Veritas storage foundation_5.0_for_unix_-_fundamentals

'symantec..
VERITAS Storage Foundation
5.0 for UNIX: Fundamentals
100·002353·A

COURSE DEVELOPERS
Gail Ade~
BilJ,(eGerrits
lECHNICAL
CONTRIBUTORS AND
REVIEWERS
Jade ArrinJ,(lolI
:Iargy Cassid~
Ro~' Freeman
Joe Gallagher
Bruce Garner
Tomer G urantz
Bill Havev
Geue Henriksen
Gerald Jackson
Haymond Karns
Bill Lehman
Boh l.ucas
Durivunc Manikhung
Chr'istlan Rahanus
Dan Rugers
Kleher Saldanha
Albrecht Scriba
"liehe! Simoni
Anaudu Sirisena
Pete 'Iuemmes
Copyright' 2006 Symamec Corporation. All rights reserved. Symantcc.
the Symanrec Logo. and "LRITAS arc trademarks or registered
trademarks uf 5) mantee Corporation Of its alfiluues in the U.S. and other
countries. Other names may be trademarks of their respective owners.
-I IllS PUBLICATION IS I'ROVIDfD "AS IS" AND ALL EXPRESS OR
IMPLllDCONIlITIONS. REPRESENTArJONS AND WARRANTIES.
INCLUDIN(i ANY 11PLlUl WARRANTY OF MFRCHANTA81L1TY.
IITNI.sS FOR A PARIICULAR PURPOSE OR NON-
INFRIN(iI:MEN r. ARL DISCLAIiIED. EXCEl'! TO THE FXTEN!
rHAT SUCH DISCLAIMERS ARE HELD TO BE LEGALLY
INVALID. SYIANTI:C CORPORATION SHAl.L NOT HE L1ABLI:
lOR INCIDENTAL OR CONSEQULNTIAL DA1AGI-.S IN
CONNECTION WITH (HI FURNISHING. PERIC)RMANCE. OR USE
OF THIS PUIlLlCAIION. TH~. INFORIATION CONTAINLD
H!:RUN IS SUBJECT ro CHANtiE WITHOUT NOTICE.
No part orthe contents ofthis hook may be reproduced or transmitted in
any torm or b) any means without the , riuen permission of the publisher.
tLRIT-l.')' ::';/orugc FOflllllulion 5.0 /iw [,i;V/.': Fundamentals
Symnnrec Corporation
~03305te ens Creek 81 U.
Cupertino. CA ()SOI4

Table of Contents
Course Introduction
What Is Storage Virtualization?.
Introducing VERITAS Storage Foundation
VERITAS Storage Foundation Curriculum ..
Lesson 1: Virtual Objects
Physical Data Storage
Virtual Data Storage
Volume Manager Storage Objects
Volume Manager RAID Levels ..
Lesson 2: Installation and Interfaces
Installation Prerequisites .
Adding License Keys.. . .
VERITAS Software Packages ..
Installing Storage Foundation.
Storage Foundation User Interfaces.
Managing the VEA Software
Lesson 3: Creating a Volume and File System
Preparing Disks and Disk Groups for Volume Creation
Creating a Volume .
Adding a File System to a Volume
Displaying Volume Configuration Information ...
Displaying Disk and Disk Group Information
Removing Volumes, Disks, and Disk Groups
Lesson 4: Selecting Volume Layouts
Comparing Volume Layouts .....
Creating Volumes with Various Layouts.
Creating a Layered Volume .
Allocating Storage for Volumes .
Lesson 5: Making Basic Configuration Changes
Administering Mirrored Volumes .
Resizing a Volume .
Moving Data Between Systems
Renaming Disks and Disk Groups
Managing Old Disk Group Versions ..
Table of Contents
Copyrigtlt ,( 2006 Svmantec Corporation All rights reserved
Intro-2
Intro-6
Intro-11
1-3
1-10
1-13
1-15
2-3
.... 2-5
. 2-7
2-10
2-16
2-21
...... 3-3
3-12
3·18
3-21
3-24
3-30
. 4-3
. 4-9
4-18
.4·25
5-3
5-10
5-16
5-21
5-23

Lesson 6: Administering File Systems
Comparing the Allocation Policies of VxFS and Traditional File Systems 6-3
Using VERITAS File System Commands 6-5
Controlling File System Fragmentation 6-9
Logging in VxFS , 6-15
Lesson 7: Resolving Hardware Problems
How Does VxVM Interpret Failures in Hardware 7-3
Recovering Disabled Disk Groups "" , 7-8
Resolving Disk Failures ,.., , 7-12
Managing Hot Relocation at the Host Level , ,... 7-22
Appendix A: Lab Exercises
Lab 1: Introducing the Lab Environment , A-3
Lab 2: Installation and Interfaces ,..,.., A-7
Lab 3: Creating a Volume and File System ,..,.., , A-15
Lab 4: Selecting Volume Layouts , A-21
Lab 5: Making Basic Configuration Changes , , A-29
Lab 6: Administering File Systems...... .. ,.., A-37
Lab 7: Resolving Hardware Problems.. ...., ,.., A-47
Appendix B: Lab Solutions
Lab 1 Solutions: Introducing the Lab Environment ,.., B-3
Lab 2 Solutions: Installation and Interfaces , , , B-7
Lab 3 Solutions: Creating a Volume and File System , ,.., B-21
Lab 4 Solutions: Selecting Volume Layouts ,.., B-33
Lab 5 Solutions: Making Basic Configuration Changes ,.." B-47
Lab 6 Solutions: Administering File Systems , " B-67
Lab 7 Solutions: Resolving Hardware Problems " "...................... B-85
Glossary
Index
VERITAS Storage Foundation 5.0 for UNIX: Fundamentals
Copyngt1t <.;; .. LUOb Symaruec Corporation All nqhts reserved.

Storage Management Issues
Human Resource E-mail
Database Server
Customer Order
Database
90% F
symantec
10% 50% Full _
Multiple-vendor hardware
Explosive data growth
Different application needs
Management pressure to
increase efficiency
I' Multiple operating systems
r , Rapid change
Budgetary constraints
Problem: Customer
order database cannot
access unutilized
storage.
Common solution: Add
more storage.
What Is Storage Virtualization?
Storage Management Issues
Storage management is becoming increasingly complex due to:
Storage hardware from multiple vendors
Unprecedented data growth
Dissimilar applications with different storage resource needs
Management pressure to increase efficiency
Multiple operating systems
Rapidly changing business climates
Budgetary and cost-control constraints
To create a truly efficient environment. administrators must have the tools to
skillfully manage large, complex, and heterogeneous environments. Storage
virtualization helps businesses to simplify the complex IT storage environment
and gain control of capital and operating costs by providing consistent and
automated management of storage.
Intro-2 VERITAS Storage Foundation 5.0 for UNIX. Fundamentals
Copynyht r; 2006 Svmantec coroorauoo All ngllts reserved

,S}lnamcc.
What Is Storage Virtualization?
.MimkM4
Consumer
& m' 'f
§$ A
Virtualizatlon:
The logical
representation of
physical storage
across the entire
enterprise
Consumer
fIiIiiIrIIi"m''fl
Consumer
sliildeM"
Application requirements from storage
• Application • Throughput • Failure
requirements • Responsiveness
resistance
• Growth otential • Recovery time
Ca acit Performance Availabilit
• Disk size • Disk seek time • MTBF
• Number of disks! • Cache hit rate • Path
path redundanc
Physical aspects of storage
Physical Storage Resources
Defining Storage Virtualization
Storage virtualization is the process of taking multiple physical storage devices
and combining them into logical (virtual) storage devices that are presented to the
operating system. applications. and users. Storage virtualization builds a layer of
abstraction above the physical storage so that data is not restricted to specific
hardware devices. creating a Ilexible storage environment. Storage virtualization
simplifies management of storage and potentially reduces cost through improved
hardware utilization and consolidation.
With storage virtualization. the physical aspects of storage arc masked to users.
Administrators can concentrate less on physical aspects of storage and more on
delivering access to necessary data.
Benefits of storage virtualization include:
Greater IT productivity through the automation of manual tasks and simplified
administration of heterogeneous environments
Increased application return on investment through improved throughput and
increased uptime
Lower hardware costs through the optimized use of hardware resources
Copyright t: 20U6 Symanter; Corporation All riqh'~ .eservco
Intro-3Course Introduction

syrnantec
Storage Virtualization: Types
Storage-Based
JfIfIJ'AIiI'AY
Servers
Host-Based
AiIII1'
Server
Network-Based
AYAYAY
Servers
Storage
~j,~
Storage
~s.,",
Storage
Most companies use a combination of these three
types of storage virtualization to support their chosen
architectures and application requirements.
How Is Storage Virtualization Used in Your Environment?
The way in which you use storage virtuulization. and the benefits derived from
storage virtualization. depend on the nature of your IT infrastructure and your
specific application requirements. Three main types of storage virtualization used
today arc:
Storage-based
Host-based
Network-based
Most companies use a combination of these three types of storage virtualization
solutions to support their chosen architecture and application needs.
The type of storage virtualization that you use depends on factors. such as the:
Heterogeneity of deployed enterprise storage arrays
Need for applications to access data contained in multiple storage devices
Importance of uptime when replacing or upgrading storage
Need for multiple hosts to access data within a single storage device
Value of the maturity of technology
Investments in a SAN architecture
Level of security required
Level of scalability needed
Inlro-4
Copynghl ·,C ;':OOb Svmantec Conioranon AU flghls reserved

Storage-Based Storage Virtualization
Storage-basedstorage virtualization refers to disks within an individual array that
are presented virtually to multiple servers. Storage is virtualized by the array itself
For example. RAID arrays virtualize the individual disks (that are contained within
the array) into logical LUNS. which are accessed by host operating systems using
the same method of addressing as a directly-attached physical disk.
This type of storage virtualization is useful under these conditions:
You need to have data in an array accessible to servers of di fferent operating
systems.
All of a server's data needs are met by storage contained in the physical box.
You are not concerned about disruption to data access when replacing or
upgrading the storage.
The main limitation to this type of storage virtualization is that data cannot be
shared between arrays. creating islands of storage that must be managed.
Host-Based Storage Virtualization
Host-basedstorage virtualization refers to disks within multiple arrays and from
multiple vendors that are presented virtually to a single host server. For example.
software-based solutions. such as VERITAS Storage Foundation. provide host-
based storage virtualizarion. Using VERlTAS Storage Foundation to administer
host-based storage virtualization is the focus of this training.
I lost-based storage virtualization is useful under these conditions:
A server needs to access data stored in multiple storage devices.
You need the flexibility to access data stored in arrays from different vendors.
Additional servers do not need to access the data assigned to a particular host.
Maturity of technology is a highly important factor to you in making IT
decisions.
Note: By combining VERITAS Storage Foundation with clustering technologies.
such as VERITAS Cluster Volume Manager. storage can be virtualized to multiple
hosts ofthe same operating system.
Network-Based Storage Virtualization
Network-basedstorage virtualization refers to disks from multiple arrays and
multiple vendors that arc presented virtually to multiple servers. Network-based
storage virtualization is useful under these conditions:
You need 10 have data accessible across heterogeneous servers and storage
devices.
You require central administration of storage across all Network Attached
Storage (NAS) systems or Storage Area Network (SAN) devices.
You want to ensure that replacing or upgrading storage does not disrupt data
access.
You want to virtualize storage to provide block services to applications.
Course Introduction Intro-5
Copyright ,~,2006 Svmantec Corporation. All nnhts reserved

syrnarucc
VERITAS Storage Foundation
VERIT AS Storage Foundation provides host-based
storage virtualization for performance, availability,
and manageability benefits for enterprise computing
environments.
High Availability
Application Soluti
Data Protection
Volume Manager
and File System
Company Business Process
VERITAS Cluster Server/Replication
ons Storage Foundation for Databases
-<
VERITAS NetBackup/Backup Exec
~ VERIT AS Storage Foundation
Hardware and Operating System
Introducing VERITAS Storage Foundation
VERITAS storage management solutions address the increasing costs of managing
mission-critical data and disk resources in Direct Attached Storage (DAS) and
Storage Area Network (SAN) environments.
Atthe heart of these solutions is VERITAS Storage Foundation, which includes
VERITAS Volume Manager (VxVM). VERITAS File System (VxFS), and other
value-added products. Independently, these components provide key benefits.
When used together as an integrated solution, VxVM and VxFS deliver the highest
possible levels of performance, availability, and manageability for heterogeneous
storage environments.
Intro-6 VERITAS Storage Foundation 5.0 for UNIX: Fundamentals
Cnpyllght ,£ 2006 Syroantec Corporation All r,ghts reserved

Users / Applications/ Databases
.................................................................................................... " ....,,
! Virtual Storage Resources
00.0 ••• 00 0 .0..(?v!2!J.t.1 '.',
VERITAS
Volume
Manager
(VxVM)
What Is VERITAS Volume Manager?
VERITAS Volume Manager, the industry-leader in storage virtualizarion. is an
easy-to-use, online storage management solution for organizations that require
uninterrupted, consistent access to mission-critical data. VxVM enables you to
apply business policies to configure, share. and manage storage without worrying
about the physical limitations of disk storage. VxVM reduces the total cost of
ownership by enabling administrators to easily build storage configurations that
improve performance and increase data availability.
Working in conjunction with VERITAS File System, VERITAS Volume Manager
creates a foundation for other value-added technologies. such as SAN
environments, clustering and failover, automated management. backup and IISM,
and remote browser-based management.
What Is VERITAS File System?
A file system is a collection of directories organized into a structure that enables
you to locate and store tiles. All processed information is eventually stored in a tile
system. The main purposes of a file system arc to:
Provide shared access to data storage.
Provide structured access to data.
Control accessto data.
Provide a common. portable application interface.
Enable the manageability or data storage.
The value of a file system depends on its integrity and performance.
Copyright os 2006 swnante- Corporation. All rigllts reserveo
Intro-7Course Introduction

svrnaruec
VERITAS Storage Foundation: Benefits
• Manageability
- Manage storage and file systems from one interface.
- Configure storage online across Solaris, HP-UX, AIX, and
Linux.
- Provide additional benefits for array environments, such as
inter-array mirroring.
• Availability
- Features are implemented to protect against data loss.
- Online operations lessen planned downtime.
• Performance
- 1/0 throughput can be maximized using volume layouts.
- Performance bottlenecks can be located and eliminated
using analysis tools.
• Scalability
- VxVM and VxFS run on 32-bit and 64-bit operating systems.
- Storage can be deported to larger enterprise platforms.
Benefits of VERITAS Storage Foundation
Commercial system availability now requires continuous uptime in many
implementations. Systems must be available 24 hours a day. 7 days a week, and
365 days a year. VERlTAS Storage Foundation reduces the cost ofownership by
providing scalable manageability, availability, and performance enhancements for
these enterprise computing environments.
Manageability
Management of storage and the tile system is performed online in real time,
eliminating the need for planned downtime.
Online volume and file system management can be performed through an
intuitive. easy-to-use graphical user interface that is integrated with the
VERITAS Volume Manager (VxVM) product.
Vx VM provides consistent management across Solaris. HP-llX, AlX, Linux,
and Windows platforms.
VxFS command operations are consistent across Solaris, HP-UX, AlX, and
Linux platforms.
Storage Foundation provides additional benefits for array environments, such
as inter-array mirroring.
Availability
Through software RAID techniques, storage remains available in the event of
hardware fai lure.
Intro-8
CopytlQtll ''",2006 Svmantec Corporation All fights reserved

1I0t relocation guarantees the rebuilding of redundancy in the case of a disk
failure.
Recovery time is minimized with logging and background mirror
resynchronization.
Logging of file system changes enables fast file system recovery.
A snapshot of a file system provides an internally consistent. read-only image
for backup. and file system checkpoints provide read-writable snapshots.
Performance
I/O throughput can be maximized by measuring and modifying volume layouts
while storage remains online.
Performance bottlenecks can he located and eliminated using YxYM analysis
tools.
Extent-based allocation of space lor files minimizes file level access time.
Read-ahead buffering dynamically tunes itself to the volume layout.
Aggressive caching of writes greatly reduces the number of disk accesses.
Direct I/O performs file [10 directly into and out of user butlers.
Scalability
YxYM runs over a 32-bit and M-hit operating system.
Ilosts can be replaced without modifying storage.
Hosts with different operating systems can access the same storage.
Storage devices can be spanned.
YxYM is fully integrated with YxFS so that modifying the volume layout
automatically modi lies the file system internals.
With YxFS. several add-on products are available for maximizing performance
in a database environment.
Course Introduction Intro-9
Copyright;;; 2006 Symalll~r. Corporation All rights .esorveo

• Reconfigure and resize storage
across the logical devices
presented by a RAID array.
• Mirror between arrays to improve
disaster recovery protection of an
array.
Use arrays and JBODs.
• Use snapshots with mirrors in
different locations for disaster
recovery and off-host processing.
Use VERITAS Volume Replicator
(VVR) to provide hardware-
independent replication services.
symantec
Storage Foundation and RAID Arrays: Benefits
With Storage Foundation, you can:
Benefits of VxVM and RAID Arrays
RAID arrays virtualize individual disks into logical LUNS which are accessed by
host operating systems as "physical devices." that is, using the same method of
addressing as a directly-attached physical disk.
VxVM virtualizes both the physical disks and the logical LUNs presented by a
RAID array. Modifying the configuration ofa RAID array may result in changes in
SCSI addresses of LUNs, requiring modification of application configurations.
VxVM provides an effective method ofrcconfiguring and resizing storage across
the logical devices presented by a RAID array.
When using VxVM with RAID arrays. you can leverage the strengths of both
technologies:
You can use Vx VM to mirror between arrays 0 improve disaster recovery
protection against the failure of an array. particularly if one array is remote.
Arrays can be of different manufacture: that is, one array can be a RAID array
and the other a J80D.
VxVM facilitates data reorganization and maximizes available resources.
VxVM improves overall performance by making 1/0 activity parallel for a
volume through more than one 110 path to and within the array.
You can use snapshots with mirrors in different locations. which is beneficial
for disaster recovery and off-host processing.
If you include VERITAS Volume Rcplicaror (VVR) in your environment,
VVR can be used to provide hardware-independent replication services.
tntro-10 VERITAS Storage Foundation 5.0 for UNIX. Fundamentals
Copyngtll' LU06Svmantec Corporaucn All nqtits reserved

Storage Foundation Curriculum Path
,S}11Hlntt'C
VERITAS Storage
Foundation for
UNIX:
Fundamentals
VERITAS Storage
Foundation for
UNIX:
Maintenance
VERITAS Storage Foundation Curriculum
VERITASStorage Foundationfor UNIX' Fundamentalstraining is designed 10
provide you with comprehensive instruction on making the most ofVERITAS
Storage Foundation.
~------------ ------------~---v--
VERIT AS Storage Foundation for UNIX
Course Introduction
Cor-yriqht © 2006 Symantec Corporauon. All rights reserved
Inlro-11

Storage Foundation Fundamentals: Overview
• Lesson 1: Virtual Objects
• Lesson 2: Installation and Interfaces
Lesson 3: Creating a Volume and File
System
• Lesson 4: Selecting Volume Layouts
• Lesson 5: Making Basic Configuration
Changes
• Lesson 6: Administering File Systems
• Lesson 7: Resolving Hardware
Problems
syrnantcc
VERITAS Storage Foundation for UNIX: Fundamentals Overview
This training provides comprehensive instruction un operating the file and disk
management foundation products: VERITAS Volume Manager (VxVM) and
VERITAS File System (VxFS). In this training. you learn how to combine file
system and disk management technology to ensure easy management of all storage
and maximum availability of essential data.
Objectives
After completing this training. you will be able to:
Identify VxVM virtual storage objects and volume layouts.
Install and configure Storage Foundation.
Configure and manage disks and disk groups.
Create concatenated. striped, mirrored. RAID-5, and layered volumes.
Configure volumes by adding mirrors and logs and resizing volumes and tile
systems.
Perform tile system administration.
Resolve basic hardware problems.
Intro-12
Copynght'~ 2006 Svmantec Corpoauon All nghl<; reserved

, S)111<1ntt'(
Course Resources
• Lab Exercises (Appendix A)
• Lab Solutions (Appendix B)
• Glossary
Additional Course Resources
Appendix A: Lab Exercises
This section contains hands-on exercises that enable you to practice the concepts
and procedures presented in the lessons.
Appendix B: Lab Solutions
This section contains detailed solutions to the lab exercises for each lesson.
Glossary
For your reference. this course includes a glossary ofterms related to V[RITAS
Storage Foundation.
Course Introduction tntro-13
Copyrigtllf' 2006 Symanter Corporation All rights reserved

Typographic Conventions Used in This Course
The following tables describe the typographic conventions used in this course.
Typographic Conventions in Text and Commands
Convention Element Examples
Courier Nell. Command input. To display the robot and drive configurauon:
bold both syntax and tpconfig -d
examples
To display disk information:
vxdisk -0 alldgs list
Courier New.
· Command output lu the output:
plain
· Command protocol mlnlmum: 40-
names. directory protocol - maximum: 60
names. tile protocol
-current: 0
names. path Locale the al tnames directory.
names. user
Go 10http: / /www.symantec.com.
names.
passwords. URLs Enler the value 300.
when used within Log on as user l.
regular text
paragraphs.
Courier New. Variables in To install the media server:
Italic. bold or command syntax, /cdrom_directory/install
plain and examples:
To access a manual page:
· Variables in
command input
man command name
-
are Italic. plain. To display detailed information lor a disk:
· Variables in vxdisk -g disk_group list
command output disk -
name
are ltulic. bold.
Typographic Conventions in Graphical User Interface Descriptions
Convention Element Examples
Arrow Menu navigation paths Select File -->Save.
Initial capitalization Buttons. menus. windows, Select the Next buuon.
options. and other interface Open the Task Sialus
clements window.
Remove the checkmark
trorn the Print File check
box.
()uotation marks lutertucc clements with Select the "Include
long names subvolumes in object view
window" check box.
Intro-14 VERITAS Storage Foundation 5.0 for UNIX: Fundamentals
Copyright ~ ;!()Clfi Svmanter: Corporation AUfights reserved

symantcc
Lesson Introduction
• Lesson 1: VirtU!!L9_bl~~ts •••
• Lesson 3: Creating a Volume and File
System
Changes
Problems
svmantec
Lesson Topics and Objectives
Topic After completing this lesson, you
will be able to:
Topic 1: Physical Data Identify the structural characteristics of
Storage a disk that are affected by placing a
disk under VxVM control.
Topic 2: Virtual Data Describe the structural characteristics
Storage of a disk after it is placed under VxVM
control.
Topic 3: Volume Manager Identify the virtual objects that are
Storage Objects created by VxVM to manage data
storage, including disk groups, VxVM
disks, subdisks, piexes, and volumes.
Topic 4: Volume Manager Define VxVM RAID levels and identify
RAID Levels virtual storage layout types used by
VxVM to remap address space.
1-2
COPYlIgl1l·.~ 2006 Symanter. Ccq-orauon All fights rnservoc
VERITAS SLorage Foundation 5.0 for UNIX: Fundamentals

, S)11Jan,'(
I
Physical Disk Structure
Physical storage objects:
• The basic physical storage device that ultimately stores
your data is the hard disk.
• When you install your operating system, hard disks are
formatted as part of the installation program.
• Partitioning is the basic method of organizing a disk to
prepare for files to be written to and retrieved from the
disk.
• A partitioned disk has a prearranged storage pattern that
is designed for the storage and retrieval of data.
Solaris I HP-UX I AIX I Linux I
Physical Disk Structure
Solaris
A physical disk under Solaris contains the partition table of the disk and the volume Table
of contents (VTOC) in the first sector 151~ bytes) of the disk. The VTOC has at least an
entry lor the backup partition on the II hole disk (partition tag 5, normally partition number
2), so the OS may work correctly with the disk. The VTOC is always a part of the backup
partition and may be part ota standard data partition. You can destroy the VTOC using the
raw device driver on that partition making the disk immediately unusable.
Sector 0 of disk: VTOC
Sector 1-15 of
/ partition: bootblock
~1US'
IPartition 2 (backup slice)
refers to the entire disk.
Partitions
(Slices)
1
Copyright 1'. 2006 Symaruec Corporation All nqtus reserved
Lesson 1 Virtual Objects

If the disk contains the partition fur the rout file system mounted on / (partition tag 2), lor
example of an OS disk, this root partition contains the bootblock for the first boot stage
after the Open Bout Prom within sector I - 15. Sector 0 is skipped, so there is no
overlapping between VTOe and boorblock. if the root partition starts at the beginning of
the disk.
The li"t sector ofa file system un Soluris cannut start before sector 16 of the partition.
Sector 16 contains the main super block of the file system. Using the block device driver
of the file system prevents VTOC and boot block from being overwritten by applicuuon
data.
Note: 011 Solaris. VxVM 4. I and later support EFI disks. EFI disks are an lntcl-based
technology that allows disks to retain BIOS code.
HP-lJX
On an HP-UX system. the physical disk is traditionally partitioned using either the whole
disk approach or Logical Volume Manager (LVM).
HP-UX Disk
cOtld4
LVM Disk
cOtld4
The whole disk approach enables you tu partition a disk in five ways: the whole disk is
used by a single file system: the whole disk is used as swap area: the whole disk is
used as a raw partition: a portion of the disk contains a file system, and the rest is used
as swap: or the boot disk contains a 2-MB special boot are". the root file system, and a
swap area.
An LVM data disk consists of four areas: Physical Volume Reserved Area (PVRA):
Volume Group Reserved Area (VGRA): user data area: and Bad Block Relocation
Area IBBRA).
AIX
A native AIX disk docs not have a partition table "I' the kind familiar on many other
operating systems. such as Solaris, Linux, and Windows. An application could use the
entire unstructured raw physical device. but the lirst 5 I 2-byte sector normally contains
intunn.uion, including a physical volume identifier (pvid) to support recognition of the
disk by AIX. An AIX disk is managed by IBM's Logical Volume Manager (LVM) by
defuult. A disk managed by LVM is called a physical volume (PV). A physical volume
consists of:
PV reserved area: A physical volume begins with a reserved area of 128 sectors
containing I'V metadatu. including the pvid.
1-4 VERITAS Storage Foundation 5.0 for UNIX: Fundamentals
COPYright ';:' 2(1()h Sytn<ll1le: Corporauon All [lgtHS reserveo

Volume Group Descrlptur Area ('G[)A): One or two copies of the V[X;/ tollows.
The V(jDA contains information describing a volume group (Vt.i ), which consists of
one or more physical volumes. Included in the mctadata in the V(j[)A is the defiuiuon
of the physical partition (PI') Sill'. nonnally-t MH,
Physlcal partitions: The remainder of the disk is divided into a number of physical
partitions, All of the PVs in a volume group have PI's of the same size. as defined in
the VGDA, In a normal VG. there can be up to.n PI's in a P'. In a big VCi. there can
be up to 12R PI's in a Pv. I
,Raw
device
,hdisk3
Physical volume
reserved area
(128 sectors)
Volume Group
Descriptor Areas
Physical partitions
(equal size,
defined in VGDA)
The term partition is used differently in different operating systems. In many kinds of
UNiX. Linux, and Windows. a partition is a variable sized portion of contiguous disk
spacethat can be formatted to contain a file system. in LVM. a PI' is mapped to a logical
partition 11..1'). and one or more LPs from any location throughout the V(j can be
combined to define a logical volume (LV). A logical '0IU111eis the entity that can be
formatted to contain a file system (by default either .IFS or .IFS2), So a physical partition
compares in concept more closely to a disk allocation cluster in some other operating
systems. and a logical volume plays the role that a partition does in some other operating
systems.
Linux
On Linux. a nonboot disk can be divided into one to lour primary partitions. One of these
primary partitions can be used to contain logical partitions. and it is called the extended
partition. The extended partition can have lip to I ~ logical partitions on a SCSi disk and lip
to 60 logical partitions on an IDE disk, You can use fdisk to set up partitions on a Linux
disk,
Lesson 1 Virtual Objects 1-5
CopyriglH '-':2006 Symantcc Corporation All fights reserveo

Primary Partition 1
/dev/sdal or/dev/hdal
Primary Partition 2
/dev/sda2or/dev/hda2
Primary Partition 3
/dev/sda3or/dev/hda3
Primary Partition 4
(Extended Partition)
/dev/sda4
/dev/hda4
On a l.inux boot disk. the boot partition must be a primary partition and is typically
located within the first 1024 cylinders of the drive. On the boot disk. you must also have a
dedicated swap partition. The swap partition can be a primary or a logical partition. and it
can be located anywhere on the disk.
Logical partitions must be contiguous. but they do not need to take up all of the space of
the extended partition. Only one primary partition can be extended. The extended partition
docs not take up any space until it is subdivided into logical partitions.
VERITAS Volume Manager 4.0 for Linux does not support most hardware RAID
controllers currently unless they present SCSI device interfaces with names of the
form / dev / sdx.
The following controllers are supported:
PERC, on the Dell 1650
MegaRAID. on the Dell 1650
ScrvcRAID. on x440 systems
Compaq array controllers that require the SmartI and CCISS drivers (w hich present
device paths. such as I dev I idal c #d#p # and I dev I cc i ssl c #d#p#) arc supported
lor normal use and for rootnbiliry.
1·6 VERITAS Storage Foundation 5.0 for UNIX: Fundamentals
Cupynqht c ;1.0011Svmanu-c Corporation All fights rcserveo

,S)l1HHlh'(
Physical Disk Naming
VxVM parses disk names to retrieve connectivity
information for disks. Operating systems have different
conventions:
Operating System Device Naming Convention Example
Solaris /dev/[rJdsk/c1t9dOs2
/dev/ [rJ dak/c3t2dO (no slice)
/dev/hdisk2 (no slice)
SCSI disks:
/dev/ade [1-4J (primary partitions)
/dev/ade [5 -16J (logical partitions)
/dev/adbN(on the second disk)
/ dev / adcN (on the third disk)
IDE disks:
/dev/hdeN. /dev/hClbN./dev/hdcN
HP-UX
AIX
Linux
Physical Disk Naming
Solaris
I
You locate and access the data on a physical disk by using a device name that
specifies the controller, target ID. and disk number. A typical device name uses the
format: c#t#d#.
c# is the controller number.
t# is the target !D.
d# is the logical unit number (LUN) of the drive attached to the target.
Ira disk is divided into partitions. you also specify the partition number in the
device name:
s# is the partition (slice) number,
For example. the device name cOtOdOsl is connected to controller number 0 ill
the system. with a target ID oro. physical disk number O. and partition number I
011 the disk.
HP-liX
You locate and access the data on a physical disk by using a device name that
specifies the controller. target ID, and disk number. A typical device name uses the
format: c#t#d#.
c# is the controller number.
t # is the target !D.
d# is the logical unit number (LUN) of the drive attached to the target.
Copyright .~..2006 Svmanter Corporauon All fights reserved
1-7Lesson 1 Virtual Objects

For example, the cOt OdO device name is connected to the controller number °in
the system, with a target 10 of 0, and the physical disk number O.
AIX
Every device in AIX is assigned a location code that describes its connection to the
system. The general format of this identifier isAB-CD-EF-GH, where the letters
represent decimal digits or uppercase letters. The first two characters represent the
bus. the second pair identify the adapter, the third pair represent the connector, and
the tinal pair uniquely represent the device. For example, a SCSI disk drive might
have a location identifier of 04 - 01- 00 - 6, O.In this example, 04 means the PCI
bus, 01 is the slot number on the PCI bus occupied by the SCSI adapter, 00 means
the only or internal connector, and the 6,0 means SCSIID 6, LUN o.
However, this data is used internally by AIX to locate a device. The device name
that a system administrator or software uses to identify a device is less hardware
dependant. The system maintains a special database called the Object Data
Manager (ODM) that contains essential definitions for most objects in the system,
including devices. Through the ODM. a device name is mapped to the location
identifier. The device names are referred to by special files found in the / dev
directory. For example, the SCSI disk identified previously might have the device
name hdisk3 (the fourth hard disk identified by the system). The device named
hdisk3 is accessed by the file name /dev/hdisk3.
If a device is moved so that it has a different location identifier, the ODM is
updated so that it retains the same device name. and the move is transparent to
users. This is facilitated by the physical volume identifier stored in the first sector
of a physical volume. This unique 128-bit number is used by the system to
recognize the physical volume wherever it may be attached because it is also
associated with the device name in the ODM.
Linux
On Linux, device names are displayed in the format:
• sdx [N]
• hdx [N]
In the syntax:
sd refers to a SCSI disk, and hd refers to an EIDE disk.
x is a letter that indicates the order of disks detected by the operating system.
For example, sda refers to the first SCSI disk, sdb refers to the second SCSI
disk. and so on.
N is an optional parameter that represents a partition number in the range I
through 16. For example. sda7 references partition 7 on the first SCSI disk.
Primary partitions on a disk are I. 2, .~.4: logical partitions have numbers 5 and up.
If the partition number is omitted, the device name indicates the entire disk.
1-8 VERITAS Storage Foundation 5.0 for UNIX.' Fundamentals
Copynqht ', ;'[)Of) Symantec Corporaucn All fights reserverr

Note: Throughout this course, the term disk is used to mean either disk or LUN.
Whatever the OS sees as a storage device, VxVM sees as a disk.
• Reads and writes
on unmanaged
physical disks
can be a slow
process .
• Disk arrays and
multipathed disk
arrays can
improve 110 speed
and throughput.
IApplications! Databases / Users
I
00PhYSi!1[!LUNS 00Disk array: A collection of physical disks used
to balance 1/0 across multiple disks
Multipathed disk array: Provides multiple ports
to access disks to achieve performance and
availability benefits
Disk Arrays
Reads and writes on unmanaged physical disks can be a relatively slow process,
because disks are physical devices that require time to move the heads to the
correct position on the disk before reading or writing. If all of the read and write
operations are performed to individual disks. one at a time. the read-write time can
become unmanageable.
A disk 1/1"'(11' is a collection of physical disks. Performing 110 operations on
multiple disks in a disk array can improve 1/0 speed and throughput.
Hardware arrays present disk storage to the host operating system as LUNs.
Multipathed Disk Arrays
Some disk arrays provide multiple ports to access disk devices. These ports.
coupled with the host bus adaptor (IIBA) controller and any data bus or 110
processor local to the array. compose multiple hardware paths to access the disk
devices. This type of disk array is called a muttipathed disk aI'I'O'.
You can connect rnultipathed disk arrays to host systems in many different
configurations. such as:
Connecting multiple ports to different controllers on a single host
Chaining ports through a single controller on a host
Connecting ports to di tferent hosts simultaneously
Cop'r'riglll'(~ 2006 Symantec Corporation All rights. roservoo

symarucc
• Volume Manager
creates a virtual
layer of data
storage.
• Volume Manager
volumes appear to
applications
to be physical disk
partitions.
• Volumes have
block and character
device nodes in the
Zdev tree:
Idev/vxl lr l dsk/ ...
Multidisk
configurations:
• Concatenation
• Mirroring
• Striping
• RAID·S
High Availability:
• Disk group
import and deport
• Hot relocation
• Dynamic
multipathing
Disk Spanning Load·Balancing
Virtual Storage Management
VER lTAS Volume Manager creates a virtual level of storage management above
the physical device level by creating virtual storage objects. The virtual storage
object that is visible to users and applications is called a 1'0111111<'.
What Is a Volume?
A volume is a virtual object, created by Volume Manager, that stores data. A
volume consists of space from one or more physical disks on which the data is
physically stored.
How Do You Access a Volume?
Volumes created by VxVM appear to the operating system as physical disks, and
applications that interact with volumes work in the same way as with physical
disks. All users and applications access volumes as contiguous address space using
special device files in a manner similar to accessing a disk partition.
Volumes have block and character device nodes in the / dev tree. You can supply
the name of the path to a volume in your commands and programs, in your file
system and database configuration files, and in any other context where you would
otherwise use the path to a physical disk partition.
Copynght '6 2006 Symantec Comor auon All fights reserved

