change the chart
format by right-clicking and selecting a different layout. Figure 13-189 is a pie chart of the
same data. The pop-ups work here also.

Figure 13-189 Pie chart selected with number of files which have modified

With these reports you can track and forecast your backups. You can also display backup
behavior for the last one, three, nine, or 12 months.


13.13.6 Database reports against DBMS size
This report shows an enterprise wide view of storage usage by all RDBMS. Select Data
Manager - Databases → Reporting → Capacity → All DBMSs → Total Instance
Storage → Network-wide and click Generate Report.
Figure 13-190 shows a sample output.

Figure 13-190 Total Instance storage used network wide

This is a quick overview database space consumption across the network. To drill down on a
particular RDBMS type, select the appropriate magnifying glass icon as in Figure 13-191.

Figure 13-191 DBMS drill down to the computer reports


The report (Figure 13-192) displays.

Figure 13-192 DBMS drill down to the computer result

Figure 13-192 shows the fields for an Oracle database. The fields for a DB2 database are as
follows:
Computer name
Total Size
Container Capacity
Container Free Space
Log File Capacity
Tablespace Count
Container Count
Log File Count


13.13.7 Database instance storage report
This report shows storage utilization by database instance. Go to Data Manager - Databases
→ Reporting → Capacity → UDB → Total Instance Storage → by Instance, select the
computer(s) of interest, and click Generate Report. Figure 13-191 shows the result.

Figure 13-193 DBMS report Total Instance Storage by Instance

Note you could select any RDBMS which is installed in your network.

The report shows the following information for each Agent with DB2, plus a total (summary):
Computer name
RDBMS instance
RDBMS type
Total size
Container Capacity
Container free space
Log file capacity
Tablespace count
Container count
Log file count

13.13.8 Database reports size by instance and by computer
The next report is based on the previous report (database Instance storage report), but in
more detail. From the report in Figure 13-193, click the magnifying glass next to a computer of
interest. Then do a further drill down on the generated report as in Figure 13-194.


Figure 13-194 Instance report RDBMS overview

Select the computer again, and click the magnifying glass. The report shows the entire DB2
environment running on computer Colorado. We have 10 DB2 UDB databases, shown in
Figure 13-195 and Figure 13-196.

Figure 13-195 Instance running on computer Colorado first part


Scroll to the right side of the panel (Figure 13-196).

Figure 13-196 Instance running on computer Colorado second part

Here we can see which databases are running in ARCHIVELOG mode.

13.13.9 Locate the LUN on which a database is allocated
This report shows you which disk or LUN is used by a database. Go to Data Manager -
Databases → Reporting → Capacity → UDB → Total Instance Storage → By
Instance, select the Agent(s) of interest, then click Generate Report. Figure 13-197 shows
the result.

Figure 13-197 LUN report selection for a database


Select an Agent, and click the magnifying glass to drill down. Figure 13-198 displays.

The report shows the following columns:
File Type
Path
File Size
Free Space
Auto Extend of an File

Figure 13-198 Database select File and Path


Select now a particular data file, and click the magnifying glass. The generated pie chart is
shown in Figure 13-199. We can see this data file is allocated on the C: drive.

Figure 13-199 Report DB2 File in a Pie Chart for DB2 File

Click the View Logical Volume button at the bottom to display the LUN report
(Figure 13-200).

Figure 13-200 LUN information


Using this procedure, we can find the LUNs where all the database data files are stored. This
information is useful for a variety of purposes, e.g. for performance planning, availability
planning, and assessing the impact of a LUN failure.

13.13.10 Finding important files on your systems
This report generates a search for specific files over all computers managed by a Data
Manager Server.

As an example, we created a text file each on Lochness and Wisla called lochness.txt and
wisla.txt, respectively. We have chosen this search for all machines because it will return a
relatively small number of results; however, any search criteria could be used.

The task requires a number of steps:
1. Define new Profile
2. Bind new Profile into a Scan
3. Generate a Report with your Profile
4. Define new Constraint
5. Generate a Report to find defined Constraint
1. Define the new Profile
First create the Profile - Data Manager → Monitoring → Profiles, right-click, and select
Create Profile. Fill out the description field accordingly, and check the Summarize space
usage by, Accumulate history, and Gather information on the fields as desired. In the
bottom half click size distribution of files, as shown in Figure 13-201.

Figure 13-201 Create Profile for own File search


Now select the File Filter tab. Click in the All files selected area and right-click to create
a new condition, as shown in Figure 13-202.

Figure 13-202 Create new Condition

Enter the desired file pattern into the Match field, and click Add to bring the condition to
the display window below, as in Figure 13-203. You can select from different conditions:
– Matches any of
– Matches none of
– Matches
– Does not match
When you have finished the condition, click OK. In our case we are matching Tivoli
Storage Manager option files.

Figure 13-203 Create Condition add


Figure 13-204 shows our newly created Condition.

Figure 13-204 Saved Condition in new Profile

Now save the new Profile with an appropriate name, (in this instance, Search for files).
The saved Profile now appears in the Profiles list, see Figure 13-205 on page 675.

Tip: We recommend choosing meaningful Profile names, which reflect the content or
function of the profile.


Figure 13-205 Listed Profiles containing Search for files

2. Bind new Profile into a Scan.
First, create a new Scan - Data Manager → Monitoring → Scans. We chose
TPCUser.Default Scan as shown in Figure 13-206 on page 675. Fill in a description for
this Scan and select the Filesystems and Computers on which the Scan will run.

Figure 13-206 Add Profile to Scan


On the Profiles tab, select the newly created Profile and add it to the Profiles to apply to
Filesystems column, as shown in Figure 13-207.

Figure 13-207 Add Profiles to apply to filesystems

Now select the schedule time when the schedule should run, save the Scan, then check
the result.
3. Generate Report with your Profile.
To view the results, select Data Manager → Reporting → Usage → Files → File Size
Distribution → By Filesystem. Select all filesystems you which to report, select the
Profile: administrator (Figure 13-208), and click Generate Report. The report contains
all the option files discovered by the Scan as in Figure 13-209.


Figure 13-208 Select Profile: tpcadmin.Search for files

Figure 13-209 Report with number of found Search for files

Note that on LOCHNESS and WISLA C drive we found 1 file each.


4. Define new Constraint.
We would like to know where specifically these files are located. To set up this search,
select Data Manager → Policy Management → Constraints → TPCUser.Orphaned
File Constraint, as shown in Figure 13-210. Enter a description, and select the
Filesystem Groups and Computers where you want to locate the files.

Figure 13-210 Create Orphaned File search

Select the Options tab, then select Edit Filter as shown in Figure 13-211.

Figure 13-211 Update the Orphaned selection


On the Edit Filter pop-up, double click the ATTRIBUTES Filter. Here we will replace the
ORPHANED condition with our own filter, since we want to actually search for the text files
we created, not orphaned files (Figure 13-212).

Figure 13-212 Update the selection with own data

Use the Del button to delete the ORPHANED condition, then select NAME from the
Attributes pull-down, and the Add button to add another Attributes condition. We will
specify to search for the text files we created, as in Figure 13-213.

Figure 13-213 Enter the file search criteria


After each file pattern entry, click Add to save it. When all search arguments are entered,
click OK to save the search. The selection is now complete as in Figure 13-214.

Figure 13-214 File Filter selection reconfirm

Click OK again. Save the search with a new description and name (File → Save As), so
that you do not overwrite the original TPCUser.Orphaned File Constraint. We called saved
the file as “File search.”
Finally, we run the Scan and check the Scan job log for correct execution, as shown in
Figure 13-215.

Figure 13-215 Scan log check


5. Generate Report to find defined Constraint
Now look for the results of the file name search. Select Data Manager → Reporting →
Usage Violations → Constraint Violations → By Computer, select all computers and
generate the report. The report will present a summary as in Figure 13-216.

Figure 13-216 Summary report of all Tivoli Storage Manager option files

To drill down, click the magnifying glass on WISLA as in Figure 13-217. This shows all the
filesystems on WISLA where matching files were found.

Figure 13-217 File selection for computer WISLA


Click the magnifying class on a filesystem (C drive, in this case). This will show all the files
found which matched the pattern, as in Figure 13-218. Note that there was 1 file reported,
which matches the summary view given in Figure 13-209 on page 677.

Figure 13-218 Report for Tivoli Storage Manager Option file searched

You can also drill down to individual files, for detailed information as in Figure 13-219.

Figure 13-219 File detail information


13.14 Creating customized reports
Customized Reporting within Data Manager is done through the My Reports option, which is
available for both Data Manager and Data Manager for Databases.

There are three main options available within My Reports:
System Reports
Reports owned by username
Batch Reports

System Reports, while included here in the customized reporting section, is in fact not
customizable currently. We will still discuss it in this section as it is part of the My Reports
group.

Reports owned by username’s Reports, where username is the currently logged in Data
Manager username, are modified versions of standard reports from the Reporting option. You
will only see reports here that you have modified and saved.

Batch Reports are reports that are typically set up to run on a schedule, although they can be
run interactively. The key difference between Batch Reports and other reporting options is
that with Batch Reports, the output will always be written to an output file rather than
displayed on the screen.

13.14.1 System Reports
These reports can, at this point in time at least, only be run as is. You cannot modify the
parameters in any way, nor can you add additional reports to the list.

These reports provide the same information than is available from running reports from the
Reporting option. The intent of these reports is to provide frequently needed information,
which can be provided quickly and repetitively without having to reenter parameters.


Data Manager
Figure 13-220 shows the available System Reports for Data Manager.

Figure 13-220 My Reports - System Reports

Figure 13-221 shows the output from running the Storage Capacity system report. We could
have generated exactly the same output by selecting Data Manager → Reporting →
Capacity → Disk Capacity → By Computer → Generate Report. Obviously, selecting
Data Manager → My Reports → Storage Capacity is a lot simpler.

Figure 13-221 My Reports - Storage Capacity


The System Reports available for Data Manager for Databases are shown in Figure 13-222.
While there are quite a few reports available, they fall into three main categories:
Database storage by database
Database storage by user
Database freespace

The only report that does not fall into one of those categories is a usage violation report.

Figure 13-222 shows the output from the All Dbms - User Database Space Usage report. We
are not so much interested in the report contents as such here, but rather in the fact that when
the report was run it produced a report for all users. You can go back to the selection tab and
select specific users if required. This capability exists for all of the System Reports.

Figure 13-222 Available System Reports for databases


13.14.2 Reports owned by a specific username
In concept this option is very similar to System Reports. You can include here those reports
that you need to run regularly, consistently and easily. The difference, compared to System
Reports, is that you get to decide what reports are included and what they look like.

However, it is important to remember that you will only see those reports that have been
created by the currently logged in TotalStorage Productivity Center for Data username.

Data Manager
We will define a report here for tpcadmin, the username that we are currently logged in as.

We will create a report that is exactly the same as the Storage Capacity system report as
shown in Figure 13-221 on page 684. In practice this is not something you would normally do
as a report already exists. However, this will demonstrate more clearly how the options relate
to each other.

We select Data Manager → Reporting → Capacity → Disk Capacity → By Computer
and click Generate Report. Once the report is produced, we save the report definition, using
the name My Storage Capacity. This is shown in Figure 13-223.

Figure 13-223 Create My Storage Capacity report


Once the report is saved you will see it available under username’s Reports for tpcadmin as

There are a few features of saved reports worth mentioning here. Firstly, characteristics such
as sort order are not saved with the report definition; however, selection criteria are saved.
Secondly, you can override the selection criteria when running your report. By default the
objects selected at the time of the save only will be reported. However, you can use the
Selection tab when running the saved report to include or exclude objects from the report. If
you change selection criteria you can resave the report, or save it under another name to
update the definition or create a new definition respectively.

Figure 13-224 My Storage Report saved


Database Reports created for specific users, in this case tpcadmin, are set up the same as in
Data Manager.