I
Volume Manager Control
When you place a disk under VxVM control, a cross-platform data sharing (CDS)
disk layout is used, which ensures that the disk is accessible on different
platforms, regardless of the platform on which the disk was initialized.
I····- ..·-------·.L~_~lOS-reserved areas
I
that contain:
" Platform blocks
I " VxVM 10 blocks
I
" AIX and HP-UX
. coexistence labels
Public
Region
Volume Manager-Controlled Disks
With Volume Manager. you enable virtual data storage by bringing a disk under
Volume Manager control. By default in VxVM 4.0 and later. Volume Manager
uses a cross-platform data sharing (CDS) disk layout. A CDS disk is consistently
recognized by all VxVM-supported UNIX platforms and consists of:
Ox-reserved area: To accommodate plat tonn-spcci fie disk usage. f 2RK is
reserved for disk labels. platform blocks. and platform-coexistence labels.
Private region: The private region stores information. such as disk headers.
configuration copies. and kernel logs. in addition to other platform-specific
management areas that VxVM uses to manage virtual objects. The private
region represents a small management overhead:
Operating System Default Block/Sector Size Default Private Region Size
Solaris 512 bytes 65536 sectors (I 024K)
HI'-UX 1024 bytes 3276Rsectors ( I024K)
AIX 512 bytes 65536 sectors (I024K)
Linux 512 bytes 65536 sectors ( I024K)
Public region: The public region consists of the remainder of the space on the
disk. The public region represents the available space that Volume Manager
can LIseto assign to volumes and is where an application stores data. Volume
Manager never overwrites this area unless specifically instructed to do so.
Copvriqht 'I' 2006 svoteotec Corporation All rights reservco

syrnantec.
Comparing CDS and Pre-4.x Disks
CDS Disk
(>4.x Default)
Private region
(metadata) and
public region (user
data) are created on
a single partition.
Suitable for moving
between different
operating systems
Not suitable for
boot partitions
Sliced Disk
(Pre-4.x Solaris Default)
Private region and
public region are
created on
separate
partitions.
Not suitable for
moving between
different operating
systems
Suitable for boot
partitions
Simple Disk
(Pre •••.x HP-UX Default)
Private region and
public region are
created on the
whole disk with
specific offsets.
Not suitable for
moving between
different operating
systems
Suitable for boot
partitions
Note: This format is
called hpdisk format
as of VxVM 4.1 on the
HP-UX platform.
Comparing CDS Disks and Pre-4.x Disks
The pre-t.v disk layouts arc still available in VxVM 4.0 and later. These layouts
are used for hringing the boot disk under VxVM control on operating systems that
support that capability,
On platforms that support bringing the boot disk under V.xVM control, CDS disks
cannot be used tor boot disks. CDS disks have specific disk layout requirements
that enable a common disk layout across different platforms, and these
requirements arc not compatible with the particular platform-specific requirements
of boot disks. Therefore, when placing a hoot disk under VxVM control. you must
use a prc-4.x disk layout (sliced on Solaris, hpdisk on HP-UX).
For non boot disks, you can convert CDS disks to sliced disks and vice versa by
using VxVM utilities.
Other disk types, working with boot disks, and transferring data across platforms
with CDS disks are topics covered in detail in later lessons.
1-12
Cupyrlght L 200t) Symantec Corporation All f,gtl1s reserved

Disk Group
Volumes
acctdg
expvol payvol
acctd 01-01
acctdgO~ - O~
llcctdg03-02
expvol-Ol
acctdg01-02 llcctdg03-01
acctdg02-01
Ploxes
payvol-Ol payvol-02
-VxVM Disks acctdg03
Subdisks
Physical Disks
I
Disk Groups
A disk group is a collection of Vx VM disks that share a common configurution.
You group disks into disk groups for management purposes. such as to hold the
data for a specific application or set of applications. For example. data for
accounting applications can be organized in a disk group called acctdg. A disk
group contigurcttion is a set olrccords with detailed information about related
Volume Manager objects in a disk group. their attributes. and their connections.
Volume Manager objects cannot span disk groups. For example. a volume's
subdisks. plexes. and disks must be derived from the same disk group as the
volume. You can create additional disk groups as necessary. Disk groups enable
you to group disks into logical collections. Disk groups and their components can
be moved as a unit from one host machine to another.
Volume Manager Disks
A Volume Manager (VxVM) disk represents the public region ota physical disk
that is under Volume Manager control. Each VxVM disk corresponds to one
physical disk. Each VxVM disk has a unique virtual disk name called a disk media
name,The disk media name is a logical name used lor Volume Manager
administrative purposes. Volume Manager uses the disk media name when
assigning space to volumes. A VxVM disk is given a disk media name when it is
added to a disk group.
Default disk media name: diskgrollplili
Copyright ~~.2Un6 Svmantec Corporanon All rights reserved
1-13Lesson 1 Virtual Objects

You can supply the disk media name or allow Volume Manager to assign a default
name. The disk media name is stored with a unique disk ID to avoid name
collision. After a VxVM disk is assigned a disk media name, the disk is no longer
referred 10 by its physical address. The physical address (for example, clltlldll or
hdiskll) becomes known <IS the disk access record.
Subdisks
A VxVM disk can be divided into one or more subdisks. A subdisk is a set of
contiguous disk blocks that represent a specific portion ofa VxVM disk, which is
mapped to a specific region of a physical disk. A subdisk is a subsection of a disk's
public region. A subdisk is the smallest unit of storage in Volume Manager.
Therefore, subdisks are the building blocks for Volume Manager objects.
A subdisk is defined by an offset and a length in sectors on a VxVM disk.
Default subdisk name: DMname-1l1l
A Vx VM disk can contain multiple subdisks, but subdisks cannot overlap or share
the same portions ofa VxVM disk. Any VxVM disk space that is not reserved or
that is not part of a subdisk is free space. You can use free space to create new
subdisks.
Conceptually, a subdisk is similar to <I partition. Both a subdisk and a partition
divide a disk into pieces defined by an offset address and length. Each of those
pieces represent a reservation of contiguous space on the physical disk. However,
while the maximum number of partitions to a disk is limited by some operating
systems, there is no theoretical limit to the number of subdisks that can be attached
to a single plex. This number has been limited by default to <I value 01'4090. If
required, this default can be changed, using the vo1_ subdisk _num tunable
parameter. For more information on tunable parameters, see the I'ERITAS Volume
.Mal/agerAdministrator '.1 Guide.
Plexes
Volume Manager uses subdisks to build virtual objects called plexes. A plex is a
structured or ordered collection of subdisks that represents one copy of the data in
a volume. A plex consists of one or more subdisks located on one or more physical
disks. The length of a plex is determined by the last block that can be read or
written on the last subdisk in the plex.
Default plcx name: volume_name-Il#
Volumes
A volume is a virtual storage device that is used by applications in a manner
similar to <I physical disk. Due 10 its virtual nature, a volume is not restricted by the
physical size constraints that apply to a physical disk. A VxVM volume can be as
large as the total of available. unreserved free physical disk space in the disk
group. A volume consists of one or more plcxes.
Default volume name: vol ul1le name##
Copyright l(; 2006 Syruantec Corporation All fights reserved

, S)111'1I1t('(
~
Volume Layouts
Volume layout: The way plexes are configured to remap the
volume address space through which 1/0 is redirected
.....Di~~~U [--R-~-;;i~;~~;1
LayeredConcatenated Striped
RAID-O
Data Redundancy
Mirrored RAID-5 Striped and
RAl0-O+1RAl0-5
Volume Manager RAID Levels
RAID
I
RAID is an acronym for Redundant Array of Independent Disks. RAID is a
storage management approach in which an army of disks is created. and part of the
combined storage capacity of the disks is used to store duplicate information about
the data in the array. By maintaining a redundant array of disks. you can regenerate
data in the case of disk failure.
RAID configuration models arc classified in terms of RAID levels. which arc
defined by the number of disks in the array. the way data is spanned across the
disks. and the method used for redundancy. Each RA ID level has speci lie features
and performance benefits that involve a trade-oil between performance and
reliability.
Volume Layouts
RAID levels correspond to volume layouts. A volume's layout refers to the
organization of plexcs in a volume. Volume layout is the way plcxes are
configured to remap the volume address space through which 110 is redirected at
run-time. Volume layouts are based on the concepts of disk spanning. redundancy.
and resilience.
Disk Spanning
Disk spanning is the combining of disk space from multiple physical disks to Iorrn
one logical drive. Disk spanning has two forms:
Copyright!': 2006 Symantec Corporation All rights reserved
1-15

Concatenation: Concatenation is the mapping of data in a linear manner
across two or more disks.
In a concatenated volume. subdisks are arranged both sequentially and
contiguously within a plex. Concatenation allows a volume to be created from
multiple regions of one or more disks if there is not enough space for an entire
volume on a single region of a disk.
Striping: Striping is the mapping of data in equally-sized chunks alternating
across multiple disks. Striping is also called interleaving.
In a striped volume. data is spread evenly across multiple disks. Stripes are
equally-sized fragments that are allocated alternately and evenly to the
subdisks of a single plcx. There must be at least two subdisks in a striped plex ,
each of which must exist on a different disk. Configured properly. striping not
only helps to balance 1/0 but also to increase throughput.
Data Redundancy
To protect data against disk failure, the volume layout must provide some form of
data redundancy. Redundancy is achieved in two ways:
Mirroring: Mirroring is maintaining two or more copies of volume data.
A mirrored volume uses multiple plcxcs to duplicate the information contained
in a volume. Although a volume can have a single plex, at least two are
required for true mirroring (redundancy of data). Each of these plexes should
contain disk space from different disks for the redundancy to be useful.
Parity: Parity is a calculated value used to reconstruct data after a failure by
doing an exclusive OR (XOR) procedure on the data. Parity information can be
stored on a disk. Ifpart ofa volume fails, the data on that portion of the failed
volume can be re-created from the remaining data and parity information.
A RAIO-S volume uses striping to spread data and parity evenly aeross
multiple disks in an array. Each stripe contains a parity stripe unit and data
stripe units. Parity can be used to reconstruct data if one of the disks fails. In
comparison to the performance of striped volumes, write throughput of RAI D-
S volumes decreases, because parity infonnauon needs to be updated each time
data is accessed. However. in comparison to mirroring, the use of parity
reduces the amount of space required.
Resilience
A resilient volume, also called a layered volume, is a volume that is built on one or
more other volumes. Resilient volumes enable the mirroring of data at a more
granular level. For example. a resilient volume can be concatenated or striped at
the top level and then mirrored at the bottom level.
A layered volume is a virtual Volume Manager object that nests other virtual
objects inside of itself. Layered volumes provide better fault tolerance by
mirroring data at a more granular level.
1-16 VERITAS Storage Foundation 5.0 for UNIX' Fundamentals
CopYrlyhl '& 2006 Symantec Corporauon AU nqhts reserved

, syrnanrec
I
Lesson Summary
• Key Points
This lesson described the virtual storage objects
that VERITAS Volume Manager uses to manage
physical disk storage, including disk groups,
VxVM disks, subdisks, plexes, and volumes.
• Reference Materials
VERITAS Volume Manager Administrator's Guide
'symantl'C
Labs and solutions for this lesson arc located on the following pages:
Appendix A provides complete lab instructions. "Lib I' IrilJ"duc'ing tile: !,;11
Lnvironmctu." p:I',!C i-,~
Appendix B provides complete lab instructions and solutions, "I .ab 1 S"IUlioI1S:
1,llrodliCIII;2 the 1:11) 1..11in1J1!l1CIlI," rage' B-,;
Lab 1
Lab 1: Introducing the Lab Environment
In this lab, you are introduced to the lab
environment, system, and disks that you will use
throughout this course.
For Lab Exercises, see Appendix A.
For .t:cabSoluti~!!s, see Appendix B,
Copyright'~ 2006 Svmantec Corporation All rights reserved
1-17

Copynghi .~ 2006 Svmantec Corporation. AlIlIghl!'i resorvec

Lesson 2
Installation and Interfaces

syrnantec
Lesson Introduction
~_Lesson~'!..sta~lationandInterface.s_~"
System
Changes
Problems
"~, .AS:, #'lli~1[ii-Jl L , svmantcc
Topic After completing this lesson, you will
be able to:
Topic 1: Installation Identify operating system compatibility and
Prerequisites other preinstallation considerations.
Topic 2: Adding License Keys Obtain license keys, add licenses by using
vxlic inst, and view licenses by using
vxlicrep.
Topic 3: VERITAS Software Identify the packages that are included in the
Packages Storage Foundation 5.0 software.
Topic 4: Installing Storage Install Storage Foundation interactively, by
Foundation using the installation utility.
Topic 5: Storage Foundation Describe the three Storage Foundation user
User Interfaces interfaces.
Topic 6: Managing the VEA Install, start, and manage the VEA server.
Server
2-2
Copynyhl:- 2006 Svmantec Corporaucn. All fights reserveu

'''l!t ..,~. ~ .. ", ,,,
~[ , S}ll1;!n1CC.
as Compatibility
The VERITAS Storage Foundation product line
operates on the following operating systems:
SF Solaris HP-UX AIX Linux
Version Version Version Version Version
5.0 8,9,10 11i.v2 (0904) 5.2,5.3
RHEL 4 Update 3,
SLES 9 SP3
4.1 8,9,10, x86 11i.v2 (0904) No release
RHEL 4 Update 1 (2.6),
SLES 9 SP1
4.0 7,8,9 No release 5.1, 5.2, 5.3 RHEL 3 Update 2 (1686)
3.5.x 2.6,7,8 11.11i (0902) No release No release*
• Note: Version 3.2.2 on Linux has
functionality equivalent to 3.5 on Solaris.
I
Installation Prerequisites
OS Version Compatibility
Before installing Storage Foundation. ensure that the version of Storage
Foundation that you are installing is compatible with the version of the operating
system that you are running. You may need to upgrade your operating system
before you install the latest Storage Foundation version.
VERITAS StorageFoundation 5.0 operates on the following operating systems:
Solaris 8 (SPARe Platform 32-bit and 64-bil)
Solaris 9 (SPARe Platform 32-bit and M-bil)
Solari, 10 (SPARe Platform M-bil)
September 2004 release of HP-UX II i version 2.0 or later
AIX 5.2 ML6 (legacy)
AIX 5.3 TL4 with SP4
Red Hat Enterprise l.inux 4 (RIIEL 4) with Update 3 (2.6.9-34
kernel) on AMD Optcron or Intel Xeon EM64T (xX6_(4)
SUSE Linux Enterprise Server <) (SLES 9) with SP3 (2.6.5-7.244.
252 kernels) on AMD Opteron or Intel Xcon EM64T (xR6_(4)
Check the /'F:R1TAS Storag« Foundation Release No/es Ior additional operating
Solaris
HP-UX
AIX
Llnux
system requirements.
Lesson 2 Installation and Interfaces 2-3
Copyright b 2006 svoantcc Corporation All nqhts reserved.

symantec.
Support Resources
·Il;'l,q! Storag~ Founnanonfor U~~IX
Products I
Search for Technotes
I Support services ~~E:~~,~""""~
1Wf'l'tIttlfUialln.
r~'-"-'''-''-F-"-,,,,-,-.,,--~----rp~a~tc~h~e~s=;I--------'---------------
"un •.•.••,"tlf~d•• "·, lI.IM ••••••••.'."". hll ••• " •.•••lo.11<,11'.
'->."';' ~:.;tn ~J ""~': .• '!It-:.," •• s. ~~, "'.1)00;<1':><;"< Jo·1'.~$ot> ,~'~.;~~'''' ""~,!~~.•.,,,~ ,.~.•~~~,,<;<'"
r·,.,,~,',1""rH .'.'''~'''''' ~,~.•••.",~~'·""··'''''.''''f· C.h"·"-" ':l'""c•.- •.."'!!....•••...•I"*"""~-~,~ 'w,'<q- •.•••"""'~r.·j'f'·
".""' ...••
.,~,.'"'At.~~'·""~'>''
~1'''fl''t'"F''''''''~I'r.-~,t>.U'
"," ~" •• It••• ,"',.,.I! ••• ,,- •••••
'''' .••'~ .'" :£.:..••,..,.. F<.'A>u<III._ ••
•( •. "-1)'
http://support.veritas.com(',ur4'''·'',IIa",I.il<·''''I''"'·
" •• ~~: ..,_''' ••.•• ,'''~ •• ·d ••• l01·••• ,.,.., ••••••••••••• "'"'".;>
:t"".~.1":1""'-. ""'.1,1-:" "''to ..••..., .,~•••r""· •.'" +"" ,~._.
"'j<.('.o.:', "~"""'''''~ "',-.~ .•.1'
Version Release Differences
With each new release of the Storage Foundation software. changes are made that
may affect the installation or operation of Storage Foundation in your
environment. By reading version release notes and installation documentation that
are included with the product, you can stay informed of any changes,
For more information about specific releases ofYERITAS Storage Foundation,
visit the YERITAS Support Web site at: http: / / support. veri tas. com
This site contains product and patch information, a searchable knowledge base of
technical notes, access to product-specific news groups and c-mail norification
services, and other infonnation about contacting technical support staff.
Note: If you open a case with YERITAS Support. you can view updates at:
http://support.veritas.com/viewcase
You can access your case by entering the e-mail address associated with your case
and the case number.
2-4
Copynqhtc. 2006 Svmaruec Ccmoranon All fights reserved
VERITAS Storage Foundation 5,0 for UNIX: Fundamentals

,S)111illllt'L
I
Storage Foundation Licensing
• Licensing utilities are contained in the VRTSvlic
package, which is common to all VERITAS products.
• To obtain a license key:
- Create a vLicense account and retrieve license keys online.
vLicense is a Web site that you can use to retrieve and
manage your license keys.
or
- Complete a License Key Request form and fax it to
VERITAS customer support.
• To generate a license key, you must provide your:
- Software serial number
- Customer number
- Order number
Note: You may also need the network and RDBMS platform, system
configuration, and software revision levels.
Adding License Keys
You must have your license key before you begin installation. because you are
prompted for the license key during the installation process. A new license key is
nut necessary if you are upgrading Storage Foundation from a previously licensed
version of the product.
lfyou have an evaluation license key, you must obtain a permanent license key
when you purchase the product. The VER[TAS licensing mechanism checks the
system date to verify that it has not been set back. [I' the system date has been reset.
the evaluation license key becomes invalid.
Obtaining a License Key
License keys arc delivered on Software License Certificates to you at the
conclusion of the order fulfillment process. The certificate specifics the product
keys and the number of product licenses purchased. A single key enables you to
install the product un the number and type of systems for which you purchased the
license.
License key arc non-node locked.
In a non-node locked model. one key can unlock a product on different servers
regardless ufllost ID and architecture type.
In a nude locked model. a single license is tied to a single specific server. For
each server. you need a di fferent key.
Lesson 2 Installationand Interfaces
Copyright C 2006 Symamec Corporation. All nqhts reserved
2-5

symaruec
Generating License Keys
[http://vli-~~~-s-e-.v~-;i-ta-s~.-c~Jf-~---'
I·········
I. Access automatic
I
, license key generation
and delivery.
• Manage and track
I
i license key inventory
and usage.
I· Locate and reissue lost
license keys.
I· Report, track, and
1 resolve license key
I issues online.
I· Consotidate and share
license key information
with other accounts.
• To add a license key:
vxlicinst
• License keys are installed in:
/etc/vx/licenses/lic
• To view installed license key
information:
vxlicrep
Displayed information includes:
- License key number
- Name of the VERIT AS
product that the key enables
- Type of license
- Features enabled by the key
Generating License Keys with vLicense
VERITAS vl.icense (v L icense. veri tas. com) is a self-service online license
management system.
vl.iccnsc supports production license keys only. Temporary. evaluation. or
demonstration keys must be obtained through your VERITAS sales representative.
Note: The VRTSvl ic package can coexist with previous licensing packages. such
as VRTSIic. If you have old license keys installed in /etc/vx/eIm.leave this
directory on your system. The old and new license utilities cun coexist.
2-6 VERITAS Storage Foundation 5.0 for UNIX. Fundamentals
Copvnqht i; 2006 Svmantec Corporation An fights reserved

,S)11Hlnt.?( .
I
~
What Gets Installed?
In version 5.0, the default installation behavior is to
install all packages in Storage Foundation Enterprise HA.
In previous versions, the default behavior was to only
install packages for which you had typed in a license key.
In 5.0, you can choose to install:
• All packages included in Storage Foundation Enterprise HA
or
• All packages included in Storage Foundation Enterprise HA,
minus any optional packages, such as documentation
and software development kits
VERITAS Software Packages
When you install a product suite. the component product packages are installed
automatically. When installing Storage Foundation. be sure to follow the
instructions in the product release notes and installation guides.
Package Space Requirements
Before you install any of the packages. confirm that your system has enough free
disk space to accommodate the installation. Storage Foundation programs and files
are installed in the /. /usr. and / opt tile systems. Refer to the product
installation guides for a detailed list of package space requirements.
Solaris Note
VxFS often requires more than the default RK kernel stack size. so entries are
added to the jete/system file. This increases the kernel thread stack size of the
system to 24K. The original / ete/ system file is copied to
/ete/fs/vxfs/system.preinstall.
COpyrig!lt:fj 2006 Symanlec Corporation. All rights reserved

symantec
Optional Features
VERITAS FlashSnap
- Enables point-in-time copies of data with minimal
performance overhead
- Includes disk group split/join, FastResync, and
storage checkpointing (in conjunction with VxFS)
VERITAS Volume Replicator
- Enables replication of data to remote locations
- VRTSvrdoc: VVR documentation
• VERITAS Cluster Volume Manager
Used for high availability environments
Features are
Included In the
VxVM package,
but they require a
separate license.
Features are
Included In the
VxFS package,
but they require a
separate license.
VERITAS Quick 1/0 for Databases
Enables applications to access prealiocated VxFS files
as raw character devices
• VERITAS Cluster File System
Enables multiple hosts to mount and perform file
operations concurrently on the same file
Dynamic Storage Tiering
Enables the support for multivolume file systems by
managing the placement of files through policies that
control both initial file location and the circumstances
under which existing files are relocated
Storage Foundation Optional Features
Several optional features do not require separate packages, only additional
licenses. The following optional features are built-in to Storage Foundation that
you can enable with additional licenses:
VERITAS Flashxnap: FlashSnap facilitates point-in-time copies of data,
while enabling applications to maintain optimal performance, by enabling
features, such as FastResync and disk group split and join functionality.
FlashSnap provides an efficient method to perform offline and off-host
processing tasks, such as backup and decision support.
VERITAS Volume Replicator: Volume Rcplicator augments Storage
Foundation functionality to enable you to replicate data to remote locations
over any IP network. Replicated copies of data can be used for disaster
recovery, off-host processing, off-host backup, and application migration.
Volume Replicator ensures maximum business continuity by delivering true
disaster recovery and flexible off-host processing.
Cluster Functionality: Storage Foundation includes optional cluster
functionality that enables Storage Foundation to be used in a cluster
environment.
A cI/lSII!I' is a set of hosts that share a set of disks. Each host is referred to as a
node in a cluster. When the cluster functionality is enabled, all of the nodes in
the cluster can share VxVM objects. The main benefits of cluster
configurations are high availability and off-host processing.
VERITAS Cluster Server (VCS): ves supplies two major components
integral to eFS: the Low Latency Transport (LLT) package and the Group
2-8
CQPynyht'~ 2006 Syrn;'Jnlt:r.: Corporation. All fights reserveu

Membership and Atomic Broadcast (GAB) package. LLT provides node-
to-node communications and monitors network communications. GAB
provides cluster state. configuration. and membership service. and it
monitors the heartbeat links between systems to ensure that they are active.
VERIT AS Cluster File System (CFS): CFS is a shared file system that
enables multiple hosts to mount and perform IIle operations concurrently
on the same file.
VERITAS Cluster Volume Manager (CVM): CVM creates the cluster
volumes necessary for mounting cluster file systems.
VERIT AS Quick 1/0 for Databases: VERITAS Quick 1/0 for Databases
(referred to as Quick 1;0) enables applications to access preallocated VxFS
tiles as raw character devices. This provides the administrative benefits of
running databases on file systems without the performance degradation usually
associated with databases created on file systems.
Dynamic Storage Tiering (DST): DST enables the support for multivolume
file systems by managing the placement of files through policies that control
both initial tile location and the circumstances under which existing files are
relocated.
Lesson 2 Installation and Interfaces
Copyrigh1 It 200{J Svmentec Corporation. All nqtus reserved.
I
2-9

syrnantec
Installation Menu
Storage Foundation and High Availability Solutions 5.0
SYMANTEC Product Version Installed Licensed
Veritas Cluster Server
Veritas File System
Veritas Volume Manager
Veri tas Volume Replicator
Veritas Storage Foundation
Veritas Storage Foundation for Oracle
Veri tas Storage Foundation for DB2
Veritas Storage Foundation for Sybase
Veri tas Storage Foundation Cluster File System
Veritas Storage Foundation for Oracle RAe
no no
no no
no
no no
no no
no no
no
no no
no no
Task Menu:
[2> Install/Upgrade a Product
L) License a Product
U) Uninstall a Product
Q) Quit
C) Configure an Installed Product
P) Perform a Preinstallation Check
0) View a Product Description
?) Help
Enter a Selection: [I,C,L,P,U,D,Q,?]
Installing Storage Foundation
The Installer is a menu-based installation utility that you can use to install any
product contained on the VERITAS Storage Solutions CD-ROM. This utility acts
as a wrapper for existing product installation scripts and is most useful when you
are installing multiple VERI rAS products or bundles, such as VERITAS Storage
Foundation or VERITAS Storage Foundation tor Databases.
Note: The example on the slide is from a Solaris platform. Some of the products
shown on the menu may not be available on other platforms. For example,
VERITAS File System is available only as part of Storage Foundation on HP-liX.
Note: The VERITAS Storage Solutions CD-ROM contains an installation guide
that describes how to use the installer utility. You should also read all product
installation guides and release notes even if you are using the installer utility.
To add the Storage Foundation packages using the installer utility:
1 Log on as supcruscr.
2 Mount the VERITAS Storage Solutions CD-ROM.
3 Locate and invoke the installer script:
cd / cdrom/ CD_name
./installer
4 If the licensing utilities are installed. the product status page is displayed. This
list displays the VERITAS products on the CD-ROM and the installation and
licensing status of each product. If the licensing utilities are not installed, you
receive a message indicating that the installation utility could not cletermine
product status.
Copynghl & 2006 Svrnaotoc Corporation All flyhls reserved

5 Type I to install a product. Follow the instructions to select the product that
you want to install. Installation begins automatically.
When you add Storage Foundation packages by using the installer utility. all
packages are installed. lfyou want to add a specific package only. for example.
only the VRTSvrndoc package. then you must add the package manually from the
command line.
After installation. the installer creates three text files that can be used for auditing
or debugging. The names and locations or each file are displayed at the end or the
installation and are located in / opt/VRTS/ install / logs:
IFile Description
Installation log file Contains all commands executed during installation. their output.
and any errors generated by the commands: used for debugging
installation problems and for analysis by VERITAS Support
Responsefile Contains configuration information enteredduring the procedure:
can be used for future installation procedures when using the
installer script with the -responsefile option
Summary file Contains the output of Vf:RITAS product installation scripts:
shows products that were installed. locations of log and response
files, and installation messagesdisplayed
Methods for Adding Storage Foundation Packages
A first-time installation or Storage Foundation involves adding the software
packages and configuring Storage Foundation fur first-time use. You can add
VERITAS product packages by using one of three methods:
Method Command Notes
VLRITAS installer Installs multiple VERITAS
Installation Menu products interactively.
Installs packagesand conligures
Storage Foundation (or first-time
use.
Product installation installvm Install individual VFRITAS
scripts installfs products internctively.
installsf Installs packagesand configures
Storage Foundation lor first time
use.
Native operating pkgadd (Solaris) Install individual packages. for
system package swinstall (iIP-UX) example. when using your 0'11
installation
installp (AIX)
custom installation scripts,
commands First-time Storage Foundation
rpm (Linux )
configuration must be run as a
Then. to configure SF: separatestep.
vxinstall
Copynqht "( 2006 Svmanter. Corporation All rights reserved

Default Disk Group
o You can set up a system-
wide default disk group to
which Storage Foundation
commands default if you do
not specify a disk group .
o If you choose not to set a
default disk group at
installation, you can set the
default disk group later from
the command line.
Note:In StorageFoundation
4.0and later,the rootdg
requirementno longerexists.
symaru«,
Configuring Storage Foundation
: When you i~~taIISt~r~g~F~~~dation, y~u are asl<edify~~'V~~tt~~~~:l
Enclosure-Based Naming
HostJ
"-cl
Disk
Enclosures
.,.II '·c' 2
. ~.I. enc
j' -;;ncl
encO
o Standard device naming is based on
controllers, for example, cltOdOs2.
o Enclosure-based naming is based on
disk enclosures, for example, encO.
Configuring Storage Foundation
When you install Storage Foundation, you are asked if you want to configure it
during installation. This includes deciding whether to use enclosure-based naming
and a default disk group.
What Is Enclosure-Based Naming'!
An enclosure.or disk enclosure,is an intelligent disk array. which permits hot-
swapping of disks. With Storage Foundation. disk devices can be named for
enclosures rather than for the controllers through which they are accessed as with
standard disk device naming (for example. eOtOdO or hdisk2).
Enclosure-based naming allows Storage Foundation to access enclosures as
separate physical entities. By configuring redundant copies of your data on
separate enclosures, you can safeguard against failure of one or more enclosures.
This is especially useful in a storage area network (SAN) that uses Fibre Channel
hubs or fabric switches and when managing the dynamic multipathing (DMP)
feature of Storage Foundation. For example, if two paths (el t 99dO and
e2t99dO) exist to a single disk in an enclosure, VxVM can use a single DMP
metanode, such as eneO 0, to access the disk.
What Is a Default Disk Group'!
The main benefit of creating a default disk group is that Storage Foundation
commands default to that disk group if you do not specify a disk group on the
command line. defaul tdg specifies the default disk group and is an alias for the
disk group name that should be assumed if a disk group is not specified ill a
command.
COp,'!v;)ht G" 2()06 Svmantec COIl-'or<lIlOI1. All flI)t1ls reservcc

I
Storage Foundation Management Server
Storage Foundation 5.0 provides central
management capability by introducing a
Storage Foundation Management Server (SFMS).
With SF 5.0, it is possible to configure a SF host as
a managed host or as a standalone host during
installation.
A Management Server and Authentication Broker
must have previously been set up if a managed
host is required during installation.
To configure a server as a standalone host during
installation, you need to answer "n" when asked if
you want to enable SFMS Management.
You can change a standalone host to a managed
host at a later time.
Note: This course does not cover SFMS and managed hosts.
Storage Foundation Management Server
Storage Foundation 5.0 provides central management capability by introducing a
Storage Foundation Management Server (SFMS). For more information. refer to
the S/Or"geFoundation ManagementS(,I"I·(,I" Administrator's Guide.
Lesson 2 Installationand Interfaces
Copyright «; 2006 Svmantec Corporation All fignls rese-veo
2-13

Sularis
symantec
Verifying Package Installation
To verify package installation, use OS-specific
commands:
• Solaris:
pkginfo -1 VRTSvxvm
• HP-UX:
sw1ist -1 product VRTSvxvm
• AIX:
lslpp -1 VRTSvxvm
• Linux:
rpm -qa VRTSvxvm
Verifying Package Installation
llyou are not sure whether VERITAS packagesare installed, or if you want to
verify which packagesare installed on the system,you can view information about
installed packagesby using Ox-specific commands to list package information.
To list all installed packageson the system:
pkginfo
To restrict the list to installed VERITAS packages:
pkginfo I grep VRTS
To display detailed information about a package:
pkginfo -1 VRTSvxvm
HP-UX
To list all installed packageson the system:
sw1ist -1 product
To restrict the list to installed VERITAS packages:
sw1ist -1 product I grep VRTS
To display detailed information about a package:
sw1ist -1 product VRTSvxvm
2-14
Cqpyngtll (,. 2006 Symautec Corporauon All nqhts reserved

AIX
To list all installed packages on the system:
lslpp
To restrict the list to installed VERITAS packages, type:
lslpp -1 'VRTS*'
To verify that a particular Iileset has heen installed, use its name, for example:
lslpp -1 VRTSvxvrn
To verify package installation on the system:
rpm -qa I grep VRTS
To verify a specific package installation on the system:
rpm -q[i] package_name
For example, to verify that the VRTSvxvm package is installed:
rpm -q VRTSvxvrn
The - i option lists detailed information about the package.
ILinux
Copyrighi ~ 2006 Svrnantec Corporation All rigl1;; reserved

svmantec
Storage Foundation User Interfaces
Storage Foundation supports three user interfaces:
• VERITAS Enterprise Administrator (VEA):
A GUI that provides access through icons, menus,
wizards, and dialog boxes
Note: This course only covers using VEA on a standalone
host.
• Command-Line Interface (CLI): UNIX utilities that
you invoke from the command line
• Volume Manager Support Operations
(vxdiskadm): A menu-driven, text-based interface
also invoked from the command line
Note: vxdiskadm only provides access to certain disk and
disk group management functions.
Storage Foundation supports three user interfaces, Volume Manager objects
created by one interface are compatible with those created by the other interfaces.
YERITAS Enterprise Administrator (YEA): VERITAS Enterprise
Administrator (VEA) is a graphical user interface to Volume Manager and
other VERITAS products. VEA provides access to Storage Foundation
functionality through visual clements, such as icons, menus. wizards, and
dialog boxes. Using VEA, you can manipulate Volume Manager objects and
also perform common tile system operations.
Command-Line Interface (CLI): The command-line interface (ell) consists
of UNIX utilities that you invoke from the command line to perform Storage
Foundation and standard UNIX tasks. You can use the ell not only to
manipulate Volume Manager objects. but also to perform scripting and
debugging functions. Most of the ell commands require supcruser or other
appropriate privileges. The ell commands perform functions that range from
the simple to the complex, and some require detailed user input.
Volume Manager Support Operations (vxdiskadm): The Volume
Manager Support Operations interface, commonly called vxdiskadm, is a
menu-driven, text-based interface that you can use for disk and disk group
administration functions. The vxdiskadm interface has a main menu from
which you can select storage management tasks.
A single VEA task may perform multiple command-line tasks.
2-16
Copyuqtu 'c 2(1)6 Syn.autec Corporauon All fights reserved

-
I
, syrnaruec.
; Menu Bart t"'_ ~.,... •....•~ .•...•....;
VEA: Main Window
Quick
Access
Bar
; Toolbar
lit 0
<_l ;::""ot.,,.._.! t_'f-"'W"" ~"'br-'" ,..,,,,,-.,.,-.l~<C"'><» ""cO._"",," ~<W1'A<'.~ "'•.•••".I>t.~l
,,: _.,._.,n
~ •.•.•••• ">.
t loi,-,q , '>
ll"''''
~ u; ,-
..."
:: t,.i:.,,;~~'
.'_•....••.
Three ways to
access tasks:
1. Menu bar
2. Toolbar
3. Context menu
(right-click)
""'_==":;":'-':'...1 ~';':"'"d,.·'.·'. '" '~.•.,~
Using the VEA Interface
The VERITAS Enterprise Administrator (VEA) is the graphical user interface for
Storage Foundation and other VERITAS products. You can use the Storage
Foundation features of VEA to administer disks, volumes, and file systems on
local or remote machines.
VEA is a Java-based interface that consists of a server and a client. You must
install the VEA server on a UNIX machine that is running VERITAS Volume
Manager. The VEA client can Tunon any machine that supports the Java 1.4
Runtime Environment, which can be Solaris. IIP-UX, AIX, Linux, or Windows.
SOllie Storage Foundation features ofVEA include:
Remote Administration
Security
Multiple Host Support
Multiple Views of Objects
Setting VE! Preferences
You can customize general VEA environment auributes through the Preferences
window (Select Tools - --Prcfcrcncev).
CopYllght ttl2006 Symanter. Corporation. All rights resetveo

symaruec
VEA: Viewing Tasks and Commands
To view underlying command
lines, double-click a task.
, *-'(:~<;,<j~"
"'¥i:t'w. " , 'I" ~ ,.__"
'-'--""';"--+-1 ~;;~~i':::J~~;~'~
fr<l flnlC" l!<;.u,t.H~"
t>O!k!'lIillIll
('jU! •• r~ltr ••O<.IP
T,"~1."l
(1"~t<i.uU!<)o~
."OI$tUi.
'oWN ,",!Jg eso 'I'+! 1(!)t;rrr", ~fnOl
•••••~.I¢"!I"'I(iK<:,.,~::.'l'
UU(
kl'~-":-:O''''''''''----'--'-----'
i~~lr4"~IIfll
Viewing Commands Through the Task Log
The Task Log displays a history of the tasks performed in the current session. Each
task is listed with properties. such as the target object of the task. the host, the start
time, the task status, and task progress.
Displaying the Task Lug window: To display the Task Log, click the Logs
tab at the left of the main window.
Clearing the Task History: Tasks are persistent in the Task History window.
To remove completed tasks from the window, right-click a task and select
Clear All Finished Tasks.
Viewing ell Commands: To view the command lines executed for a task,
double-click a task. The Task Log Details window is displayed tor the task.
The CLI commands issued are displayed in the Commands Executed field of
the Task Details section.
VERITAS Storage Foundation 5.0 for UNIX: Fundamentals2-18
COlJyflgtll '"' 2006 Symantec Coporauon All lights reservcc

, S)111.1I1lt'l.
I
Command-Line Interface
You can administer CLI commands from the UNIX shell
prompt.
Commands can be executed individually or combined
into scripts.
Most commands are located in /usr/sbin. Add this
directory to your PATH environment variable to
access the commands.
Examples of CLI commands include:
vxassist
vxprint
vxdg
vxdisk
Creates and manages volumes
Lists VxVM configuration records
Creates and manages disk groups
Administers disks under VM control
Using the Command-Line Interface
The Storage Foundation command-line interface (CLl) provides commands used
for administering Storage Foundation from the shell prompt on a UNIX system.
CLl commands can be executed individually for specific tasks or combined into
scripts.
The Storage Foundation command set ranges from commands requiring minimal
user input to commands requiring detailed user input. Many of the Storage
Foundation commands require an understanding of Storage Foundation concepts.
Most Storage Foundation commands require supcruser or other appropriate access
privileges.
CLI commands are detailed in manual pages.
Accessing Manual Pages for CLI Commands
Detailed descriptions ofVxVM and VxFS commands. the options for each utility.
and details on how to use them are located in VxVM and VxFS manual pages.
Manual pages are installed by default in / opt/VRTS/man. Add this directory to
the MANPATI I environment variable. if it is not already added.
To access a manual page. type man command name.
Examples:
man vxassist
man mount vxfs
Linux Note
On Linux. you must also set the MANSECT and ~1ANPATH variables.
Lesson 2 Installation and Interfaces
Copyrigtlt l~, 2U06 Symamec Corpotauon All rights .csorvoo
2-19

symantcc
The vxdi skadm Interface
vxdiskadm
Volume Manager Support Operations
Menu: volumeManager/Disk
1 Add or initialize one or more disks
2 Encapsulate one or more disks
3 Remove a disk
4 Remove a disk for replacement
5 Replace a failed or removed disk
list List disk information
Display help about menu
?? Display help about the menuing system
q Exit from menus
Note: This example is from a Solaris platform. The options may be
slightly different on other platforms.
Using the vxdiskadm Interface
The vxdiskadm command is a CLI command that you can use to launch the
Volume Manager Support Operations menu interface. You can use the Volume
Manager Support Operations interface, commonly referred to as vxdiskadm. to
perform common disk management tasks. The vxdiskadm interface is restricted
10 managing disk objects and does not provide a means of handl ing all other
VxVM objects.
Each option in the vxdiskadm interface invokes a sequence ofCLI commands.
The vxdiskadm interlace presents disk management tasks to the user as a series
of questions. or prompts.
To start vxdiskadm. you type vxdiskadm at the command line to display the
main menu.
The vxdiskadm main menu contains a selection of main tasks that you can use to
manipulate Volume Manager objects. Each entry in the main menu leads you
through a particular task by providing you with information and prompts. Default
answers arc provided for many questions, so you can select common answers.
The menu also contains options for listing disk information, displaying help
information. and quilling the menu interface.
The tasks listed in the main menu are covered throughout this training. Options
available in the menu differ somewhat by platform. See the vxdiskadm (1m)
manual page for more details on how to use vxdiskadm.
Note: vxdiskadm can be run only once per host. A lock file prevents multiple
instances from running: /var / spool / locks/ .DrSKADO. LOCK.
2-20
Copynqht F; ';:006 Svmamcc Corporation All rights reserved
VERITAS Storage Foundation 5.0 for UNIX. Fundamentals

I
Installing VEA
Installation
administration
file (Solaris only):
VRTSobadmin
Windows
Client packages:
• VRTSobgui, VRTSat, VRTSpbx,
VRTSicsco (UNIX)
Server packages:
• VRTSob
• VRTSobc33
• VRTSaa
VRTSccg
• VRTSdsa
• VRTSvail
• VRTSvmpro
• VRTSfspro
'UN/X VRTSddlpr
--_.__._---_._--..-
r-;, Install the VEA
I server on a UNIX
I machine running
I
I. Storage Foundation.
Install the VEA
client on any
I machine that
,I
supports the Java
1.4 Runtime
Environment (or
I later).
• windows/VRTSobgui .rosi (Windows)
VEA is installed automatically when you run the SF installation
scripts. You can also install VEA by adding packages manually.
Managing the VEA Software
YEA consists of a server and a client. You must install the YEA server on a UNIX
machine that is running YERITAS Volume Manager. You can install the YEA
client on the same machine or on any other UNIX or Windows machine that
supports the Java 1.4 Runtime Environment (or later),
Installing the VEA Server and Client on UNIX
If you install Storage Foundation by using the Installer utility. you arc prompted to
install both the ,[A server and client packages automatically. If you did not install
all of the components by using the Installer. you can add the YEA packages
separately.
It is recommended that you upgrade YEA to the latest version released with
Storage Foundation in order to take advantage of new functionality built into YEA.
You can use the YEA with 4.1 and later to manage 3.5.2 and later releases.
When adding packages manually. you must install the Volume Manager
(VRTSvl ie. VRTSvxvrn) and the infrastructure packages (VRTSat. VRTSpbx.
VRTSieseo) before installing the YEA server packages. After installation. also
add the YEA startup scripts directory. / opt/VRTSob/bin. to the PATH
environment variable.
Lesson 2 Installationand Interfaces 2-21
Copyright ,°,2006 Symanrec Corporation. Anuqhts resorvoo

syrnanrec
Starting the VEA Server and Client
Once installed, the VEA server starts up automatically
at system startup.
To start the VEA server manually:
1. Log on as superuser.
2. Start the VEA server by invoking the server program:
/opt/VRTSob/bin/vxsvc (on Solaris and HP-UX)
/opt/VRTSob/bin/vxsvcctrl (on Linux)
When the VEA server is started:
/var /vx/ isis/vxis is. lock ensures that only one instance
of the VEA server is running.
/var/vx/isis/vxisis .log contains server process log
messages.
To start the VEA client:
On UNIX: /opt/VRTSob/bin/vea
On Windows: Select Start->Programs->VERIT AS->
VERITAS Enterprise Administrator.
Starting the VEA Server
In order to use YEA. the YEA server must be running on the UNIX machine to be
administered. Only one instance of the VEA server should be running at a time.
Once installed. the YEA server starts up automatically at system startup. You can
start the YEA server manually by invoking vxsvc (on Solaris and HP-UX).
vxsvcctrl (on Linux ), or by invoking the startup script itself, for example:
Solaris
/etc/rc2.d/S73isisd start
~IP-LJX
/sbin/rc2.d/S700isisd start
The YEA client call provide simultaneous access to multiple host machines. Each
host machine must be running the VEA server.
Note: Entries for your user name and password must exist in the password file or
corresponding Network Information Name Service table on the machine to be
administered. Your user name must also be included in the YERITAS
administration group (v r t s adm, by default) in the group tile or NIS group table.
If the vrtsadm entry does not exist. only root can run YEA.
You can contigure YEA to connect automatically to hosts when you start the YEA
client. In the YEA main window. the Favorite Hosts node can contain a list of
hosts that arc reconnected by default at the startup of the YEA client.
2-22
Copyright .; 2006 Svrnantec Corpcrahon. AU nqnts reserved

, symanrec.
Managing VEA
The VEA server program is:
/opt/VRTSob/bin/vxsvc (Solaris and HP-UX)
/opt/VRTSob/bin/vxsvcctrl (Linux)
To confirm that the VEA server is running:
vxsvc -m (Solaris and HP-UX)
vxsvcctrl status (Linux)
To stop and restart the VEA server:
/etc/init.d/isisd restart (Solaris)
/sbin/init.d/isisd restart (HP-UX)
To kill the VEA server process:
vxsvc -k (Solaris and HP·UX)
vxsvcctrl stop (Linux)
To display the VEA version number:
vxsvc -v (Solaris and HP-UX)
vxsvcctrl version (Linux)
Managing the VEA Server
I
Monitoring VEA Event and Task Logs
You can monitor VEA server events and tasks from the [vent Log and Task Log
nodes in the VEA object tree. You can also view the VEA log file. which is located
at /var /vx/ isis/vxisis. log. This tile contains trace messages for the V[A
server and VEA service providers.
Copylight <E2006 Symantec Corporation Anucnts reserved
2-23Lesson 2 Installation and Interfaces

symantcc
Labs and solutions for this lesson are located on the following pages:
Appendix A provides complete lab instructions, "lab ::: lnstatl.uiou and
hucrruccs." I'a~l' , ..7
Appendix B provides complete lab instructions and solutions, "Lab 2 Solutiuns:
lnstullation and lnrcriucc-." page n 7
Lesson Summary
• Key Points
In this lesson, you learned guidelines for a first-
time installation of VERITAS Storage Foundation,
as well as an introduction to the three interfaces
used to manage VERITAS Storage Foundation.
- VERITAS Volume Manager Administrator's Guide
- VERITAS Storage Foundation Installation Guide
- VERITAS Storage Foundation Release Notes
- Storage Foundation Management Server
Administrator's Guide
2-24
Lab 2
Lab 2: Installation and Interfaces
In this lab, you install VERITAS Storage
Foundation 5.0 on your lab system. You also
explore the Storage Foundation user interfaces,
including the VERITAS Enterprise Administrator
interface, the vxdiskadm menu interface, and the
command-line interface.
For Lab Solutions, see Appendix B,
Copynqt-t ( 20DI'}Svrnantec Lorpcratton All riqhts leserved

Lesson 3
Creating a Volume and File System

svmantec
Lesson Introduction
• Lesson 3: Creating a Volume and File "",',
System
Changes
• Lesson 7: Resolving Hardware Problems
~ ~~'" ,~,"';'~;i!1I.. , symantcc
Topic After completing this lesson, you will be
able to:
Topic 1: Preparing Disks and Initialize an OS disk as a VxVM disk and
Disk Groups for Volume create a disk group by using VEA and
Creation command-line utilities.
Topic 2: Creating a Volume Create a concatenated volume by using VEA
and from the command-line,
Topic 3: Adding a File System Add a file system to and mount an existing
toa Volume volume.
Topic 4: Displaying Volume Display volume layout information by using
Configuration Information VEA and by using the vxprint command.
Topic 5: Displaying Disk and View disk and disk group information and
Disk Group Information identify disk status.
Topic 6: Removing Volumes, Remove a volume, evacuate a disk, remove a
Disks, and Disk Groups disk from a disk group, and destroy a disk
group.
3-2
Cop;lfIyhl'~ 2006 Svrnantec COrpOI(ltloll All rights reserved

."-*..,,.,.dt~., d"U1I;:M
Selecting a Disk Naming Scheme
Types of naming schemes:
• Traditional device naming: OS-dependent and based on
physical connectivity information
• Enclosure-based naming: OS-independent, based on the
logical name of the enclosure, and customizable
You can select a naming scheme:
• When you run Storage Foundation installation scripts
• Using vxdiskadm, "Change the disk naming scheme"
Enclosure-based named disks are displayed in three
categories:
Enclosures: enclosurenarne #
Disks: Disk #
Others: Disks that do not return a path-independent identifier
to VxVM are displayed in the traditional OS-based format.
Preparing Disks and Disk Groups for Volume Creation-
Here are some examples of naming schemes:
Naming Scheme Example
Traditional Solaris: /dev/ l r l dsk/clt9dOs2
HP-UX: /dev/ l r ] dsk/c3t2dO (no slice)
AIX: /dev/hdisk2
I.inux: /dev/sda. /dev/hda
Enclosure-based senaO-
1.senaO
-
2,senaO- 3..
Enclosure-based Customized englab2.hrl.boston3
I
Benefits of enclosure-based naming include:
Easier fault isolation: Storage Foundation can more effectively place data and
metadata to ensure data availability.
Device-name independence: Storage Foundation is independent of arbitrary
device names used by third-party drivers.
Improved SAN management: Storage Foundation can create better location
identification information about disks in large disk limns and SANs.
Improved cluster management: In a cluster environment. disk array names
on all hosts in a cluster can be the same.
Improved dynamic multipathing (DMP) management: With multipathcd
disks. the name of a disk is independent of the physical communication paths.
avoiding confusion and conflict.
Copyrighl;~ 2006 Symantec Corporation. All nqtus reserved
3-3Lesson 3 Creating a Volume and File System

symantec
~
Stage 1: ;
Initialize disk. J
~ !
Uninitialized :
Disk ;
Stage 2:
Assign disk
to disk group.
Before Configuring a Disk for Use by VxVM
In order to use the space ofa physical disk to build VxVM volumes, you must
place the disk under Volume Manager control. Before a disk can be placed under
volume Manager control, the disk media must be formatted outside ofVxVM
using standard operating system formatting methods. SCSI disks arc usually
prcformaued. After a disk is formatted. the disk can be initialized for use by
Volume Manager. In other words. disks must be detected by the operating system,
before VxVM can detect the disks.
Stage One: Initialize a Disk
A formatted physical disk is considered uninitialized until it is initialized for use
by VxVM. When a disk is initialized. the public and private regions are created.
and VM disk header information is written to the private region. Any data or
partitions that may have existed on the disk are removed.
These disks are under Volume Manager control but cannot be used by Volume
Manager until they are added to a disk group.
Note: Encapsulation is another method of placing a disk under VxVM control in
which existing data on the disk is preserved. This method is covered in a later
lesson.
Changing the Disk Layout
To display or change the default values that are used for initializing disks, select
the "Change/display the default disk layouts" option in vxdiskadm:
COPYright ,~, 2006 Svmantec Cornorauon All fights reserved

For disk initialization. you can change the default format and the default length
of the private region. If the attribute settings for initializing disks are stored in
the user-created tile. / etc/ defaul t /vxdi sk, they apply to all disks to be
initialized
On Solaris for disk encapsulation. you can additionally change the offset
values for both the private and public regions. To make encapsulation
parameters different from the default VxVM values. create the user-detined
/ etc/ defaul t /vxencap tile and place the parameters in this tile.
On HP-UX when converting LVM disks. you can change the default format
and the default private region length. The attribute settings are stored in the
/etc/default/vxencap file.
Stage Two: Assign a Disk to a Disk Group
When you add a disk to a disk group. VxVM assigns a disk media name to the disk
and maps this name to the disk access name.
Disk media name: A disk media name is the logical disk name assigned to a
drive by VxVM. VxVM uses this name to identify the disk for volume
operations. such as volume creation and mirroring.
Disk access name: A disk access name represents all UNIX paths to the
device. A disk access record maps the physical location to the logical name
and represents the link between the disk media name and the disk accessname.
Disk accessrecords arc dynamic and can be re-created when vxdctl enable
is run.
The disk media name and disk access name. in addition to the host name. are
written to the private region of the disk. Space in the public region is made
available for assignment to volumes. Volume Manager has full control of the disk.
and the disk can be used to allocate space tor volumes. Whenever the VxVM
configuration daemon is started (or vxdctl enable is run). the system reads the
private region on every disk and establishes the connections between disk access
names and disk media names.
A tier disks are placed under Volume Manager control. storage is managed in terms
of the logical configuration. File systems mount to logical volumes. not to physical
partitions. Logical names. such as
/dev/vx/ l r l dsk/diskgroup/volume_name. replace physical locations.
such as /dev/ [rl dsk/ device_name.
The free space in a disk group refers to the space on all disks within the disk group
that has not been allocated as subdisks, When you place a disk into a disk group.
its space becomes part or the tree space pool of the disk group.
Stage Three: Assign Disk Space to Volumes
When you create volumes. space in the public region of a disk is assigned to the
volumes. Some operations. such as removal of a disk from a disk group. are
restricted itspace on a disk is ill use by a volume.
Lesson 3 Creating a Volume and File System
Copyright c· 2006 Symantcc Corporation All rignls reserved
I
3-5

Disk Group Purposes
sym.uuec
Disk groups enable
you to:
• Group disks into logical
collections for a set of
users or applications.
• Easily move groups of
disks from one host to
another.
• Ease administration of
high availability
environments through
deport and import
operations.
sysdg
L[~~r[j[j[j[j
VM disks
VM disks
LiJ:~=:Ifqfigill [j [j
[j [j [j [j [j [j
VM disks VM disks
What Is a Disk Group?
A disk group is a collection of physical disks, volumes, plexes, and subdisks that
are used for a common purpose. A disk group is created when you place at least
one disk in the disk group. When you add a disk to a disk group. a disk group entry
is added to the private region header of that disk. Because a disk can only have one
disk group entry in its private region header. one disk group does not "know
about" other disk groups, and therefore disk groups cannot share resources, such as
disk drives, plexes, and volumes.
A volume with a plcx can belong to only one disk group. and subdisks and plexes
ofa volume must be stored in the same disk group. You can never have an "empty"
disk group, because you cannot remove all disks from a disk group without
destroying the disk group.
Why Are Disk Groups Needed?
Disk groups assist disk management in several ways:
Disk groups enable the grouping of disks into logical collections for a
particular set of users or applications.
Disk groups enable data. volumes. and disks to be easily moved from one host
machine to another.
Disk groups ease the administration of high availability environments. Disk
drives can be shared by two or more hosts. but they can be accessed by only
one host at a time. If one host crashes. the other host can take over its disk
groups and therefore its disks.
Copynght@ 2006 Svmantec Corporanon All flghLS reservec

System-Wide Reserved Disk Groups
Reserved
I System A I
[iJ
acc~
names:
LEi8J• bootdg bootdg • sysdg
• defaultdg defaultdg • acctdg mJ[j EJEH"HJI
• nodg
1::---'1I System B
bootdg • nodg
defaultdg • nodg
noc1g1s the default value tor
bootdg and defaultdg.
To display what is set as bootdg or defaul tdg:
vxdg bootdg
vxdg defaultdg
To set the default disk group after VxVMinstallation:
vxdctl defaultdg diskgroup ISystem-Wide Reserved Disk Groups
VxVM has reserved three disk group names that are used to provide boot disk
group and default disk group functionality. The names "bootdg, .,
"de f au I tdg, " and "nodq" arc system-wide reserved disk group names and
cannot be used as names for any of the disk groups that you set up.
If you choose to place your boot disk under VxVM control. VxVM assigns
bootdg as an alias for the name of the disk group that cuntains the volumes that
are used to boot the system.
de£aultdg is an alias for the disk group name that should be assumed if the -g
option is not specified to a command. You can set defaul tdg when you install
VERITAS Volume Manager or anytime alter installation.
By default. both bootdg and defaul tdg arc set to nodg.
Notes
The definitions ofbootdg and defaul tdg are written to the volboot file.
The definition ofbootdg results in a symbolic link from the named bootdg
in /dev/vx/dsk and /dev/vx/rdsk.
The rootdg disk group name is no longer a reserved name for VxVM
versions after 4.0. If you arc upgrading from a version ofVolume Manager
earlier than 4.0 where the system disk is encapsulated in the rootdg disk
group, the bootdg is assigned the value of rootdg automatically.
Copyright'f 2006 Symantec Corporation All rights reserved
3-7Lesson 3 Creatinga VolumeandFileSystem

syrnantcc
To create a disk group, you add a disk to a disk group.
• You can add a single disk or multiple disks.
• You cannot add a disk to more than one disk group.
• Default disk media names vary with the interface used to
add the disk to a disk group, but they are conventionally
in the format diskgroup##, such as datadgOO,
datadgOl, and so on.
• Disk media names must be unique within a disk group.
• Adding a disk to a disk group makes the disk space
available for use in creating Volume Manager volumes.
Creating a Disk Group
A disk must be placed into a disk group before it can be used by VxVM. A disk
group cannot exist without having at least one associated disk. When you create a
new disk group. you specify a name for the disk group and at least one disk to add
to the disk group. The disk group name must be unique for the host machine.
Adding Disks
To add a disk to a disk group, you select an uninitialized disk or a free disk. If the
disk is uninitializcd, you must initialize the disk before you can add it to a disk
group.
Disk Naming
When you add a disk to a disk group, the disk is assigned a disk media name. The
disk media name is a logical name used tor VxVM administrative purposes.
Notes on Disk Naming
You can change disk media names after the disks have been added to disk groups.
However. if you must change a disk media name, it is recommended that you make
the change before using the disk for any volumes. Renaming a disk does not
rename the subdisks on the disk, which may be confusing.
Assign logical media names. rather than use the device names. to facilitate
transparent logical replacement of failed disks. Assuming that you have a sensible
disk group naming strategy, the VEA or vxdiskadm default disk naming scheme
is a reasonable pol icy to adopt.
3-8
COPYright ~ 2006 Symantec Corporation. All rights reserveo

I
Create a disk group or add disks using vxdiskadm:
"Add or initialize one or more disks"
Initialize disks:
vxdisksetup -i device_tag [attributes)
vxdisksetup - i Disk 1 (Enclosure-based naming)
vxdisksetup -i c2tOdO (Solaris and HP-UX)
vxdisksetup -i hdisk2 (AIX)
vxdisksetup - i sda2 (Linux)
Initialize the disk group by adding at least one disk:
vxdg init diskgroup disk_name=device_tag
vxdg init datadg datadgOl=Disk_l
Add more disks to the disk group:
vxdg -g diskgroup adddisk disk_name=device_tag
vxdg -g datadg adddisk datadg02=Disk_2
From the vxdiskadm main menu, select the "Add or initialize one or more disks"
option. Specify the disk group to which the disk should be added. To add the disk
to a new disk group, you type a name for the new disk group. You use this same
menu option to add additional disks to the disk group.
To verify that the disk group was created, you can use vxdisk list.
When you add a disk to a disk group, the disk group configuration is copied onto
the disk, and the disk is stamped with the system host !D.
Creating a Disk Group: vxdiskadm
Copyright ,~ 2006 Symantec Corporation. All rights reserved
3-9

svrnantcc
New Disk Group Wizard Ei1Ercet e unique neme for this di::.f 9rouP, and then select .5iiI~
the disk: to include .11
.-~====~-------,
Av~iIabIe disks: Add >
'3 Qts~_O ..!.l
Add AI »
<3Dtsk_,
JGi DIS'_S ir":·~'~·:~~:~""·11. oJ:;i Dts,_6
I~Olsk 7
..:..I «Remo'w'e AI
Actions->Add Disk to Disk Grou~~ ~
Creating a Disk Group: VEA
Select: Disk Groups folder, or a free or uninitialized disk
Navigatlun path: Acrions=-c-Ncw Disk Group
Input: Group Name: Type the name of the disk group to be created.
Available/Selected disks: Select at least one disk to be placed in
the new disk group.
Disk names: To specity a disk media name for the disk that you
arc placing in the disk group. type a name in the Disk name field.
Itno disk name is specified. VxVM assigns a default name. If you
arc adding multiple disks and specify only one disk name. VxVM
appends numbers to create unique disk names.
Organization Principle: In an Intelligent Storage Provisioning
(ISP) environment. you can organize the disk group based on
policies that you setup. This option is covered in a later lesson.
Comment: Any user comments
Create cluster group: Displayed on HP-UX plauorms: to create
a shared disk group. mark this check box: only applicable in a
cluster environment.
Activation 'lode: Displayed on HP-UX platforms: applies to
cluster environments: possible values are on: Read write. Read
only: the default selling is Read write for non-cluster
environments.
Notc: When working in a SAN environment. or any environment in which
3-10 VERITAS Storage Foundation 5,0 for UNIX: Fundamentals
Copynght':;. ;WOf)Symantec copcreuou. All rights reserved

multiple hosts may share access to disks. it is recommended that you perform a
rescan operation to update the YEA view of the disk status before allocating any
disks. From the command line, you can run vxdctl enable.
Adding a Disk: VEA
Select: A free or unuutinlized disk
Navlgation path: Aclions->AJd Disk 10 Disk Group
Input: Disk Group name: Select an existing disk group.
New disk group: Click the ~ew disk group button 10 add the disk
to a new disk group.
Select the disk to add: You can move disks between the Selected
disks and Available disks fields by using the Add and Remove
buttons.
Disk Name(s): By default. Volumc Manager assignsa disk media
name that is basedon the disk group name of"a disk. You can
assign a different name to the disk by Iyping a name in the Disk
namcts) field. lfyou are adding more than one disk. place a space
between each name in the Disk nametsj field.
Comment: Any user comments
IWhen the disk is placed under YxYM control. the Type property changes to
Dynamic. and the Status property changes to Imported.
Lesson 3 Creating a Volume and File System 3-11
Copyright,f;; 2006 swnemec Corporaunn. All rigtlts reserved.

syrnantec
Creating a Volume: CLI
To create a volume:
vxassist -g diskgroup make volume name
length [attributes]
For example:
vxassist -g datadg make datavol 100m
Block and character (raw) device files are set up that
you can use to access the volume:
• Block device file for the volume:
/dev/vx/dsk/diskgroup/volume_name
• Character device file for the volume:
/dev/vx/rdsk/diskgroup/volume_name
To display volume attributes, use:
vxassist -g diskgroup help showattrs
Creating a Volume
Creating a Volume
When you create a volume using VEA or ell commands, you indicate the desired
volume characteristics, and VxVM creates the underlying plexcs and subdisks
automatically. The VxVM interfaces require minimal input if you use default
settings. For experienced users, the interfaces also enable you to enter more
detailed specifications regarding all aspects of volume creation.
Before You Create a Volume
Before you create a volume. ensure that you have enough disks to support the
layout type.
A striped volume requires at least two disks.
A mirrored volume requires at least one disk for each plex. A mirror cannot be
on the same disk that other plcxes of the same volume arc using.
Creating a Volume: CLI
To create a volume from the command line. you use the vxassist command. In
the syntax:
Use the -g option to specify the disk group in which to create the volume.
make is the keyword for volume creation.
volume_name is a name you give to the volume. Specify a meaningful name.
length specifies the number of sectors in the volume, You can specify the
length by adding an rn, k. g, or t to the length.
3-12
Copynqbt : 2(106 Symal118C:couorauon. All fights resorvec

I
, synuuucc
Creating a Volume: VEA: Assigning Disks
r······························ _ .
! Actions->New Volume I
, Step 1: Select disks to use for the volume. i
Name
..J Controllers
iil.9 0:3
,.j UIS'S
~ c3t2 .. , 16.9..
c:3tJ.. , 16.9 ..
--=:..J Excluded:
~ Enclosures
......31 DISk
r Mirror Across: 1 . ::J
r stripe Across: I"'I-"--::J--'
r Ordered
Creating a Volume: VEA
Select: A disk group
Navigation path: Actions->:--Jcw Volume
Input: Disks for this volume: Let VxVM decide (detuult). or manually
select disks to use.
Volume attributes: Specify a volume name. the size of the
volume. the type of volume layout. and other layout
characteristics. Assign a meaningful name to the volume that
describes the data stored in the volume.
File system: Create a tile system on the volume and set file
system options.
New Volume Wizard Step I: Assigning Disks to Use for a New Volume
By default. VxVM locates available space on all disks in the disk group and
assigns the space to a volume automatically based on the layout you choose.
Alternatively. you can choose specific disks. mirror or stripe across controllers.
trays. targets. or enclosures. or implement ordered allocation. Ordered allocation is
a method of allocating disk space to volumes based on a specific set "fVxVM
rules.
Copyriglll 1'. 2006 Svmantec Corporation All rigtlts reserved

Creating a Volume: VEA: Setting Volume
Attributes
_
S~te~p~2~:~s~p~e~C~ifY~V~O~lu~m~e~a~tt~r~ib~u~te~s~'~==::::-__ ---j Default options
change based on
!olume name: :d~la'~IOli Ithe layout type
comment i:.·~";·:·::·:"~··~":"·"""""""""·"""""""""·"""""""""·" .J•• ~ !you setect.
Si<e: :: . . ... _ _i. GB .1 I Ma~ Size:
Layout Mlrrorln!o
nMirrored
svmuntcc
New Volume Wizard Step 2: Specifying Attributes for a New Volume
Volume name: Assign a meaningful name to the volume that describes the data
stored in the volume.
Size: Specify a size for the volume. The default unit is GB. If you click the Max
Size button, YxYM determines the largest size possible for the volume based on
the layout selected and the disks to which the volume is assigned.
Select a size for the volume based on the volume layout and the space available
in the disk group. The size of the volume must be less than or equal to the
available tree space on the disks.
The size specified in the Size field is the usable space in the volume. For a
volume with redundancy (RAID-5, mirrored), YxYM allocates additional tree
space for the volume's parity information (RAID-5) or additional plexcs
(mirrored).
The free space available for constructing a volume of a specific layout is
generally less than the total free space in the disk group unless the layout is
concatenated or striped with no mirroring or logging.
Layout: Select a layout type from the group of options. The default layout is
concatenated.
Concatenated: The volume is created using one or more regions of specified
disks.
Striped: The volume is striped across two or mort: disks. The default number
uf columns across which the volume is striped is two. and the default stripc
unit size is 128 sectors (64K) on Solaris, AIX. and l.inux: 128 sectors (128K)
on HP-UX. You can specify different values,
I!..Concatenatell
o ~triped
C; BAlO-S
,.~, CQncatenated Mirrored
() Strined Mirrored
nEnable [astResl'11c
'..1 jnnialtze zero
3-14
Copynqbt - ;:OG6Svrnantec Corporation. All rlghlS ieeesvea

Concatenated Mirrored and Striped Mirrored: These options denote
layered volume layouts.
Mirror Info:
Mirrored: Mirroring is recommended, To mirror the volume, mark the
Mirrored check box. Only striped or concatenated volumes can be mirrored.
RAID-5 volumes cannot he mirrored.
Total mirrors: Type the total number of mirrors for the volume. A volume can
have up to 32 plcxcs: however. the practical limit is 31. One plex is reserved by
YxYM to perform restructuring or relocation operations.
Enable logging: To enable logging. mark the Enable logging check box. If you
enable logging. a log is created that tracks regions of the volume that are currently
being changed by writes. In case ofa system failure. the log is used to recover only
those regions identified in the log.
YxYM creates a dirty region log or a RAD-5 log. depending on the volume
layout. If the layout is RAID-5. logging is enabled by default. and YxYM adds an
appropriate number of logs to the volume.
Enable Fastltesync:To enable FastResync. mark the Enable FastResync check
box. This option is displayed only if you have licensed the FastResync option.
Initialize zero: To clear the volume before enabling it for general use. mark the
Initialize zero check box. For security purposes. you can use the Initialize Zero
option to overwrite all existing data in the volume area. However, this is time
consuming due to all the space that has to be written.
No layered volumes: To prevent the creation of a layered volume, mark the No
layered volumes check box. This option ensures that the volume has a nonlaycred
layout. If a layered layout is selected. this option is ignored.
Copyriqht ts" 2006 Symilntec: Corporanoo AI; right., rcsprvcri
I
3-15

symantec
Creating a Volume: VEA: Adding a File System
Durin Volume Creation
r No file system
r. Create a fie system I
~--------~------~~Create OptIons
'-1,,'-;-' ".,-" -::J'~
6Iockslze: IDefault (I " ::J
Now F'. System Det ...
MountOptions
Mountp... r:11-rnn-w-' 1-----
P Create mount posit
r Read only
P Honor setuid
P i<M=t9f;;;;:~g~lrit~
r Mountat boot
fsck pess: r
Mount file System Det"'I s.,;
New Volume Wizard Step 3: Creating a Snapshot Cache Volume
A storage cache may be named and shared among several volumes in the same
disk group, This is used only for point-in-time copies.
New Volume Wizard Step 4: Creating a File System on a New Volume
When you create a volume, you can place a file system on the volume and specify
options for mounting the tile system. You can place a tile system on a volume
when you create a volume or any time after creation.
The default option is "No tile system:' To place a tile system on thc volume, select
the "Create a tile system" option and specify:
File system type: Specify the tile system type as either vxfs (VERITAS File
System: or other OS-supported file system types (U FS on Solaris; I-IFS on H P-
UX; on AIX. JFS and .IFS2 arc not supported on VxVM volumes), To add a
VERITAS tile system. the VxFS product must be installed with appropriate
licenses,
Create Options:
Compress: If your platform supports tile compression. this option
compresses the files on your tile system (not available on Solaris/HP-UX).
Allocation unit or Block size: Select an allocation unit size (for OS-
supported tile system types): or a block size (for VxFS tile systems).
New File System Details: Click this button to specify additional tile-
system-specific mkf s options. For VxFS, the only explicitly available
additional options are large file support and log size. You can specify other
options in the Extra Options field,
COPYrlghl'.," 2006 Svrnantec Corporation All fight':. reservcc