We will show one brief example here. We will take one of the reports that we created earlier in
our discussion on Reporting (in this case Figure 13-100 on page 609) the Monitored Tables
by RDBMS Type report and set it up to be able to run more easily.
First we run the report by choosing Data Manager - Databases → Reporting → Usage →
All DBMSs → Tables → Monitored Tables → By RDBMS Type and click Generate
Report. We then saved the report definition, naming it Monitored Tables by RDBMS Type.
This is shown in Figure 13-225.

Figure 13-225 Monitored Tables by RDBMS Types customized report

The report is more easily run now by choosing IBM Tivoli SRM for Databases → My
Reports → username’s Reports → Monitored Tables by RDBMS Type.

13.14.3 Batch Reports
In this section we will show how we set up some Batch Reports. All of the reports were set up
in the same way so we will use only one as an example. The process is the same whether the
report is for Data Manager or Data Manager for Databases.


Data Manager
To set up a new report Data Manager → My Reports → Batch Reports right-click Batch
Reports and select Create Batch Report. You will then see the screen shown in
Figure 13-226.

Figure 13-226 Create a Batch Report

Now, it is a simply a matter of specifying what has to be reported, plus when and what the
output should be. In this case we are going to create a system uptime report. As shown in
Figure 13-227, we entered our report description of System Uptime and have then selected
Availability → Computer Uptime → By Computer and clicked >>. Our selection is then
moved into the right hand panel, Current Selections.

Figure 13-227 Create a Batch Report - report selection


We then selected the Selection tab, which is shown in Figure 13-228. Here we are able to
select a subset of available data by either reporting for a specified time range or a subset of
available systems. We took the defaults here.

Figure 13-228 Create a Batch Report - selection

On the Options tab, we specified that the report should be executed and generated on the
Agent called COLORADO, which is our Data Manager server. We selected HTML for Report
Type Specification and then changed the rules for the naming of the output file under Output
File Specification.

By default the name will be {Report creator}.{Report name}.{Report run number}. In this
case we do not really care who created the report and having a variable like report run
number, which changes every time a new version of the report is created and makes it
difficult to access the file from a static Web page. So we changed the report name to be
{Report name}.html.

The report will be created in <install-directory>logData-agent-namereports on the Agent
system where the report job is executed. There is no ability to override the directory name.
For example, C:Program FilestivoliepsubagentsTPCDatalogcoloradoreports on our
Windows 2000 Data Manager server COLORADO or /usr/tivoli/tsrm/log/brazil/reports on an
AIX Data Manager Agent called BRAZIL.

The Option tab is shown in Figure 13-229.

Note here that it possible to run a script after the report is created to perform some type of
post-processing. For example, you might need to copy the output file to another system if
your Web server is on a system that is not running a Data Manager Agent.


Figure 13-229 Create a Batch Report - options

On the When to REPORT tab we specified when the report should be generated. We chose
REPORT Repeatedly and then selected a time early in the morning (3:00 AM) and specified
that the report should be generated every day. This is shown in Figure 13-230.

Figure 13-230 Create a Batch Report - When to REPORT


We left the Alert tab options as default, but it is possible to generate an Alert through several
mechanisms including e-mail, an SNMP trap, or the Windows event log should the generation
of the report fail.

Finally, we saved the report, calling it System Uptime, as shown in Figure 13-231.

Figure 13-231 Create a Batch Report - saving the report


We will use the same example here as we used in Figure 13.14.2 on page 686, that is a
Monitored Tables by RDBMS Type, but here we will save it in HTML format.

We choose Data Manager - Databases → My Reports → Batch Reports, right-click
Batch Reports and select Create Batch Report as shown in Figure 13-232.

Figure 13-232 Create a database Batch Report

Figure 13-233 shows the Report tab. We expanded in turn Usage → All DBMS’s → Tables
→ Monitored Tables → By RDBMS Type and clicked >>. We also entered a Description of
Monitored Tables by RDBMS Type.


Figure 13-233 Create a database Batch Report - Report tab

We accepted the defaults on the Selection tab, which is to report on all RDBMS types and
then went to the Options tab, shown in Figure 13-234. We set the Agent computer, which will
run the report to COLORADO.

Note that the system that you run the report on must be licensed for each type of database
that you are reporting on. If we were to run the report on COLORADO, the Data Manager
server system, we would need to have the Data Manager for Databases licences for Oracle
and SQL-Server licences loaded there even though COLORADO does not run these
databases.

We also set the report type to HTML and changed the output file name to be {Report
name}.html. This is shown in Figure 13-234.


Figure 13-234 Create a database Batch Report - Options tab

On the When to Report tab, shown in Figure 13-235, we chose REPORT Repeatedly and
set a start time.

Figure 13-235 Create a database Batch Report - When to Report tab

We did not change anything in the Alert tab. We saved the definition with the name Monitored
Tables by RDBMS Type as shown in Figure 13-236.


Figure 13-236 Create a database Batch Report - save definition

We can now run the report by choosing Data Manager - Databases → My Reports →
Batch Reports and then right-clicking tpcadmin.Monitored Tables by RDBMS Type and
choosing Run Now.

Figure 13-237 shows the output from the report execution.

Figure 13-237 Monitored Tables by RDBMS Type batch report output


13.15 Setting up a schedule for daily reports
Data Manager can produce reports according to a schedule. In our lab environment, we set
up a number of Batch Reports as shown in Figure 13-238. Note that the name of each of the
reports is prefixed by tpcadmin. This is the Windows username that we used to log into Data
Manager. Even though the reports were created by a particular user, other Data Manager
administrative users still have access to the reports (Data Manager non-administrative users
can only look at the results).

It is possible to generate output from Batch Reports in various formats including HTML,CSV,
(comma separated values) and formatted reports. For all of the reports that we set up, we
specified HTML as the output type, and also set them to run on a daily schedule. That way it
is easy to use a browser to quickly look at the state of the organization’s storage. It also
means that anyone can look at the reported data through their browser, without having
access to, or indeed, knowing how to use Data Manager. Obviously, if unrestricted access to
this data was not desirable some sort of password based security could be included within the
Web page.

Currently, all of the HTML output from Batch Reports is in table format - graphs cannot be
produced. There is also no ability to affect the layout of the reports in terms of sort order,
nominating the columns to be displayed or the column size. Using the interactive reporting
capability of the product does allow graphs to be produced and gives you some additional
capability in determining what the output looks like. To go further than that you can export to a
CSV file, and then use a tool such as Lotus 1-2-3® or Microsoft Excel to manipulate the
output.

Figure 13-238 Batch Reports listing

The next section shows how to develop the Web site.


13.16 Setting up a reports Web site
Since Data Manager can easily generate reports in HTML format, it is a logical extension to
set up a Web site where the reports can be easily viewed.

Since Data Manager itself is easy to install and use, we likewise took a fairly simplistic view to
creating the Web site. We used the Microsoft Word Web Page Wizard to create the basic
layout of the page as shown in Figure 13-239.

The main page has two frames. In the left hand frame we have created links to each of the
report files. The right hand frame is where the reports are displayed.

As additional Batch Reports are needed, it is a relatively simple process of editing the HTML
source and including another hot link.

Obviously, this could be made more sophisticated. An example would be to have the browser
list all HTML files within the report directory.

Figure 13-239 MS Word created Web page

We then used the Virtual Directory Creation Wizard within Microsoft Internet Information
Server (IIS) to set up access to the reports as shown in Figure 13-240. Detailed information
on using IIS is shown in 8.2.2, “Using Internet Information Server” on page 299.


Figure 13-240 Setting up a Virtual Directory within IIS

We could then access the reports through a Web browser as shown in Figure 13-241.

Figure 13-241 Reports available from a Web browser


13.17 Charging for storage usage
Through the Data Manager for Chargeback product, Data Manager provides the ability to
produce Chargeback information for storage usage. The following items can have charges
allocated against them:
Operating system storage by user
Operating system disk capacity by computer
Storage usage by database user
Total size by database-tablespace

For each of the Chargeback by user options, a Profile needs to be specified. Profiles are
covered in “Probes” on page 527.

Data Manager can directly produce an invoice or create a file in CIMS format. CIMS is a set of
resource accounting tools that allow you to track, manage, allocate, and charge for IT
resources and costs. For more information on CIMS see: http://www.cims.com.

Figure 13-242 shows the Parameter Definition screen. The costs allocated here do not
represent any real environment, but represent an example, based on these assumptions:
Disk hardware costs, including controllers and switches. is $0.50 per MB
Hardware costs are only 20% of the total cost over the life of the storage = $2.50 /MB
On average only 50% of the capacity is used = $5.00 /MB used
The expected life of the storage is 4 years - $5.00 /48 = 0.1042 /MB /month
The figures used are for monthly Chargeback
Chargeback is for cost recovery only, no profit

Figure 13-242 Chargeback parameter definition

In this example we have chosen to perform Chargeback by computer. It is possible to
separately charge for database usage and use a different rate from the computer rate. To do
this you would need to set up a Profile that excluded the database data, otherwise, it would
be counted twice.


Chargeback is useful, even if you do not actually collect revenue from your users for the
resources consumes. It is a very powerful tool for raising the awareness within the
organization of the cost of storage, and the need to have the appropriate tools and processes
in place to manage storage effectively and efficiently.

Figure 13-243 shows the Chargeback Report being created. Currently, it is not possible to
have the Chargeback Report created automatically (that is, scheduled).

Figure 13-243 Create the Chargeback Report

Example 13-5 shows the Chargeback Report that was produced.

Example 13-5 Chargeback Report
Data Manager - Chargeback page 1
Computer Disk Space Invoice Aug 23, 2005

tpcadmin.Linux Systems

NAME SPACE COST
GB 0.104/GB

klchl5h 0 0.00

group total 0 0.00


tpcadmin.Windows DB Systems

NAME SPACE COST
GB 0.104/GB

colorado 69 7.19
senegal 0 0.00

group total 69 7.19



tpcadmin.Windows Systems

NAME SPACE COST
GB 0.104/GB

gallium 59 6.15
lochness 75 7.82
wisla 75 7.82

group total 209 21.79


TPCUser.Default Computer Group

NAME SPACE COST
GB 0.104/GB

Cluster Group.DB2CLUSTER.ITSOSJNT 137 14.28

group total 137 14.28

Run Summary Aug 23, 2005

Computer Disk Space Invoice 415 GB 43.26

run total 43.26

Example 13-6 shows the Chargeback Report in CIMS format.

Example 13-6 Chargeback Report in CIMS format
TSAOC1,20050823,20050823,17:55:00,17:55:59,,3,tpcadmin,Linux Systems,klchl5h,1,0
TSAOC1,20050823,20050823,17:55:00,17:55:59,,3,tpcadmin,Windows DB
Systems,colorado,1,71687000
TSAOC1,20050823,20050823,17:55:00,17:55:59,,3,tpcadmin,Windows DB Systems,senegal,1,0
TSAOC1,20050823,20050823,17:55:00,17:55:59,,3,tpcadmin,Windows Systems,gallium,1,61762720
TSAOC1,20050823,20050823,17:55:00,17:55:59,,3,tpcadmin,Windows Systems,lochness,1,78156288
TSAOC1,20050823,20050823,17:55:00,17:55:59,,3,tpcadmin,Windows Systems,wisla,1,78156288
TSAOC1,20050823,20050823,17:55:00,17:55:59,,3,TPCUser,Default Computer Group,Cluster
Group.DB2CLUSTER.ITSOSJNT,1,142849536


14

Center for Fabric
In this chapter we provide an introduction to the features of TotalStorage Productivity Center
for Fabric. We discuss the following topics:
IBM Tivoli NetView navigation overview
Topology view
Data collection, reporting, and SmartSets


14.1 NetView navigation overview
Since TotalStorage Productivity enter for Fabric (formerly Tivoli SAN Manager) uses IBM
Tivoli NetView (abbreviated as NetView) for display, before going into further details, we give
you a basic overview of the NetView interface, how to navigate in it and how TotalStorage
Productivity Center for Fabric integrates with NetView. Detailed information on NetView is in
the redbook Tivoli NetView V6.01 and Friends, SG24-6019.