Mount Options:
Mount point: Specify the mount point directory on which to mount the file
system. The new file system is mounted immediately after it is created.
Leave this field empty if you do not want to mount the tile system.
Create mount point: Mark this check box to create the directory il' it does
not exist. The mount point must be specified.
Read only: Mark this check box to mount the file system as read only.
Honor setuid: Mark this check box to mount the file system with the suid
mount option. This option is marked by default.
Add to tile system table: Mark this check box to include the file system in
the /etc/vfstab file (Solaris). the /etc/fstab file (IIP-UX. Linux).
or the /etc/filesystems tile (AIX).
Mount at boot: Mark this check box to mount the file system
automatically whenever the system boots. This option is not displayed on
IIP-UX.
fack pass: Specify how many fsck passeswill be run if the file system is
not clean at mount time.
Mount File System Details: Click this button to specify additional mount
options. For YxFS. the explicitly available additional options include
disabling Quick 110. setting directory permissions and owner. and setting
caching policy options. You can specify other options. such as quota. in
the Extra options field.
Copynght if' 2006 Symantec Corporation All right~ res.erved
I
3-17

syrnanicc.
Adding a File System to a Volume
A file system provides an organized structure to facilitate the storage and retrieval
of files. You can add a tile system to a volume when you create a volume or any
time after you create the volume initially.
When a tile system has been mounted on a volume. the data is accessed
through the mount point directory.
When data is written to files. it is actually written to the block device tile:
/dev/vx/dsk/disk_group/volume_name.
When fsck is run on the file system. the raw device tile is checked:
/ dev/vx/ rdsk/ disk _group/ vol ume_ name.
Adding a File System After Volume Creation
1. CLI: Create the file system using mkf s (VxFS) or
OS-specific file system creation commands.
VEA: Select Actions->File System->New File System
2. ell: Create a mount point directory on which to mount
the file system.
VEA: Specify the mount point in the New File System
dialog box.
3. CLI: Mount the volume to the mount point by using the
mount command.
VEA: If a file system was previously created on a
volume, but not mounted, you can explicitly mount the
file system by selecting Actions->File System->
Mount File System.
Solaris I HP-UX I ~I~.J. LinuxJ
Adding a File System to a Volume: CLI
To add a file system to a volume from the command line. you must create the tile
system. create a mount point for the tile system, and then mount the tile system.
Solaris
To create and mount a YxFS tile system:
mkfs -F vxfs /dev/vx/rdsk/datadg/datavol
mkdir /data
mount -F vxfs /dev/vx/dsk/datadg/datavol /data
To create and mount a UFS file system:
newfs /dev/vx/rdsk/datadg/datavol
mkdir Idata
3-18
COpyflglll~ 2006 Symantec Corporation All rights ro servoo

mount /dev/vx/dsk/datadg/datavol /data
HI'-lIX
To create and mount a YxFS file system:
mkfs -F vxfs /dev/vx/rdsk/datadg/datavol
mkdir /data
mount -F vxfs /dev!vx/dsk/datadg/datavol /data
To create and mount an IIFS file system:
newfs -F hfs /dev/vx/rdsk/datadg/datavol
mkdir /data
mount -F hfs /dev/vx/dsk/datadg/datavol /data
AIX
To create and mount a YxFS file system using mkfs:
mkfs -v vxfs /dev/vx/rdsk/datadg/datavol
mkdir /data
mount -v vxfs /dev/vx/dsk/datadg/datavol /data
To create and mount a YxFS file system using crfs:
crfs -v vxfs -d /dev/vx/rdsk/datadg/datavol -m /data -A
yes
Notes:
An uppercase V is used with mkfs: a lowercase v is used with crfs (to avoid
conflict with another crfs option).
crfs creates the file system. creates the mount point. and updates the file
systems file (/etc/filesystems). The -A yes option requests mount at
boot.
If the file system already exists in /etc/filesystems. you can mount the
file system by simply using the syntax: mount mount_point.
Linux
To create and mount a YxFS file system using mkfs:
mkfs -t vxfs /dev/vx/rdsk/datadg/datavol
mkdir /data
mount -t vxfs /dev/vx/dsk/datadg/datavol /data
Copyright rr 2006 Sjrnantec Co-potation All riglns reserved
I
3-19

Mounting a File System at Boot
To mount the file system automatically at boot time, edit
the OS-specific file system table file to add an entry for
the file system.
Specify information, such as:
Device to mount:
Device to f sck:
• Mount point:
• File system type:
• fsck pass:
• Mount at boot:
Mount options:
/dev/vx/dsk/datadg/datavol
/dev/vx/rdsk/datadg/datavol
/data
vxfs
Mount Options
Mount pOlnt,"C":Jd-cala---f
~ Create mount ~oinl
syrnarucc.
Mounting a File System at Boot
Using CLI. if you want the tile system to be mounted at every system boot, you
must edit the tile system table file by adding an entry for the file system. If you
later decide to remove the volume. you must remove the entry in the tile system
table tile.
1
yes
C Read only
'" Honor setujd
Platform File System Tahle File
Sularis /etc/vfstab
HI'-UX /etc/fstab
AIX letc/filesystems
l.inux /etc/fstab
AIX
In AIX, you can use the following commands when working with the tile system
table tile, /etc/filesystems:
To Jew entnes lsfs mount po i nt:
To change details of an entry, use chf s. For example. to turn off mount at
boot chfs -A no mount po i nt:
111 YEA. in the Mount File System dialog. mark the "Add to tile system table" and
"Mount at boot" (not on HP-UX) check boxes. the entry is made in the file system
table tile automatically. Ifthe volume is later removed through YEA. its
corresponding tile system table tile entry is also removed automatically.
3-20
COPYJlyht& 2006 Svmantec Corporation. All nqlns reserved

S)111'l!1t('(
I
Displaying Volume Information: CLI
To display volume configuration information:
vxprint -g diskgroup [options]
-vpsd Select only volumes (v), plexes (p), subdisks (s),
or disks (d).
List hierarchies below selected records.
Display related records of a volume containing
subvolumes.
Print single-line output records that depend upon
the configuration record type.
Display all information from each selected record.
-h
-r
-t
-1
- a Display all information about each selected
record, one record per line.
-A Select from all active disk groups.
-e pattern Show records that match an editor pattern.
Displaying Volume Configuration Information
Displaying Volume Layout Information: ell
The vxprint Command
You can use the vxpr int command to display information about how a volume is
configured. This command displays records from the VxVM configuration
database.
vxprint -g diskgroup [options]
The vxpr int command can display information about disk groups. disk media.
volumes. plcxes, and subdisks. You can specify a variety of options with the
command to expand or restrict the information displayed. Only some of the
options are presented in this training. For more information about additional
options. see the vxprint (Lm) manual page.
Copvnqbt ;f' 2006 svroeruec co.oorauco All rights reserved

auto
auto
auto
auto
2048
2048
2048
2048
4191264
4191264
4191264
4191264
Displaying Volume Information: ell...- ...-..-...----.--~--.--.-- ..-....-- ..-..-~-.------.
vxprint -9 datad9 -ht I more I
OG NAME NCONFlG NLOG MINORS GROUP-ID
ST NAME STATE DM CNT SPARE CNT APPVOL_ CNT
OM NAME DEVICE TYPE PRIVLEN PUBLEN STATE
RV NAME RLINK_CNT I<STATE STATE PRIMARY DATAVOLS SRL
RL NAME RVG KSTATE STATE REM..HOSTREM_DG REM_RLNI<
CO NAME CACHEVOL KSTATE STATE
SD NAME
SV NAME
SC NAME
DC NAME
SP NAME
PLEX
PLEX
PLEX
P.ARENTVOL
SNAPVOL
DISI< DISKOFFSLENGTH [COLI] OFF DEVICE MODE
VOLNAME NVOLLAYRLENGTH [COL/JOFF AM/NM MODE
CACHE DISKOFFS LENGTH [COLI J OFF DEVICE HODE
LOGVOL
OCO
dg datadg default default 91000 1000753077.1117. train12
dm datadg01
dID datadg02
dm datadg03
dm datadg04
c1tlOdOs2
c1 tlldOs2
clt14dOs2
c1tl5dOs2
-I To interpret the
-I output, match header
-j lines with output lines.
ENABLED AC"TIVE-· "2 ii~ CONCAi~N~"~~. ·.·.~M~~ ·RW
1 datadgOl 0 21168 0 clt10dO ENA
Displaying Information for All Volumes
To display the volume. plcx, and subdisk record information for a disk group:
vxprint -g diskgroup -ht
In the output, the top few lines indicate the headers that match each type of output
line that follows. Each volume is listed along with its associated plexes and
subdisks and other VxVM objects.
dg is a disk group.
st is a storage pool (used in Intelligent Storage Provisioning).
dm is a disk.
rv is a replicated volume group (used in VERITAS Volume Rcplicator).
rl is an rlink (used in VERITAS Volume Rcplicator).
eo is a cache object.
vt is a volume template (used in Intelligent Storage Provisioning).
v is a volume.
p L is a plcx.
sd is a subdisk.
sv is a subvolumc.
se is a storage cache.
de is a data change object.
sp is a snap object.
For more information, s.:ethe vxp ri nt; ( i m) manual page.
Copynyhl :.,;,:!(Jr)f:; Svmantec Corpor auon All nghts reserved

~ S)111iHlI(,(
Object Views in the Main Window
!!!ILogs
SIZe CDS
Id~t.;dan~
rl ~t~rlgOl
c!~lcH!CI If(rrnr1p(1 1('1:'8':·8 I'€,=
(I.;jt~dll III ported I q:~ (,8 '''t?s
Highlight a volume, and click
the tabs to display details.
Displaying Volume Information: VEA
To display information about volumes in VEA. you can select from several
different views.
Object Views in the Main Window
You can view volumes and volume details by selecting an object in the object tree
and displaying volume properties in the grid:
To view the volumes in a disk group. select a disk group in the object tree and
click the Volumes tab in the grid.
To explore detailed components of a volume. select a volume in the object tree
and click each of the tabs in the grid.
Copyright t: 2006 Symantec corcoreuoo All nqh(s reserved

symantcc.
Viewing Basic Disk Information: ell
To display basic information about all disks:
DEVICE TYPE DISK GROUP STATUS
cltlOdOs2 auto:cdsdisk datadgOl datadg online
cl tlldOs2 auto:cdsdisk datadg02 datadg online
cl tl2dOs2 auto:cdsdisk online
cltl3dOs2 auto:none online invalid
cl t14dOs2 auto:none online invalid
cl tlSdOs2 auto:none online invalid
clt16dOs2 auto:none online invalid
clt17dOs2 auto:none online invalid
VxVM
Disks
..J
Free
Disk
Note: In a shared access environment, when displaying disks, run
vxdctl enable frequently to rescanfor disk changes.
Uninitialized
Disks
Displaying Disk and Disk Group Information
Displaying Basic Disk Information: ell
You use the vxdisk list command to display basic information about all disks
attached to the system. The vxdisk list command displays the:
Device names for all recognized disks
Type of disk, that is. how a disk is placed under VxVM control
Disk names
Disk group names associated with each disk
Status of each disk
In the output:
A status of onl ine. in addition to entries in the Disk and Group columns
indicates that the disk has been initialized or encapsulated, assigned a disk
media name, and added to a disk group. The disk is under Volume Manager
control and is available for creating volumes.
A status of onl ine without entries in the Disk and Group columns indicates
that the drive has been initialized or encapsulated but is not currently assigned
to a disk group.
A status of onl ine inval id indicates that the disk has neither been
initialized nor encapsulated by VxVM. The disk is not under VxVM control.
Note: On the HP-UX platform, LVM disks have a type of auto: LVM and a status
ofLVM.
Copynght'~ 200ti Svrnaruec Corporatron All rights rescrveo

J~-<
, S)111i1l1t('(
I
To display detailed information for a disk:
vxdisk -g diskgroup list disk_name
vxdisk -g datadg list datadgOl
Device: cltlOdOs2
devicetag: cltlOdO
type: auto
hostid:
disk:
group:
train12
name=datadgOl id=l000753057.1114.train12
name=datadg id=l000753077.1117.train12
To display a summary for all disks:
vxdisk -s list
To display detailed information about a disk. you use the vxdisk list
command with the name of the disk group and disk:
vxdisk -g diskgroup list disk name
In the output:
Device is the YxYM name for the device access path.
device tag is the name used by YxYM to refer to the physical disk.
type is how a disk was placed under YM control. auto is the default type.
hos t id is the name of the system that currently manages the disk group to
which the disk belongs: ifblank. no host is currently controlling this group.
disk is the YM disk media name and internal ID.
group is the disk group name and internal !D.
To view a summary of information tor all disks. you use the - s option with the
vxdisk list command.
Note: The disk name and the disk group name are changeable. The disk ID and
disk group !D are never changed as long as the disk group exists or the disk is
initialized.
Note: The detailed information displayed by this command will be discussed later
in the course
Lesson 3 Creating a Volume and Fite System 3-25
Copyright C'" 2006 Symantor; Corporation. All nqhts reserved

svmantcc
cap_'t;
d,jt,JlJ!)OI iht"d<J
Keeping Track of Your Disks
By viewing disk information. you can determine if a disk has been initialized and
added to a disk group, verify the changes that you make to disks, and keep track of
the status and configuration of your disks.
Displaying Disk Information: VEA
The status of a disk can be:
Not Initialized: The disk is not under YxYM control. The disk may be in use
as a raw device by an application.
Free: The disk is initialized by YxYM but is not in a disk group. You cannot
place a disk in this state using YEA, but YEA recognizes disks that have been
initialized through other interfaces.
Foreign: The disk is under the control uf another host.
Imported: The disk is in an imported disk group.
Deported: The disk is in a deported disk group.
Disconnected:The disk contains subdisks that are not available because of
hardware failure. This status applies to disk media records tor whieh the
hardware has been unavailable and has not been replaced within YxYM.
External: The disk is in use by a foreign manager.
Inactive/Import failed: The disk group is not imported but the disks have thc
same host ID tag as the hostnarne of the system, for example, if the disk group
is deported using the same hostname.
COPYright ~ 2006 Svmeotec Corporation. All nqhts reserved

,S)11mll1tx.
General
CDS: v~
Efl: No
5t~tus' Ilfl)OI'ted
Capacity: ';j"%"i
i Select a unit to
'I display capacity and
unallocated space in
I other units.Unalocatedsp.!Jce: ~95!
UDIO'
I
No
see-e:
Reserved:
Hot use:
AlocatOl'Dlsk:
Yes
cowrent:
Viewing Disk Properties: VEA
In YEA, you can also view disk properties in the Disk Properties window. To open
the Disk Properties window, right-click a disk and select Properties.
The Disk Properties window includes the capacity of the disk and the amount of
unallocatcd space. You can select the units for convenient display in the unit of
your choice.
3-27Lesson 3 Creating a Volume and File System
Copvnqht 'r;; 2006 Svmantec Corporation All rights reserved

symantec
To display disk groups:
vxdg list
NAME STATE ID
datadg enabled,cds 9695836l3.l025.cassius
newdg enabled, cds 971216408.l133.cassius
To display free space in a disk group, use one of these:
vxassist -g diskgroup help space
vxdg -g diskgroup free
Displaying Disk Group Information: ell
To display disk group information:
Use vxdg list to display disk group names, states, and IDs for all imported
disk groups in the system.
Use vxdg free to display tree space on each disk. This command displays
free space on all disks in all disk groups that the host can detect. Add -g
diskgroup to restrict the output to a specific disk group.
Note: This command does not show space on spare disks. Reserved disks are
displayed with an "r " in the FLAGS column.
Use vxdisk -0 alldgs 1ist to display all disk groups, including
deported disk groups. For example:
Disk - 1 auto:cdsdisk datadgOl datadg online
Disk 7 auto:cdsdisk (acctdg) online
3-28
Copynoht :. 2006 Svmaulec Corporation All rights ruserveo

I
Viewing Disk Group Information: VEA
ditadg
Right-click a disk group
and select Properties.
The object tree in the VEA main window contains a Disk Groups node that
displays all of the disk groups attached to a host. When you click a disk group. the
VxVM objects contained in the disk group are displayed in the grid.
To view additional infonnation about a disk group. right-click a disk group and
select Properties.
The Disk Group Properties window is displayed. This window contains basic disk
group properties. including:
Disk group name. status. 10. and type
Number of disks and volumes
Disk group version
Disk group size and tree space
Note: On IIP-UX. there is another attribute between the Version and Enabled
attributes. which is Shared: No.
Stalus Imported
Id 11~)094318 25.~oulsecl>lV9
CDS "es
Disks:
Volumes
ColTtlnt'l9!iion Yes
Version 140
EtlBblM 'tag
JlLl/V.ftlo"ornooe-
01"" uetacn pOll Olobal
Ol$kgroupfail POlicy Dgdlsabl~
.AJlo!pd snes
911a ConSIstency O.
sne: 3 ~-':f, ,-.p
rree eper e 1 :::~:I:, ,.,[3
Viewing Disk Group Properties: VEA
Copynqht '''' 2006 Syrnantec Corporanon All rights reserved
3-29

syrnantec
Removing a Volume
• When a volume is removed, the space used by the
volume is freed and can be used elsewhere.
• Unmount the file system before removing the
volume.
VEA:
• Select the volume that you want to remove.
• Select Actions->Delete Volume.
vxassist remove volume:
vxassist -g diskgroup remove volume volume name
vxassist -g datadg remove volume datavol
vxedit:
vxedit -g diskgrollp -rf rm volume name
vxedit -g datadg -rf rm datavol
Removing Volumes, Disks, and Disk Groups
Removing a Volume
Only remove a volume if you are sure that you do not need the data in the volume,
or if the data is backed up elsewhere. A volume must be closed before it can be
removed. For example. if the volume contains a tile system. the tile system must
be unmounted. You must edit the OS-specitic tile system table tile manually in
order to remove the entry for the tile system and avoid errors at boot. If the volume
is used as a raw device. the application. such as a database. must close the device.
Cupyllghl'f: 20(}6 S,.rn<l!lIU. Corporauon All fights reserved

, symaruec
I
Before removing a disk, you may need to evacuate data
from the disk to another disk in the disk group.
VEA:
• Select the disk that you want to evacuate.
• Select Actions->Evacuate Disk.
vxdiskadm:
"Move volumes from a disk"
ell:
vxevac -g d i ek qr oiip from disk [to_disk]
vxevac -g datadg datadgOl datadg02
To evacuate to any disk except for datadg03:
vxevac -g datadg datadg02 !datadg03
Evacuating a disk moves the contents of the volumes on a disk to another disk. The
contents of a disk can be evacuated only to disks in the same disk group that hale
sufficient free space.
Evacuating a Disk
Copyrighl '[~ 2006 Syrnantec Corporation All rights reserved
3-31

symnntrc
roup
VEA:
• Select the disk that you want to remove.
• Select Actions->Remove Disk from Disk Group.
vxdiskadm:
"Remove a disk"
ell:
vxdg -g diskgroup rmdisk disk_name Remove the disk
from the diskvxdiskunsetup [-CJ device_tag
Example:
group, and then
uninitiatize it.
vxdg -g datadg rmdisk datadg02
vxdiskunsetup Disk 2
Removing a Disk
I f you select all disks for removal from the disk group, the disk group is destroyed
automatically.
You can verify the removal by using the vxdisk list command to display disk
information. A deconfigured disk has a status ofonline invalid and no
longer has a disk media name or disk group assignment.
The vxdiskunsetup Command
After the disk has been removed from its disk group. you can remove it from
Volume Manager control completely by using the vxdiskunsetup command.
This command reverses the configuration of a disk by removing the public and
private regions that were created by the vxdisksetup command. The
vxdiskunsetup command docs not operate on disks that are active members of
an imported disk group.
This command does not usually operate on disks that appear to be imported by
some other hostv for example, a host that shares access to the disk. You can use
the - C option to force deconfiguration otthc disk, removing host locks that may be
detected.
3-32
COPYright C:'2006 Symaotec Ccrporanoo All righL<; reserved

olddg
DO
DO
___L _
Destroying a disk group:
• Means that the disk group no longer
exists
• Removes all disks
• Is the only method for freeing the last
disk in a disk group
VEA: Actions->Destroy Disk Group
ell: vxdg destroy diskgroup
Example: To destroy the disk group
olddg and place its disks in the free disk
pool:
vxdg destroy olddg
IDestroying a Disk Group
Destroying a disk group permanently removes a disk group from Volume Manager
control, and the disk group ceasesto exist. When you destroy a disk group, all of
the disks in the disk group are reinitialized as empty disks. Volumes and
configuration information about the disk group are removed.
Because you cannot remove the last disk in a disk group, destroying a disk group is
the only method to free the last disk in a disk group for reuse. A disk group cannot
be destroyed ilany volumes in that disk group arc in use or contain mounted file
systems. The bootdg disk group cannot be destroyed.
Caution: Destroying a disk group can result in data loss. Only destroy a disk
group if you are sure that the volumes and data in the disk group are not needed.
Destroying a Disk Group: VEA
Actions-c=Desrroy Disk GroupNavigation path:
The disk group to be destroyedSelect:
Input: Group name: Specify the disk group to be destroyed.
Destroying a Disk Group: ell
To destroy a disk group 11-0111the command line, use the vxdg destroy
command.
Note: You can hring back a destroyed disk group by importing it with its dgid.
Copyriglll =: 2006 Symantec Corporauoo All rights reserved
3-33

Appendix A provides complete lab instructions. "Lab :;. (I,';IIIII>! a V,)IIIlile' .uul
File ~y'k'I11."page i-I~
Appendix B provides complete lab instructions and solutions. "/';iIJ 3 Sohuion-:
Cn.:atillg 'I Volume and lrlc S'kPl." I'd!!': 1:1·21
• Key Points
In this lesson, you learned how to create a volume with a
file system. This lesson also described device-naming
schemes and how to add a disk to a disk group, in
addition to how to view configuration information for
volumes, disk groups, and disks, In addition, you learned
how to remove a volume, disk, and disk group.
- VERITAS Storage Foundation Installation Guide
3-34
svmantec.
Lab 3
Lab 3: Creating a Volume and File System
In this lab, you create new disk groups, simple
volumes, and file systems, mount and
unmount the file systems, and observe the
volume and disk properties.
The first exercise uses the VEA interface. The
second exercise uses the command-line
interface.
For Lab Exercises, see Appendix A,
For Lab Solutions, see Appendix B.
Copynqnt', 2006 Svrnanu,c Corporatron. AUuqtus reserved

Lesson 4
Selecting Volume Layouts

Lesson Introduction
" Lesson 1: Virtual Objects
" Lesson 2: Installation and Interfaces
" Lesson 3: Creating a Volume and File
System
" ..~(!!>!>()Il~~:~E!'.(!<:tiny'~~o~~~E!_~~r~~!!>_.
" Lesson 5: Making Basic Configuration
Changes
" Lesson 6: Administering File Systems
" Lesson 7: Resolving Hardware
Problems
Will antt'(
Topic After completing this lesson, you
will be able to:
Topic 1: Comparing Identify the features, advantages, and
Volume Layouts disadvantages of volume layouts
supported by VxVM.
Topic 2: Creating Volumes Create concatenated, striped, and
with Various Layouts mirrored volumes by using VEA and
from the command line.
Topic 3: Creating a Create layered volumes by using VEA
Layered Volume and from the command line.
Topic 4: Allocating Allocate storage for a volume by
Storage for Volumes specifying storage attributes and
ordered allocation.
4-2
C{JP~1I9hl :- 2006 Symanter Cbrporauon All fights reservou

~V1if.~- ·---'S)~il;~tt'(
Concatenated Layout
Disk Group
Volume
Plex
Subdisks
VxVM Disks
Subdisks
datadg
datavol
IComparing Volume Layouts
Each volume layout has different advantages and disadvantages. For example. a
volume can be extended across multiple disks to increase capacity. mirrored on
another disk to provide data redundancy. or striped across multiple disks to
improve [/0 performance. The layouts that you choose depend on the levels of
performance and reliability required by your system.
Concatenated Layout
A concatenated volume layout maps data in a linear manner onto one or more
subdisks in a plcx. Subdisks do not have to be physically contiguous and can
belong to more than one VM disk. Storage is allocated completely from one
subdisk before using the next subdisk in the span. Data is accessed in the
remaining subdisks sequentially until the end of the last subdisk.
For example. if you have 14 GB of data then a concatenated volume can logically
map the volume address space across subdisks on di fferent disks. The addresses
o GB to 8 (iB of volume address space map to the first 8-gigabytc subdisk. and
addresses 9 GB to 14 (iB map to the second 6-gigabyte subdisk. An address offset
of 12 GB. therefore. maps to an address onset of4 GB in the second subdisk.
Copyright <£ 200[, Synrantec Corporation All riqhts rcserveo
4-3Lesson 4 Selecting Volume Layouts

Disk Group
Volume
datadg
Striped Layout
Plex .,c
E
"o
(.) SU10 SU11 SU12
datavol-Ol
SU1 SU2 SU3
datavol
Subdisks
SU4 SU5 SU6
SU7 SUB SU9
VxVM Disks
Subdisks
Striped Layout
A striped volume layout maps data so that the data is interleaved, or allocated in
stripes, among two or more subdisks on two or more physical disks. Data is
allocated alternately and evenly to the subdisks of a striped plex.
The subdisks are grouped into "columns." Each column contains one or more
subdisks and can be derived from one or more physical disks. To obtain the
maximum performance benefits of striping, you should not use a single disk to
provide space for more than one column.
All columns must be the same size. The minimum size of a column should equal
the size of the volume divided by the number of columns. The default number of
columns in a striped volume is based on the number of disks in the disk group.
Data is allocated in equal-sized units, called stripe 1111 its. that are interleaved
between the columns. Each stripe unit is a set of contiguous blocks on a disk. The
stripe unit size can be in units of sectors, kilobytes. megabytes, or gigabytes. The
default stripe unit size is 64K, which provides adequate performance for most
general purpose volumes. Performance of an individual volume may be improved
by matching the stripe unit size to the 110 characteristics of the application using
the volume.
Cop)'flyhl .~ ?()(J6Svmantec Corporauou All fights reserved

~-~"'~. ~~ ..' ";'.r&~;11[~-------'»~11a;lIt~~
Mirrored Layout .
r+: --; Each piex must
Disk Group datadg have disk space
from different
Volume C datavol :=:: disks to achieve
redundancy.
Plex
Subdisks
VxVM Disks
Subdisks
datavol-Ol datavol-02
I I
datadg03-01
datadgOl-02 datadgO~-02
...!. ----..J!L
r-datadg02-- r-datadg03--
dat4c",;;; ;;]. datadg03-01
datadg02..02 d~tadg')? 02
J&.tadti'.!;' :;.~ ...J,"lt.il~i~~t)_~~ 0
datadgOl
datadgOl-02
Mirrored Layout
By adding a mirror to a concatenated or striped volume, you create a mirrored
layout. A mirrored volume layout consists of more than one plcx that duplicate the
information contained in a volume. Each plcx in a mirrored layout contains an
identical copy of the volume data. In the event of a physical disk failure and when
the plex on the failed disk becomes unavailable, the system can continue to operate
using the unaffected mirrors.
Although a volume can have a single plcx. at least 111'0 plexes are required 10
provide redundancy of data. Each of these plcxes must contain disk space trom
di tlcrcnt disks to achieve redundancy.
Volume Manager uses true mirrors. which means that all copies of the data are the
same at all rimes. When a write occurs to a volume. all plexcs must receive the
write before the write is considered complete.
Distribute mirrors across controllers to eliminate the controller as a single point of
failure.
Lesson 4 Selecting Volume Layouts
Copyright f 2006 Symantcc. Corporation All rights reserved
I
4-5

RAID-5 Layout
Disk Group
Volume
Plex
Subdisks
VxVM DisksI,---I~~~,..J~~!I.,
Subdisks
P = Parity; a
calculated value used 1---------------'
to reconstruct data
after disk failure.
RAIO-5 Layout
A RAID-5 volume layout has the same attributes as a striped plex, but it includes
one additional column of data that is used for parity. Parity provides redundancy.
Parity is a calculated value used to reconstruct data after a failure. While data is
being written to a RAID-5 volume. parity is calculated by performing an exclusive
OR (XOR) procedure on the data. The resulting parity is then written to the
volume. If a portion ofa RAIO-5 volume fails, the data that was on that portion of
the failed volume can be re-created from the remaining data and parity
information.
RAID-5 volumes keep a copy of the data and calculated parity in a plex that is
striped across multiple disks. Parity is spread equally across columns. Given a
five-column RAID-5 where each column is I GB in size, the RAID-5 volume size
is 4 GB, One column of space is devoted to parity, and the remaining four I-G8
columns arc used tor data.
The default stripe unit size for a RAID-5 volume is 16K. Each column must be the
same length but may be made from multiple subdisks of variable length. Subdisks
used in different columns must nut be located on the same physical disk.
RAIO-5 requires a minimum of three disks for data and parity. When implemented
as recommended. an additional disk is required for the log.
RAIO-5 cannot be mirrored.
4-6
COPYJlghl : 2()06 Syrl1antec Cnrporauon AU fights reserved

,S)111<1ntt'(
Comparing Volume Layouts
Concatenation Striping Mirroring RAID·5
• Removessize • Parallel data • Improved • Redundanc¥
restrictions transfer reliability and through panty
., availability • Requires lessCIl
• Better • Load-balanclng01
space than
~ utilization of • Improved read
c free space • Improved performance mirroring
10 performance
> • Improved read"0 • Simplified (if properly • Fast recovery performance0(
administration configured) through
logging • Fast recovery
through logging
., • No redundancy • No redundancy • Requires more • Slower writeCIl
01
• Single disk • Single disk
disk space performance
s than mirroring
e failure causes failure causes • Slightly
10
volume failure. volume failure. slower write • Poor>"0 performance performance10
., after a disk
0 failure
Comparing Volume Layouts
IConcarcnatlon: Advantages
Removes size restrictions: Concatenation removes the restriction on size or
storage devices imposed by physical disk size.
Better utlllzarlon of free space: Concatenation enables better utilization of
free space on disks by providing for the ordering of available discrete disk
space on multiple disks into a single addressable volume.
Simplified administration: System administration complexity is reduced
because making snapshots and mirrors uses any size space. and volumes can be
increased in size by any available amount.
Concatenation: Disadvantages
No protection against disk failure: Concatenation does not protect against disk
failure. A single disk failure results in the failure of the entire volume.
Striping: Advantages
Improved performance through parallel data transfer: Improved
performance is obtained by increasing the effective bandwidth of the I '0 path
to the data. This may be achieved by a single volume JiO operation spanning
across a number of disks or by multiple concurrent volume 1/0 operations to
more than one disk at the same time.
Load-balancing: Striping is also helpful in balancing the I/O load from
multiuser applications across multiple disks.
Copyriqht '. 20(16 Syrnantec Corporation All dgtlls reserved
4-7Lesson 4 Selecting Volume Layouts

Striping: Disadvantages
No redundancy: Striping alone offers no redundancy or recovery features.
Disk failure: Striping a volume increases the chance that a disk failure results
in failure of that volume, For example, if you have three volumes striped
across two disks, and one of the disks is used by two of the volumes, then if
that one disk goes down, both volumes go down.
Mirroring: Advantages
Improved reliability and availability: With concatenation or striping, failure
of anyone disk makes the entire plex unusable. With mirroring, data is
protected against the failure of anyone disk. Mirroring improves the reliability
and availability of a striped or concatenated volume.
Improved read performance: Reads benefit from having multiple places
from which to read the data.
Mirroring: Disadvantages
Requires more disk space: Mirroring requires twice as much disk space,
which can be costly for large configurations. Each mirrored plex requires
enough space for a complete copy of the volume's data.
Slightly slower write performance: Writing to volumes is slightly slower,
because multiple copies have to be written in parallel. The overall time the
write operation takes is determined by the time needed to write to the slowest
disk involved in the operation.
The slower write performance ofa mirrored volume is not generally significant
enough to decide against its use. The benefit of the resilience that mirrored
volumes provide outweighs the performance reduction.
RAIO-5: Advantages
Redundancy through parity: With a RAID-5 volume layout data can be re-
created from remaining data and parity in case ofthe failure of one disk.
Requires less space than mirroring: RAIO-5 stores parity information, rather
than a complete copy of the data.
Improved read performance: RAID-5 provides similar improvements in read
performance as in a normal striped layout
Fast recovery through logging: RAID-5 logging minimizes recovery time in
case of disk failure.
RAIO-5: Disadvantages
Slow write performance: The performance overhead for writes can be
substantial, because a write can involve much more than simply writing to a
data block. A write can involve reading the old data and parity, computing the
new parity, and writing the new data and parity. If you have more than twenty
percent writes, do not use RAID-5.
Very poor performance after a disk failure: After one column fails, all 1/0
performance goes down. This is not the case with mirroring, where a disk
failure does not have any significant effect on performance.
Couvnqnt ,~ 2006 Symantec Corpcratron All nqhts reserved

Selecting a Layout Type: VEA
Specify volume attributes.
Creating Volumes with Various Layouts
You can create volumes with a variety or layouts. In V[i in the Specify Volume
Attributes window. select:
Layout: Select a layout type from the group of options. The default layout is
concatenated.
Concatenated:The volume is created using one or more regions of
specified disks.
Striped: The volume is striped across two or more disks. The default
number of columns across which the volume is striped is two. and the
default stripe unit size is 12X sectors (64K) on Solaris. iIX. and Linux:
128 sectors (128K) on I II'-UX. You can specify different values.
ConcatenatedMirrored and Striped Mirrored: These options denote
layered volume layouts.
Mirror Info:
Mirrored: Mirroring is recommended. To mirror the volume. mark the
Mirrored check box. Only striped or concatenated volumes can be
mirrored. RiID-S vol limes cannot be mirrored.
Total mirrors: Type the total number of mirrors for the volume. i volume
can have up to 32 plexcs: however. the practical limit is 31. One plex is
reserved by VxVM to perform restructuring or relocation operations.
~olume name: ~~':;;";:~""~""~~~C~~"""'~'""
cornmgnt
Si~e'
Layout Minor Info
.~. Concatenatel!
~! Qtrlped
-:' fiAID-5
C; CQn(atenatE'd Mirrored
C..StrlQed Mirrored
r:-JEna~le [astResync
f] !nltiallze zero
COP/right '~~.2006 Svmantoc Corporation Ail right~ reserved
I
4-9

svrnantec
Concatenated Volume: CLI
To create a concatenated volume:
vxassist -g datadg make datavol 109
Disk group name
This command creates a concatenated volume
called datavol with a length of 10 gigabytes, in
the disk group datadg, using any available disks.
Creating a Concatenated Volume: CLI
By default. vxassist creates a concatenated volume that uses one or more
sections of disk space. The vxassist command attempts to locate sufficient
contiguous space on one disk for the volume. However. if necessary, the volume is
spanned across multiple disks. Vx VM selects the disks on which to create the
volume.
Note: To guarantee that a concatenated volume is created. include the
Layout e nost ri pe attribute in the vxassist makecommand. Without the
layout attribute. the default layout is used that may have been changed by the
creation of the letc/default/vxassist file.
For example:
vxassist -g datadg make datavol 109 layout=nostripe
If you want the volume to reside on specific disks. you can designate the disks by
adding the disk media names to the end of the command. More than one disk can
he specified.
vxassist [-g diskgroupJ make volume name length
[disks ... J
4-10
Copyright :,2006 Symantec Lorporatron All rights reserved

To create a striped volume. you add the layout type and other attributes to the
vxassist makecommand.
layout=stripe designates the striped layout.
ncol =n designates the number of stripes. or columns. across which the
volume is created. This attribute has many aliases. For example. you can also
use ns t r ipe e n or s t r ipe s e n.
The minimum number or stripes in a volume is 2 and the maximum is R. You
can edit these minimum and maximum values in /ete/default/
vxassist using the min _eol umns and max_eol umns attributes.
stripeunit=size specifics the size of the stripe unit to he used. The
default is MK.
To stripe the volume across specific disks. you can specify the disk media
names at the end of the command. The order in which disks are listed on the
command line does not imply any ordering of disks within the volume layout.
To exclude a disk or list of disks. add an exclamation point ( ! ) before the disk
media names. For example. !datadgOl specifies that the disk datadgOl
should not be used to create the volume.
, 5)1 1;1111 <;'c .
Striped Volume: ell
To create a striped volume:
vxassist -g diskgroup make volume_name length
layout=stripe [ncol=n] [stripeunit=size]
[disks ... J
Examples:
vxassist -g acctdg make payvol 2g
layout=stripe ncol=3 !acctdg04
vxassist -g acctdg make expvol 2g
layout=stripe ncol=3 stripeunit=256k
acctdgOl acctdg02 acctdg03
Creating a Striped Volume: CLI
Lesson 4 Setecting Volume Layouts
Cop yeiqht ;;; 2006 Swnantec Corporation All rights r"!S(,lIIed
I
4-11

syrnarucc.
To mirror a concatenated volume, you add the Layout erni r r o r attribute in the
vxassist command.
To specify more than two mirrors, you add the nmirror attribute.
When creating a mirrored volume. the volume initialization process requires
that the mirrors be synchronized. The vxassist command normally waits for
the mirrors to be synchronized before returning to the system prompt. To run
the process in the background. you add the -b option.
Mirrored Volume: ell
To create a mirrored volume:
vxassist -g diskgroup [-b] make volume name
length layout=mirror [nmirror=number]
,Ex~,!,plel!:
! Concatenated
i and mirrored
i Specify three
t mirrors.
vxassist -g datadg make datavol Sg
layout;mirror
vxassist -g datadg make datavol Sg
layout=stripe,mirror nmirror=3
Creating a Mirrored and Logged Volume: CLI
When you create a mirrored volume, you can add a dirty region log by adding the
Loq t ype e d r I attribute:
vxassist -g diskgroup [-b] make volume_name length
layout=mirror logtype=drl [nlog=n]
Specify Loqr.ypee d r I to enable dirty region logging. A log plex that consists
of a single subdisk is created.
If you plan to mirror the log. you can add more than one log plex by specifying
a number of logs using the nLoq e n attribute. where n is the number of logs.
To create a concatenated volume that is mirrored and logged:
vxassist -g datadg make datavol Sm layout=mirror
logtype=drl
Note: Dirty regions logs are covered in a later lesson.
vxassist -g datadg -b make datavol
Sg layout=stripe,mirror nmirror=3
4-12
; Run process in
lbackground.
Creating a Mirrored Volume: CLI
Copynght ,; ~006 Symantec Corporauoo All fights reserved