14.1.1 NetView interface
NetView uses a graphical interface to display a map of the IP network with all the components
and interconnect elements that are discovered in the IP network. As your Storage Area
network (SAN) is a network, TotalStorage Productivity Center for Fabric uses NetView and its
graphical interface to display a mapping of the discovered storage network.

14.1.2 Maps and submaps
NetView uses maps and submaps to navigate in your network and to display deeper details
as you drill down. The main map is called the root map while each dependent map is called a
submap. Your SAN topology will be displayed in the Storage Area Network submap and its
dependents. You can navigate from one map to its submap simply by double-clicking the
element you want to display.

14.1.3 NetView window structure
Figure 14-1 shows a basic NetView window.

submap window

submap
stack

child submap area

Figure 14-1 NetView window

704 IBM TotalStorage Productivity Center: Getting Started

The NetView window is divided into three parts:
The submap window displays the elements included in the current view. Each element can
be another submap or a device
The submap stack is located on the left side of the submap window. This area displays a
stack of icons representing the parent submaps that you have already displayed. It shows
the hierarchy of submaps you have opened for a particular map. This navigation bar can
be used to go back to a higher level with one click
The child submap area is located at the bottom of the submap window. This submap area
shows the submaps that you have previously opened from the current submap. You can
open a submap from this area, or bring it into view if it is already opened in another
window on the window.

14.1.4 NetView Explorer
From the NetView map based window, you can switch to an Explorer view where all maps,
submaps and objects are displayed in a tree scheme (similar to the Microsoft Windows
Explorer interface). To switch to this view, right-click a submap icon and select Explore as

Figure 14-2 NetView Explorer option

Chapter 14. Using TotalStorage Productivity Center for Fabric 705

Figure 14-3 shows the new display using the NetView Explorer.

Figure 14-3 NetView explorer window

From here, you can change the information displayed on the right pane by changing to the
Tivoli Storage Area Network Manager view on the top pull-down field. The previously
displayed view was System Configuration view. The new display is shown in Figure 14-4.

Figure 14-4 NetView explorer window with Tivoli Storage Area Network Manager view


Now, the right pane shows Label, Name, Type and Status for the device. You may scroll right
to see additional fields.

14.1.5 NetView Navigation Tree
From the any NetView window, you can switch to the Navigation Tree by clicking the tree icon
circled on Figure 14-5.

Figure 14-5 NetView toolbar

NetView will display, with a tree format, all the objects contained in the maps you have
already explored. Figure 14-6 shows the tree view.

Figure 14-6 NetView tree map

You can see that our SAN — circled in red — does not show its dependent objects since we
have not yet opened this map through the standard NetView navigation window.

You can click any object and it will open its submap in the standard NetView view.

14.1.6 Object selection and NetView properties
To select an object, right-click it. NetView displays a context-sensitive menu with several
options including Object Properties as shown in Figure 14-7.


Figure 14-7 NetView objects properties menu

The Object Properties for that device will display (Figure 14-8). This will allow you to change
NetView properties such as the label and icon type of the selected object.

Figure 14-8 NetView objects properties

Important: As TotalStorage Productivity Center for Fabric runs its own polling and
discovery processes and only uses NetView to display the discovered objects, each
change to the NetView object properties will be lost as soon as TotalStorage Productivity
Center for Fabric regenerates a new map.


14.1.7 Object symbols
TotalStorage Productivity Center for Fabric uses its own set of icons as shown in Figure 14-9.
Two new icons have been added for Version 1.2 - ESS and SAN Volume Controller.

Figure 14-9 Productivity Center for Fabric icons

14.1.8 Object status
The color of a symbol or the connection represents its status. The colors used by Productivity
Center for Fabric and their corresponding status are shown in Table 14-1.

Table 14-1 Productivity Center for Fabric symbols color meaning
Symbol color Connection color Status Status meaning

Green Black Normal The device was detected in at least one of the
scans

Green Black New The device was detected in at least one of the
scans and a new discovery has not yet been
performed since the device was detected

Yellow Yellow Marginal (suspect) Device detected - the status is impaired but still
functional

Red Red Missing None of the scans that previously detected the
device are now reporting it

IBM Tivoli NetView uses additional colors to show the specific status of the devices, however
these are not used in the same way by Productivity Center for Fabric (Table 14-2).

Table 14-2 IBM Tivoli NetView additional colors
Symbol color Status Status Meaning

Blue Unknown Status not determined

Wheat (tan) Unmanaged The device is no longer monitored for topology
and status changes.

Dark green Acknowledged The device was Missing, Suspect or Unknown.
The problem has been recognized and is being
resolved

Gray (used in NetView Unknown Status not determined
Explorer left pane)

If you suspect problems in your SAN, look in the topology displays for icons indicating a
status of other than normal/green. To assist in problem determination, Table 14-3 provides an
overview of symbol status with possible explanations of the problem.


Table 14-3 Problem determination
Display Agents Device Link Non-ISL ISL explanation
explanation

Any Normal Marginal One or more, but One or more, but
(green) (yellow) not all links to the not all links
device in this between the two
topology are switches is missing
missing.

Any Normal Critical All links to the All links between
(green) (red) device in this the two switches
topology are are missing, but
missing, while other the out-of-band
links to this device communication to
in other topologies the switch is
are normal. normal

Any Critical Critical All links to the All links between
(red) (red) device in this the two switches
topology are are missing, and
missing, while all the out-of-band
other links to communication to
devices in other the switch is
topologies are missing or
missing (if any) indicates that the
switch is in critical
condition

Both Critical Normal All in-band agents This condition
(red) (black) monitoring the should not happen.
device can no If you see this on
longer detect the an ISL where
device. For switches on either
example, a server side of the link
reboot, power-off, have an
shutdown of agent out-of-band agent
service, Ethernet connected to your
problems, and SAN Manager,
soon. then you are
having problems
with your
out-of-band agent.

Both Critical Marginal At least one link to This condition
(red) (yellow) the device in this should not happen.
topology is normal If you see this on
and one or more an ISL where
links are missing. In switches on either
addition, all in-band side of the link
agents monitoring have an
the device can no out-of-band agent
longer detect the connected to your
device SAN Manager,
then you are
having problems
with your
out-of-band agent.


14.1.9 Status propagation
Each object has a color representing its status. If the object is an individual device, the status
shown is that of the device. If the object is a submap, the status shown reflects the summary
status of all objects in its child submap. Status of lower level objects is propagated to the
higher submap as shown in Table 14-4.

Table 14-4 Status propagation rules
Object status Symbols in the child submap

Unknown No symbols with status of normal, critical, suspect or unmanaged

Normal All symbols are normal or acknowledged

Suspect (marginal) All symbols are suspect or
Normal and suspect symbols or
Normal, suspect and critical symbols

Critical At least one symbol is critical and no symbol are normal

14.1.10 NetView and Productivity Center for Fabric integration
Productivity Center for Fabric adds a SAN menu entry in the IBM Tivoli NetView interface,
shown in Figure 14-10. The SAN pull-down menu contains the following entries:
SAN Properties to display and change object properties, such as object label and icon
Launch Application to run a management application
ED/FI Properties to view ED/FI events
ED/FI Configuration to start, stop, and configure ED/FI
Configure Agents to add and remove agents
Configure Manager to configure the polling and discovery scheduling
Set Event Destination to configure SNMP and TEC events recipients
Storage Resource Manager to launch TotalStorage Productivity Center for Data
Help

Figure 14-10 SAN Properties menu


All those items will subsequently be described in more detail.

14.2 Walk-through of Productivity Center for Fabric
This section takes you through the Productivity Center for Fabric. It steps through different
views to help you understand how to use different panels.

To anyone familiar with NetView, this is similar. Productivity Center for Fabric uses NetView
to display your SAN along with your IP network. In the first view, you see three icons: IP
Internet, SmartSets, and SAN. We focus on the SAN icon.

Figure 14-11 shows the root display window when you first launch NetView. The green
background on the SAN icon indicates that all is well in that environment.

Figure 14-11 NetView root display


There are three different types of views in Productivity Center for Fabric: Device Centric view,
Host Centric view, and SAN view. In our configuration, the NetView display (Figure 14-12)
shows two separate SANs that we are monitoring: TPC SAN and TSM SAN.

Figure 14-12 SAN view

14.2.1 Device Centric view
The first view is the Device Centric view. From this view, you can drill down to see the device
point of view. In this example, we have a view of two IBM FAStT devices. The one we are
using is labeled FAStT-1T14859668. We drill down on that device to see which systems are
using LUNs from the FAStT.

Here we see that two LUNs are available. As we drill down on LUN1, we see that a host
named PDQDISRV has been assigned that LUN. If we go further, we can see that this
system is a Windows 2000 system.


14.2.2 Host Centric view
Now we investigate the Host Centric view.

Note: Only the systems that have the Productivity Center for Fabric agent installed on
them are displayed in this view.

The Host Centric view displays all host systems and their logical relationships to local and
SAN-attached devices. Here again we see a system called PQDISRV. If we drill down on this
system, we can see that this is a Windows 2000 system that has four file systems defined on
it. We can also look at the properties of those file systems. This enables us to see such
information as the type of file system, mount point, total amount of space and how much free
space is available. As we drill down further, we can see the logical volume or volumes behind
those file systems.

14.2.3 SAN view
The SAN view displays one symbol for each SAN. You can see from Figure 14-12 on
page 713 that there are two SANs. When we double-click the SAN icon labelled TPC SAN,
we see the underlying submap (Figure 14-13). From the submap, you can choose either the
Topology View or Zone View.

Figure 14-13 SAN subview


First we explore the Zone View. The Zone View displays information by zone groupings.
Figure 14-14 displays the information about the three zones that have been setup on the
Fibre Channel switch: the Colorado, Gallium and PQDI zones.

Figure 14-14 Zone View


We can drill down in each zone and see which system and devices have been assigned to
that specific zone. Figure 14-15 shows the Colorado zone in which there is one host and a
FAStT disk subsystem.

Figure 14-15 Colorado zone contents


Now we look at the Topology View (Figure 14-16). The Topology View draws a picture of how
the SAN is configured, which devices are connected to which ports, and so on. As we drill
down in the Topology View, we first see the interconnect elements. This shows you the
connection between any switches. In our small environment, we have only one switch, so the
only device connected is the itsosw4 switch, which is an IBM 2109-F16 switch.

Figure 14-16 Topology View of switches


If we had two switches in our SAN, we would see a switch icon on either side of the
Interconnect elements icon. As we drill down on the switch, we see what devices and
systems are directly attached to it. Figure 14-17 shows five hosts, the FAStT device, and the
IBM switch in the middle.

Figure 14-17 SAN topology

From here, we show you several features of Productivity Center for Fabric such as:
How to configure the manager and what happens when things go wrong
Properties of a host with the Productivity Center for Fabric agent installed
How to configure SNMP agents


We begin by showing what happens when things go wrong. Figure 14-18 shows that the
FAStT disk system has a redundant connection. Let’s see what happens when one
connection goes down.

Figure 14-18 FAStT dual connections


In Figure 14-19 you notice on the left, that all of the parent icons have turned yellow. This
indicates that something has happened in your SAN environment. You can then drill down,
following the yellow trail until you find the problem. Here we can see that one of the
connections to the FAStT disk system has gone down.

Figure 14-19 Failed resource

This gives an administrator a place to start looking. After they determine what the problem is,
they can take corrective action. The FAStT icon has turned Red, not because it has failed, but
so you can see that it is affected. In our case, we lost access to one of the controllers of the
FAStT, because it was the only path to that controller. If you right-click the FAStT icon and
then select Acknowledge, it changes back to Green if the device itself is OK. The path to the
icon still remains Yellow. When the problem is corrected, the topology is updated to reflect the
resolution.


Now let us see the kind of information that we can view from a host that has a Productivity
Center for Fabric agent installed on it. You select the required host, and then click
SAN →SAN Properties.

A Properties window (Figure 14-20) opens. It shows such information as IP address,
operating system, host bus adapter type driver versions, and firmware levels.

Figure 14-20 Properties of host GALLIUM


When you click the Connection tab on the left, you see the port on the switch to which the
specific host is connected as shown in Figure 14-21.

Figure 14-21 GALLIUM connection

Now that you have seen the agents and where you can define them, let’s look at the manager
configuration. The manager configuration is simple and enables you set the polling intervals.
Figure 9-46 on page 349 shows the polling setup, in which you specify how often you want
your agents to poll the SAN. You can set this to minutes, hours, days, weeks, which days you
want to poll on, and the exact time.

Or you can poll manually by clicking the Poll Now button. The Clear History button changes
the state of an object that previously had a problem but is back up. The state appears as
yellow, but the Clear History button changes it back to normal or green.


14.2.4 Launching element managers
Productivity Center for Fabric also has the ability to launch element managers. By element
manager, we are referring to applications that vendors use to configure their hardware.
Figure 14-22 shows the Productivity Center for Fabric launching the element manager for the
IBM 2109 Fibre Channel switch.

Figure 14-22 Launching an element manager


Figure 14-23 shows the management tool for the IBM 2109 after being launched from
Productivity Center for Fabric.

Figure 14-23 Switch management


14.2.5 Explore view
Along with the Productivity Center for Fabric Topology View, you can view your SAN
environment with a Windows Explore type view. By clicking the Submap Explorer button in
the center of the toolbar, you see a view like the example in Figure 14-24. The Navigation
Tree button shows a flowchart-type view of the Productivity Center for Fabric views.

Figure 14-24 Explorer view

14.3 Topology views
The standard IP-based IBM Tivoli NetView root map contains IP Internet and SmartSets
submaps. Productivity Center for Fabric adds a third submap, called Storage Area Network,
to allow the navigation through your discovered SAN. Figure 14-25 shows the NetView root
map with the addition of Productivity Center for Fabric.


Figure 14-25 IBM Tivoli NetView root map

The Storage Area Network submap (shown in Figure 14-26) displays an icon for each
available topology view. There will be a SAN view icon for each discovered SAN fabric (three
in our case), a Device Centric View icon, and a Host Centric View icon.

Figure 14-26 Storage Area Network submap


You can see in this figure that we had three fabrics. They are named Fabric1, Fabric3, and
Fabric4, since we have changed their label using SAN → SAN Properties as explained in
“Properties” on page 736.

Figure 14-27 shows the complete list of views available. In the following sections we will
describe the content of each view.

Topology views
Tivoli NetView root map

Storage Area Network

SAN view Device Centric view Host Centric view

Devices
Topology view Zone view Hosts
(storage servers)

Switches Interconnect elements Zones LUNs Platform

Elements Elements (switches) Elements Host Filesystems

Platform Volumes

Figure 14-27 Topology views

14.3.1 SAN view
The SAN view allows you to see the SAN topology at the fabric level. In this case we clicked
the Fabric1 icon shown in Figure 14-26 on page 726. The display in Figure 14-28 appears,
giving access to two further submaps:
Topology view
Zone view

Figure 14-28 Storage Area Network view


Topology view
The topology view is used to display all elements of the fabric including switches, hosts,
devices, and interconnects. As shown on Figure 14-29, this particular fabric has two switches.

Figure 14-29 Topology view

Now, you can click a switch icon to display all the hosts and devices connected to the
selected switch (Figure 14-30).

Figure 14-30 Switch submap


On the Topology View (shown in Figure 14-29 on page 728) you can also click Interconnect
Elements to display information about all the switches in that SAN (Figure 14-31).

Figure 14-31 Interconnect submap

The switch submap (Figure 14-30), shows that six devices are connected to switch
ITSOSW1. Each connection line represents a logical connection. Click a connection bar twice
to display the exact number of physical connections (Figure 14-32).
We now see that, for this example, SOL-E is connected to two ports on the switch ITSOSW1.

Figure 14-32 Physical connections view


When the connection represents only one physical connection (or, if we click one of the two
connections shown in Figure 14-32), NetView displays its properties panel (Figure 14-33).

Figure 14-33 NetView properties panel

Zone view
The Zone view submap displays all zones defined in the SAN fabric. Our configuration
contains two zones called FASTT and TSM (Figure 14-34).

Figure 14-34 Zone view submap


Click twice on the FASTT icon to see all the elements included in the FASTT zone
(Figure 14-35).

Figure 14-35 FASTT zone

In lab1, the FASTT zone contains five hosts and one storage server. We have installed
TotalStorage Productivity Center for Fabric Agents on the four hosts that are labelled with
their correct hostname (BRAZIL, GALLIUM, SICILY and SOL-E). For the fifth host, LEAD, we
have not installed the agent. However, it is discovered since it is connected to the switch.
Productivity Center for Fabric displays it as a host device, and not as an unknown device,
because the QLogic HBA drivers installed on LEAD support RNID. This RNID support gives
the ability for the switch to get additional information, including the device type (shown by the
icon displayed), and the WWN. The disk subsystem is shown with a question mark because
the FAStT700 was not yet fully supported (with the level of code available at the time of
writing) and Productivity Center for Fabric was not able to determine all the properties from
the information returned by the inband and outband agents.

14.3.2 Device Centric View
You may have several SAN fabrics with multiple storage servers. The Device Centric View
(accessed from the Storage Area Network view, as shown in Figure 14-26 on page 726),
displays the storage devices connected to your SANs and their relationship to the hosts. This
is a logical view as the connection elements are not shown. Because of this, you may prefer
to see this information using the NetView Explorer interface as shown in Figure 14-36. This
has the advantage of automatically displaying all the lower level items for Device Centric View
listed in Example 14-27 on page 727 simultaneously, such as LUNs and Host.


Figure 14-36 Device Centric View

In the preceding figure, we can see the twelve defined LUNs and the host to which they have
been allocated. The dependency tree is not retrieved from the FAStT server but is
consolidated from the information retrieved from the managed hosts. Therefore, the
filesystems are not displayed as they can be spread on several LUNs and this information is
transparent to the host. Note that the information is also available for the MSS storage server,
the other disk storage device in our SAN.

14.3.3 Host Centric View
The Host Centric View (accessed from the Storage Area Network view, as shown in
Figure 14-26 on page 726) displays all the hosts in the SAN and their related local and
SAN-attached storage devices. This is a logical view that does not show the interconnect
elements (and runs across the fabrics). Since this is also a logical view, like the Device
Centric View, the NetView Explorer presents a more comprehensive display (Figure 14-37).


Figure 14-37 Host Centric View for Lab 1

We see our four hosts and all their local filesystems whether they are locally or SAN-attached.
NFS-mounted filesystems and shared directories are not displayed. Since no agent is running
on LEAD, it is not shown in this view.

14.3.4 iSCSI discovery
For this environment we will reference SAN Lab 2 (“Lab 2 environment” on page 752).

Starting discovery
You can discover and manage devices that use the iSCSI storage networking protocol
through Productivity Center for Fabric using IBM Tivoli NetView. Before discovery, SNMP and
the iSCSI MIBs must be enabled on the iSCSI device, the Tivoli NetView IP Discovery must
be enabled. See 14.11, “Real-time reporting” on page 786 for enabling IP discovery.

The IBM Tivoli NetView nvsniffer daemon will discover the iSCSI devices. Depending on
the iSCSI operation chosen, a corresponding iSCSI SmartSet will be created under the IBM
Tivoli NetView SmartSets icon. By default, the nvsniffer utility runs every 60 minutes. Once
nvsniffer discovers a iSCSI device, it creates an iSCSI SmartSet located on the NetView
Topology map at the root level.

The user can select what type of iSCSI device is discovered. From the menu bar, click Tools
→ iSCSI Operations menu and select Discover All iSCSI Devices, Discover All iSCSI
Initiators or Discover All iSCSI Targets, as shown in Figure 14-38.

For more details about iSCSI, refer to 14.12, “Productivity Center for Fabric and iSCSI” on
page 810.


Figure 14-38 iSCSI discovery

Double-click the iSCSI SmartSet icon to display all iSCSI devices.

Once all iSCSI devices are discovered by NetView, the iSCSI SmartSet can be managed from
a high level. Status for iSCSI devices is propagated to the higher level, as described in 14.1.9,
“Status propagation” on page 711. If you detect a problem, drill to the SmartSet icon and
continue drilling through the iSCSI icon to determine what iSCSI device is having the
problem. Figure 14-39 shows an iSCSI SmartSet.

Figure 14-39 iSCSI SmartSet

14.3.5 MDS 9000 discovery
The Cisco MDS 9000 is a family of intelligent multilayer directors and fabric switches that
have such features as: virtual SANs (VSANs), advanced security, sophisticated debug
analysis tools and an element manager for SAN management.

Productivity Center for Fabric has enhanced compatibility for the Cisco MDS 9000 Series
switch. Tivoli NetView displays the port numbers in a format of SSPP, where SS is the slot
number and PP is the port number. The Launch Application menu item is available for the
Cisco switch. When the Launch Application is selected, the Cisco Fabric Manager application
is started. For more details, see 14.7.1, “Cisco MDS 9000 discovery” on page 745.


14.4 SAN menu options
In this section we describe some of the menu options contained under the SAN pull-down
menu option for Productivity Center for Fabric.

14.4.1 SAN Properties
As shown in Figure 14-40, select an object and use SAN → SAN Properties to display the
properties gathered by Productivity Center for Fabric. In this case we are selecting a
particular filesystem (the root filesystem) from the Agent SOL-E.

Figure 14-40 SAN Properties menu

This will display a SAN Properties window that is divided into two panes. The left pane always
contains Properties, and may also contain Connection and Sensors/Events, depending on
the type of object being displayed. The right pane contains the details of the object.

These are some of the device types that give information in the SAN Properties menu:
Disk drive
Hdisk
Host file system
LUN
Log volume
OS
Physical volume
Port
SAN


Switch
System
Tape drive
Volume group
Zone

Properties
The first grouping item is named Properties and contains generic information about the
selected device. The information that is displayed depends on the object type. This section
shows at least the following information:
Label: The label of the object as it is displayed by Productivity Center for Fabric. If you
update this field, this change will be kept over all discoveries.
Icon: The symbol representing the device type. If the object is of an unknown type, this
field will be in read-write mode and you will be able to select the correct symbol.
Name: The reported name of the device.

Figure 14-41 shows the Properties section for a filesystem. You can see that it displays the
filesystem name and type, the mount point, and both the total and available space. Since a
filesystem is not related to a port connection and also does not return sensor events, only the
Properties section is available.

Figure 14-41 Productivity Center for Fabric Properties — Filesystem

Figure 14-42 shows the Properties section for a host. You can see that it displays the
hostname, the IP address, the hardware type, and information about the HBA. Since the host
does not give back sensor related events, only the Properties and Connections sections are
available.


Figure 14-42 Productivity Center for Fabric Properties — Host

Figure 14-43 shows the Properties section for a switch. You can see that it displays fields
including the name, the IP address, and the WWN. The switch is a connection device and
sends back information about the events and the sensors. Therefore, all three item groups
are available (Properties, Connections, and Sensors/Events).

Figure 14-43 Productivity Center for Fabric Properties — Switch


Figure 14-44 shows the properties for an unknown device. Here you can change the icon to a
predefined one by using the pull-down field Icon. You can also change the label of a device
even if the device is of a known type.

Figure 14-44 Changing icon and name of a device

Connection
The second grouping item, Connections shows all ports in use for the device. This section
appears only when it is appropriate to the device displayed — switch or host.

On Figure 14-45, we see the Connection tab for one switch where six ports are used. Port 0 is
used for the Inter-Switch Link (ISL) to switch ITSOSW2. This is a very useful display, as it
shows which device is connected on each switch port.

Figure 14-45 Connection information


Sensors/Events
The third grouping item, Sensors/Events, is shown in Figure 14-46. It shows the sensors
status and the device events for a switch. It may include information about fans, batteries,
power supplies, transmitter, enclosure, board, and others.

Figure 14-46 Sensors/Events information

14.5 Application launch
Many SAN devices have vendor-provided management applications. Productivity Center for
Fabric provides a launch facility for many of these.


14.5.1 Native support
For some supported devices, Productivity Center for Fabric will automatically discover and
launch the device-related administration tool. To launch, select the device and then click SAN
→ Launch Application.