I
~synlilnti'(.
Estimating Volume Size: ell
To determine the largest possible size for a volume:
vxassist -g diskgroup maxsize attributes
Example:
vxassist -g datadg maxsize layout=raid5
Maximum volume size: 376832 (184Mb)
To determine how much a volume can expand:
vxassist -g diskgroup maxgrow volume name
Example:
vxassist -g datadg maxgrow datavol
Volume datavo1 can be extended by 366592 to
1677312 (819Mb)
Estimating Volume Size: ell
The vxassist command can determine the largest possible size for a volume that
can currently be created with a given set of attributes. vxass ist can also
determine how much an existing volume can be extended under the current
conditions.
This maxs i ze command does not create the volume but returns an estimate ofthe
maximum volume size. The output value is displayed in sectors, by default.
II'the volume with the speci lied attributes cannot be created, an error message is
returned:
VxVM vxassist ERROR V-S-1-7S2 No volume can be created
within the given constraints
The maxgrow command docs not rcsizc the volume but returns an estimate or
how much an existing volume can be expanded. The output indicates the amount
by which the volume can be increased and the total size to which the volume can
grow. The output is displayed in sectors, by default.
Copynqht 'h- 2006 Svmanter. Corporauon All !lgnts reservec
4-13

symantcc
Observing Volume Layouts in VEA
i Highlighta ~~I~;;;~-~~d-sel~~t-Acti~~s--->-L-ay-o-u-t-V-i-e-w-·-I
[jle liew
/olurn€" (j~tav(lI02
T~...pe ~:ttlpe(i
Size 2097152
#(:01 c·
State Healthy
#1!1irrors 1
Stripe Sz.1 28
ptex ciatavol(l2·01
Type ~::tripeli
I r-r- Siale: P.,ltached
r' PrefenecP No
~;tripe Sz 128
#Col: 2
Sul)(Jlsk (latadgO"1·02
Size "1048576
Colurnnl
Offset 20~1T152
Plex Offset (I
Usage. Strlpeci
Subdlsk datadg(l2-01
t:izel(l48576
Column 0
Offset 0
F'lex Offset (I
Usage Sltiped
I SelectVieW->Hori~~;:;t~I~~----I'
View->Verticalto changethe
orientationof thediagram.
Volume Layout Window
The Volume Layout window displays a graphical view of the selected volume's
layout, components, and properties. You can select objects or perform tasks on
objects in the Volume Layout window. This window is dynamic. so the objects
displayed in this window arc updated automatically when the volume's properties
change.
To display the Volume Layout window, highlight a volume and select
Actions--->Layout View.
The View menu changes the way objects are displayed in this window. Select
Vicw->Horizontalto display a horizontal layout and View->Vertical to display
a vcrticul layout.
4-14
Copyright ~ 2006 Svmantec Corporation AlIllghls reserveo

, S)111ill1ltX .••••••
Volume to Disk Mapping Window
Dilk Group: daudg
Volumes
iii
Click a triangle to
hide subdisks.
* j~..!.~_. .__ ._._ ...__ ._'~.,) Dilk Group: datadg
o datodgOl-0l (0: 20111152) ~ Volumes
datodgOl.02 (2091152: 1048516)
0dg02.
datodg02·01(0: 1048516)
datodg02.01(0: 1048576)
~ ;' idatadg01
6 dol_'-01 (0: 2(97152)
dol 1·02(2097152: 1048576)
Click a dot to
highlight an
intersecting row
and column.
• ',I
IVolume to Disk Mapping Window
The Volume to Disk Mapping window displays a tabular view of volumes and their
relationships to underlying disks. To display the Volume to Disk Mapping window.
highlight a disk group and select Actions vc-Disk/Volumc Map.
To view subdisk layouts. click the triangle button to the left of the disk name. or
select View-- >Expand All.
To help identify thc row and column headings in a large grid, click a dot in the grid
to highlight the intersecting row and column.
Lesson 4 Selecting Volume Layouts 4-15
Copyrigh1'~ 200b Sjmaruec Corporation, All nghlS reserved

symarucc
Volume View Window
: Highlight a volume and select Actions->Volume View. "'-1"
1:.id:EX~~J'-"" .. ..............
AddatavoI01!"':'!TYPe: Concat Size. 1.000 GB Mirrors 1 Logged: No
IJJ I
.:I3t3dg01.(l'1
II
1.(100 G8
LJdatavoI02!' IIType. Striped Size. 1.000 OB Mirrors 1 Logged. No
jJ datadg02·01 dal.ldg01-02
512 000 MB 512.000 M8
columnO cctvmrr 1
Volume View Window
The Volume View window displays characteristics of the volumes on the disks. To
display the Volume View window. select a volume or disk group and select
Aetions->Volumc View.
Display options in the Volume View window include:
Expand: Click the Expand button to display detailed information about
volumes.
New volume: Click the New Volume button to invoke the New Volume
wizard.
4-16
Copyright 'h- 21106 Symantec Coenor.auon All fights reserved

,S)111iHllt'(.
The Disk View window displays a close-up graphical view of the layout of
subdisks in a volume. To display the Disk View window. select a volume or disk
group and select Actions->Disk View.
Display options in the Disk View window include:
Expand: Click the Expand button to display detailed information about all
disks in the Disk View window.
Vol Details: Click the Vol Details button to include volume names, layout
types. and volume status for each subdisk.
Projection: Click the Projection button to highlight objects associated with a
selected subdisk or volume. Projection shows the relationships between objects
by highlighting objects that are related to or part ofa specific object.
Caution: You can move subdisks in the Disk View window by dragging subdisk
icons to different disks or to gaps within the same disk. Moving subdisks
reorganizes volume disk space and must be performed with care.
Disk View Window
: Highlight a volume and select Actions->Disk View.
Volume Overview ISubdisk Details ID(I~ Expand I~ Vol Details Projection ~
@datadgldatadg01:t1t2d6s2 Size: 4.961 G8 Free: 3.461 G8 (70%)
datavcln t (Concatenated) I ;,1 av,,102(SI';p,d)
I3061GB I1000 <38 of 1 000 G8 5-12.000 MB of 1.000 GB
Health, He attbv
Q datadgldatadg02.c1 t2d7s2 Size' 4 961 GB Free. 4.461 GB (90%)
datavol02 (Stnpe d)
14061 GB I
512000 M8 of 1.000 08
Health.,
Disk View Window
Copyright r. 2006 Syrnantec Corporation Att rights reserved
I
4-17

Top-Level Volume
How Do Layered Volumes Work?
Subvolumes
Underlying
Disks
• Volumes are
constructed from
subvolumes .
• The top-level volume
is accessible to
applications.
Advantages
• Improved
redundancy
• Faster recovery
times
Disadvantages
Requires more
VxVM objects
Volume Volume
ISUbdiS1
Plex
ISUbdIS, FUbdiSkl
Plex Plex
Creating a Layered Volume
What Is a Layered Volume?
Vx VM provides two ways to mirror your data:
Original 'xVM mirroring: With the original method of mirroring, data is
mirrored at the plex level. The loss of a disk results in the loss of a complete
plcx. A second disk failure could result in the loss ofa complete volume if the
olume has only two mirrors. To recover the volume, the complete volume
contents must be copied from backup.
Enhanced mirroring: VxVM 3.0 introduced support for an enhanced type of
mirrored volume called a layered volume, A layered volume is a virtual
Volume Manager object that mirrors data at a more granular level. To do this,
VxVM creates subvolumes from traditional bottom-layer objects, or subdisks.
These subvolumes function much like volumes and have their own associated
plexcs and subdisks.
With this method of mirroring, data is mirrored at the column or subdisk level.
Loss of a disk results in the loss of a copy of a column or subdisk within a plex.
Further disk losses may occur without affecting the complete volume. Only the
data contents of the column or subdisk affected by the loss of the disk need to
be recovered. This recovery can be performed from an up-to-date mirror of the
failed disk.
Note: Only VxVM versions 3.0 and later support layered volumes. To create a
layered volume. you I11UStupgrade the disk group that owns the layered
volume to version 60 Dr later.
4-18
COPYright r:; ;m06 Symantec Corporation. All rlgtllS reserved

How Do Layered Volumes Work?
In a regular mirrored volume. top-level plcxcs consist of subdisks. In a layered
volume. these subdisks are replaced by subvolumcs. Each subvolumc is associated
with a second-level volume. This second-level volume contains second-level
plexes, and each second-level plcx contains one or more subdisks.
In a layered volume. only the top-level volume is accessible as a device for use by
applications.
Note: You can also build a layered volume from the bottom up by using the
vxmake command. For more information. see the vxmake (1m) manual page.
Layered Volumes: Advantages
Improved redundancy: Layered volumes tolerate disk failure better than
nonlayered volumes and provide improved data redundancy.
Faster recovery times: If a disk in a layered volume filils. a smaller portion of the
redundancy is lost. and recovery and resynchronization times are usually quicker
than for a nonlayered volume that spans multiple drives.
For a stripe-mirror volume. recovery of a single subdisk failure requires
resynchronization of only the lower plex, not the top-level plcx. For a mirror-stripe
volume. recovery of a single subdisk failure requires resynchronization of the
entire plex (full volume contents) that contains the subdisk.
Layered Volumes: Disadvantages
Requires more VxVM ohjccts: Layered volumes consist of more VxVM objects
than nonlayered volumes. Therefore. layered volumes Illay Ii II up the disk group
configuration database sooner than nonlaycrcd volumes. When the configuration
database is full. you cannot create more volumes in the disk group.
With SF 5.0. the default size of the private region is 32 MB. Each VxVM object
requires about 25() bytes.
Note: On the Solaris platform. in prc=l.v tonnat, the private region size is rounded
up 10 the cylinder boundary. With modern disks with large cylinder sizes. this size
can be quire large.
The private region can be made larger when a disk is initialized. The size cannot be
changed once disks have been initialized.
Copyrigtll 2006 Symantec Corporation. All rights reservoc
I

Traditional Mirroring
Plex
i sd = subdisk
diskOl disk02 disk03 disk04 Volume Status
Down
u
Down
Comparing Regular Mirroring with Enhanced Mirroring
To understand the purpose and benefits of layered volume layouts, compare
regular mirroring with the enhanced mirroring of layered volumes in a disk failure
scenario.
Regular Mirroring
The example illustrates a regular mirrored volume layout called a mirror-stripe
layout. Data is striped across two disks, diskOl and disk03, to create one plex,
and that plex is mirrored and striped across two other disks, disk02 and disk04.
If two drives fail, the volume survives 2 out 01'6 (1/3) times. As more subdisks are
added to each plcx, the odds of a traditional volume surviving a two-disk failure
approach (but never equal) 50 percent.
II'a disk fails in a mirror-stripe layout, the entire plex is detached, and redundancy
is lost on the entire volume. When the disk is replaced, the entire plcx must be
brought up-to-dare. or resynchronized.
C0IJynyhi ic 2006 Svmantec Corporauon. All nqhts reserved
VERITAS Storage Foundation 5.0 for UNIX.' Fundamentals4-20

Layered Volumes
The example illustrates a layered volume layout called a stripe-mirror layout. In
this layout, VxVM creates underlying volumes that mirror each subdisk. These
underlying volumes are used as subvolumes to create a top-level volume that
contains a striped plex of the data.
If two drives fail. the volume survives 4 out of6 (2/3) times. In other words, the
use of layered volumes reduces the risk of failure rate by 50 percent without the
need for additional hardware. As more subvolumes are added. the odds of a
volume surviving a two-disk failure approach 100 percent. For volume failure to
occur, both subdisks that compose a subvolume must tail. If a disk tails. only ihe
failing subdisk must be detached, and only thai pori ion of the volume loses
redundancy. When the disk is replaced, only a portion ofthe volume needs 10 be
recovered, which takes less time.
~~~.~-..i-' " ,~f---- --'~)~l;;;;l-(~'~'
Layered Volumes
disk01 disk02 disk03 disk04
IS = su IS
Volume Status
X X Down
X X Up
X X Up
X X Up
• When two disks
X X Up : fail, the volume
X X Down _ survives 4/6, or
X = failed disk I i 2/3 times.
Failed Volume Status
Subdisks
Stripe-Mirror (Layered) Mirror-Stripe (Nontavered)
I and 2 Down J)own
I and J Up Up
I and 4 Up Down
2 and J Up Down
2 and 4 Up Up
3 and 4 Down Down
Copyright ~: 2006 Svrnantec Corporation All rigtlls reserved
I
4-21

Terminology for Mirrored Layouts
The four types of mirroring in VxVM:
• mirror-concat (Non-layered, RAID-O+1)
- The top-level volume contains more than one plex
(mirror).
- Plexes are concatenated.
• mirror-stripe (Non-layered, RAID-O+1)
- The top-level volume contains more than one plex
(mirror).
- Plexesare striped.
• concat -mirror (Layered, RAID-1+0)
- The top-level volume is a concatenated plex.
- Subvolumes are mirrored.
• stripe-mirror (Layered, RAID-1+0)
- The top-level volume is a striped plex.
- Subvolumes are mirrored.
symantcc
Layered Volume Layouts
In general, use regular mirrored layouts for smaller volumes and layered layouts
for larger volumes. By default in VxVM, a volume larger than I GB is created as a
layered volume, unless you specify otherwise. Before you create layered volumes,
you need to understand the terminology that defines the different types of mirrored
layouts in VxVM.
mirror-concat: This layout mirrors data across concatenated plexes. The
concatenated plexes can consist of subdisks of different sizes. When you create
a simple mirrored volume that is less than I GB in size. a nonlayercd mirrored
volume is created by default.
mirror- stripe: This layout mirrors data across striped plexes. The striped
plexes can consist of different numbers of subdisks.
concat -mirror: This volume layout contains a single plcx consisting ofone
or more concatenated subvolurnes. Each subvolume consists of two
concatenated plexes (mirrors), which consist of one ur more subdisks. II' you
have two subdisks in the top-level plex, a second subvolumc is created, which
is used as the second concatenated subdisk of the plex. In the VEA interface,
the Gl.ll tcrm used for a layered. concatenated layout is Concatenated
Mirrored. These volumes require at least two disks.
stripe-mirror: This volume layout stripes data across mirrored volumes.
The difference between stripe-mirror and concat-mirror is that the top-level
plex is striped rather than concatenated. Each mirrored subvolumc must have
the same number of disks. In the VEA interface. the GUI term used for a
layered. striped layout is Striped M inured. Striped Mirrored volumes require at
least four disks.
Copy light ,~: ;!Ullti Syruauter. CorPOfil!IOJL AlIlIght<; reserved
4-22 VERITAS Storage Foundation 5.0 for UNIX.' Fundamentals

, symurnec
Creating Layered Volumes
VEA:
In the New Volume Wizard, select Concatenated Mirrored or
Striped Mirrored as the volume layout.
vxassist make:
I
layout=stripe-mirror
layout=concat-mirror
Note:
To create simple mirrored volumes (nonlayered), you can use:
• layout=mirror-concat
• layout=mirror-stripe
Creating a Layered Volume: VEA
In the New Volume wizard. select one of the two layered volume layout types:
Concatenated Mirrored: The Concatenated Mirrored layout refers to a
concat-mirror volume.
Striped Mirrored: The Striped Mirrored layout refers to a stripe-mirror
volume.
Creating a Layered Volume: CLI
In the vxassist make syntax. you can specify any ofthe following layout
types:
To create layered volumes:
layout=concat-mirror
layout=stripe-mirror
To create simple mirrored volumes:
layout=mirror-concat
layout=mirror-stripe
For striped volumes. you can specify other attributes, such as
ncol=number_of_columns and s t ri peun i t e s i z e.
Copyriqht <t 2006 Symantec Corporation All rights reserved

syrnantec
Viewing Layered Volumes
vxprint -rt volOl
Top-level v volOl ENABLED ACTIVE ...
volume and plex pl volOl-03 volOl ENABLED ACTIVE ...
Subvolume, sv volOl-SOl volOl-03 volOl-LOl 1. ..
second-level v2 volOl-LOl ENABLED ACTIVE ...
volume, plex, p2 volOl-POl volOl-LOl ENABLED ACTIVE ...
and subvolume s2 datadg05-02 volOl-POl datadg05 0 ...
p2 volOl-P02 volOl-LOl ENABLED ACTIVE ...
s2 datadg03-02 volOl-P02 datadg03 O...
sv volOl-S02 vol01-03 volOl-L02 1. ..
Viewing a Layered Volume: VEA
To view the layout of a layered olume, you can use any of the methods tor
displaying volume information. including the:
Object views in the main window
Disk View window
Volume View window
Volume to Disk Mapping window
Volume Layout window
Viewing a Layered Volume: ell
To view the contiguration of a layered volume from the command line, you use the
-r option of the vxprint command. Thc - r option ensures that subvolume
configuration information for a layered volume is displayed. The - L option is also
useful for displaying layered volume information when used with - r. -L displays
related records of a volume containing subvolumcs, but grouping is performed
under any volume.
4-24
Copyright ,::. 200fi Svmautec Corporation All nqhts reserved

'S)111ilI1hX.
With storage attributes, you can specify:
• Which storage devices are used by the volume
• How volumes are mirrored across devices
When creating a volume, you can:
• Include specific disks, controllers, enclosures,
targets, or trays to be used for the volume.
• Exclude specific disks, controllers, enclosures,
targets, or trays from being used for the volume.
• Mirror volumes across specific controllers,
enclosures, targets, or trays. (By default, VxVM
mirrors across different disks.)
Allocating Storage for Volumes
Specifying Storage Attributes for Volumes
YxYM selects the disks on which each volume resides automatically. unless you
specify otherwise. To create a volume on specific disks. you can designate those
disks when creating a volume. By specifying storage attributes when you create a
volume. you can:
Include specific disks. controllers. enclosures. targets. or trays to be used for
the volume.
Exclude specific disks. controllers. enclosures. targets. or trays from being
used lor the volume.
Mirror volumes across specific controllers. enclosures. targets. or trays. (By
default. YxYM does not permit mirroring on the same disk.)
By specifying storage attributes. you can ensure a high availability environment.
For example. you can only permit mirroring of a volume on disks connected to
different controllers and eliminate the controller as a single point of failure.
Note: When creating a volume. all storage attributes that you specify for use must
belong to the same disk group. Otherwise. YxYM does not use these storage
attributes to create a volume.
Lesson 4 SelectingVolumeLayouts 4-25
Copynghl 20:)6 Symaruer. Corporation All nqtus reserved
I

syrnantec
Storage Attributes: Methods
VEA: In the New Volume wizard, select "Manually select
disks for use by this volume" and select the disks and
storage allocation policy.
ell: Add storage attributes to vxassist make:
vxassist -g diskgroup make volume_name length
[layout=layoutJ [mirror=ctlr Ienclr I target]
[I] [storage_attributes •.. J
~
• Disks: datadg02 • Mirror across controllers:
C t II mirrormctlr
. Exclude. : E~:I~:U~::~ ctlr: c2 • Mirror across enclosures:
mirror=enclrenclr:emcl
• Targets: target: c2t4 • Mirror across targets:
mirrormtarget
• Trays: c2tray2
For example, to exclude all disks that are on controller cz:
vxassist -g datadg make datavol 5g !ctlr:c2
Specifying Storage Attributes: VEA
You can specify that the volume is to be mirrored or striped across controllers.
enclosures. targets, or trays .
.-J M!rrorAcross: (.QnlrCi!ff I (I
~ ~IrJpe Across: ,co';;trOlier r(J
.-J Qrdered CQl11rolier -J'
Tray
1-~---~-ITarget
Enclosure
Note: A tray is a set of disks within certain Sun arrays, Note that this option may
not be available 011 other platforms.
To exclude a disk. controller. enclosure. target. or tray. you add the exclusion
symbol ( I ) before the storage attribute. For example. to exclude datadg02 from
volume creation. yuu use the format: ! datadg02.
For example. to create a volume on specific disks by creating a 5-GB volume
called datavol on datadg03 and datadg04:
vxassist -g datadg make datavol 5g datadg03 datadg04
Copyflghl 200b Symantec Corporation, All nqtus reservec

, syrnanrec
Ordered Allocation
Ordered allocation enables you to control how columns
and mirrors are laid out when creating a volume.
With ordered allocation, storage administrators can
override the built-in allocation defaults.
VEA:
In the New Volume wizard, select "Manually select disks for use by
this volume." Select the disks and the storage allocation policy and
mark the Ordered check box. '" go""''' I
CLl:
Add the -0 ordered option:
vxassist -9 diskgrotlp[-o ordered] make
volume name length [layout=layout] ...
Specifying Ordered Allocation of Storage for Volumes
In addition 10 specifying which storage devices YxYM uses10 create a volume.
you can also specify how the volume is distributed on the specified storage. By
using the ordered allocation feature ofYxYM. you can control how volumes are
laid out on specified storage.
For example. if you are creating a three-column mirror-stripe volume using six
specified disks. YxYM creates column I on the first disk. column 2 on the second
disk. and column 3 on the third disk. Then. the mirror is created using the fourth,
fifth, and sixth specified disks.
Without the ordered allocation option. Vx VM selects disks in several ways.
including the following:
vxconf igd selects a disk in the group which has no subdisks.
vxconf igd selects subdisks for a striped plcx from disks already associated
into striped plcxes rather than disks associated into concat plcxcs.
vxconf igd selects a disk with an existing log plex for the log plcx of another
volume.
YxYM has default methods for space allocation. as indicated by the f sgen
UTYPEin the vxprint output. Storage administrators can override the built-in
defaults:
First, YxYM concatenates subdisks in columns.
Secondly. YxVM groups columns in striped plcxcs.
Finally. YxYM forms mirrors.
Use the -0 ordered option to the vxassist make command.
Copyri9ht:f; 20U6 Syrnantec Corporation. All rights reserved
I
4-27

symantec
Ordered Allocation: Example
Specifying the order of columns:
vxassist -g datadg -0 ordered
make datavol 2g layout=stripe
neol=3 datadg03 datadg02 datadgOl
Without using ordered allocation:
(No guarantee of disk order)
vxassist -g datadg
make datavol 2g layout=stripe
neol=3 datadg03 datadg02 datadgOl
Example I: Order of Columns
To create a IO-GB striped volume, called datavol, with three columns striped
across three disks:
vxassist -g datadg -0 ordered make datavol 109
layout=stripe neol=3 datadg03 datadg02 datadgOl
Because the -0 ordered option is specified, column I is placed on datadg03,
column 2 is placed on datadg02. and column 3 is placed on datadgOl.
Without the - 0 ordered option. column I would be placed on da tadgO 1. and
so on.
Example 2: Order of Mirrors
To create a mirrored volume using datadg02 and datadg04:
vxassist -g datadg -0 ordered make datavol 109
layout=mirror datadg04 datadg02
Because the -0 ordered option is specified, the first mirror is placed on
datadg04, and the second mirror is placed on datadg02. Without this option.
the first mirror could be placed on either disk.
Note: There is no logical difference between the mirrors. However, by controlling
the order of mirrors, you can allocate plex names with specific disks (for example,
datavol- 01with datadg02 and datavol- 02 with datadg04). This level of
control is significant when you perform mirror breakoff and disk group split
operations. You can establish conventions that indicate to you which specific disks
are used for the mirror breakoff operations.
4-28
Copyright ,c 2006 Svmautec Corporation AlllIghls reservec

, sym.uuec.
I
Lesson Summary
• Key Points
This lesson described the advantages and
disadvantages of volume layouts supported by
VxVM. You learned how to create concatenated,
striped, mirrored, and layered volumes. In addition,
you learned how to allocate storage for a volume
by specifying storage attributes and ordered
allocation.
'symantct:,
Lab 4
Lab 4: Selecting Volume Layouts
In this lab, you create simple concatenated
volumes, striped volumes, and mirrored
volumes.
You also practice creating a layered volume and
using ordered allocation while creating volumes.
Labs and solutions tor this lesson arc located on the following pages:
Appendix A provides complete lab instructions, "1,Clh .1: Sc:kC:llli" ',,:UI11<:
lavoutv." P;I1!C ,·21
Appendix B provides complete lab instructions and solutions, "1;11, -+ '.;olnt;"I',:
Sl..'lecting v'olumc IJOlih," pdg;,:lL ~:~
Copyright c: 2006 Syrnantec Corporation. Ali right" rc serveo
4-29

4-30 VERITAS Storage Foundation 5.0 for UNIX· Fundamentals
CopyJlytl1 ~ 20Dti Svrnantec Corporatron. All flotlls reservoc

Lesson 5
Making Basic Configuration Changes

symantcc
Lesson Introduction
System
• Lesson 5: Making Basic Configuration ~
gfJi:lrJ{1f!~W""""W"_~
Problems
~~'.%~'k& .Ab'l[dl! .• l($;tl!c. 'symanlLT
be able to:
Topic 1: Administering Add a mirror to and remove a mirror from an
Mirrored Volumes existing volume, add a log, and change the
volume read policy.
Topic 2: Resizing a Volume Resize an existing volume by using VEA and
from the command line.
Topic 3: Moving Data Between Deport a disk group from one system and
Systems import it on another system.
Topic 4: Renaming Disks and Rename disks and disk groups.
Disk Groups
Topic 5: Managing Old Disk Upgrade disk groups and convert non-CDS
Group Versions disk groups to CDS.

Example Array Structure
LUNs are a virtual presentation. Therefore, you have to
take the array configuration into account to understand
where the ac;:~.!!~datais placed. Array with
~it~~UUUUUU00 ;~'::~'~'
100[j0[j[1 ~;:a~~D groups
r",,,· • • •

Twelve array- "
based LUNs. ~
-.
Administering Mirrored Volumes
Example Array Structure
In an array. the LUNs are a virtual presentation. Therefore. you cannot know
where in the array the actual data will be put. That means you have no control over
the physical conditions.
The array in the slide contains slots for 14physical disks. and the configuration
places 12 physical disks in the array. These physical disks arc paired together into
(, mirrored RAID groups. In each RAID group. 12 logical units. or LUNs, arc
created. These LUNs appear to hosts as SAN-based SCSI disks. The remaining
two disks are used as spares in case one of the active disks fails.
I
Copyright ':92006 Syrnantec Corporation All rights reserved
5-3Lesson 5 Making Basic Configuration Changes

syrn.mtec
When to Add a Mirror to a Volume
• To add redundancy if it is
not provided at the I SAN I
hardware level. 1 Old Array I" I (/1 I
• To eliminate the disk J LJ"~I,..N..!ew. Array.
array as a single point ~._
of failure (SPOF) by
mirroring across arrays.
• To provide disaster
recovery across sites when
there is a SAN connecting two
or more sites.
Original Data Mirror of Data
I Migrating Data I
• To improve concurrent read performance by adding
mirrors with different 1/0 paths.
• To migrate data from one array to another.
When to Add a Mirror to a Volume
Without Storage Foundation, moving data from one array to another requires
downtime. Using Storage Foundation. you can mirror 10 a new array, ensure it is
stable, and then remove the plexes from the old array, No downtime is necessary,
These are the steps for migrating data using Storage Foundation:
1 Add new array to SAN.
2 Mirror volumes to new array.
3 Remove plexes/LUNs from old array.
4 Remove old array.
This is useful in many situations, for example, if a company purchases a new array.
With Storage Foundation, yuu:
1 Add the new array to the SAN.
2 Zone for the server to see the LUNS.
3 Rcscan with VEA.
4 Add the LUNs from the new array to the disk group.
S Mirror the volumes to the new army.
6 Remove the plcxes on the old array.
7 Remove the LUNS that are on the old array from the disk group.
This method does not require downtime.

II'a volume was not originally created as a mirrored volume. or iI'you want to add
additional mirrors. you can add a mirror to an existing volume.
By default, a mirror is created with the same plex layout as the plex already in the
volume. For example. assume that a volume is composed of a single striped plex.
If you add a mirror to the volume. VxYM makes that plex striped. as well. You can
specify a different layout using YEA or from the command line.
A mirrored volume requires at Icast two disks. You cannot add a mirror to a disk
that is already being used by the volume. A volume can have multiple mirrors. as
long as each mirror resides on separate disks.
Only disks in the same disk group as the volume can be used to create the new
mirror. Unless you specify the disks to be used for the mirror. Vx Vlvl
automatically locates and uses available disk space to create the mirror.
A volume can contain up to .,2 plcxcs (mirrors): however. the practical limit is : I.
One plex should be reserved for use by YxYM for background repair operations.
Adding and Removing Mirrors to a Volume
Adding a Mirror:
• Only concatenated or striped volumes can be mirrored.
• By default, a mirror is created with the same plex layout
as the original volume.
• Each mirror must reside on separate disks. ~ ~
• All disks must be in the same disk group. ~~
A volume can have up to 32 plexes, or mirrors.
• Adding a mirror requires plex synchronization.
Removing a Mirror:
When a mirror is removed, the space
occupied by that mirror can be used
elsewhere.
When a mirror (plex) is no longer needed. you can remove it. You can remove a
mirror to provide free space. to reduce the number of mirrors. to remove a
temporary mirror.
Caution: Removing a mirror results in loss of data redundancy. If a volume only
has two plcxes. removing one ofthem leaves the volume unmirrorcd.
Adding and Removing Mirrors
Removing a Mirror
Lesson 5 Making Basic Configuration Changes
Copyright 1: 2006 Svmantec Corporation All rights reserved
I
5-5

symantec
Adding/Removing Mirrors
VEA:
• SelectActions->Mirror->Add.
• SelectActions->Mirror->Remove.
vxassist mirror:
vxassist -g diskgroup mirror volume name
[layout=layout~_ type] [di sk _ name]
vxassist -g datadg mirror datavol
vxassist remove mirror:
vxassist -g cti ekoroup remove mirror vojume name L! l dm~nal1le
To remove the plex that contains a subdisk from the disk datadg02:
vxassist -g datadg remove mirror datavol !datadg02
To remove the plex that uses any disk except datadg02:
vxassist -g datadg remove mirror datavol datadg02
Adding a Mirror: VEA
Select: Ihe volume to be mirrored
Navigation path: Actions- ~>Mirror->;dd
Input: Number of mirrors to add: Type a number. Default is I.
Choose the layout: Select from Concatenated or Striped.
Select disks to use: Vx VM can select the disks. or you can
choose specific disks. You can also mirror or stripe across
controllers. trays. targets. or enclosures.
To verify that a new mirror was added. view the total number of copies of the
volume as displayed in the main window. The total number of copies is increased
by the number of mirrors added.
Adding a Mirror: CLI
To add a mirror onto a specific disk. you specify the disk name in the command:
vxassist -g datadg mirror datavol datadg03
Removing a Mirror: CLI
To remove a mirror. use vxassist remove mirror:
vxassist -9 diskgroup remove mirror volume name
You can also use vxplex:
vxplex -g diskgroup -0 rm dis plex_name
COPYllgtH ';/2006 Symantec Corpor aucn. All nghts reserveo

~. symarue«
Adding a Dirty Region Log (DRL) to a Volume
• Log keeps track of changed regions.
• If the system fails, only the changed regions of the volume
must be recovered.
• Not enabled by default. When enabled, one log is created.
• You can create additional logs to mirror log data.
VEA:
• Actions->Log->Add
• Actions->Log->Remove
vxassist
vxassist -g diskgroup addleg volume name [logtype=drl]
{nlo9:::on) [attributes)
vxassist -9 diskgroup remove log vol ume
vxassist -9 datadg addlog datavol logtype~drl
vxassist -9 datadg remove log nlog=2 datavol
Adding a Log to a Volume
Logging in VxVM
By enabling logging. YxYM tracks changed regions of a volume. Log information
can then be used to reduce plcx synchronization times and speed the recovery of
volumes after a system failure. Logging is an optional feature, but is highly
recommended. especially for large volumes.
Dirty Region Logging
Dirty region logging (DRL) is used with mirrored volume layouts. DRL keeps
track of the regions that have changed due to JlO writes to a mirrored volume.
Prior to every write. a bit is set in a log to record the area of the disk that is being
changed. In case of system failure. DRL uses this information to recover only the
portions of the volume that need to be recovered.
IfDRL is not used and a system failure occurs, all mirrors of the volumes must be
restored to a consistent state by copying the full contents of the volume between its
mirrors. This process can be lengthy and I/O intensive.
When you enable logging on a mirrored volume. one log plex is created by default.
The log plex uses space from disks already used for that volume, or you can
specify which disk to use. To enhance performance. you should consider placing
the log plex on a disk that is not already in use by the volume.
To create a volume that is mirrored and logged:
vxassist -g datadg make datavol Sm layout=mirror
logtype=drl
Lesson 5 MakingBasicConfigurationChanges
CDpyright F-2006 swneotec Corporation. All riqtllS reserved
I
5-7

Volume Read Policies
I Round Robin I
S
Olume
Read .. . .. Read
110 I/O
Rea
110
Selected Plex
Read r-, Volume...,
110
Is there a 1'-'striped plex? .~
IDefault Method I "- .-'
IPreferred Plex I
Siteread I
Read I/O ,.... .G~~~~~~.....,from host: Site R ~H Site :
at Site A i A t=J f; t::::j B :
!.... ~ ..... j
svrnantcc.
Volume Read Policies with Mirroring
One of the benefits of mirrored volumes is thai you have more than one copy of the
data from which to satisfy read requests. The read policy for a volume determines
the order in which plexes are accessed during 1/0 operations.
VxVM has three read policies the YOIl can specify to satisfy read requests:
Round robin: VxVM reads each plcx in turn in "round-robin" manner lor
each nonsequential 1:0 detected. Sequential access causes only one plex to be
accessed in order to take advantage of drive or controller read-ahead caching
policies. If a read is within 256K ofthe previous read. then the read is sent to
the same plcx.
Preferred plcx: Vx VM reads first from a plex that has been named as the
preferred plcx. Read requests are satisfied from one specific plex, presumably
the plcx with the highest performance. Ifthe preferred plex fails, another plex
is accessed. For example. if you are mirroring in a campus environment and
the local plcx would he faster than the remote one. setting the local plex as the
preferred plcx would increase performance.
Selected plex: This is the default read policy. Under the selected plex policy,
Volume Manager chooses an appropriate read policy based on the plex
configuration to achieve the greatest 1/0 throughput. If the mirrored volume
has exactly one enabled striped plex, the read policy defaults to that plex;
otherwise. it defaults to a round-robin read policy.
Sitcrcad: VxVM reads preferentially from plcxcs at the locally defined site.
This is the default policy for volumes in disk groups where site consistency has
been enabled.
CUPYrlght e 20% Svmantec Corporation All n91'16reserved

, symamec.
Setting the Volume Read Policy
VEA:
• Actions->Set Volume Usage
• Select from Based on layouts, Round robin, or Preferred.
vxvol rdpol:
vxvol -g diskgroup rdpol policy volume name [plexl
Examples:
• To set the read policy to round robin:
vxvol -g datadg rdpol round datavol
• To set the read policy to read from a preferred plex:
vxvol -g datadg rdpol prefer datavol datavol-02
• To set the read policy to select a plex based on layouts:
vxvol -g datadg rdpol select datavol
Changing the Volume Read Policy: VEA
Select: A volume
Navigation path: Actions->Set Volume Usage
Input: Volume read policy: Select Basedon layouts (default: the
selected plex method). Round robin. Site local read, or Preferred.
Iryou select Preferred. then you can also select the preferred plcx
from the list of available plexes,
Changing the Volume Read Policy: CLI
vxvol -9 diskgroup rdpol round volume name
vxvol -9 diskgroup rdpol prefer volume name
preferred_plex
vxvol -9 diskgroup rdpol select volume name
I
Copyright:~ 2006 Symantec coro.euoo All flqllt~ -eserveo
5-9Lesson 5 Making Basic Configuration Changes

symantec
Resizing a Volume
To resize a volume, you can:
• Specify a desired new volume size.
• Add to or subtract from the current volume size.
g-Disk space must be available.
- VxVM assigns disk space, or you can specify disks.
L_
~t.....__..-J
Shrinking a volume enables you to use space
elsewhere. VxVM returns space to the free
space pool.
If a volume is resized, its file system must also be resized.
VxFS can be expanded or reduced while mounted.
• UFS/HFS can be expanded, but not reduced. HFS needs to be
unmounted to be expanded.
• Ensure that the data manager application supports resizing.
Resizing a Volume
Resizing a Volume
If users require more space on a volume. you can increase the size of the volume.
Ifa volume contains unused space that you need to use elsewhere, you can shrink
the volume.
When the volume size is increased. sufficient disk space must be available in the
disk group. When increasing the size of a volume. VxVM assigns the necessary
new space from available disks. By default, VxVM uses space from any disk in the
disk group, unless you define specific disks.
Resizing a Volume with a File System
Volumes and file systems arc separate virtual objects. When a volume is resized,
the size of the raw volume is changed. Ifa tile system exists that uses the volume,
the file system must also be resizcd.
When you resize a volume using VEA or the vxres i ze command, the tile system
is also resized.
Resizing Volumes with Other Types of Data
For volumes containing data other than file systems. such as raw database data.
you must ensure that the data manager application can support the resizing of the
data device with which it has been configured.
5-10
Copyright ;: 2006 Svmantec Corporauon AU lights reserved

, symamec
Resizing a Volume: Methods
Method What Is Reslzed?
VEA Both volume and file system
To resize a volume from the command line. you can use either the vxassist
command or the vxresize command. Both commands can expand or reduce a
volume 10 a specific size or bv a specified amount of space. with one significant
difference:
vxres i ze automatically resizes a volume's f Ie system.
vxassist does not resin? a volume's tile system.
When using vxass i s t ,you must resize the tile system separately by using the
f sadmcommand.
vxresize Both volume and file system
IWhen you expand a volume. both commands automatically locate mailable disk
space unless you designate specific disks to use. When you shrink a volume, the
unused space becomes free space in the disk group.
When you resizc a volume. you can specify the length of a new volume in sectors.
kilobytes. megabytes. or gigabytes. The unit of measure is added as a suffix to the
length (s, k, m, or g). Ifno unit is specified, the default unit is sectors.
vxassist Volume only
fsadm File system only (VxFS only)
Resizing a Volume and File System: Methods
Copyriyhl~' 2006 Symantec Corporation All rights reserved
5-11

syruantcc
Resizing a Volume: VEA
I Highlight a volume, and select Actions->Resize Volume.
Volume name: datavoiOI Specify the amount
Current volume size: 1500 ~
of space to add or
subtract, or specify a
Add by: 1 fMB-::J new volume size.
5ubtract by: I fMB-::J
New volume size;
I fMB-::J Max Size
I
r. Let VolumeManager decide what disks to use for this volume [If. desired, specify
r Manually select disks for use by this volume J disks to be used
lfor the additional
space.
Resizing a Volume and File System: VEA
Select: The volume to be resized
Navigation path: Actions -->Resize Volume
Input: Add by: To increase the volume size by a specific amount of
'pace. input how much space should be added to the volume.
Subtract by: To decreasethe volume size by a specific amount of
space. input how much 'pace should be removed.
New volume size: To specify a new volume size. input the size.
Max Size: To determine the largest possible size. click Max Size.
Select disks for use by this volume: You can select specific
disks to use and specify mirroring and striping options.
Force: You can force the rcsize if the size is being reduced and
the volume is active.
Notes: When you resize a volume. if a YERITAS file system (YxFS) is mounted
on the volume. the file system is also rcsized. The tile system is not resized if it is
unmounted.

,S)111ank'C.
Resizing a Volume and File System: vxresize
The new_length operand can begin with a plus sign (+) to indicate that the new
length is added to the current volume length. ; minus sign ( - ) indicates that thc
new length is subtracted. - b runs the process in the background.
The ability to expand or shrink a file system depends on the tile system type and
whether the tile system is mounted or unmounted.
Resizing a Volume: vxresize
vxresize [-bJ fs_tlpe -g diskgroup volume narne
[+ I-J new_length
Original volume size: 10 MB
~
vxresize -g mydg myvol SOm
vxresize -g mydg myvol +lOm
vxresize -g mydg myvol 40m
vxresize -g mydg myvol -10m
File System Type Mounted FS Unmounted FS
VxFS Expand and shrink '101 allowed
UFS (Solaris) Expand only Expand only
HFS (llP-UX) /01 allowed Expand only
I
Example: The size of the volume myvol is 10MB. To extend myvol to 50 MB:
To extend myvol by an additional 10 MB:
vxresize -g mydg myvol +10m
To shrink myvol back to a length 01'40 MB:
To shrink myvol by an additional 10MB:
vxresize -g mydg myvol -10m
Copyrigtll:E 2006 Syrnantcc Corporation. All rights reserved.
5-13

syrnantec
Resizing a Volume: vxassist
vxassist -g diskgrollp {growtolgrowbylshrinktol
shrinkby} volume_name size
Original volume size: 20 MB
CD vxassist -g datadg growto datavol 40m
~
vxassist -g datadg growby datavol 10m
vxassist -g datadg shrinkto datavol 30m
vxassist -g datadg shr nkby datavol 10m
Resizing a Volume Only: vxassist
q rowt o
growby
shrinkto
shrinkby
Increases volume (0 specified length
Increases volume hi' specified amount
Reduces volume (0 specified length
Reduces volume 1>.1' specified amount
Resizing a File System Only: fsadm
You may need to resize a tile system to accommodate a change in use-for
example, when there is an increased need for space in the file system. You may
also need to resize a tile system as part of a general reorganization of disk
usage-for example, when a large tile system is subdivided into several smaller
file systems. You can resize a VxFS tile system while the tile system remains
mounted by using the f sadm command:
fsadm [-b newsizel [-r rawdevl mount point
Using fsadm to resizc a tile system does not automatically resize the underlying
volume. When you expand a tile system. the underlying device must be large
enough to contain the new larger file system.
5-14
CCP'i'flght:, ,nn6 Symantcc Corporation All fights reserved

S}11liHH.X
Resizing a Dynamic LUN
• If you resize a LUN in the hardware, you should resize the
VxVM disk corresponding to that LUN.
• Disk headers and other VxVM structures are updated to
reflect the new size.
• Intended for devices that are part of an imported disk group.
VEA:
• Select the disk that you want to expand.
• Select Actions->Resize Disk.
ell:
vxdisk [-fJ -g diskgrollp resize dm nAme
Example:
vxdisk -g datadg resize datadgOl
Resizing a Dynamic LUN
When you resize a LUN in the hardware. you should resize the YxYM disk
corresponding to that LUN. You can use vxdi sk res i ze to update disk headers
and other VxYM structures to match a new LUN size. This command does not
resize the underlying LUN itself.
I
Lesson 5 Making Basic Configuration Changes 5-15
Copyright tt' 2006 Symantec Corpo-auon All riqhts reserved

r: Computer B1, Computer A I
,------;-----,
acctdg
l§'CJJ
engdg
, Additional Disks
Example: Disk Groups and High Availability
The example in the diagram represents a high availability environment.
In the example, Computer A and Computer B each have their own bootdg on
their own private SCSI bus. The two hosts are also on a shared SCSI bus. On the
shared bus, each host has a disk group, and each disk group has a set of VxVM
disks and volumes. There are additional disks on the shared SCSI bus that have not
been added to a disk group.
If Computer A fails. then Computer B. which is on the same SCSI bus as disk
group acctdg. can take ownership or control of the disk group and all of its
components.
5-16
Copyright - /.u1J6Syruantec Corporation All flghls reserved

What is a deported disk group?
• The disk group and its volumes are
unavailable.
• The disks cannot be removed.
• The disk group cannot be accessed
until it is imported.
Before deporting a disk group:
• Unmount file systems.
• Stop volumes.
When you deport a disk
group, you can specify:
• A new host
• A new disk group name
~lk.•."&i ,~wP .)~
Deporting a Disk Group
Deporting a Disk Group
A deported disk group is a disk group over which management control has been
surrendered. The objects within the disk group cannot be accessed. its volumes are
unavailable. and the disk group configuration cannot be changed. (You cannot
access volumes in a deported disk group because the directory containing the
device nodes for the volumes are deleted upon deport.) To resume management of
the disk group. it must be imported.
A disk group cannot be deported irany volumes in that disk group arc in use.
Before you deport a disk group. you must unmount file systems and stop any
volumes in the disk group.
Deporting and Specifying a New Host
I
When you deport a disk group using YEA or CLI commands, you have the option
to speeify a new host to which the disk group is imported at reboot. Iryou know
the name of the host to which the disk group will be imported. then you should
specify the new host during the operation. If you do not specify the new host. then
the disks could accidentally be added to another disk group. resulting in data loss.
You cannot specify a new host using the vxdiskadm utility.
Deporting and Renaming
When you deport a disk group using YEA or CLI commands. you also hale the
option to rename the disk group when you deport it. You cannot rename a disk
group when deporting using the vxdiskadm utility.
Copyright G 2006 Svrnantec Corporation. Alilighis reserved
5-17

symaruec
. Select Actions->Deport Disk Group.
: Disk group to
be deported
datadg
lid Dellor! options
New name: r+: r-······
N8¥/hOst:
~I
vxdiskadm: "Remove access to (deport) a disk group"
vxdg [-n new_name] [-h hostname] deport diskgroup
Deporting a Disk Group:
Disks that were in the disk group now have a state of Deported. If the disk group
was deported to another host. the disk state is Foreign.
Note: If you amine the disks. you must manually online the disks before you
import the disk group. To online a disk. use vxdiskadm option "Enable (online) a
disk device."
Before deporting a disk group. unmount all tile systems used within the disk group
that is to be deported. and stop all volumes in the disk group:
umount mount po~nt
vxvo1 -g diskgroup stopa11
5-18
Copyngt1 C' 21106 Svmautec Corporauon All nqbts rescrveo

Importing a Disk Group
Importing a disk group
reenables access to the
disk group.
When you import a disk
group, you can:
• Specify a new disk group
name.
• Clear host locks.
• Import as temporary.
• Force an import.
Importing a Deported Disk Group
All volumes are stopped by default after importing a disk group and must be
started before data can be accessed.
Importing and Renaming
A deported disk group cannot be imported ifanother disk group with the same
name has been created since the disk group was deported. You can import and
rename a disk group at the same time.
Importing and Clearing Host Locks
When a disk group is created. the system writes a lock on all disks in the disk
group. The lock ensures that dual-ported disks (disks that can be accessed
simultaneously by two systems) are not used by both systems at the same time. If a
system crashes. the lucks stored on the disks remain. and if you try to import a disk
group containing those disks. the import fails.
Importing As Temporary
A temporary import does not persist across reboots. A temporary import can be
useful, for example. if you need to perform administrative operations on the
temporarily imported disk group. VEA docs not support temporary import.
Forcing an Import
A disk group import fails if the VxVM configuration daemon cannot rind all ofthe
disks in the disk group. If the import tails because a disk has failed. you can force
the disk group to be imported. Forcing an import should always be performed with
caution.
Lesson 5 Making Basic ConfigurationChanges
Copyright rs 20u6 Symantec Corporation All rights reserved
I
5-19

symantec
Importing a Disk Group
, Select Actions->Import Disk Group.
Group name: datadg,1155041Sn64.coursedev6
Options
blewname: IClalaclg
=cc"-----
SileName G
[] Clear hosllD
Options include:
Clearing host IDs
at import
Forcing an import
• Starting all volumes
Importing a Disk Group:
By default, when you import a disk group by using VEA, all vol limes in the disk
group are started automatically. By default. the vxdiskadm import option starts
all volumes in the disk group. When yuu import a disk gruup from the command
line, you must manually start all volumes
A disk group must be deported from its previous system before it can be imported
to the new system. During the import operation. the system checks for host import
lucks. Ifany lucks are found. you are prompted to clear the locks.
To temporarily rename an imported disk group. you use the - t option. This option
imports the disk group temporarily and docs not set the autoimport flag, which
means that the import cannot survive a reboot.
To display all disk groups, including deported disk groups:
[] Eorce
~ Start all ~olumes
STATUS
online
online
5-20
vxdiskadm: "Enable access to (import) a disk group"
vxdg [-ftCl [-n nelv_namelimport dJskgroup
vxvol -g diskgroup startall
VXdlSk -0 alldgs list
DEVICE
clt2:iOs2
clt2dls2
TYPE
auto:cdsdisk
auto:cdsdisk
DISK
datadgOl
GROUP
datadg
(acctdg)
CopYrlght'~ 2006 Syruanter Ccrpo.auoo All lights reserved
VERITAS Storage Foundation 5,0 for UNIX' Fundamentals

Renaming Disks and Disk Groups
Changing the Disk Media Name
VxVM creates a unique disk media name for a disk when you add a disk to a disk
group. Sometimes you may need to change a disk name to reflect changes or
ownership or use or the disk. Renaming a disk does not change the physical disk
device name. The new disk name must be unique within the disk group.
VEA:
o Select the disk that you want to rename.
o Select Actions->Rename Disk.
o Specify the original disk name and the new name.
vxedit rename:
vxedit -g diskgroup rename old name new name
Example:
vxedit -g datadg rename datadgOl datadg03
Notes:
o The new disk name must be unique within the disk group.
o Renaming a disk does not automatically rename subdisks on
that disk.
IBefore you rename a disk, you should carefully consider the change. VxVM
names subdisks based on the disks on which they are located. J disk named
datadgOl contains subdisks that are named datadgOl-Ol, datadgOl-02,
and so on. Renaming a disk does not automatically rename its subdisks. Volumes
are not affected when subdisks arc named differently trom the disks.
Before You Rename a Disk
Lesson 5 Making Basic ConfigurationChanges
Copyrtght ~ 2006 Svrnantec Corporation AII'iqlm. reserved
5-21

symantcc.
Host A
Deport "
I In VEA, select Actions->
: Rename Disk Group
"~
--------
vxdg -n new~nafile deport
vxdg import new name
vxdg deport old name
vxdg -n new_name import
Renaming a Disk Group
You cannot import or deport a disk group when the target system already has a disk
group of the same name. To avoid name collision or to provide a more appropriate
name for a disk group, you can rename a disk group.
To rename a disk group when moving it from one system to another, you
specify the new name during the deport or during the import operations.
To rename a disk group without moving the disk group, you must still deport
and reimport the disk group on the same system.
The YEA interface has a Rename Disk Group menu option. On the surface, this
option appears to be simply renaming the disk group. However, the option works
by deporting and rcimporting the disk group with a new name.
Using the CLI, for example, 10 rename the disk group datadg to mktdg:
vxdg -n mktdg deport datadg
vxdg import mktdg
vxvol -g mktdg startall
or
vxdg deport datadg
vxdg -n mktdg import datadg
vxvol -g mktdg startall
From the command line, you must restart all volumes in the disk group:
vxvol -9 new name startall
Cu~ynyht:; 20l~,':':".panlec Corporauou All r.glils rescr-•..eo

'symantt'(
All disk groups have a version number based on the Storage
Foundation release.
Each disk group version supports a set of features. You
must upgrade old disk group versions in order to use new
features.
SF Disk Group Supported Disk
Release Version Group Versions
3.2,3.5 90 20-90
4.0 110 20-110
4.1 120 20-120
5.0 140 20-140
To upgrade the disk group version:
In VEA, select the disk group to be upgraded then select
Actions->Upgrade Disk Group Version.
In CLI, type:
vxdg [-T version] upgrade diskgroup
Managing Old Disk Group Versions
Upgrading a Disk Group
All disk groups have an associated version number. Each VxVM release supports a
specific set of disk group versions and can import and perform tasks on disk
groups with those versions. Some new features and tasks only work on disk groups
with the current disk group version. so you must upgrade existing disk groups in
order to perform those tasks.
Once you upgrade a disk group. the disk group becomes incompatible with earlier
releases ofVxVM that do not support the new version. Upgrading the disk group
version is an online operation. You cannot downgrade a disk group version.
IDisplaying the Disk Group Version:
In the VEA Disk Group Properties window. if the Current version property is Yes.
then the disk group version is current.
In CLI. type:
vxdg list newdg
Group: newdg
dgid: 97121640B.1133.cassius
version: 140
Lesson 5 MakingBasic ConfigurationChanges 5-23
Copyriglll '£, 2006 Symantec Corporation. All righls reserved

symantcc
CDS Disk Groups
CDS disk groups are used for seamless transfer of data
between different platforms. For example, for moving
copies of data to a backup server that is on a different OS.
• CDS disk groups are created by default as of VxVM 4.x.
• Disk groups created before version 4.x are non-CDS.
·····_··_··················1CDS attribute: c ds eon
II DG version: version=110 (or higher)~~~~
Idatadg I
A CDS disk group cannot have non-CDS disks in it.
However, a CDS disk can be added to a non-CDS disk
group as long as the disk group version supports it.
Requirements for CDS Disk Groups
Thc CDS attribute indicates that the disk group can be shared across platforms.
CDS disk groups have fields indicating which platform-type created the disk group
and which platform-type last imported the disk group, in addition to device quotas.
5-24
Ccpynyht ': 200b Symautec Cosporanco. All fights reserved

'symanl('(
Converting a Non-CDS Disk Group to a CDS Disk Group
• The disk group must be in good condition .
• Disk groups can be converted while online or offline.
Use the CDS conversion utility vxcdsconvert to convert a
VxVM non-CDS disk group to a CDS disk group:
vxcdsconvert [-A] [-d defaultsfile] -g diskgroup
[-0 novolstop]
alignmentlalldisksldisk namelgroup [attribute]
For example, to convert the disk group olddg to a CDS
disk group while its volumes are still online, type:
vxcdsconvert -g olddg -0 novolstop group
Converting a Non-CDS Disk Group to a CDS Disk Group
Requirements for Converting a Non-CnS Disk Group to a CDS Disk Group
The disk group must be in good condition:
No dissociated or disabled objects
No sparse plexes
No volumes requiring recovery or having pending snapshot operations
No objects in an error state
Disk groups can be converted online or online:
Performing the conversion online. while use of the disk group continues. may
greatly increase the amount of time required for conversion.
Performing the conversion offline requires minimal online time.
What Happens When a Disk Group Is Convertcd?
The following are some other (actors to consider when converting a disk group:
II non-CDS disk group is upgraded (using the vxdg upgrade command).
II' the non-CDS disk group has one or more disks that are not CDS disks. these
disks are converted to CDS disks.
If the non-CDS disk group does not have a CDS-compatible disk group
alignment. the objects go through relayout so that they are CDS-compatible.
Applications using disks that require format conversion are terminated for the
duration of the disk conversion process (unless novol stop is used).
Using novolstop may require objects to be evacuated and then unrelocatcd.
COPYright e.2006 Syrnantec Corporation. All rigtw:; reserved.
I
5-25

symantec
5-26