This will launch the Web application associated with the device. In our case, it launches the
Brocade switch management Web interface for the switch ITSOSW1, shown in Figure 14-47.

Figure 14-47 Brocade switch management application

14.5.2 NetView support for Web interfaces
For devices that have not identified their management application, IBM Tivoli NetView allows
you to manually configure the launch of a Web interface for any application, by doing the
following actions:
Right-click the device and select Object Properties from the context-sensitive menu.
On the dialog box, select the Other tab (shown in Figure 14-48).
Select LANMAN from the pull-down menu.
Check isHTTPManaged.
Enter the URL of the management application in the Management URL field.
Click Verify, Apply, OK.


Figure 14-48 NetView objects properties — Other tab

After this, you can launch the Web application by right-clicking the object and then selecting
Management Page, as shown in Figure 14-49.

Figure 14-49 Launch of the management page

Important: This definition will be lost if your device is removed from the SAN and
subsequently rediscovered, since it will be a new object for NetView.


14.5.3 Launching TotalStorage Productivity Center for Data
The TotalStorage Productivity Center for Data interface can be started by using TotalStorage
Productivity Center for Fabric NetView console. To do this, select SAN → Storage Resource
Manager, as shown in Figure 14-50.

Figure 14-50 Launch Tivoli Storage Resource Manager

The user properties file contains an SRMURL setting that defaults to the fully qualified host
name of Tivoli Storage Area Network Manager. This default assumes that both TotalStorage
Productivity Center for Disk and TotalStorage Productivity Center for Fabric are installed on
the same machine. If TotalStorage Productivity Center for Data is installed on a separate
machine, you can modify the SRMURL value to specify the host name of the TotalStorage
Productivity Center for Data machine. For instructions on how to do this, please refer to the
manual IBM Tivoli Storage Area Network Manager User’s Guide, SC23-4698.

If the following conditions are true, you can start the TotalStorage Productivity Center for Data
graphical interface from the Tivoli NetView console:
TotalStorage Productivity Center for Data or the TotalStorage Productivity Center for Data
graphical interface is installed on the same machine as TotalStorage Productivity Center
for Fabric, or the SRMURL value specifies the hostname of TotalStorage Productivity
Center for Data.
The TotalStorage Productivity Center for Fabric is currently running.

For more information on TotalStorage Productivity Center for Data, see the redbook IBM Tivoli
Storage Resource Manager: A Practical Introduction, SG24-6886.

14.5.4 Other menu options
For the other options on the SAN pull-down menu:
Configure Agents is covered in *“Configuring the outband agents” on page 346 and
“Checking inband agents” on page 348.
Configure Manager is covered in “Performing an initial poll and setting up the poll interval”
on page 349.
Set Event Destination is covered in “Configuring SNMP” on page 342
ED/FI Properties and ED/FI Configuration are covered in “Configuration for ED/FI - SAN
Error Predictor” on page 818.


14.6 Status cycles
Figure 14-51 shows the typical color change status cycles which reflect normal operation as a
device goes down and comes up. Table 14-1 on page 709 and Table 14-2 on page 709 list the
meanings of the different colors.

NEW
GREEN
Clear History

Device down NORMAL
GREEN
Device down

Device up
MISSING
RED

Figure 14-51 IBM Tivoli SAN Manager — normal status cycle

If you do not manually use NetView capabilities to change status, the status of a Tivoli SAN
Manager object goes from green to red and from red to green.

Note that the only difference between an object in the NORMAL/GREEN and NEW/GREEN
status is in the Status field under SAN Properties (see Figure 14-42 on page 737 for an
example). A new object will have New in the field and a normal object will show Normal. The
icon displayed in the topology map will look identical in both cases.

You can encounter situations where your device is down for a known reason such as an
upgrade or hardware replacement and you don’t want it displayed with a missing/red status.
You can use the NetView Unmanage function to set its color as tan to avoid having the yellow
or red status reported and propagated in the topology display. See Figure 14-52.


IBM Tivoli SAN Manager status cycle (with Unmanage)
NORMAL
TAN
Manage /
Unmanage

NORMAL
GREEN
Device up Device down

MISSING
RED
Clear History
Manage /
Unmanage

MISSING NOT DISCOVERED
Clear History / NOT DISPLAYED
TAN

Figure 14-52 Status cycle using Unmanage function

However, when a device is unmanaged and you do a SAN → Configure Manager → Clear
History to remove historical data, the missing device will be removed from the Productivity
Center for Fabric database and will no longer be reported until it is up back with a new/green
status. If you have changed the label of the device, and it is re-discovered after a Clear
History, it will reappear with the default generated name, as this information is not saved.
See Figure 14-53.

IBM Tivoli SAN Manager status cycle (with Acknowledge)
NORMAL
GREEN

Device
down

MISSING Device
up
RED

Ack/Unack

MISSING
DARK GREEN

Figure 14-53 Status cycle using Acknowledge function

You can use the NetView Acknowledge function to specify that you have been notified about
the problem and that you are currently searching for more information or for a solution. This
will set the device’s color as dark green to avoid having the yellow or red status reported and
propagated in the topology display. Subsequently, you can use the Unacknowledge function
to return in the normal status and colors cycle. When the device becomes available, it will
automatically return to the normal reporting cycle.


14.7 Practical cases
We have re-created some typical errors that can happen in a production environment to see
how and why Productivity Center for Fabric reacts to them.

We have also used different configurations of the inband and outband agents and correlated
the results with the explanations.

14.7.1 Cisco MDS 9000 discovery
In this section we discuss the discovery of the Cisco MDS 9509, which is part of the MDS
9000 family. Our MDS 9509 is a multilayer switch/director with a 6 slot configuration. We have
one 16 port card and one 32 port card running at 2GB/s. Discovery of the MDS 9509 is
performed using inband management.

Figure 14-54 is the lab environment used to demonstrate the following discovery. We will call
this Lab environment 3.

Cisco 9509
Sanan
ITSANM Agent

Intranet

Lochness
SAN Manager Sanxc1

Sanxc2

Sanan3

Figure 14-54 Lab environment 3

We first deployed an Productivity Center for Fabric Agent to SANAN. Once the agent was
installed, it registered with the Productivity Center for Fabric - LOCHNESS and discovered
the CISCO1 (MDS 9509). The topology in Figure 14-55 was displayed after deploying the
agent.

Note: In order to discover the MDS 9000, at least one Productivity Center for Fabric Agent
must be installed on a host attached to the MDS 9000. Outband management is not
supported for the MDS 9000.


Figure 14-55 Discovery of MDS 9509

To display the properties of CISCO1, right-click the CISCO1 icon to select it and select SAN
→ SAN Properties. See Figure 14-56.

Figure 14-56 MDS 9509 properties


The Connection option (Figure 14-57) displays information about the slots and ports where
the hosts SANXC1, SANXC2 and SANXC3 are connected, as well as the status of each port.

Figure 14-57 MDS 9509 connections

14.7.2 Removing a connection on a device running an inband agent
Next, we removed the FC link between the host SICILY and the switch ITSOSW1.
Productivity Center for Fabric does not show that the device is missing but shows that the
connection is missing. As the host was running an in-band management agent, the host
continues to report its configuration to the manager using the IP network. However, the
attached switch sends a trap to the manager to signal the loss of a link. You can use Monitor
→ Events → All to view the trap received by NetView. Double-click the trap coming from
ITSOSW1 to see details about the trap as shown in Figure 14-58.

Figure 14-58 Trap received by NetView

We see that ITSOSW1 sent a trap to signal that FCPortIndex4 (port number 3) has a status of
2 (which means Offline).


The correlation between the inband information and the trap received is then made correctly
and only the connection is shown as missing. You can see in Figure 14-59 that the connection
line has turned red, using the colors referenced in Table 14-1 on page 709.

Figure 14-59 Connection lost

We then restored the connection, and following the status cycle explained in Figure 14-51 on
page 743, the connections returned to normal (Figure 14-60).

Figure 14-60 Connection restored


Next, we removed one out of the two connections from the host TUNGSTEN to ITSOSW3.
One link is lost, so the connection is now shown as suspect (yellow) – Figure 14-61.

Figure 14-61 Marginal connection

NetView follows its status propagation rules in Table 14-4 on page 711. This connection links
to a submap with the two physical connections. The bottom physical connection is missing
(red), and the other (top) one is normal (black), resulting is propagated status of (yellow) on
the parent map (left hand side). See Figure 14-62.

Figure 14-62 Dual physical connections with different status


14.7.3 Removing a connection on a device not running an agent
A device with no agent is only detected via its connection to the switch. If the connection is
broken, the host cannot be discovered. In this case, we unplugged the FC link between the
host LEAD and the switch ITSOSW2. LEAD is not running either an inband or an outband
agent — as we can see using SAN → Agents configuration, shown in Figure 14-63.

Figure 14-63 Agent configuration

After removing the link on LEAD and we received a standard Windows missing device popup
(Figure 14-64) indicating it could no longer see its FC-attached disk device.

Figure 14-64 Unsafe removal of Device


Productivity Center for Fabric shows the device as Missing (the icon changes to red — see
the color status listing in Table 14-1 on page 709) — as it is no longer able to determine the
status of the device (see Figure 14-65).

Figure 14-65 Connection lost on a unmanaged host

In Figure 14-66, the host is Unmanaged (tan) status since we decided to unmanage it.

Figure 14-66 Unmanaged host


We finally select SAN → Configure Manager → Clear History (See Figure 14-67).

Figure 14-67 Clear History

After the next discovery, as explained in Figure 14-52 on page 744, the host is no longer
displayed (Figure 14-68), since it has been removed from the Productivity Center for Fabric
database.

Figure 14-68 NetView unmanaged host not discovered

14.7.4 Powering off a switch
In this test we power off a SAN switch and observe the results.

Lab 2 environment
For demonstration purposes in the following sections, this lab is referenced as Lab 2. The
configuration consists of:
Two IBM 2109-S08 (ITSOSW1 and ITSOSW2) switches with firmware V2.6.0g


One IBM 2109-S16 (ITSOSW3) switch with firmware V2.6.0g
One IBM 2109-F16 (ITSOSW4) switch with firmware V3.0.2
One IBM 2107-G07 SAN Data Gateway
Two pSeries 620 (BANDA, KODIAK) running AIX 5.1.1 with:
– Two IBM 6228 cards
One IBM pSeries F50 (BRAZIL) running AIX 5.1.1ML4 with:
– One IBM 6227 card with firmware 02903291
– One IBM 6228 card with firmware 02C03891
One HP Server running HP-UX 11.0
– One FC HBA
Four Intel® servers (TONGA, PALAU, WISLA, LOCHNESS)
Two Intel servers (DIOMEDE, SENEGAL) with:
– Two QLogic QLA2200 card with firmware 8.1.5.12
One IBM xSeries 5500 (BONNIE) with:
– Two QLogic QLA2300 card with firmware 8.1.5.12
One IBM Ultrium Scalable Tape Library (3583)
One IBM TotalStorage FAStT700 storage server

Figure 14-69 shows the SAN topology of our lab environment.

Senegal - Win2k Srv sp3

LTO 3583 lochness Win2k Srv sp3
SAN Manager
SDG

banda - AIX 5.1
PO W ERFAU L DA T AL AR M
T A

Easter - HPUX
tonga - Win2k Srv sp3 11

sw4
wisla - Win2k Srv sp3

bonnie - Win2k Srv sp3

sw2
palau- Win2k Srv sp3
sw1

fastT700
Kodiak - AIX
5.1 brazil - AIX 5.1

sw3

iSCSI

clyde - Win2k Srv sp3 diomede- Win2k Srv sp3
NAS 200

Figure 14-69 SAN lab - environment 2


We have powered off the switch ITSOSW4, with managed host SENEGAL enabled. The
topology map reflects this as shown in Figure 14-70. The switch and all connections change
to red.

Figure 14-70 Switch down Lab 2

The agent running on the managed host (SENEGAL) has scanners listening to the HBAs
located in the host. Those HBAs detect that the attached device, ITSOSW4, is not active
since there is no signal from ITSOSW4. The information is retrieved by the scanners and
reported back to the manager through the standard TCP/IP connection. Since the switch is
not active, the hosts can no longer access the storage servers.