Lesson Summary
• Key Points
This lesson described how to add a mirror to and
remove a mirror from an existing volume, change
the volume read policy, and resize an existing
volume. You also learned how to rename disks
and disk groups, upgrade disk groups, and
convert non-CDS disk groups to CDS.
svmantec
Lab 5
Lab 5: Making Basic Configuration Changes
This lab provides practice in making basic
configuration changes.
In this lab, you add mirrors and logs to
existing volumes, and change the volume
read policy. You also resize volumes, rename
disk groups, and move data between systems.
[
Labs and solutions fur this lesson are located on the following pages:
Appendix A provides complete lab instructions. "Lilt>:': fhklllg Basic
('on jl.~Ul ~t!ion changc-, l! pilgC :,<2(j
Appendix B provides complete lab instructions and solutions. "1.;iI) :; Soluuons
Mak iru; I$:'"IC ('unfi"ULlli()l1 C'hallf!l'';,'' pdgC n·-I"!
CuPyrlght::: 7000 Svmantec COfP(;f;;j(I(JI) All rlyht5 reserved

Lesson 6
Administering File Systems

symantec
Lesson Introduction
System
Changes
• Lesson 6: Adm;nister;nflEil~~y~~e,!,~'"
Problems
~~%.,.#k., '~~r:_~£~~ , symautcc
be able to:
Topic 1: Comparing the Describe the benefits VxFS extent-based
Allocation Policies of VxFS allocation over traditional block-based
and Traditional File Systems allocation.
Topic 2: Using VERITASFile Apply the appropriate VxFS commands
System Commands from the command line.
Topic 3: Controlling File Defragment a VxFS file system.
System Fragmentation
Topic 4: Logging in VxFS Perform logging in VxFS by using the
intent log and the file change log.
6-2
Copyright I~;2006 Svmantec Corporauon. All rl~l'lh,' .erved

Traditional Block-Based Allocation
Block-based allocation:
• Allocates space to the next
rotationally adjacent block
• Allocates blocks at random
from a free block map
• Becomes less effective
as the file system fills
• Requires extra disk 110
to write metadata
n+8
n+13
n+20
n+21
S)11!antC<':.
Comparing the Allocation Policies of VxFS and Traditional File
Systems
Both VxFS and traditional UNIX file systems, such as UFS. use index tables to
store information and location information about blocks used for tiles. However.
VxFS allocation is extent-based, while other file systems are block-based.
Block-based allocation: File systems that US!! block-based allocation assign
disk space to a file one block at a time.
Extent-based allocation: File systems that use extent-based allocation assign
disk space in groups of contiguous blocks, called extents.
Example: UFS Block-Based Allocation
UFS allocates space for files one block at a time. When allocating space to a tile.
UFS uses the next rotationally adjacent block until the file is stored.
UFS can perform at a level similar to an extent-based file system on scquential l/O
by using a technique called block clustering. In UFS, the maxcontig file system
tunable parameter can be used to cluster reads and writes together into groups of
multiple blocks. Through block clustering. writes are delayed so that several small
writes are processed as one large write. Sequential read requests can be processed
as one large read through read-ahead techniques.
Block-based allocation requires extra disk 1/0 to write file system block structure
information, or metadata. Metadata is always written synchronously to disk. which
can significantly slow overall file system performance. Over time, block-based
allocation produces a fragmented file system with random file access.
Lesson 6 Administering File Systems
Copyright ,r;:,2006 Symantec Corporation. All rights reservC(1
I
6-3

n+9 n+10 n+11 n+12 n+13n+14 n+1 n+16 n+17 n+37 n+38 n+39
symantec
VxFS Extent-Based Allocation
n n+1 n+2 n+3 n+4 n+5 n+6 n+7 n+8
Address-length
pair consists of:
• Starting block
• Length of extent
Extent: A set
of contiguous
blocks
• Extent size is based on
the size of I/O write
requests.
• When a file expands,
another extent is allocated.
• Additional extents are
progressively larger, reducing
the total number of extents
used by a file.
n+4 n+41n+42
VxFS Extent-Based Allocation
VERITAS File System selects a contiguous range of tile system blocks. called an
extent. for inclusion in a tile. The number of blocks in an extent varies and is based
on either the I/O pattern of the application. or explicit requests by the user or
programmer. Extent-based allocation enables larger I/O operations to be passed to
the underlying drivers.
VxFS attempts to allocate each tile in one extent of blocks. If this is not possible.
VxFS attempts to allocate all extents for a tile close to each other.
Each file is associated with an index block. called an inode. In an inode, an extent
is represented as an address-lengthpair, which identifies the starting block address
and the length of the extent in logical blocks. This enables the tile system to
directly access any block of the file.
VxFS automatically selects an extent size by using a default allocation policy that
is based on the size of I/O write requests. The default allocation policy attempts to
balance two goals:
Optimum 1/0 performance through large allocations
Minimal tile system fragmentation through allocation from space available in
the file system that best fits the data
The first extent allocated is large enough for the first write to the file, Typically, the
first extent is the smallest power of 2 that is larger than the size of the first write.
with a minimum extent allocation of 8K. Additional extents are progressively
larger. doubling the size of the file with each new extent. This method reduces the
total number of extents used by a single tile.
C()Py'l~lh 200f) Symamec ccrpo-auco All ngtlls reserved

Using VERITAS File System Commands
You can generally use VERITAS File System (Vxf'S) as an alternative to other
disk-based, OS-specilic file systems, except for the lile systems used to boot the
system. File systems used to boot the system are mounted read-only in the boot
process, before the VxFS driver is loaded.
VxFS can be used in place of:
UNIX File System (UFS) on Solaris, except for root, /usr, /var, and /opt.
Hierarchical File System (IIFS) on IIP-UX, except for /stand .
Journaled File System (.IFS) and Enhanced Journalcd File System (JFS1) on
AJX, except for root and /usr.
Extended File System Version 2 (EXT1) and Version 3 (EXT3) on Linux,
except for root, /boot, / et c, /1 .i b, /var, and /usr.
Using VxFS Commands
• VxFS can be used as the basis for any file system except for
file systems used to boot the system.
• Specify directories in the PATH environment variable to
access VxFS-specific commands.
• VxFS uses standard file system management syntax:
command [fs type] [generic opt.iOIlS]
[-0 VxFS_optiollS] [specialTmount_poillr]
• Use the file system switchout to access VxFS-specific
versions of standard commands.
Without the file system switchout. the file system type is
taken from the default specified in the default file system file.
To use VxFS as your default. change this file to contain vxfs.
Solaris I HP-UX I A1X I Linu~1
Location of VxFS Commands IPlatform Location of VxFS Commands
Solaris /opt/VRTSvxfs/sbin,/usr/lib/fs/vxfs,/etc/fs/vxfs.
'opl/VRTS/bin
HP-UX /opt/VRTS/bi~/sbin/fs,/usr/lbin/fs
AIX /opt/VRTSvxfs/sbin,/usr/lib/fs/vxfs./etc/fs/vxfs
Linux /sbin./usr/lib/fs/vxfs
Specify these directories in the PATH environment variable.
Cop ynqht i: 2006 Syrnantec. Corporation. All rights reserved
6-5

General File System Command Syntax
To access VxFS-specific versions, or wrappers, of standard commands, you use
the Virtual File System switchout mechanism followed by the file system type,
vxfs. The switchout mechanism directs the system to search the appropriate
directories for Vxf Svspccific versions of commands.
Platform File System Switchour
Soluris -F vxfs
HI'-UX -F vxfs
AIX -v vxfs (or-v vxf s when used with er f s)
l.inux -t vxfs
Using VxFS Commands by Default
If you do not use the switchout mechanism, then the file system type is taken from
the default specified in the OS-specitic default file system file. If you want
VERITAS File System to be your default tile system type, then you change the
default tile system tile to contain vxfs.
Platform Defuu It File System Fite
-
Solaris /ete/default/fs
HI'-UX /ete/default/fs
1--
AIX /ete/vfs
Linux /ete/default/fs
Copyuqltt ;, 20(l6 Svmantec Corporation All fights reserved

VxFS-Specific mkf s Options
'synlilntt?c.
-0 bsize_n
Sets logical block size
Default: 1024 bytes (lK) for most
Cannot be changed after creation
In most cases, the default is best.
Resizing the file system does not
change the block size.
-0 logsize_n
Sets size of logging area
Default depends on file system size.
Default is sufficient for most
workloads.
Log size can be changed after
creation using fsadm.
mkfs [fs_type] [-0 specific_options] special
-0 N
Provides information only
• Does not create the file
system
-0 largefilesl
nolargefiles
Supports files >= 2 gigabytes
(or >= 8 million files)
Default: largefiles
-0 version-n
Specifies layout version
Valid values are 4, 5, 6,
and 7.
Default: Version 7
Using mkfs Command Options
YOLI can set a variety of file system properties when you create a V[RIT!S file
system by adding VxFS-spccific options to the mkfs command.
I
6-7
Copyright if; 2006 Svmantec Corporauon. All rights reserved

symantec
Other VxFS Commands
Mount options:
mount -r ...
mount -v
mount -p
Mounts as read only
Displays mounted file systems
Displays in file system table format
(Not on Linux)
Mounts all in file system tablemount -a
Unmount options:
umount Imydata
umount -a
umount -0 force Irnydata
Unmounts a file system
Unmounts all mounted file systems
Forces an unmount
Display file system type:
fstyp -v Idev/vx/dsk/datadg/datavol
Display free space:
df -F vxfs /mydata
Identifying File System Type
If you do not know the tile system type of a particular tile system. you can
determine the tile system type by using the fstyp command. You can use the
fstyp command to describe either a mounted or unmounted tile system.
In YEA, right-dick a tile system in the object tree, and select Properties. The tile
system type is displayed in the File System Properties window.
Identifying Free Space
To report the number of free disk blocks and inodcs fur a YxFS File System, you
use the d f command. The d f command displays the number of free blocks and
free inodes in a tile system or directory by examining thc counts kept in the
superblocks. Extents smaller than XK may not be usable for all types of allocation,
so the df command docs not count tree blocks in extents below 8K when reporting
the total number of free blocks.
In YEA. right-click a file system. and select Properties to display tree space and
usage information.
6-8
COpy light 't: 2006 Svmautnc Corporation All rights reserved

,S)ll1<lntt'l
~ ..~~
Fragmentation
Degree of fragmentation depends on:
• File system usage
• File system activity patterns
IInitial Allocation I ~~ IDefragmented I
II 0 00 II
a... ..0 ... _•••• --.080.--. •••
••••• 00.. ..000
00000 ••• 00000
Fragmentation types:
• Directory
• Extent
Controlling File System Fragmentation
In a VERITAS file system, when free resources are initially allocated to files, they
are aligned in the most efficient order possible to provide optimal performance. On
an active file system, the original order is lost over time as files are created,
removed, and resized. As space is allocated and deallocated from f les, the
available free space becomes broken up into fragments. This means that space has
to be assigned to files in smaller and smaller extents. This process is known as
fragmentation. Fragmentation leads to degraded performance and availability.
VxFS provides online reporting and optimization utilities to enable you to monitor
and defragment a mounted tile system. These utilities are accessible through the
file system administration command, f sadm.
Types of Fragmentation
VxFS addresses two types of fragmentation:
Directory fragmentation: As files are created and removed, gaps are left in
directory inodes. This is known as directory fragmentation. Directory
fragmentation causes directory lookups to become slower.
Extent fragmentation: As files are created and removed, the free extent map
for an allocation unit changes from having one large tree area to having many
smaller free areas. Extent fragmentation occurs when files cannot be allocated
in contiguous chunks and more extents must be referenced to accessa file. III a
case of extreme fragmentation, a file system may have free space, none or
which can be allocated.
Lesson 6 Administering Fife Systems
Copyligh! <S.2006 Syrnantec Corporation. All nghts reserved
I
6-9

Monitoring Fragmentation
To monitor directory fragmentation:
fsadm -D frontl
Dirs Total
Searched Blocks
total 486 99
Immed Immeds Dirs to Blocks
Dirs to Add Reduce to Reduce
388 6 6 6
A high total in the Di rs to Reduce
column indicates fragmentation.
To monitor extent fragmentation:
fsadm -E fhome
in VEA, select File Syslem->
Properties->Statistics
% Free blocks in extents smaller than 64 blks: 8.35
% blks allocated to extents 64 blks or larger: 45.81
Output displays percentages of free
and allocated blocks per extent size.
Running Fragmentation Reports
You can monitor fragmentation in a VERITAS file system by running reports that
describe fragmentation levels, You use the f sadm command to run reports on
both directory and extent fragmentation. The d f command, which reports on tile
system tree space. also provides information useful in monitoring fragmentation.
Interpreting Fragmentation Reports
In general. for optimum performance. the percentage of tree space in a tile system
should not fall below 10 percent. A tile system with 10 percent or more free space
has less fragmentation and better extent allocation.
A badly fragmented file system will have one or more of the following
characteristics:
Greater than 5 percent of tree space in extents of less than 8 blocks in length
More than 50 percent of tree space in extents of less than 64 blocks in length
Less than 5 percent of the total tile system size available as free extents ill
lengths of 64 or more blocks
6-10
CopYright :;··2006 S}'ma!H€C Corporation All rights reserved

symaruec..
(';I1I'.1##J;.....'-'.
~
Defragmenting a File System
fsadm I-d] I-D] I-e] I-E] I-t time] I-p passes] mount point
During extent reorganization: During directory reorganization:
In VEA, highlight a file system, and select Actions->Defrag File System,
• Small files are made contiguous.
Large files are built from large
extents.
• Small, recent files are moved
near the Inodes.
Large, old files are moved to the
end of the allocation unit.
Free space Is clustered In the
center of the allocation unit.
Example:
fsadm -e -E -s /mntl
Valid entries are moved to the
front.
Free space is clustered in the
center of the allocation unit.
Directories are packed Into Inode
area.
Directories are placed before
other files.
• Entries are sorted by access time.
Example:
fsadm -d -D /mntl
VxFS Defragmentation
You can use the oniinc administration utility fsadm to defragment. or reorganize.
file system directories and extents. The f sadm utility detragmcnts a file system
mounted for read/write access by:
Removing unused space from directories
Making all small files contiguous
Consolidating free blocks for file system use
Only a privileged user can reorganize a file system.
Defragmenting Extents
Entries are sorted by the time of last access.
Other f sadm Defragmentation Options
If you specify both -d and -e. directory reorganization is always completed
before extent reorganization.
II'you use the - D and - E with the - d and - e options. fragmentation reports are
produced both before and after the reorganization.
You can use the - t and - p options to control the amount of work performed by
f sadm. either in a specificd time or by a number of passes. By default. f sadm
runs five passes. If both - t and - p are specified. f sadm exits if either of the
terminating conditions is reached.
Copynqtn,; 2006 Svrnantcc ccrro.auoo All rights reserved
I
6-11

symantec
Scheduling Defragmentation
• The frequency of defragmentation depends on
usage, activity patterns, and the importance of
performance.
• Run defragmentation on demand or as a cron job:
- Daily or weekly for frequently used file systems
- Monthly for infrequently used file systems
• Adjust defragmentation intervals based on reports.
• To defragment using VEA, highlight a file system
and select Actions->Defrag File System.
Scheduling Defragmentation
The best way to ensure that fragmentation does not become a problem is to
defragment the file system on a regular basis. The frequency of dcfragmeruation
depends on file system usage. activity patterns. and the importance of file system
performance. In general, follow these guidelines:
Schedule dctragmeruarion during a time when the file system is relatively idle.
For frequently used tile systems. you should schedule dctrugmentauon daily or
weekly.
For infrequently used file systems. you should schedule dcfragmenration at
least monthly,
Full tile systems tend to fragment and are difficult to defragment. You should
consider expanding the tile system.
To determine the dcfragmcniation schedule that is best for your system, select
what you think is an appropriate interval for running extent reorganization and run
the fragmentation reports both before and after the reorganization, If the degreeof
tragmcruation is approaching the bad fragmentation figures. then the interval
between fsadm runs should be reduced. If the degree offragmentation is low, then
the interval between f sadm runs can be increased.
You should schedule directory reorganization for file systems when the extent
reorganization is scheduled. The fsadm utility call run on demand and can be
scheduled regularly as a cronjob.
The dcfragmcutation process can take some time. You receive an alert when the
process is complete.

vxbench~latform -w workload [options) file name
'synmn1t'C.dl1
Testing Performance Using vxbench
Example: Sequential write
vxbench_placform -w write -i iosize-8,iocount-131072 ,
/mnt/testfileOl
total: 0.560 Bee 14623.53 KB/s cpu: 0.10 sys 0.01 user
The output disptays etapsed time in seconds, throughput in
KB/second, and CPU time for the system and the user in seconds.
Example: Random write
vxbench_platform -w rand_write -i iosize=8,
iocount=131072,maxfilesize=1048576
/mnt/testfileOl
Note: Separate suboptions using commas with no spaces.
Benchmarking Using vxbench
What Is Benchmarking'!
Benchmarking is a testing technique that enables you to measure performance
based on a set of standards. or benchmarks. You can use benchmarking techniques
to try to predict the performance of a new file system configuration or to analyze
the performance of an existing tile system.
What Is vxbench'?
VERITAS engineering developed a benchmarking tool called vxbench that
enables you to create different combinations of I/O workloads.
The vxbench program is installed as part of the VRTSspt software installation
and exists under the / opt/VRTSspt/ FS/VxBench directory.
Notes on Testing Performance
The vxbench program applies a workload to a file system and measures
performance based on how long file system operations take. II'anything else is
using the tile system at the same time. then the vxbench performance reports
are affected.
For sequential workloads: iosize x iocount = size of the file.
The ios ize and maxf ilesi ze parameters are defined in units of I K:
therefore. .i os i ze e s defines a size of8K.
Copynght If' 2006 Symantec Corporation All rights reserved
I
6-13

The vxbenchyl at form Command
In the syntax, you specify the command followed by a type of workload. Valid
workloads are:
read
write
rand read
rand write
rand mixed
mmap read
mmap_write
Performs a sequential read of the test tiles
Performs a sequential write of the test tiles
Performs a random read of the test files
Performs a random write of the test files
Performs a mix ofrandom reads and writes
Uses mmapto read the test tiles
Uses mmapto overwrite the test tiles
After specifying the type of workload. you can add specific options that
characterize the test that you want to perform.
Finally. you specify the name of the tile on which to run the test. If you specify
multiple filenames, vxbench_platform runs tests in parallel to each tile,
which simulates multiple simultaneous users. If you use the option that specifies
multiple threads. then each simulated user runs multiple threads. The total number
of 110 threads is the number of users multiplied by the number of threads.
Command Options
By adding options to the vxbench _pl atform command. you can simulate a
wide variety of I/O environments. The following table describes some of these
options and their uses. You can display a complete list ofvxbenchylatform
command options by typing vxbenchylatform -h.
Option Use
-h Prints a detailed help message
-p Uses processes for users and uses threads for multithreaded 1,'0
(This is the default option.)
-p Uses processes lor users and for multithreaded I/O
-t Uses threads for users and lor multithreaded 110
-m Locks 1:'0 butlers in memory
-s For multiuser tests. only prints summary results
-v For multithreaded tests. prints per-thread results
-k Prints throughput in kilobytes/second (This is the default option.)
-M Print, throughput in megabytes/second
-i [suboptions] SPCCitlCSsuboprions describing the test you want to perform
vxbench is included in the VRTSspt package.
6-14
Copyright ~ 2006 SYI11~-I!l!eC Corporation All rights reserved
VERITAS Storage Foundetion 5.0 for UNIX: Fundamentals

---- -'~)~;~~mlt:;
The intent log records
pending file system
changes before metadata
is changed.
Structural IntentLo
Files
Logging in VxFS
If the system
crashes, the
intent log is
replayed by
VxFS fsck.
Role of the Intent Log
A file system may be left in an inconsistent state after a system failure. Recovery
of structural consistency requires examination of file system metadara
structures.VERITAS File System provides fast file system recovery after a system
failure by using a tracking feature called ill/em /oggillg or journaling. Intent
logging is the process by which intended changes to file system metadata are
written to a log before changes are made to the file system structure. Once the
intent log has been written. the other updates to the tile system can be written in
any order. In the event of a system failure. the VxFS f sck utility replays the intent
log to nullify or complete file system operations that were active when the system
failed.
Traditionally. the length oftime taken for recovery using fsck was proportional to
the size of the file system. For large disk configurations. running f sck is a time-
consuming process that checks. verities. and corrects the entire file system.
The VxFS version of the fsck utility performs an intent log replay to recover a
file system without completing a full structural check of the entire tile system. The
time required for log replay is proportional to the log size. not the file system size.
Therefore. the file system can be recovered and mounted seconds after a system
failure. Intent log recovery is not readily apparent to users or administrators. and
the intent log can be replayed multiple times with no adverse effects.
Note: Replaying the intent log may not completely recover the damaged file
system structure if the disk suffers a hardware failure. Such situations may require
a complete system check using the VxFS fsck utility.
Lesson 6 AdministeringFile Systems
Copyright ~i;:2006 Symarrtec Corporation All nqtus reserved
I
6-15

svmarucc
Maintaining VxFS Consistency
To check file system consistency by using the intent log for
the VxFS on the volume datavol:
fsck [fs_type] /dev/vx/rdsk/datadg/datavol
To perform a full check without using the intent log:
fsck [fs_type] -0 full,nolog
/dev/vx/rdsk/datadg/datavol
To check two file systems in parallel using the intent log:
fsck [fs_type] -0 p /dev/rdsk/clt2dOs4
/dev/rdsk/cltOdOs5
To perform a file system check using the VEA GUI, highlight an
unmounted file system, and select Actions->Check File System.
Maintaining File System Consistency
You use the YxFS-specitic version of the fsck command to check the consistency
of and repair a Y xFS file system. The f sck utility replays the intent log by
default, instead of performing a full structural tile system check. which is usually
sufficient to set the tile system state to CLEAN. You can also use the f sck utility to
perform a full structural recovery in the unl ikcly event that the log is unusable.
The syntax for the f sck command is:
fsck [fs_type] [genenc opUons] [-yl-Y] [-nl-N]
[-0 full,nolog] special
For a complete list of generic options. see the f sck (1m) manual page. Some of
the generic options include:
Option Description
-m Checks. but docs not repair. a tile system before mounting
-niN Assumes a responseof no to all prompts by fsck (This option
does not replay the intent log and performs a full Isck.)
-v Echoesthe expanded command line but does not execute the
command
-yly Assumes a responseof yes to all prompts by fsck (I r the lile
system requires a lull tsck after the log replay. then a full tsck is
performed.j
- 0 p can only be run with log f sck. not with full f sck.
6-16
Copyright:: 211Ot; Syroante.. Comoraunn All notus rvserveo

, syrnanrcc.
Resizing the Intent Log
• Intent log size can be changed using fsadm:
fsadm [-F vxfsl ·0 logaic •• size [,logvol_vol_namel
mount_point Specify a Place log on a
naw log alze. separate device.
• Use the fsadm -L mount poi nt: command to get detailed
information on the current intent log
• Larger log sizes may improve performance for intensive
synchronous writes, but
may increase:
- Recovery time
- Memory requirements
- Log maintenance time
..--.-... . .
Highlight a file system, and select "
Actions->Set Intent Log Options.
File system Jmnt!
Intent logsee:
I Not on HP-UX r
Resizing the Intent Log
The VxFS intent log is allocated when the file system is first created. The size of
the intent log is based on the size of the file system-the larger the file system, the
larger the intent log.
Default log size: Based on file system size: in the range of256K to 64 MB
Default maximum log size: 64 !'vIB (Version 6 and 7 layout): 16 MB (Versions
4 and 5 layout)
With the Version 6 disk layout. you can dynamically increase or decrease the intent
lug size using the log option of the f sadmcommand. The allocation can be
directed to a specified intent logging device, as long as the device exists and
belongs to the same volume set as the file system.
Increasing the size of the intent log (an improve system performance because it
reduces the number of times the log wraps around. However, increasing the intent
log size can lead to greater times required for a log replay if there is a system
failure. / large log provides better performance on metadata-intensive workloads.
Memory requirements for log maintenance increase as the log size grows. The log
size should not be more than 50 percent of the physical memory size otthc system.
/ small log uses less space on the disk and leaves more room for file data. For
example. selling a log size smaller than the default log size may be appropriate for
a small floppy device. On small systems, you should ensure that the log size is not
greater than half the available swap space.
Note: The Loqvo I.« option 10 place the intent log on a separate volume can only
be used with multi-volume file systems (file systems on volume sets).
Copyright ~ 2006 Svmantec Corporation All nqhts reserveo
I
6-17

symantec
Logging mount Options
mount -F vxfs [-0 specific_options]
Most logging delayed; great
All structural performance improvement,
changes logged but changes could be lost
-0 log -0 tmplog
IIntegrity
. Performance
-0 delaylog
Default; some logging
delayed; improves
performance
Controlling Logging Behavior
VERITAS File System provides VxFS-spccific logging options that you can use
when mounting a file system to alter default logging behavior. By default, when
you mount a VERITAS file system, the -0 delaylog option is used with the
mount command. With this option, some system calls return before the intent log
is written. This logging delay improves the performance of the system, and this
mode approximates traditional UNIX guarantees for correctness in case of system
failures. You call specify other mount options to change logging behavior to
further improve performance at the expense of reliabi lity.
Selecting mount Options for Logging
You can add VxFS-specific mount options to the standard mount command using
- 0 in the syntax:
mount [-F vxfs] [generic_options] [-0 specific_options]
special mount_point
Logging mount options include:
• - 0 log
• -0 delaylog
• -0 tmplog
Cr;[JYrlgt1! ,0 :!006 SYJT1W'lec co-ooreuoo All r!gh!<; reserved

, symanrec.
Logging and Performance
To select the best logging mode for your environment:
• Understand the different logging options.
• Test sample loads and compare performance results.
• Consider the type of operations performed in addition to
the workload.
• Performance of 1/0 to devices can improve if writes are
performed in a particular size, or in a multiple of that size.
To specify an 1/0 size to be used for logging, use the
mount option:
-0 logiosize=size
• Place the intent log on a separate volume and disk.
Logging and VxFS Performance
In environments where data reliability and integrity is of the highest importance.
logging is essential. However, logging does incur performance overhead. If
maximum data reliability is less important than maximum performance. then you
can experiment with logging mount options. When selecting mount options for
logging to try to improve performance. follow these guidelines:
Copyright If, 2006 Symaruec Corporation All nqnts, reserved
I
6-19

syrnantec
File Change Log
The VxFS File Change Log (FCL) is another type of log that tracks changes to
tiles and directories in a tile system. Applications that can make use ofthe FCL arc
those that are typically required to scan an entire tile system to discover changes
since the last scan, such as backup utilities, wcbcrawlers, search engines, and
replication programs. The File Change Log records tile system changes such as
creates, links, unlinks, renaming. data appended, data overwritten, data truncated,
extended attribute modifications, holes punched, and other tile property updates.
Note: The FCL records only that data has changed, not the actual data. It is the
responsibility of the application to examine the tiles that have changed data to
determine which data has changed.
FCL. stores changes in a sparse file in the tile system namespace. The FCL log tile
is always located in mount_point/lost+found/changelog.
File Change Log
Tracks changes to files and directories in a file system for use
by backup utilities, webcrawlers, search engines, and
replication programs.
In contrast to the intent log, the FCL is not used for recovery.
Location: mount_point/lost+found/changelog
To activateldeactivate an FCL for a mounted file system:
fcladm on I off mount_pOint (Default is off.)
To remove an FCL (FCL must be off first):
fcladm rm mount_point
To obtain the current FCL state for a mounted file system:
fcladm state mount_point
To print the file change log:
fcladm print Olprint x mount_point
• To translate the log entries in inodes to full paths:
vxlsino inode number mount_point
Comparing the Intent Log and the File Change Log
The intent log is used to speed recovery of the tile system after a crash. The FCL
has no such role. Instead. the FCL is used to improve the performance of
applications. For example: your IT department mandates that all systems undergo
a virus scan once a week. The virus scan takes some time and your system takes a
performance hit during the scan. To improve this situation, an FCL. could be used
with the virus scanner. The virus scanner. if using an FCL, could read the log, find
all tiles on your system that arc either new or that have been modified, and scan
only those files.
FCL is used with NctBackup to greatly improve the speed of incremental backups.
Cup, IIgl11 21)06 Svmantec CorPOIi:lIIOIl All rights reserveo

, symaruec
I
Lesson Summary
• Key Points
This lesson describes how to administer file
systems using VERITAS File System (VxFS). You
learned how to defragment a file system and use
the logging capabilities in VxFS.
- VERITAS File System Administrator's Guide
'symalltec
Appendix B provides complete lab instructions and solutions. "lab (, Solulion,:
,dllllllislCring Fik S,knl"c." j1ag~B-(,7
Lab 6: Administering File Systems
• In this lab, you practice file system
administration, including defragmentation
and administering the file change log.
~ Lab Exercises, see Appendix A.
L~ Lab Solutions, see Appendix B.
Labs and solutions for this lesson arc located on the following pages:
Appendix A provides complete lab instructions. "l.ab (,: :dlllilli:':lcril1)! File>
SY'ilCIli'." I':;:,:C /-.17
Copyright ~ 2006 Symanter. Corporation. All rights reserved
6-21

Copyright ': 20()6 Svmantec Corporauon. All f191"1S re<;ervN

Lesson 7
Resolving Hardware Problems

syrnantcc.
Lesson Introduction
System
Changes
• Lesson 7: Res~/ving_,!~rdware Proble~~~
svrnantec
Topic After completing this lesson,
..
you will be able to:
Topic 1: How Does VxVM Interpret failures in hardware.
Interpret Failures in Hardware
Topic 2: Recovering Disabled Recover disabled disk groups.
Disk Groups
Topic 3:Resolving Disk Resolve disk failures.
Failures
Topic 4: Managing Hot Manage hot relocation at the host
Relocation at the Host Level level.
7-2
Copynytd ;. 2006 Svmantec Coruorauon All fights reservoo

Potential Failures in a Storage Environment
ITemporary Failures I
t
m'."·,'··'j ~
"
"
Disk
Arrays
• Power cut
• Fibre connection
failure
Complete SAN failure
• SAN switch failure
• HBA card/port failure
r:A§(i, f~ i :-LUNlDisk fail~;;--"l
/' (+~(~ (j . C~mplete disk array
~ JBOD failure
• Site Failure
,_ _...1 ,
: Can be Permanent
L~ Temporary _
How Does VxVM Interpret Failures in Hardware
YxYM interprets failures in hardware in a variety of ways. depending on the type
of failure.
I
Lesson 7 ResolvingHardwareProblems 7-3
Copyngtl! ,~~2006 Symanter; Corporation All nqbts reserved

syrnarucc
1/0 Error Handling
If the LUN/disk cannot be accessed at all, dynamic multipathing (DMP)
disables the path. If there is only one path, the DMP node is disabled.
Identifying 1/0 Failure
Disk Failure
Data availability and reliability are ensured through most failures if you are using
VxVM redundancy features, such as mirroring or RAID-5. If the volume layout is
not redundant. loss of a drive may result in loss of data and may require recovery
from backup. For 1/0 failure on a nonrcdundant volume. VxVM reports the error.
bUI it docs not take any further action.
Disk Failure Handling
When a drive becomes unavailable during an I/O operation or experiences
uncorrcctablc I/O errors, the operating system detects SCSI failures and reports
them 10 VxVM. The method that VxVM uses 10 process the SCSI failure depends
on whether the failure occurs on a nonrcdundant or a redundant volume.
FAILING vs. FAILED Disks
Volume Manager differentiates between FAILING and FAILED drives:
FAILING: Ifthere arc uncorrcctable 1/0 failures on the public region of the
drive. bUI VxVM can still access the private region of the drive. the disk is
marked as FAILING.
FAILED: IfVxVM cannot access the private region or the public region. the
disk is marked as FAILED.
The condition flags and object slates arc described in detail in the Maintenance
course.
Copyllght •. 2006 Symaotec Corporauor All rlgt1ls rcsorveo

,S)11mmC(_
Identifying Disabled Disk Groups
VxVM disk and disk group records before the failure:
r;:;-xdlsk-iist--------------------------------------------------
iDEVICE TYPE DISK GROUP STATUS
IdiSkO_l auto:cdsdisk datadgOl datadg online
ldd ak O--
2 auto:cdsdisk datadg02 datadg online
!diskO 3 auto:none online invalid
ivxdg list
!NAME STATE ID
jdatadg enabled. cds 11S0193039.S8_trainl
jvxdisk list
IDEVICE TYPE DISK GROUP
datadg
datadg
STATUS
IdiSkO 1
idiskO - 2
idiskO=3
:-;xdg list
!NAME
Idatadg
auto: cdsdisk datadgOl
auto:cdsdisk datadg02
auto
online dgdisabled
online dgdisabled
error
STATE
disabled
ID
11S0193039.S8.trainl
Identifying Disabled Disk Groups
When disk groups aredisabled. the statuschangesto dqd i s ab I ed.
I
Copyngh! (~';2nO!) Symanter. Corporanor, All rights reserved
7-5Lesson 7 Resolving Hardware Problems

syrnantec
Identifying Failed Disks
STATUS
online
online
online
vxdisk list
diskO -0 sliced rootdisk sysdg online
diskO -1 auto:cdsdisk datadgOl datadg online
diskO
-
2 auto error
diskO -3 auto error
diskO
-
4 auto error
datadg02 datadg failed was:diskO 2
Identifying Failure: Disk Records
When VxVM detaches the disk, it breaks the mapping between the VxVM
disk-s-disk media record (datadg02)-and the disk drive (diskO 2).
However. information on the disk. media record, such as the disk media name, the
disk group, the volumes, plexcs. and subdisks on the Vx VM disk, and so on, is
maintained in the configuration database in the active private regions of the disk
group.
The output ofvxdisk list displays the tailed drive as online until the
VxVM configuration daemon is forced to reread all the drives in the system and to
reset its tables.
To force the VxVM configuration daemon to reread all the drives in the system:
vxdctl enable
After you run this command, the drive status changes to error for the failed
drive, and the disk media record changes to fai led. The disk is immediately
marked as error state, when the public region is not accessible.
Ccpvnqht '; 2006 Svmaotr-r Corporation All nqhts reserved

symamec
Permanent versus Temporary Failures
• Temporary Failure
- Data on the LUN/disk is still there, only
temporarily unavailable.
- When the hardware problem is resolved, in most
cases recovery can make use of the pre-existing
data.
• Permanent Failure
- The data on the LUN/disk is completely
destroyed.
- If the volumes were not redundant, data needs
to be restored from backup.
- However, the VxVM objects and the disk group
configuration information can be restored.
Disk Failure Types
The basic types of disk failure are permanent and temporary.
Permanent disk failures are failures in which the data on the drive can no
longer be accessed for any reason (that is. uncorrectable). In this cascothe data
on the disk is lost.
Temporary disk failures are disk devices that have failures that are repaired
some lime later. This type offailure includes a drive that is powered ofland
back on. or a drive that has a loose SCSI connection that is fixed later. In these
cases. the data is still on the disk. bUI it may not be synchronized with the other
disks being actively used in a volume.
I
Lesson 7 Resolving Hardware Problems 7-7
Copyright (["':;2006 Syrnantec Corporation. All right" reserved

syrnantec ,
Device Recovery
• As soon as the hardware problem is resolved, the OS
recognizes the disk array and the disks.
• DMP automatically detects the change, adds the disk
array to the configuration, and enables the DMP paths.
This may take up to 300 seconds. If you want to make it
faster, you can execute the vxdctl enablecommand
immediately after resolving the hardware problem.
• Relevant messages are logged to the system log.
June 13 12':06:25 train1 vxdmp: [10 803759 kern.notice] NOTICB:
VxVM vxdmp V-5-0-34 added disk array D60JODDA. datype =
HDS9500-ALUA
June 13 12:06:25 train1 vxdmp: [10 736771 kern.notice] NOTICE:
VxVM vxdmp V-5-0-148 enabled path 32/0xaO belonging to the
dmpnode 253/0x10
June 13 12:06:25 train1 vxdmp: [10 899070 kern.notice] NOTICB,
VxVM vxdmp V-5-0-147 enabled dmpnode 253/0x10 Solari. Example
Recovering Disabled Disk Groups
Device Recovery
As soon as the hardware problem is resolved. the OS recognizes the disk array and
the disks, DMP automatically detects the change. adds the disk array to the
configuration. and enables the DMP paths,
Relevant messages arc logged to the system log.
CopYrlght~' 2006 Svmantec Corporation All flgnl" reserved

I
Recovering From Temporary Disk Group Failures
The disks still have their private regions. Therefore,
there is no need to recover the disk group
configuration data.
Recover the disk group as follows:
1. Unmount any disabled file systems in the disk group.
2. Deport the disk group.
3. Make sure that the DMP paths are enabled using:
4. Import the disk group.
5. Start the volumes in the disk group using:
vxvol -g diskgroup startall
Note that mirrored volumes may go through a
synchronization process at the background if they were
open at the time of the failure.
6. Carry out file system checks.
7. Mount the file systems.
Recovering From Temporary Disk Group Failures
The disks still have their private regions. so there is no need to recover the disk
group configuration data.
Recover the disk group as described in the slide.
Lesson 7 Resolving Hardware Problems
Copyrigtll 'G 2006 Symaetec Corporation. All rights reserved
7-9

Recovering From Permanent Disk Group Failures
DMP recovery is again automatically done as in
temporary failures. However, this time the disks do
not have any private region that has the disk group
configuration data.
After the DMP paths are enabled, recover the disk
group as follows:
1. Unmount any disabled file systems in the disk group.
2. Deport the disk group.
At this point all disk group information is lost except
for the configuration backups.
3. Restore the disk group configuration data.
Note that mirrored volumes will go through a
synchronization process at the background.
4. Re-create the file systems if necessary.
5. Restore data from a backup.
Recovering From Permanent Disk Group Failures
DMP recovery is again automatically done as in temporary failures. However. this
time the disks do not have any private region that has the disk group configuration
data.
After the DMP paths are enabled. recover the disk group as described in the slide.
7-10
COPYlIgh!' 2006 Svmantec Corporauoo A.IInqhts «isorveo

, symamec .
Disk Group Configuration Backup and Restore
engdg engdg
Commit
I-c diskgroup
Protecting the VxVM Configuration
The disk group configuration backup and restoration feature enables you to back
up and restore all configuration data for disk groups, and for volumes that are
configured within the disk groups. The vxconf igbackupd daemon monitors
changes to the YxYM configuration and automatically records any configuration
changes that occur.
The vxconf igbackup utility is provided for backing up and restoring a YxYM
configuration for a disk group.
The vxconf igres tore utility is provided for restoring the configuration. The
restoration process has two stages: precommit and commit. In the precomrnit
stage. you can examine the configuration of the disk group that would be restored
from the backup. The actual disk group con figuration is not permanently restored
until you choose 10 commit the changes.
By default. YxYM configuration data is automatically backed up to the files:
• /etc/vx/cbr/bk/diskgroup.dgid/dgid.dginfo
• /etc/vx/cbr/bk/diskgroup.dgid/dgid.diskinfo
• /etc/vx/cbr/bk/diskgroup.dgid/dgid.binconfig
• /etc/vx/cbr/bk/diskgroup.dgid/dgid.cfgrec
IConfiguration data from a backup enables you 10 reinstall private region headers of
YxYM disks in a disk group, re-create a corrupted disk group configuration, or re-
create a disk group and the YxYM objects within it.
This process is handled automatically by the vxconf igbackupd daemon.
Back Up
I
Precommit
1vxconfigbackup Jiskgroup
vxconfigrestore -p diskgroup
Copyright K: 2006 Symauter. Corporation, All rrghts -ese-veo
7-11

symantcc.
Disk Failure: Volume States After the Failure
vxprint -g datadg -ht
: datadg02 is the failed disk.
DG NAME NCONFIG NLOG MINORS GROUP-IO
OM NAME DEVICE TYPE PRIVLEN PUBLEN STATE
RV NAME RLINK_CNT KSTATE STATE PRIMARY DATAVOLS SRL
RL NAME RVG KSTATE STATE REM HOST REM __DG REM RLNK
V NAME RVG KSTATE STATE LENGTH READPOL PREFPLEX UTYPE
PL NAME VOLUME KSTATE STATE LENGTH LAYOUT NeOL/WID MODE
so NAME FLEX DISK DISKOFFS LENGTH [COLI] OFF DEVICE MODE
SV NAME PLEX VOLNAME NVQLLAYR LENGTH [COLIl OFF AM/NM MODE
dg datadg default default 64000 954250803.2005. train06
dm datadgOl diskO
-1 auto:cdsdiak 1519 4152640
dm datadg02 NODEVICi:
iv volOl ENABLED ACTIVE 204800 SELECT fagen
1 p L volOl-Ol volOl ENABLED ACTIVE 205200 CONCAT RW
I sd datadgOl-Ol volOl-Ol datadgOl 0 205200 0 diskO 1 ENA
p I volOl-02 volOl DISABLED NODKVICE 205200 CONCAT RW
sd datadg02-0l volOl 02 datadg02 0 205200 0 RLOC
i v vel02 DISABLED ACTIV1i 204800 SELECT fsgen
! pl vel02-0l vol02 DISABLRD NODBVICE 205200 CONCAT RW
I sd datadg02-02 velO2-0l datadg02 205200 205200 NDliV
Resolving Disk Failures
Volume States After the Failure
As soon as YxYM detects the disk failure, it detaches the disk media record from
the disk access record, the corresponding plex states change to NODEYICE as
shown 011 the slide. At this point, YxYM docs not differentiate between a
permanent failure and a temporary failure.
Cupyngtll c; 2006 Symantec Cornorauoo All ';'11,,<;-oservec
VERITAS Storage Foundation 5.0 for UNIX' Fundamentals7-12

Disk Replacement Tasks
Replacing a failed or corrupted disk involves both physically replacing the disk
and then logically replacing the disk and recovering volumes in VxVM:
Disk replacement:When a disk fails, you replace the COITuptdisk with a new
disk. The replacement disk cannot already be in a disk group. If you want to
use a disk that exists in another disk group, then you must remove the disk
from the disk group before you can use it as the replacement disk.
Volume recovery:When a disk fails and is removed for replacement. the plex
on the failed disk is disabled, until the disk is replaced. Volume recovery
involves starling disabled volumes, resynchronizing mirrors, and
resynchronizing RAID-5 parity.
After successful recovery, the volume is available for use again. Redundant
(mirrored or RAID-5) volumes can be recovered by VxVM. Nonredundant
(unmirrored) volumes must be restored from backup.
Disk Replacement Tasks
~ CDPhys;c.,Replacement
. Replace corrupt disk With
~ a new disk.