The active agent (SENEGAL) sends the information to the manager which triggers a new
discovery. Since the switch does no longer responds to outband management, Productivity
Center for Fabric will correlate all the information and as a result, the connections between
the managed hosts and the switch, and the switch itself, are shown as red/missing. The
storage server is shown as green/normal because of a second Fibre Channel connection to
ITSOSW2. ITSOSW2 is also green/normal because of the outband management being
performed on this switch.

The active agent host is still reported as normal/green as it sends its information to the
Manager through the TCP/IP network. Therefore the Manager can determine that only the
agent’s switch connections, not the host itself, is down.


Now, we powered the switch on again. At startup, the switch sends a trap to the manager.
This trap will cause the manager to ask for a new discovery. The result is shown in
Figure 14-71.

Figure 14-71 Switch up Lab 2

Now, following the status propagation detailed in 14.6, “Status cycles” on page 743, all the
devices are green/normal.


14.7.5 Running discovery on a RNID-compatible device
When you define a host for inband management, the topology scanner will launch inband
queries to all attached HBAs. The remote HBAs, if they support RNID, will send back
information such as device type. On switch ITSOSW2 is a Windows host, CLYDE, with a
QLogic card at the requested driver level. There is no agent installed on this host. We see
however that it is discovered as a host rather than as an Unknown device, as shown in
Figure 14-72, because of the HBA RNID support.

Figure 14-72 RNID discovered host

You can see under the SAN Properties window, Figure 14-73, that the RNID support only
provides the device type (Host) and the WWN. Compare this with the SAN Properties
window for a managed host, shown in Figure 14-42 on page 737.

Figure 14-73 RNID discovered host properties


To have a more explicit map, we put CLYDE in the Label field (using the method shown in
Figure 14-44) and the host is now displayed with its new label.

Figure 14-74 RNID host with changed label


14.7.6 Outband agents only
To see what happens if there are only outband agents, that is, with no Productivity Center for
Fabric agents running, we stopped all the running inband agents, cleared the Productivity
Center for Fabric configuration, by using SAN → Configure Agent → Remove button, and
then re-configured the outband agents on the switches ITSOSW1, ITSOSW2, and ITSOSW4

Figure 14-75 Only outband agents

When configuring the agents, we also used the Advanced button to enter the administrator
userid and password for the switches. This information is needed by the scanners to obtain
administrative information such as zoning for Brocade switches.


Productivity Center for Fabric discovers the topology by scanning the three registered
switches. This is shown in Figure 14-76. The information about the attached devices is limited
to the WWN of the device since this information is retrieved from the switch and there is no
other inband management. Note the ‘-’ signs next to Device Centric and Host Centric Views
— this information is retrieved only by the inband agent so is not available to us here.

Figure 14-76 Explorer view with only outband agents

Figure 14-77 shows the information retrieved from the switches (SAN Properties).

Figure 14-77 Switch information retrieved using outband agents


14.7.7 Inband agents only
For this practical case, we first unplugged all Fibre Channel connections from all agents and
we removed all the outband agents from the configuration using SAN → Configure Agents
→ Remove tab. We then forced a new poll. As expected, the agents returned only
information about the node and the local filesystems, shown in Figure 14-78. Note the ‘-’ sign
in front of /data01 for host SICILY. The filesystem is defined but not mounted, as the Fibre
Connections are not active.

Figure 14-78 Inband agents only without SAN connections

We reconnected the Fibre Channel connections to all agents into the switch and forced a new
polling. We now see that all agents reported information about their filesystems. Since the
agents are connected to a switch, the inband agents will retrieve information from it using
inband management. That explains why we see all the devices including those without agents
installed. Figure 14-79 shows that:
Our four inband agents (BRAZIL, GALLIUM, SICILY, SOL-E) are recognized.
The two switches ITSOSW1 and ITSOSw2 are found, since agents are connected to
them.
Device 1000006045161FF5 is displayed since it is connected to the switch ITSOSW1. The
device type is Unknown, as there is no inband nor outband agent on this device.


Figure 14-79 Inband agents only with SAN connections

We can also display SAN Properties as shown in Figure 14-80.

Figure 14-80 Switches sensor information

We now have no zoning information available since this is retrieved from the switch outband
Agent for the 2109 switch. This is indicated by the — sign next to Zone View on Figure 14-79.


14.7.8 Disk devices discovery
The topology scanner will launch inband queries to all attached HBAs. The Attribute scanner
will then do a SCSI request to get attribute information about the remote devices. Due to LUN
masking, the storage server will deny all requests if there are no LUNs defined for the
querying host. Figure 14-81 shows how our SAN topology is mapped when there is an IBM
MSS storage server but with no LUNs defined or accessible for the hosts in the same fabric.
The storage server is shown as an Unknown device because the inband agents were not
allowed to do SCSI requests to the storage servers as they had no assigned LUNs.

Figure 14-81 Discovered SAN with no LUNS defined on the storage server


Figure 14-82 shows that the host CRETE is not included in the MSS zone (we have enabled
the outband agent for the switch in order to display zone information). This zone includes
TUNGSTEN, which has no LUNs defined on the MSS.

Figure 14-82 MSS zoning display

We changed the MSS zone to include the CRETE server. We run cfgmgr on CRETE so that it
scans its configuration and finds the disk located on the MSS as shown in Example 14-1.

Example 14-1 cfgmgr to discover new disks
# lspv
hdisk0 00030cbf4a3eae8a rootvg
hdisk1 00030cbf49153cab None
hdisk2 00030cbf170d8baa datavg
hdisk3 00030cbf170d9439 datavg
# cfgmgr
# lspv
hdisk0 00030cbf4a3eae8a rootvg
hdisk1 00030cbf49153cab None
hdisk2 00030cbf170d8baa datavg
hdisk3 00030cbf170d9439 datavg
hdisk4 00030cbf8c071018 None


Now, the agent on CRETE is able to run SCSI commands on the MSS and discovers that it is
a storage server. Productivity Center for Fabric maps it correctly in Figure 14-83.

Figure 14-83 MSS zone with CRETE and recognized storage server

14.7.9 Well placed agent strategy
The placement of inband and outband agents will determine the information displayed:
For a topology map, you need to define inband and outband agents on some selected
servers and switches in order to discover all your topology. Switch zoning and LUN
masking may restrict access to some devices.
For a complete topology map, including correct device icons, you need to define inband
and outband on all servers and switches, except on those supporting RNID.
For information on zones, you need to define the switches as outband agents and set the
user ID and password on the Advanced properties.
For a complete Device Centric and Host Centric views, you need to place inband agents
on all servers you want to be displayed.

Before implementing inband and outband agents, you should have a clear idea of your
environment and the information you want to collect. This will help you to select the agents
and may minimize overhead caused by inband and outband agents.

In our configuration, we decided to place one agent on GALLIUM which is connected to the
two fabrics and has LUNs assigned on the FAStT storage server (Figure 14-84).


Figure 14-84 “Well-placed” agent configuration

The agent will use inband management to:
Query the directly attached devices.
Query the name server of the switches to get the list of other attached devices.
Launch inband management to other devices to get their WWN and device type (for RNID
compatible supported drivers).
Launch SCSI request to get LUN information from storage servers.

You can see in Figure 14-85 that the agent on GALLIUM has returned information on:
Directly attached switches (ITSOSW1 and ITSOSW4)
Devices attached to those switches (if they are in the same zones)
LUNs defined on the FAStT for this server
Its own filesystems

Because of the other hosts, only CLYDE runs with RNID compatible drivers, all other devices
— excluding switches and FAStT storage server — are displayed with an unknown device
icon. However, we have shown how we can get a complete map of our SAN by deploying just
one inband agent.


Figure 14-85 Discovery process with one well-placed agent

14.8 Netview
In this section we describe how to use the NetView program’s predefined performance
applications and how to create your own applications to monitor the Storage Area Network
performance. The NetView program helps you manage performance by providing several
ways to track and collect Fibre Channel MIB objects. You can use performance information in
any of the following ways:
Monitoring the network for signs of potential problems
Resolving network problems
Collecting information for trend analysis
Allocating network resources
Planning future resource acquisition

The data collected by the NetView program is based on the values of MIB objects. The
NetView program provides applications that display performance information:
NetView Graph displays MIB object values in graphs.
Other NetView tools display MIB object values in tables or forms.


14.8.1 Reporting overview
The NetView MIB Tool Builder enables you to create applications that collect, display, and
save real-time MIB data. The MIB Data Collector provides a way to collect and analyze
historical MIB data over long periods of time to give you a more complete picture of your
network’s performance. We will explain the SNMP concepts and standards, demonstrate the
creation of Data Collections and the use of the MIB Tool Builder as it applies to SAN network
management.

Figure 14-86 lists the topics we cover in this overview section.

NetView Reporting Overview
Understanding SNMP and MIBs

Configuring
MIBs (copying and loading)
IBM 2109
NetView

MIB Data Collector

MIB Tool Builder

NetView Graphing tool
Figure 14-86 Overview

14.8.2 SNMP and MIBs
The Simple Network Management Protocol (SNMP) has become the de facto standard for
internet work (TCP/IP) management. Because it is a simple solution, requiring little code to
implement, vendors can easily build SNMP agents for their products. SNMP is extensible,
allowing vendors to easily add network management functions to their existing products.
SNMP also separates the management architecture from the architecture of the hardware
devices, which broadens the base of multi vendor support. SNMP is widely implemented and
available today.

SNMP network management system contains two primary elements:
Manager — This is the console through which the network administrator performs network
management functions.
Agents — These are the entities that interface to the actual device being managed.
Switches and directors are examples of managed devices that contain managed objects

Important:
In our configuration, the SNMP manager is NetView and the SNMP agents are IBM
2109 Fibre Channel Switches.

These objects are arranged in what is known as the Management Information Base (MIB).
SNMP allows managers and agents to communicate for the purpose of accessing these
objects. Figure 14-87 provides an overview of the SNMP architecture.


SNMP architecture

iSCSI
MIB

Application Server
iSCSI Initiator

SNMP agent

IP Storage

Ethernet
iSCSI Target

Desktop
iSCSI Initiator
2109 Fibre Channel switch

SNMP agent
FC switch MIB
FA/FC MIB
FE MIB

Tivoli SAN Manager
“NetView”

Figure 14-87 SNMP architecture overview

A typical SNMP manager performs the following tasks:
Queries agents
Gets responses from agents
Sets variables in agents
Acknowledges asynchronous events from agents

A typical SNMP agent performs the following tasks:
Stores and retrieves management data as defined by the MIB
Signals an event to the manager

MIBs supported by NetView
NetView supports the following types of MIBs:
Standard MIB: All devices that support SNMP are also required to support a standard set
of common managed object definitions, of which a MIB is composed. The standard MIB
object definitions, MIB-I and MIB-II, enable you to monitor and control SNMP managed
devices. Agents contain the intelligence required to access these MIB values.
Enterprise-specific MIB: SNMP permits vendors to define MIB extensions, or
enterprise-specific MIBs, specifically for controlling their products. These
enterprise-specific MIBs must follow certain definition standards, just as other MIBs must,
to ensure that the information they contain can be accessed and modified by agents. The
NetView program provides the ability to load enterprise-specific MIBs from a MIB
description file. By loading a MIB description file containing enterprise-specific MIBs on an
SNMP management station, you can monitor and control vendor devices.

Note: We are using the Brocade 2.6 enterprise-specific MIBs for SAN network
performance reporting and the IBM TotalStorage IP Storage 200i iSCSI MIB


MIB tree structure
MIB objects are logically organized in a hierarchy called a tree structure. Each MIB object has
a name derived from its location in the tree structure. This name, called an object ID, is
created by tracing the path from the top of the tree structure, or the root, to the bottom, the
object itself. Each place where the path branches is called a node. A node can have both a
parent and children. If a node has no children, it is called a leaf node. A leaf node is the actual
MIB object. Only leaf nodes return MIB values from agents. The MIB tree structure is shown
in Figure 14-88. Note the leaf entry for bcsi which has been added into the tree.

For more information regarding SNMP MIB tree structures. See the following Web sites
relating to SNMP RFCs.
http://silver.he.net/~rrg/snmpworld.htm
http://www.cisco.com/univercd/cc/td/doc/cisintwk/ito_doc/snmp.htm

TOP

CCITT (0) JOINT-ISO-CCITT (2)
ISO (1)

STD (0) REG AUTHORITY
(1)
ORG (3)
MEMBER BODY
(2)

DOD (6)

INTERNET (1)

PRIVATE (4)
DIRECTORY (1) MGMT (2) EXPERIMENTAL (3)

ENTERPRISE (1)
MIB (1)

RESERVED (0) IBM (2) bcsi (1588) iSCSI

Figure 14-88 MIB tree structure

14.9 NetView setup and configuration
In this section we provide step by step details for copying and loading the Fibre Channel and
iSCSI MIBs into NetView. We then describe the FE MIB and SW MIB in the Brocade 2109
Fibre Channel switch and also describe the FC (Fibre Alliance) MIB in the IBM TotalStorage
IP Storage 200i device.

Note: The FC (Fibre Alliance) MIB is shipped with most Fibre Channel switch vendors.
Brocade Communications provides limited support for the FC MIB.

14.9.1 Advanced Menu
In order to enable certain advanced features in NetView, we must first enable the Advanced
Menu feature in the Options pull-down menu as shown in Figure 14-89. Shut down and
restart NetView for the changes to take effect.


Figure 14-89 Enabling the advanced menu

14.9.2 Copy Brocade MIBs
Before MIBs can be loaded into NetView, they must first be copied into the osovsnmp_mibs
directory. All vendor specific MIBS are located here.

We accessed the Brocade MIBs from the Web site:
http://www.brocade.com/support/mibs_rsh/index.jsp

We downloaded the MIBs below and copied them to the directory.
v2_6trp.mib (Enterprise Specific trap)
v2_6sw.mib (Fibre Channel Switch)
v2_6fe.mib (Fabric Element)
v2_6fa.mib (Fibre Alliance)

Note: If you have unloaded all the MIBs in the MIB description file (usrovsnmp_mibs),
you must load MIB-I or MIB-II before you can load any enterprise-specific MIBs. These are
loaded by default in NetView.

In Example 14-2 we show the usrovsnmp_mibs directory listing with our newly added MIBs.

Example 14-2 MIB directory
Directory of C:usrovsnmp_mibs

04/13/2002 09:33a 81,253 v2_6FA.mib
08/27/2002 02:45p 79,095 v2_6FE.mib
04/13/2002 09:33a 60,139 v2_6SW.mib
04/13/2002 09:33a 5,240 v2_6TRP.mib
4 File(s) 225,727 bytes


0 Dir(s) 6,595,670,016 bytes free

C:usrovsnmp_mibs>

14.9.3 Loading MIBs
After copying the MIBs to the appropriate directory, they must then be loaded into NetView.

IBM 2109
The IBM 2109 comes configured to use the MIB II-private MIB (TRP-MIB), FC Switch MIB
(SW-MIB), Fibre Alliance MIB (FA-MIB) and Fabric Element MIB (FE-MIB). By default, the
MIBs are not enabled. Here is a description of each MIB and their respective groupings.

MIB II-private MIB (v2_6trp.mib or TRP-MIB)
The object types in MIB-II are organized into the following groupings:
The System Group
The Interfaces Group
The Address Translation Group
The IP Group
The ICMP Group
The TCP Group
The UDP Group
The EGP Group
The Transmission Group
The SNMP Group

FC_MGMT (Fibre Alliance) MIB (v2_6fa.mib or FA-MIB)
The object types in FA-MIB are organized into the following groupings. Currently Brocade
does not write any performance related data into the OIDs for this MIB.
Connectivity
Trap Registration
Revision Number
Statistic Set

Fabric Element MIB (v2_6fe.mib or FE-MIB)
The object types in FE-MIB are organized into these groupings:
Configuration
Operational
Error
Accounting
Capability

FC Switch MIB (v2_6sw.mib or SW-MIB)
The object types in SW-MIB are organized into the following groupings:
swSystem
swFabric
swActCfg
swFCport
swNs
swEvent
swFwSystem
swEndDevice


To enable the MIBs for the IBM/Brocade switch, log into the switch via a telnet session, using
an ID with administrator privilege (for example, the default admin ID). We enabled all four of
the above MIBS using the snmpmibcapset command. The command can either disable or
enable a specific MIB within the switch. Example 14-3 shows output from the snmpmibcapset
command.

Example 14-3 snmpmibcapset command on IBM 2109
itsosw2:admin> snmpmibcapset
The SNMP Mib/Trap Capability has been set to support
FE-MIB SW-MIB FA-MIB SW-TRAP FA-TRAP SW-EXTTRAP
FA-MIB (yes, y, no, n): [yes]
SW-TRAP (yes, y, no, n): [yes]
FA-TRAP (yes, y, no, n): [yes]
SW-EXTTRAP (yes, y, no, n): [yes]
no change
itsosw2:admin>

NetView
The purpose of loading a MIB is to define the MIB objects so the NetView program’s
applications can use those MIB definitions. The MIB you are interested in must be loaded on
the system where you want to use the MIB Data Collector or MIB Tool Builder. Some
vendor’s specific MIBs are already loaded into NetView.

Since we want to collect performance MIB objects types for the Brocade 2109 switch, we will
load its MIB. On the NetView interface, select Tools → MIB → Loader SNMP V1. This will
launch the MIB Loader interface as shown in Figure 14-90.

Figure 14-90 MIB loader interface


Each MIB that you load adds a subtree to the MIB tree structure. You must load MIBs in order
of their interdependencies. We loaded the v2_6TRP.MIB first by clicking Load then selecting
the TRP.MIB from the usrovsnmp_mibs directory — see Figure 14-91.

Figure 14-91 Select and load TRP.MIB

Click Open and the MIB will loaded into NetView. Figure 14-92 shows the MIB loading
indicator.

Figure 14-92 Loading MIB


We then loaded the v2_6 SW.MIB, v2_6FE.MIB and v2_6FA.MIBs in turn using the same
process. You must load the MIBs in order of their interdependencies. A MIB is dependent on
another MIB if its highest node is defined in the other MIB.

After the MIBs are loaded, we now verify that we are able to traverse the MIB tree and select
objects from the enterprise-specific MIB. We used the NetView MIB Browser to traverse the
branches of the above MIBs. Click Tools → MIB → Browser SNMP v1. to launch the MIB
browser and use the Down Tree button to navigate down through a MIB (see Figure 14-93).

Figure 14-93 NetView MIB Browser

14.10 Historical reporting
NetView provides a graphical reporting tool that can be used against real-time and historical
data. After loading the Brocade (IBM 2109) MIBs into NetView, we demonstrate how to
compile historical performance data about the IBM 2109 by using the NetView MIB Data
Collector and querying the MIB referred to in 14.9.3, “Loading MIBs” on page 771. This tool
enables us to manipulate data in several ways, including:
Collect MIB data from the IBM 2109 at regular intervals.
Store MIB data about the IBM 2109.
Define thresholds for MIB data and generate events when the specified thresholds are
exceeded. Setting MIB thresholds enables us to automatically monitor important SAN
performance parameters to help report, detect and isolate trends or problems.


Brocade 2109 MIBs and MIB objects
We now need to understand what MIB objects to collect. The IBM 2109 has four MIBs loaded
and enabled, described in 14.9.3, “Loading MIBs” on page 771. We selected the MIB object
identifiers in Figure 14-94 because of their importance in managing SAN network
performance. SAN network administrators may want to specify other MIB object identifiers to
meet their own requirements for performance reporting. You should consult your
vendor-specific MIB documentation for details of the objects in the MIB. We will describe how
to create a MIB Data Collector for the following object identifiers in the following MIBs, shown
in Figure 14-94 and Figure 14-95.

FE-MIB Error Group

fcFXPortLinkFailures - Number of link failures detected by this FxPort

fcFXPortSyncLosses - Number of loss of synchronization detected by
the FxPort

fcFXPortSigLosses - Number of signal losses detected by the FxPort.

Figure 14-94 FE-MIB — Error Group

SW-MIB Port Table Group
swFcPortTXWords - Number of FC words transmitted by the port

swFcPortRXWords - Number of FC words received by the port

swFcPortTXFrames - Number of FC frames transmitted by the port

swFcPortRXFrames - Number of FC frames received by the port

swFcPortTXC2Frames - Number of Class 2 frames received by the port

swFcPortTXC3Frames - Number of Class 3 frames received by the port

Figure 14-95 SW MIB — Port Table Group

14.10.1 Creating a Data Collection
Our first Data Collection will target the MIB object swFCPortTxFrames. The
swFCPortTxFrames counts the number of Fibre Channel frames that the port has
transmitted. It contains group information about the physical state, operational status,
performance and error statistics of each Fibre Channel port on the switch for example,
F_Port, E_Port, U_Port, FL_Port.


Figure 14-96 describes the MIB tree where this object identifier resides. The root of the tree,
bcsi, stands for Brocade Communication Systems Incorporated.

The next several pages describe the step-by-step process for defining a Data Collection on
the swFcPortTxFrames MIB object identifier using NetView.

bcsi (1588)

commDev (2)

Fibre channel (1)

fcSwitch (1)

sw (1)

swFCPort (6)

swFCPortTable (2)

IBM 2109
swFCPortEntry (1)

private MIB tree
swFCPortTxFrames (13)

Figure 14-96 Private MIB tree for bcsi

1. To create the NetView Data Collection, select Tools → MIB → Collect Data from the
NetView main menu. The MIB Data Collector interface displays (Figure 14-97). Select
New to create a collection.

Figure 14-97 MIB Data Collector GUI


2. If creating the first Data Collection, you will also see the pop-up in Figure 14-98 to start the
Data Collection daemon. Click Yes to start the SNMPCollect daemon.

Figure 14-98 starting the SNMP collect daemon

3. The Data Collection Wizard GUI then displays (Figure 14-99). This is the first step in
creating a new Data Collection. By default NetView has navigated down to the Internet
branch of the tree (.iso.org.dod.internet). See Figure 14-88 on page 769 for the overall
tree structure. Highlight private and click Down Tree to navigate to the private MIB.

Figure 14-99 internet branch of MIB tree

We have now reached the private branch of the MIB tree (.iso.org.dod.internet.private).
See Figure 14-100.


Figure 14-100 Private arm of MIB tree

4. Continue to navigate down the enterprise branch of the tree by clicking Down Tree.
Figure 14-101 shows the enterprise branch of the tree
(.iso.org.dod.internet.private.enterprise).

Figure 14-101 Enterprise branch of MIB tree


5. We reach the bcsi branch of the tree by clicking Down Tree. Figure 14-102 shows the bcsi
(Brocade) branch of the tree (.iso.org.dod.internet.private.enterprise.bcsi).

Figure 14-102 bcsi branch of MIB tree


6. We continue to navigate down the tree, using the path shown in Figure 14-96,
and, as shown in Figure 14-103 on page 780, eventually reaching:
.iso.org.dod.internet.private.enterprise.bcsi.commDev.fibrechannel.fcSwitch.sw.swFCport.
swFCPort.swFCPortEnrty.swFCPortTxFrames.

Figure 14-103 swFCPortTxFrames MIB object identifier

7. We selected swFCPortTxFrames and clicked OK. We received the following pop-up
(Figure 14-104) from the collection wizard. This pop-up occurs because this will be the
first node added to this collection. NetView then adds the swFCTxFrames MIB Data
Collection definition as a valid data collector entry.

Figure 14-104 Adding the nodes


This launches the Add Nodes to the Collection Dialog, which is the second step in creating
a new Data Collection. See Figure 14-105.

Figure 14-105 Add Nodes to the Collection Dialog

8. We proceeded to customize the section Collect MIB Data from fields, using the following
steps:
a. We entered the switch node name for which we wanted to collect performance data (in
this case, ITSOSW2.ALMADEN.IBM.COM) and clicked Add Node. You can add a
node either by selecting it on the topology map or typing in the field as the IP address
or hostname for the device. Also, you can select multiple devices on the topology map
and click Add Selected Nodes from Map. This adds all the selected nodes selected
on the topology map to the Collect MIB Data From field. We also added several
nodes to the collection by adding one device at a time in the Node field and clicking
Add Node. To remove the node, just click the node name in the list and click Remove.
b. We then customized the section Set the Polling Properties for these Nodes, using
the following steps:
i. We changed the Poll Nodes Every field to 5 minutes. This specifies the frequency
in which the nodes are polled.

Important: Before setting the polling interval, you should have a clear
understanding of available and used bandwidth in your network. Shorter polling
intervals generate more SNMP data on the network.

ii. We checked Store MIB Data. This will store the MIB data that is collected to
C:/usr/ov/databases.
iii. The Check Threshold if box was checked. This will define the arm threshold. We
want to collect data and signal an event each time more than 200 frames are sent
on a particular port. Since we checked this box, we will be required to define the
trap value and rearm number fields.
iv. The option then send Trap Number was configured. We used the default setting,
which is the MIB-II enterprise-specific trap.


v. We then configured and rearm When. We specified a rearm value of greater than
or equal to 75%. of the arm threshold value. This means that a trap will be
generated and sent when the number of TX frames reaches 150. Note that these
traps are NetView-specific traps (separate from Productivity Center for Fabric traps)
and will therefore be sent to the NetView console.
9. Click OK to create the new Data Collection, shown in Figure 14-106. Select the
swFCPortTxFrames Data Collection and click Collect.

Figure 14-106 Newly added Data Collection for swFCTxFrames

Note: It could take up to 2 minutes before the newly defined Data Collection is being
collected by NetView. To verify that data is being captured, navigate to:
c:usrovdatabasessnmpcollect. If there are files present, then the Data Collection is
functioning properly.

10.Click Close and the Stop and restart Collection dialog is displayed as in Figure 14-107.
Click Yes to recycle the snmpcollect daemon. At this point the Data Collection status
(Figure 14-106 above) should change from Suspended to To be Collected

Figure 14-107 Restart the collection daemon

We are now collecting the data swFCTxFrames on ITSOSW2. Depending upon the level of
granularity that is required for your reporting needs, you may want to collect data over shorter
or longer periods. In our lab we collected every 5 minutes, but you may want to collect data
once every hour for a week or once every hour for a month.

We will now use the NetView Graph tool to display the data collected as described in 14.10.4,
“NetView Graph Utility” on page 784.


Note: We followed the same procedure to add the remaining metrics for Data Collection
swFCRxFrames, swFCTxErrors, and swFCRxErrors. For demonstration purposes we used
a of 50 for an arm threshold and a value of 75% for re-arm. Your values for arm/re-arm may
differ from what we used.

14.10.2 Database maintenance
You can periodically purge the Data Collection entries by selecting Options → Server
Setup, the Files tab page, then select Schedule SNMP Files to Delete from the drop-down
list. See Figure 14-108. Select the Purge day at a specific time.

Figure 14-108 Purge Data Collection files

Important: There are documented steps on how to perform important maintenance of
Tivoli NetView. Refer to the IBM Redbook Tivoli NetVIew and Friends, SG24-6019.


14.10.3 Troubleshooting the Data Collection daemon
If you find data is not being collected, ensure that the snmpCollect daemon is running and
that there is space available in the collection file system usrovdatabasessnmpcollect. The
daemon can stop running if there is no filesystem space.

To verify that the daemon is running, type ovstatus snmpcollect from the DOS command
prompt (see Example 14-4).

Example 14-4 snmpcollect daemon running
C:>ovstatus snmpcollect
object manager name: snmpcollect
behavior: OVs_WELL_BEHAVED
state: RUNNING
PID: 1536
last message: Initialization complete.
exit status: -

Done

C:>

If the snmpcollect daemon is not running, you will see a state value of NOT RUNNING from the
ovstatus snmpcollect command as shown in Example 14-5.

Example 14-5 snmpcollect daemon stopped
C:>ovstatus snmpcollect
object manager name: snmpcollect
behavior: OVs_WELL_BEHAVED
state: NOT RUNNING
PID: 1536
last message: Exited due to user request.
exit status: -

Done

C:>

The snmpcollect daemon can be started manually. At a command prompt, we typed in
ovstart snmpcollect. You will see the output shown in Example 14-6. We then issued an
ovstatus snmpcollect for verification, as shown in Example 14-4.

Example 14-6 snmpcolllect started
C:>ovstart snmpcollect
Done

C:>

Note: If no Data Collections are currently defined to the MIB Data Collector tool, the
snmpcollect daemon will not run.

14.10.4 NetView Graph Utility
We used the NetView graph utility to display the MIB object data that we collected in 14.10.1,
“Creating a Data Collection” on page 775.


We used the NetView Graph tool to display the collected data. This provides a convenient
way to display numerical performance information on collected data.

We now show how to display the collected data from the previous Data Collection that was
built for ITSOSW2 (swFCPortTxFrames). We start by single-clicking ITSOSW2 on the
NetView topology map (Figure 14-109).

Figure 14-109 Select ITSOSW2

Select Tools → MIB → Graph Data to launch the graph utility This will report on the historical
data that has been collected on ITSOSW2. After selecting this, NetView takes some time to
process the data and present it in the graphical display. The graph build time depends on the
amount of data collected. Figure 14-110 shows the progress indicator.

Figure 14-110 Building graph

After the graph is built, it displays the swFCTxFrames data that was collected (Figure 14-111).
Note there are multiple instances of the object ID mapped — that is, swFCPortTxFrames.1,
swFCPortTxFrames.2 and so on. In this case they represent the data collected for each port
in the switch.


Figure 14-111 Graphing of swFCTxFrames

For viewing purposes, we adjusted the x-axis for Time by clicking Edit → Graph Properties
in the open graph window. This allowed us to zoom into shorter time periods. See
Figure 14-112.

Figure 14-112 Graph properties

Any MIB object identifier that has been collected using the NetView MIB Data Collector can
be graphed using the NetView Graph facility using the above process.

14.11 Real-time reporting
In section we introduce the NetView MIB Tool Builder for real-time reporting. Figure 14-113
provides an overview.


Describe the MIB Tool Builder

Use of the Tool Builder
build
modify
delete
Figure 14-113 Real-time reporting — Tool Builder overview

Important: Depending on the configuration, some advanced functionality may be initially
disabled in NetView under Tivoli SAN Manager. This section requires this functionality to
be enabled. To enable all functionality required, in NetView, click Options → Polling and
check the Poll All Nodes field. This is shown in Figure 14-114.

Figure 14-114 Enabling all functions in NetView

14.11.1 MIB Tool Builder
In section we introduce the NetView MIB Tool Builder. The Tool Builder enables you to build,
modify, and delete MIB applications. MIB Applications are programs used by NetView to
monitor the network. The Tool Builder allows you to build MIB applications without
programming. The MIB Application monitors the real-time performance of specific MIB
objects on a regular basis and produces output such as forms, tables, or graphs.


We will demonstrate how to build a a MIB application that will query the swFCPortTxFrames
MIB object identifier in the SW-MIB. This process can used to query any SNMP enabled
device using NetView.

With the switch ITSOSW2 selected, we start building the MIB Application by launching the
Tool Builder. Select Tools → MIB → Tool Builder → New. The MIB Tool Builder interface
is launched — as in Figure 14-115. Click New to create a new Tool Builder entry for collecting
data on ITSOSW2.

Figure 14-115 MIB tool Builder interface

The Tool Builder Wizard Step1 window is displayed (Figure 14-116). We entered
FCPortTxFrames in the Title field and clicked in the Tool ID field to auto populate the
remaining fields. We clicked Next to continue with the wizard.

Figure 14-116 Tool Wizard Step 1


The Tool Wizard Step 2 interface displays. You can see our title of FCPortTxFrames has
carried over. We are now ready to select the display type. We can choose between Forms,
Tables, or Graphs. We will choose Graph and click New as shown in Figure 14-117.

Figure 14-117 Tool Wizard Step 2

The NetView MIB Browser is now displayed. We will use the MIB Browser to navigate down
to the FCPortTxFrames object identifier. Use the Down Tree button to navigate through the
MIB tree. Figure 14-118 shows the path through the SW-MIB error table. Click OK to add the
object identifier.

SW MIB - Port Table group

private...
enterprise...
bcsi...
commDev...
fibrechannel...
fcSwitch...
sw...
swFcPort...
swFcPortTable...
swFCPortTxFrames
Figure 14-118 SW-MIB — Port Table


The newly created MIB application is displayed in the Tool Builder Step 2 of 2 window. See
Figure 14-119 for the completed MIB Application. Click OK to complete the definition.

Figure 14-119 Final step of Tool Wizard

Now, the final window for the Tool Builder is displayed. It shows the newly created MIB
application in the window, Figure 14-120. Click Close to close the window. The new MIB
Application has been successfully created.

Figure 14-120 New MIB application — FXPortTXFrames


14.11.2 Displaying real-time data
Now that we have a MIB application, we want to collect real-time data from the switch. Select
ITSOSW2 from the NetView topology map by single clicking the ITSOSW2 symbol, then
select Monitor → Other → FCPortTXFrames. Our MIB application FCPortTXFrames has
been added to the menu (shown in Figure 14-121).

Figure 14-121 Monitor pull-down menu

Clicking on the FCPortTXFrames option, launches a graph utility, shown in Figure 14-122.

Figure 14-122 NetView Graph starting

The collection of MIB data starts immediately after selecting the swFCPortTXFrames MIB
application from the Monitor → Other menu. Figure 14-123 shows the data being collected
and displayed for each MIB instance of the ITSOSW2.


Figure 14-123 Graph of FCPortTXFrames

The polling interval of the application can be controlled using the Poll Nodes Every field
located under Edit → Graph Properties. See Figure 14-124.

Figure 14-124 Graph Properties


This launches a dialog to specify how often NetView Graph receives real-time data for
graphing, shown in Figure 14-125. This determines how often the nodes are asked for data.

Figure 14-125 Polling Interval

We continued to use the Tool Builder process defined in 14.11.1, “MIB Tool Builder” on
page 787 to build additional MIB applications for real-time performance monitoring. We used
the following MIB objects:
swFcPortTXWords
swFcPortRXC2Frames
swFCPortRXC3Frames
fcFXPortLinkFailures
fcFXPortSyncLosses
fcFXPortSigLosses

Figure 14-126 shows the newly defined MIB Applications as they appear in the Tool Builder.

Figure 14-126 Tool Builder with all MIB objects defined


Figure 14-127 shows all the above MIB objects as they appear in the NetView Monitor
pull-down menu. Note we have abbreviated the names of the MIB applications listed in the
Monitor → Other menu for ease of use.

Figure 14-127 All MIB objects in NetView

14.11.3 SmartSets
With Productivity Center for Fabric (Tivoli SAN Manager) providing the management of the
SAN, we can further extend the management functionality of the SAN from a LAN and iSCSI
perspective. NetView SmartSets gives us this ability. This section describes the concept of
the NetView SmartSet. See Figure 14-128 below. For an overview, we provide details on how
to group and mange your SAN attached resources from an TCP/IP (SNMP) perspective.

By default, the iSCSI SmartSet is created by Productivity Center for Fabric when nvsniffer is
enabled. SmartSets for iSCSI “initiators” and “targets” can be created using the process
described here.

What is a SmartSet?

Why SmartSets?

Defining a SmartSet

SmartSets and Data Collections
Figure 14-128 SmartSet Overview

794 IBM TotalStorage Productivity

Ibm total storage productivity center v2.3 getting started sg246490

More Related Content

What's hot (16)

Similar to Ibm total storage productivity center v2.3 getting started sg246490 (20)

More from Banking at Ho Chi Minh city (20)

Recently uploaded (20)

Ibm total storage productivity center v2.3 getting started sg246490