0L09;C.' Repl.cemen, 8~:'~m.
• Replace the disk in VxVM. ~ ~
• Start disabled volumes. ~ Sf!!]
• Resynchronize redundant volumes. ~
I
Copvnqht '.i;:; 200(; Svruanter. Corporation. All rights reserved
7-13

Physically Replacing a Disk
1. Connect the new disk.
2. Ensure that the operating system recognizes the disk.
3. Get VxVM to recognize the disk:
vxdctl enable
4. Verify that VxVM recognizes the disk:
Note: In VEA, use Actions->Rescan to run disk setup
commands appropriate for the OS and ensure that VxVM
recognizes newly attached hardware.
Adding a New Disk
1 Connect the new disk.
2 Get the operating system to recognize the disk:
Platform Ox-Specific Commands to Recognize a Disk
Soluris devfsadm
prtvtoc /dev/dsk/device_ name
HP-UX ioscan -fC disk
insf -e
;IX cfgmgr
lsdev -C -1 device name
Linux blockdev --rereadpt /dev/xxx
3 Get YxYM to recognize that a failed disk is now working again:
vxdctl enable
4 Verify that Yx YM recognizes the disk:
vxdisk list
After the operating system and YxYM recognize the new disk, you can then use
the disk as a replacement disk.
Note: In YEA. use the Actions=-c-Rcscun option to run disk setup commands
appropriate for the operating system. This option ensures that YxYM recognizes
newly attached hardware.

, symanrec$L,
Logically Replacing a Disk
VEA:
• Select the disk to be replaced.
• Select Actions->Replace Disk.
vxdiskadm:
"Replace a failed or removed disk"
CLl:
vxdg -k -g diskgroup adddisk disk_name=device_name
The -k option forces VxVM to take the disk media name of
the failed disk and assign it to the new disk. Use with caution.
Example:
vxdg -k -g datadg adddisk datadgOl=cltldO
Note: You may need to initialize the disk prior to running the vxdg
adddisk command: vxdisksetup -i device name
Replacing a Disk: VEA
Select the disk to be used as the new (replacement) disk.Input:
Acuons=-c-Replacc Disk
Select: The disk to be replaced
Navigation path:
YxYM replaces the disk and attempts to recover volumes.
Replacing a Failed Disk: vxdiskadm
To replace a disk that has already failed or that has already been removed. you
select the "Replace a failed or removed disk" option. This process creates a public
and private region on the new disk and populates the private region with the disk
media name of the failed disk.
Replacing a Disk: ell
IThe - k switch forces YxYM to take the disk media name of the failed disk and
assign it to the new disk. For example. if the failed disk datadgOI in the datadg
disk group was removed. and you want to add the new device cl t IdO as the
replacement disk:
vxdg -k -g datadg adddisk datadgOl=cltldO
Note: If the disk failure was temporary. the disk still has the private region that
would enable YxYM to recognize it. In this case you can use the vxreat tach
command instead of the vxdg - k adddi sk command to reattach the failed
disk.
Copy!lgl1( rc' 20U6 Symantec Corporation All rights reserved

svmaruec
Recovering a Volume
VEA:
• Select the volume to be recovered.
• Select Actions->Recover Volume.
CLl:
vxreattaeh [-ber] [device_tag]
• Reattaches disks to a disk group if disk has a transient failure,
such as when a drive is turned off and then turned back on
• -r attempts to recover stale plexes using vxreeover.
vxreeover [-bnpsvV] t-s diskgroup]
[volume_name I disk_name]
vxreeover -b -g datadg datavol
Recovering a Volume
The vxreat tach Command
The vxrea t tach utility reattaches disks to a disk group and retains the same
media name. This command attempts to find the name of the drive in the private
region and to match it to a disk media record that is missing a disk access record.
This operation may be necessary if a disk has a transient failure-for example, if a
drive is turned offand then back on. or if the Volume Manager starts with some
disk drivers unloaded and unloadable.
The vxrecover Command
To perform volume recovery operations from the command line, you use the
vxrecover command. The vxrecover program performs plex attach, RAID-S
subdisk recovery, and resynchronizc operations for specified volumes
(volume _name), or tor volumes residing on specified disks (disk_name). You
can run vxrecover any time to resynchronizc mirrors.
For example, utter replacing the tailed disk datadgOl in the datadg disk group.
and adding the new disk cltldOs2 in its place, you can attempt to recover the
volume datavol:
vxreeover -bs -g datadg datavol
To recover. in the background, any detached subdisks or plexes that resulted trom
replacement of the disk datadgOl in the datadg disk group:
vxrecover -b -g datadg datadgOl
7-16
Cupyflght 'C, 200£ SYll'iintec corcoauon. All rights reserved
VERITAS Storage Foundation 5.0 for UNIX' Fundamentals

.~ ~ .;~~~~ , SYIl1:lI1h'( .
Resolving Disk Failures - Summary
Permanent Disk Failure Temporary Disk Failure
1. Fix the hardware problem. (Replace disks, re-cable, change HBA, ... j
2 Ensure that the OS recognizes the device
3. Force VxVM to scan for added devices: vxdctl enable
4-a. Initialize a new drive.
vxdisksetup -i device_name
4-b. Attach the disk media name 4. Reattach the disk media name
to the new drive. to the disk access name.
vxdg -g diskgroup -k adddisk vxreattach
disk nemevoev i ce name
5. Recover the redundant volumes.
vxrecover
6. Start any non-redundant volumes.
vxvol -9 diskgroup -£ start volume
7. Restore non-redundant volume 7. Check data for consistency.
data from backup. fsck -F vxfs
/dev/vx/rdsk/ diskgroup/vol wne
Resolving Disk Failures: Summary
Disk failures can be resolved by following the process described in the slide.
Copyright if; 2006 Svmanter, Corporation All rights reserved
I
7-17

syrnantec
Disk Failure: Volume States After Attaching the Disk
vxprint -g datadg -ht
DG NAME NCONFIG NLOG MINORS GRQUP- 10
DM NAME DEVICE TYPE PRIVLEN PUBLEN STATE
RV NAME RLINK eNT KSTATE STATE PRIMARY DATAVOLS SRL
RL NAME RVG KSTATE STATE REM _HOST REM DG REM RLNK
PL NAME VOLUME KSTATE STATE LENGTH LAYOUT NeOL/WID MODE
SD NAME PLEX DISK DISKOFFS LENGTH [COLI] OFF DEVICE MODE
SV NAME PLEX VOLNAME NVOLLAYR LENGTH [COLI J OFF AM/NM MODE
dg datadg default defaul t 64000 95425080 3.2005. train06
dm datadgOl diskO 1 auto: cdsdisk 1519 4152640 -
dm datadg02 di.:kO_2 auto:cdadiak 1519 4152640 -
v volOl ENABLED ACTIVE 204900 SELECT fsgen
pl volOI-O! volOI ENABLED ACTIVE 205200 CONCAT RW
ad datadgO1- 01 voIOI-O! datadgOl 0 205200 0 diskO _1 ENA
p I volOl-02 volOI DlSABLRD IOPAIL 205200 CONCAT RW
ad datadg02- 01 volOl-02 datadg02 0 205200 0 diskO
-2 KNA
v vol02 DISABLBD ACTIVE 204900 SELECT fsgen
p L vo102-01 vol02 DISABLlID RECOV2R 205200 CONCAT RW
ad datadg02- 02 vo102-01 datadg02 205200 205200 0 diskO -2 ENA
Volume States After Attaching the Disk Media
After reattaching the disk. volume and plex states are as displayed in the slide.
Notice the different states of volOl and vol02. The volOl volume can still
receive I/O and contains a plcx in the IOFAIL state. This indicates that there was a
hardware failure underneath the plcx while the plcx was online.
Also notice that the only plcx ofvol02 has a state of RECOVER. This state
means that Vx VM believes that the data in this plex will need to be recovered. In a
temporary disk failure, where the disk may have been turned oil' during an I/O
stream, the data on that disk may still be valid. Therefore. do not always interpret
the RECOVER state in terms of bad data on the disk.
Copvnoht ,; 2006 Symantec Corporation All nqt-ts reserved

, synuuuec
Disk Failure: Volume States During Recovery
Ivxprint -g datadg -ht
I DG NAME NCONFIG NLOG MINORS GROUP-ID
DM NAME DEVICE TYPE PRIVLEN PUSLEN STATE
RV NAME RLINK _CNT KSTATE STATE PRIMARY DATAVOLS SRL
I RL NAME RVG KSTATE STATE REM HOST REM DG REM RLNK
I PL NAME VOLUME KSTATE STATE LENGTH LAYOUT NeOL/WID MODE
SD NAME FLEX DISK DISROFFS LENGTH (COLI] OFF DEVICE MODE
I SV NAME PLEX VOLNAME NVOLLAYR LENGTH [COLli OFF AM/NM MODE
I .,: d~tadg default default 64000 95425 08 03.2005. train06
I dm datadgOl diskO 1 auto:cdsdisk 1519 4152640 -
Ielm datadg02 diskO 2 auto: cdsdisk 1519 4152640
v volOl ENABLED ACTIVE 204800 SELECT fsgen
pl volOl-Ol volOl ENABLED ACTIVE 205200 CONCAT RW
i sd datadgOl- 01 volDl-Ol datadgOl 0 205200 0 diskO 1 ENA
I pi vo101-02 volOl ENABLED STALE 205200 CONCAT WO
, sd datadg02 - 01 volOl-02 datadg02 0 205200 0 diskO _2 ENA
Iv vol02 DISABLED ACTIVE 204900 SELECT fsgen
p L voI02 -01 voI02 DISABLBD RECOVER 205200 CONCAT RW
j sd datadg02- 02 vol02-01 datadg02 205200 205200 0 diskO 2 ENA
Volume States After Recovering Redundant Volumes
When you start the recovery on redundant volumes. the plcx that is not
synchronized with the mirrored volume has a state of ENABLED and STALE.
During the period of synchronization. the stale plex is write-only (WO). Aller the
synchronization is complete. the plcx state changes to ENABLED and ACTIVE.
and it becomes read-write (RW).
I
Copyright '1: 2006 Symautec Coero-auon. All rights reserved

symantec
Intermittent Disk Failures
• VxVM can mark a disk as failing if the disk is experiencing 1/0
failures but is still accessible.,~.----.
I~~~~~~list TYPE
1 . .-
idiskO 1 auto:cdsdisk datadgOl datadg online
IdiskO:2_ .._.. au~o.:".<::'.<:Ii.."~..~~.".~dg02 datadg online
• Disks marked as failing are not used for any new volume space.
• To resolve intermittent disk failure problems:
- If any volumes on the failing disk are not redundant, attempt to
mirror those volumes:
• If you can mirror the volumes, continue with the procedure for
redundant volumes.
• If you cannot mirror the volume, prepare for backup and restore.
- If the volume is mirrored:
• Prevent read 1/0 from accessing the failing disk by changing the
volume read policy.
Remove the failing disk.
Replace the disk.
Set the volume read policy back to the original policy.
DISK STATUSGROUP
Intermittent Disk Failure
Intermittent disk failures are failures that occur off and on and involve problems
that cannot be consistently reproduced. Therefore, these types of failures are the
most difficult for the operating system to handle and can cause the system to slow
down considerably while the operating system attempts to determine the nature of
the problem. If you encounter intcrmiucnt failures, you should move data offofthe
disk and remove the disk from the system to avoid an unexpected failure later.
The method that you use to resolve intermittent disk failure depends on whether
the associated volumes are redundant or nonrcdundanl.
7-20
Copyright ';' 2006 Sy!llar.(ec Corporation All rights resorveo

, symaruec.
Ifvolumes are performing writes and each write is taking a long time to succeed
because of the intermittent failures, then the system may slow down significantly
and fall behind in its work, II' this scenario occurs, you may need to forcibly
remove the disk and not evacuate the data:
Use the vxdiskadm option. "Remove a disk for replacement." With this
option, VxVM treats the drive as though it has already failed.
The problem with using this command is that all volumes that have only two
mirrors (or that have a RAID-5 layout for redundancy) and that are using this
drive are no longer redundant until you replace the drive. During this period, if
a bad block occurs on the remaining disk, you cannot easily recover and may
have to restore from backup. You must also restore allnonredundant volumes
using the drive from backup.
2 After you remove the drive. you must replace the drive in the same way as
when a drive completely fails. To replace a drive. you can use the vxdiskadm
option, "Replace a failed or removed disk."
Note: The state uf the disk is set to REMOVED when you use the vxdiskadm
option "Remove a disk for replacement." In terms of fixing the drive, the
REMOVED state is the same as NODEVICE. You must use the vxdiskadm option,
"Replace a failed or removed disk," to replace the drive.
Forced Removal
To forcibly remove a disk and not evacuate the data:
1, Use the vxdiskadm option, "Remove a disk for
replacement." VxVM handles the drive as if it has
already failed,
2. Use the vxdiskadm option, "Replace a failed or
removed disk."
Using the command line:
vxdg -k -g diskgroup rmdisk Idisk_namel
vxdisksetup -i Inew_device_namel
vxdg -k -g diskgroup adddisk
Idisk_namel:lnew_device_namel
Forced Removal
I
Copvriqht '.c 2006 Svmantec Corporation. All rights reservoo
7-21

symantec
What Is Hot Relocation?
Hot Relocation: The system automatically reacts to 1/0
failures on redundant VxVM objects and restores
redundancy to those objects by relocating affected
subdisks.
Spare Disks
VM Disks
Subdisks are relocated to disks designated as
spare disks or to free space in the disk group.
Managing Hot Relocation at the Host Level
What Is Hot Relocation?
Hot relocation is a feature ofYxYM that enables a system to automatically react to
I/O failures on redundant (mirrored or RAID-5) YxYM objects and restore
redundancy and access to those objects. YxYM detects I/O failures on objects and
relocates the affected subdisks. The subdisks arc relocated to disks designated as
spare disks or 10 free space within the disk group. YxYM then reconstructs the
objects thai existed before the failure and makes them redundant and accessible
again.
Note: YxYM hot relocation is applicable when working with both physical disks
and hardware arrays. For example, even with hardware arrays if you mirror a
volume aeross LUN arrays, and one array becomes unusable, it is better 10
reconstruct a new mirror using Ih..:remaining array than to do nothing.
Partial Disk Failure
When a partial disk failure occurs (that is, a failure affecting only some subdisks
on a disk), redundant data on the failed pori Ion of the disk is relocated. Existing
volumes on the unaffected portions of the disk remain accessible. With partial disk
failure, the disk is not removed from YxYM control and is labeled as FAILING,
rather than as FAILED. Before removing a FAI LING disk for replacement. you
must evacuate any rcmaining olumcs on the disk.
Note: Hot relocation is only performed for redundant (mirrored or RAID-5)
subdisks on a tailed disk. Nonrcdundant subdisks on a failed disk are not relocated,
but the system administrator is notified of the failure.

~--a;:.'·~·-----'S}l1~lnll'l.
Hot-Relocation Process
Volumes
VM Disks
1. vxrelocd detects disk failure.
2. Administrator is notified bye-mail.
3. Subdisks are relocated to a spare.
4. Volume recovery is attempted.
How Does Hot Relocation Work?
The hot-relocation feature is enabled by default. No system administrator action is
needed to start hot relocation when a failure occurs.
The vxrelocd daemon starts during system startup and monitors VxVM for
failures involving disks, plexes, or RIID-5 subdisks. When a failure occurs.
vxrelocd triggers a hot-relocation attempt and notifies the system administrator,
through e-mail, of failures and any relocation and recovery actions.
The vxrelocd daemon is started from the S95vxvm- recover file (on Solaris).
/ etc/ rc. d/ rc2 . d/ S02VXVID-r ecover file (on Linux). or / sbin/
rc2 . d/S096vxvm-recover (on IIP-UX).
The argument to vxrelocd is the list of people to e-mail notice ofa relocation
(default is root). To disable vxrelocd, you can place a "W' in front of the line in
the corresponding start-up file.
1 successful hot-relocation process involves:
1 Failure detection: Detecting the failure of a disk, plex, or RAID-5 subdisk
2 Notification: Notifying the system administrator and other designated users
and identifying the affected Volume Manager objects
3 Relocation: Determining which subdisks can be relocated, finding space for
those subdisks, and relocating the subdisks (The system administrator is
notified of the success or failure of these actions. Ilot relocation does not
guarantee the same layout of data or the same performance after relocation.)
4 Recovery: Initiating recovery procedures, if necessary. to restore the volumes
and data (Again, the system administrator is notified of the recovery auempt.)
Copyright ~ 2006 Symantec Corporation All rights reserved
I
7-23

symantec
How Is Space Selected?
• Hot relocation attempts to move all subdisks from a
failing drive to a single spare destination disk.
• If no disks have been designated as spares, VxVM
uses any available free space in the disk group in
which the failure occurs.
• If there is not enough spare disk space, a
combination of spare disk space and free space is
used.
• Free space that you exclude from hot relocation is
not used.
How Is Space Selected for Relocation?
When relocating subdisks. YxYM aucmpts to select a destination disk with the
fewest differences from the failed disk. YxYM:
Attempts to relocate to the same controller. target. and device as the failed
drive
2 Attempt to relocate to the same controller and target, but to a different device
3 Attempts to relocate to the same controller, but to any target and any device
4 Attempts to relocate to a different controller
5 Potentially scaners the subdisks to different disks
A spare disk must be initialized and placed in a disk group as a spare before it can
be used for replacement purposes.
Hot relocation attempts to move all subdisks from a failing drive to a single
spare destination disk. if possible.
If no disks have been designated as spares. YxYM automatically uses any
available free space in the disk group not currently on a disk used by the
volume.
II' there is not enough spare disk space, a combination of spare disk space and
tree space is used. Free space that you exclude li'OI11 hot relocation is not used.
In all cases, hot relocation attempts to relocate subdisks to a spare in the same disk
group, which is physically closest to the failing or failed disk.
When hot relocation occurs, the failed subdisk is removed from the configuration
database. The disk space used by the failed subdisk is not recycled as free space.
7-24
Copvnqt-t ~.2006 Svmantoc Corporauon. All right'>reserved

, symnruec.
Managing Spare Disks
VEA:
Actions->Set Disk Usage
vxdiskadm:
• "Mark a disk as a spare for a disk group"
• "Turn off the spare flag on a disk"
• "Exclude a disk from hot-relocation use"
• "Make a disk available for hot-relocation use"
ell:
To designate a disk as a spare:
vxedit -g diskCiroup set spare=onloff disk name
To exclude/include a disk for hot relocation:
vxedit -g diskgroup set nohotuse=onloff disk name
To force hot relocation to only use spare disks:
Add spare=only to /etc/default/vxassist
Managing Spare Disks
When you add a disk to a disk group, you can specify that the disk be added to the
poul of spare disks available to the hot relocation feature ofVxVM. Any disk in
the same disk group can use the spare disk. Try to provide at least one hot-
relocation spare disk per disk group. While designated as a spare, a disk is not used
in creating volumes unless you specifically name the disk on the command line.
I
Copyright © 2006 Svmantec Ccrporauon. All rights reserved
7-25Lesson 7 ResolvingHardwareProblems

symantec
Unreloeating a Disk
VEA:
• Select the disk to be unrelocated.
• Select Actions->Undo Hot Relocation.
vxdiskadm:
"Unrelocate subdisks back to a disk"
ell:
vxunreloc [-fl [-g diskgroupl [-t tasktagl
[-n disk_namel orig_disk_name
• orig_disk_Ilal1le Original disk before relocation
• -n disk name Unrelocates to a disk other
than the original
Forces unrelocation if exact
offsets are not possible
• - f
Unrelocating a Disk
The vxunreloc Utility
The hot-relocation feature detects I/O failures in a subdisk, relocates the subdisk.
and recovers the plex associated with the subdisk.
YxYM also provides a utility that unrelocatcs a disk-that is, moves relocated
subdisks back to their original disk. After hot relocation moves subdisks tj'OI11 a
failed disk to other disks. you can return the relocated subdisks to their original
disk locations after the original disk is repaired or replaced.
Unrelocaiion is performed using the vxunreloc utility. which restores the system
to the same configuration that existed before a disk failure caused subdisks to be
relocated.
7-26
Copyright G ;:006 Svmantec Corporation. AU fights reserved

,S}11Janti'l
Lesson Summary
• Key Points
This lesson described how to interpret failures in
hardware, recover disabled disk groups, resolve
disk failures, and manage hot relocation at the host
level.
,syrnalltl"
Lab 7
I
Lab 7: Resolving Hardware Problems
In this lab, you practice recovering from a
variety of hardware failure scenarios, resulting
in disabled disk groups and failed disks.
First you recover a temporarily disabled disk
group, and then you use a set of interactive
lab scripts to investigate and practice
recovery techniques.
IFor Lab Exercises, see Appendix A.
~r Lab Solutions, see Appendix B.
Appendix A provides complete lab instructions. "l.uh 7: [{c',)hill~. l lurdv arc
Problems." l)<lgl' : -~ 7
Appendix B provides complete lab instructions and solutions. "I ab 7 Solutions
[{C,,,l,'lllg I lardwarc Problems." i',lC'C H-~,"
Copyright (i;;, 2006 Symanter; Corporation All rights reserved

Copyright _ 2006 Svmautec Corporation All [191'IS reserved

A-2 VERITAS Storage Foundation 5.0 for UNIX: Fundamentals

I
~ sym.uu«;
Lab 1
For Lab Exercises, see Appendix A. JFor Lab Solutions, see Appendix B.
In this lab. you are introduced to the lab environment. the systems. and disks that
you will use throughout this course. You will also record some prerequisite
information that will prepare you for the installation ofVERITAS Storage
Foundation and the labs that follow throughout this course.
The Lab Snlilt;"lh Ior Ihi, Llh ,n'c' Inc';lIcd Oil the 1,1110illl2 p;lge:
"Lab I C;"iUlioll': Illimdllc';ilg !lIe Lab Lnviroumcut." "age 1-3
A-3Lab 1: Introducing the Lab Environment
Copyright 1-"-,2006 Symantcc Corporation. All rights reserved

Lab Environment Introduction
The instructor will describe the classroom environment, review the configuration
and layout of the systems. and assign disks for you to use. The content of this
activity depends on the type of classroom, hardware. and the operating system(s)
deployed.
Lab Prerequisites
Record the following information to be provided by your instructor:
Object Sample Value Your Value
root password veritas
Host name trainl
Domain name classrooml. int
Fully qualified hostname trainl.classrooml
(FQHI') .int
Host name of the system train2
sharing disks ith my
system
(my partner system)
ly Boot Disk: Solaris: cOt OdO
HP-UX: clt15dO
AIX: hdiskO
l.inux: hda
2nd Internal Disk: Solaris: cOt2dO
HP-UX: c3t15dO
AIX: hdiskl
l.inux: hdb
ly Data Disks: Solaris: c It #dO -
clt#d5
HP-UX c4tOdO -
c4tOd5
AIX:hdisk21-
hdisk26
Liuux: sda - sdf
Copynght .; ?()06 Svmantec Copornnon AU rights reserveu

Location of Storage /student/
Foundation 5.0 software/sf/sf50
Software:
Location of Lab Scripts: /student/labs/sf/
sf 50
Location of the fp /student/labs/sf/
program: sf50/bin
Location of VERITAS Istudent/
Storage Foundation software/license/
license keys: sf 50 -
entr -
lic.txt
Copynqht C 2006 Symantec Corporation. All rights te servr-rt
I
A-5

Instructor Classroom Setup
Perform the following steps to enable zoning configurations for the Storage
Foundation 5-day course (not required for High Availability Fundamentals):
1 Use coursesetup script: Select Classroom.
(Setup scripts are all included in Classroom SAN configuration Version 2).
Select Function To Perform:
- Select Zoning by Zone Name
2 - Select Zoning and Hostgroup Configuration by Course Name
3 - Select/Check Hostgroup Configuration
2 Select option 3 - Select/Check Hostgroup Configuration.
Select HostGroup Configuration to be Configured:
1 - Standard Mode: 2 or 4 node sharing, No DMP
2 DMP Mode: 2 node sharing, switchable between 1 path and 2
path access
3 - Check active HDS Hostgroup Configuration
3 Select option 2 - DMP Mode. Wait and do not respond to prompts.
4 Exit to first level menu,
5 Select option I - Select Zoning by Zone Name.
Select Zoning Configuration Required:
1 - Mode 1: 6 sets of 2 Systems sharing 12 LUNs, no Tape Library
available (HDS DMP Mode 6 x 2 Systems - Single Path to 12
LUNs)
Mode 2: 3 sets of 4 Systems sharing 24 LUNs, no Tape Library
available IHDS DMP Mode - 6 x 2 Systems - Dual Paths to 12 LUNs)
6 Select option I - Mode I (single path to 12 LUNs).
7 Select option 4 - Sularis as the OS.
S Exit out of the course_setup script.
9 Rebout each system using reboot - - - r.
Copyrlght'~ 2006 Symautec Corpor anon All riglil'> ies.uvoa

'symam('(.
Lab 2
command-line interface .
. For Lab Exercises, see APpendix-A:!
For Lab Solutions, see Appendix ~~
I
In this exercise. you install VERITAS Storage Foundation S.Uon yuur lab system.
You also explore the VxVM user interfaces, including the VERITAS Enterprise
Administrator interlace. the vxdiskadm menu interface. and the command line
interface.
The Lal.' S"llIilUlh Ill] this Ldl dr,' I,',·alui Oil the 1(11IuIIH2rage:
"Lab 2 S"luilOIl< IlhlalLuinil :Illd Illll'llucl·'." P:lg,' H·'
Prerequisite Setup
To perform this lab. you need a lab system with the appropriate operating system
and patch sets pre-installed. At this point there should be no Storage Foundation
software installed on the lab system. The lab steps assume that the system has
access to the Storage Foundation S.Osoftware and thai you have a Storage
Foundation S.OEnterprise demo license key that can be used during installation.
Copyright <l'.2006 Syrnantec Corporation All nqtus rese-veo
A-7

Classroom Lab Values
In preparation for this lab, you will need the following information about your lab
environment. For your reference. you may record the information here. or refer
back to the first lab where you initially documented this information.
Host name trainl
Domain name c La s s room t .int
(FQHl') .int
My Boot Disk: Solaris: cOtOdO
HP·UX: clt15dO
AIX:hdiskO
Linux: hda
Software:
Location of VERITAS /student/
license keys: sf50 - entr - lic.txt
sf50
Copynqht ',:,2006 SY'1l8111ecCornoranon All nqhts reserved

Preinstallation
Determine if there are any VRTS or SYMC packages currently installed on your
system.
I2 Before installing Storage Foundation. save the following important system
files into backup files named with a". preVM" extension. Also. save your
boot disk information to a file lor later use (do not store the file in /tmp). You
may need the boot disk information when you bring the boot disk under V.xVM
control in a later lab.
3 Are any VERITAS license keys installed on your system'! Check for installed
licenses.
4 To test if ONS is configured in your environment. check if nslookup
resolves the hostname to a fully qualified hostname by typing nslookup
hostname.If there is no ONS or if the hosmamc cannot be resolved to a fully
qualified hostname, carry out the following steps:
a Ensure that the fully qualified hostnamc is listed in the
/ ete/hosts file. For example.
eat jete/hosts
192.168.xxx.yyy tl·ain#.domain train#
where domainis the domain name used in the classroom. such as
elassrooml. into
If the fully quali ficd hostname is not in the / et e/hos t s file. add it as an
alias to hostname.
b Change to the directory containing lab scripts and execute the
prepare_ns script. This script ensures that your lab system only uses
local files for name resolution.
Lab 2 Installation and Interfaces A-9
Copyright <h 20(10Symantec cc.roreuoo All fignls resorvon

Installing VERITAS Storage Foundation
Navigate to the directory containing the Veritas Storage Foundation software.
Ask your instructor for the location of the installer script. Using the
installer script, run a prechcck to determine if your system meets all
preinstallauon requirements. If any requirements (other than the license
software not being installed) arc not met, follow the instructions to take any
required actions before you continue. Note that you can look into the log tile
created to see the details of the checks the script performs.
2 Navigate to the directory containing the Veritas Storage Foundation software.
Install and perform initial configuration of Storage Foundation (Vx VM and
VxFS) using the following steps:
a Start the installer script.
b Select I for InstaJIIUpgrade a Product option.
c Select the Veritas Storage Foundation software to install.
On the HP-UX platform, confirm that you wish to continue the
installation of this version.
d Enter the name of your system when prompted.
e Obtain a license key from your instructor and record it here. Type the
license key when prompted.
License Key:
Enter n when you are asked if you want to enter another license key.
Select to install All Veritas Storage Foundation packages when prompted.
9 Press Return to scroll through the list of packages.
h Accept the default of y to configure SF.
IIP-lIX On the HP-UX platform. the installer script starts the software
installation without asking any configuration questions. When the
software installation is complete, it prompts you to reboot your
system. Continue with the configurarion usiug . / installer
- conf igure alter the system is rebooted.
Do not set lip enclosure-based naming tor Volume Manager.
VERITAS Storage Foundation 5.0 for UNIX: FundamentalsA-10
Copvnqbt ('0 ;.1006 Symantec Corporation. All rights reserved

Do not set up a default disk group.
Obtain the domain name from your instructor and type the fully qualified
host name of your system when prompted. For example:
train5.classrooml.int
I
k Ifan error message is displayed that the fully-qualified host name could not
be queried. press return to continue.
m Do not enable Storage Foundation Management Server Management. The
system will he a standalone host.
n Select y to start Storage Foundation processes,
o Wait while the installation proceeds and processes are started.
p When the installation script completes. you will be asked to reboot your
system. Perform the next lab step (lab step 3) to modify the root profile
before rebooting your system.
q This step is only for North American Mobile Academy lab environment. If
you are working in a different lab environment. skip this step.
If you are working in a North American Mobile Academy lab environment
with iSCSI disk devices. change to the directory containing the lab scripts
and execute the iscsi_setup lab script. This script disables DMP
support for iSCS[ disks so that they can be recognized correctly by Volume
Manager.
Only if you are working in a North American Mobile Academy lab
environment:
cd /location_of_lab_scripts
./iscsi setup
3 Check in /. profile to ensure that the following paths are present.
Note: Your lab systems may already be configured with these environment
'ariable settings. However. in a real-Ii fe environment you would need to carry
out this step yourself.
4 Reboot your system.
Lab 2: Installation and Interfaces A-11
Ccpyrigl11 ~ 2006 Symantec Corporation. All riqhts reserved

Setting Up VERITAS Enterprise Administrator
Is the YEA server running" Ifnot. start it.
2 Start the YEA graphical user interface.
Note: On some systems, you may need to configure the system to use the
appropriate display. For example. if the display is pel: O. before you run
YEA, type:
DISPLAY=pcl:O
export DISPLAY
It is also important that the display itself is configured to accept connections
from your client, If you receive permission errors when you try to start YEA.
in a terminal window on the display system, type:
xhost system or xhost +
where system is the hostname of the client on which you are running the
vea command.
3 In the Select Protile window. click Manage Profiles button and contigure YEA
to always start with the Default profile.
4 Click "Connect to a Host or Domain link" and connect to your system as
root. Your instructor provides you with the password.
5 On the left pane (object tree) view. drill down the system and observe the
various categories ofYxYM objects.
6 Select the Assistant perspective on the quick access bar and view tasks for
systemname/StorageAgent.
7 Using the System perspective find out what disks are available to the OS.
8 Execute the Disk Scan command and observe the messages on the console
view. Click on a message to see the details.
9 What commands were executed by the Disk Scan task?
10 Exit the YEA graphical interface.
11 Create a root equivalent administrative account named adminl for use of
YEA.
Cop~nght ,~ 2006 Svmantec CDJpOnllOIl All rights resorvco

12 Test the new account. After you have tested the new account, exit VEA.
Exploring vxdiskadm
From the command line. invoke the text-based VxVM menu interface.
2 Display information about the menu or about specific commands.
3 What disks are available to the OS·)
4 Exit the vxdiskadm interface.
Copyright 'E 2006 Svmantec Corporation All riglll" reserved
I
A-13

Optional Lab: Accessing CLI Commands
Note: This exercise introduces several commonly used VxVM commands. These
commands and associated concepts arc explained in detail throughout this course.
If you have used Volume Manager before, you may already be familiar with these
commands. If you are new to Volume Manager, this exercise aims to show you the
amount of information you can get from the manual pages. Note that you do not
need to read all of the manual pages for this exercise.
From the command line. invoke the VxVM manual pages and read about the
vxassist command.
2 What vxassist command parameter creates a VxVM volume?
3 From the command line, invoke the VxVM manual pages and read about the
vxdisk command.
4 What disks arc available to VxVM'.'
5 From the command line, invoke the Vx VM manual pages and read about the
vxdg command.
6 How do you list locally imported disk groups'.'
7 From the command line, invoke the VxVM manual pages and read about the
vxprint command.
Optional Lab: More Installation Exploration
When does the VxVM license expire?
2 What is the version and revision number of the installed version of V x VM'.'
3 Which daemons are running after the system boots under VxVM control?
A-14 VERITAS Storage Foundation 5.0 for UNIX. Fundamentals
C[)pynf~hlc lOOA Svrnantec Cornoranon All nqhls .r-sorvco

'synmntc(.
Lab 3
interface.
For Lab Solutions, see App.!llldix ,!!._
I
In this lab. you create new disk groups. simple volumes. and file systems. mount
and unmount the file systems. and observe the volume and disk properties. The
first exercise uses the VEA interface. The second exercise uses the command line
interface.
I'hc lal' S"llllie"b lor lili, hh .uc Ipealed on the i(lllu in!! p:Jgc:
"Lab"" S,.[utiolls: (rc;rlillg;1 volume ami Fil,: SlsICIH." page H-.:'I
If you use object names other than the ones provided. substitute the names
accordingly in the commands.
Caution: In this lab. do not include the boot disk in any of the tasks.
Prerequisite Setup
To perform this lab. you need a lab system with Storage Foundation pre-installed.
configured and licensed. In addition to this. you also need four empty and unused
external disks to be used during the labs.
Note: Although you should not have to perform disk labeling. here are some tips
that may help if your disks are not properly formatted:
On Solaris, use the forma t command to place a label on any disks that are not
properly labeled for use under Solaris, Ask the instructor for details.
On Linux, if you have problems initializing a disk. you may need to run this
command: fdisk /dev/disk.
Use options - 0 and -w to write a new DOS partition table. (The disk may have
previously been used with Solaris.)
Copyright'~ 2006 Symantec Corporation All rlght<;reserved
A-15Lab 3: Creating a Volume and File System

environment. For your reference, you may record the information here, or refer
Host name trainl
'Vly Data Disks: Solaris: c lt #dO -
clt#d5
HP-UX: c4 t OdO -
c4tOd5
AIX: hdisk21-
hdisk26
l.inux: sda - sdf
Prefix to be used with name
object names
COp'rl!Jhl',~2006 Svmanter cope-anon All flghlS reserved

Creating a Volume and File System: VEA
Run and log on to the VEA interface as the root user.
I2 View all the disk devices on the system. What is the status of the disks assigned
to you for the labs?
3 Select an uninitializcd disk and initialize it using the VEA. Observe the change
in the Status column. What is the status of the disk now')
4 Create a new disk group using the disk you initialized in the previous step.
Name the new disk group namedgl.Observe the change in the disk status.
Note: If you arc sharing a disk array. make sure that the prefix you are using
for the disk group names is unique.
5 Using VEA create a new volume of size Igin namedgl.Name the new
volume namevoll. Create a fi Ie system on it and make sure that the file
system is mounted at boot time to / name1 directory.
6 Check if the file system is mounted and verify that there is an entry for this file
system in the file system table.
7 View the properties of the disk in the namedgldisk group and note the
Capacity and the Unallocated space fields.
8 Try to create a second volume, namevo12.in the namedgl and specify a
size slightly larger than the unallocated space on the existing disk in the disk
group. for example 4g in the standard Symantcc classroom systems. Do not
create a file system on the volume. What happens')
9 Add a disk to the namedgl disk group.
10 Create the same volume. namevo12.in the namedgl disk group using the
same size as in step R. Do not create a file system.
11 Observe the volumes by selecting the Volumes object in the object tree. Can
you tell which volume has a mounted file system'!
12 Create a VxFS file system on namevo12and mount it to / name2directory.
Ensure that thc lile system is not mounted at boot time. Check if the /name2
file system is currently mounted and verify that it has not been added to the file
system table.
Lab 3: Creating a Volume and File System A-17
Copyright @ 2006 Symanlec Corporation All figtlls reserved

13 Observe the commands that were executed by VEA during this section of the
lab.
Creating a Volume and File System: CLI
View all the disk devices on the system. What is the status of the disks assigned
to you for the labs?
2 Select an uninitializcd disk and initialize it using the CLI. Observe the change
in the Status column. What is the status of the disk now'?
Name the new disk group namedg2. Observe the change in the disk status.
Note: If you arc sharing a disk array, make sure that the prefix you are using
4 Using the vxassist command, create a new volume of size Ig in namedg2.
Name the new volume namevo13.
5 Create a Vcritas file system on thc namevo13 volume, mount the file system
to the / name3 directory.
Make sure that the file system is mounted at boot time.
6 Unmount the / name3 tile system. verify the unmount, and remount using the
mount -a command to mount all file systems in the file system table.
7 Identify the amount of free spacein the namedg2 disk group. Try to create a
volume in this disk group named namevo14 with a size slightly larger than
the available free space. lor example 5g on standard Symantec classroom
systems. What happens'!
Note: The disk sizes in Syrnantcc Virtual Academy lab environments are
slightly less than 2g. Ensure that you use the correct value suitable to your
environment instead of the 5g example used here.
8 Initialize a new disk and add it to the namedg2 disk group. Observe the
change in tree space.
9 Create the same volume. namevo14, in the namedg2 disk group using the
same size as in step 7.
CO~y!lght ,::;200fi Svmanrcc Corporanoo All rIghts reserved

10 Display volume information for namedg2disk group using the
vxprint -g namedg2 -htr command. Can you identify which disks arc
used for which volumes')
I11 List the disk groups on your system using the vxdg list command.
Removing Volumes, Disks and Disk Groups: eLl
Unmount the / name3file system and remove it from the file system table.
2 Remove the namevo14volume in the namedg2disk group. Observe the disk
group configuration information using the vxprint -g namedg2 -htr
command.
3 Remove the second disk (namedg2 02) from the namedg2 disk group.
Observe the change in its status.
4 Destroy the namedg2disk group.
5 Observe the status of the disk devices on the system.
Removing Volumes, Disks and Disk Groups: VEA
Unmount both / namel and / name2 file systems using YEA. Accept to
remove the file systems from the tile system table ifprompted. Check if the file
systems are unmounted and veri fy that any corresponding entries have been
removed from the file system table.
a Select the File Systems node in the object tree and select / narnel file
system.
b Select Actions->Unmount File System.
c Confirm the unmount and select Yes when prompted to remove it
from the file system table.
d Selectthe /name2 file system. SelectActions->Unmount File System.
Confirm the unmount.
Lab 3: Creating a Volume and File System A-19
Copyright 'G 2006 svmeotcc Corporation. All nqhts reserved

Both file systems should disappear from the file system list in VEA. You
can use the command line tu verify the changes as follows:
Solaris mount
cat /etc/vfstab
HP-UX. mount
Linux cat /etc/fstab
The / namel and / name2 file systems should nut be among the mounted
fill' systems. and the file system table should nut contain any entries
currespunding tu these tile systems.
2 Remove the namevo12volume in the namedgl disk group.
a Select the Volumes node in the object tree and select namevo12
volume.
b Select Actiuns->Delete Volume. Confirm when prompted.
3 Select the Disk Groups node in the object tree and observe the disks in the
namedgl disk group. Can you identify which disk is empty'?
The %Used column should show 0%for the unused disk which is the
second disk in the disk gruup (namedgl02).
4 Remove the disk yuu identified as empty from the namedg1 disk group.
Select the empty disk and select Actions->Remove Disk From Disk
Gruup.
5 Observe all the disks on the system. What is the status of the disk you removed
from the disk group')
Select the Disks node in the object tree and observe the disks in the right
pane view.
The disk removed in step 4 shuuld be in Free state.
6 Destroy the namedgl disk group.
a Select the Disk Gruups node in the object tree and the namedgl disk
group in the right pane view.
b Select Actions->Destroy Disk Group. Confirm when prompted.
7 Observe all the disks on the system. What is the status of the disks?
pane view.
If you have followed all the lab steps, you should have 4 disks in Free state;
they are already initialized but nut in a disk group.
Copyright 20(}6 Symantec Corporation AU ".!t'l~ reserveo

I
, symaruec
Lab 4
volumes.
In this lab. you create simple concatenated volumes. striped volumes. and mirrored
volumes. You also practice creating a a layered volume and using ordered
allocation while creating volumes.
The Lab Solution-, tor ihi-, lab arc locakd on the 1()llp ing J1dgc:
'Tab -l SuIUl!PIlS: S,icctlllg volume layouts." pageB-.~.
Prerequisite Setup
A-21Lab 4: Selecting Volume Layouts
Copyng"1 .:; 2006 svroentec Corporation. All fights reserved.

Object Sample Value Your Value_._--
1'001 password veritas
Host name trainl
;Iy Data Disks: Solaris: c It #dO -
clt#d5
HP-UX: c-i t odo -
c4tOd5
AIX: hdisk21-
hdisk26
Linux: sda - sdf
object names
.:., reserved

Creating Volumes with Different Layouts: CLI
Add four initialized disks to a disk group called namedg. Verify your action
using vxdisk -0 alldgs list.
Note: If you are sharing a disk array. make sure that the prefix you are using
for the disk group name is unique.
I2 Create a 50-MB concatenated volume in namedg disk group called
namevoll with one drive.
3 Display the volume layout. What names have been assigned to the plex and
subdisks?
4 Remove the volume.
5 Create a 50-MB striped volume on two disks in namedg and specify which
two disks to use in creating the volume. Name the volume namevo12.
What names have been assigned to the plex and subdisks?
6 Create a 20-MB. two-column striped volume with a mirror in namedg. Set the
stripe unit size to 256K. Name the volume namevo13.
7 Create a 20-MB. two-column striped volume with a mirror. Set the stripe unit
size to 12RK. Select at least one disk that you should not use. Name the volume
namevo14.
Was the volume created?
8 Create a 20-MB 3-column striped volume with a mirror. Specify three disks to
be used during volume creation. Name the volume namevo14.
Was the volume created')
9 Create the same volume specified in the previous step. but without the mirror.
What names have been assigned to the plcx and subdisks?
10 Remove the volumes created in this exercise.
11 Remove the disk group that was used in this exercise.
Lab 4: Selecting Volume Layouts A-23
Copyright~' 2U06 Symantec Corporation. All rights rcsorveo

Creating Volumes with Different Layouts: VEA
If you had exited out of VEA, start it and connect back to your system.
2 Add four initialized disks to a disk group called namedg. Verify your action in
the main window.
3 Create a 50-MB concatenated volume in namedg disk group called
4 Display the volume layout. Notice the naming convention ofthe plex and
subdisk.
6 Create a 50-MI3 striped volume on two disks in namedg, and specify which
View the volume.
7 Create a 20-MB, two-column striped volume with a mirror in namedg. Set the
View the volume. Notice that you now have a second plcx.
8 Create a 20-MI3, two-column striped volume with a mirror. Set the stripe unit
size to 128K. Select at least one disk you should not use. Name the volume
namevo14.
9 Create a 20-MI3 3-column striped volume with a mirror. Specify three disks to
be used during volume creation. Name the volume namevo14.
Was the volume created'!
10 Create the same volume specified in step 9, but without the mirror.
Note: If yo II did not cancel out of the previous step, then just uncheck the
mirrored option and continue the wizard.
11 Delete all volumes in the namedg disk group.
COPYrighl ,.;: 2(1(1t} Svo-anrec Corporation AIlIlY!1IS reserved

12 View the commands executed by VEA during this section of the lab.
IYou can complete this exercise using either the VEA or the CLI interface.
Solutions arc provided lor both.
Note: In order to perform the tasks in this exercise. you should have at least four
disks in the disk group that you are using.
First. ensure that any volumes created in the previous Jabs arc removed from
the namedgdisk group.
2 Create a IOO-MB Striped Mirrored volume with 110 logging. Name the volume
namevoll.
3 Create a Concatenated Mirrored volume with no logging called namevo12.
The size of the volume should be greater than the size of the largest disk in the
disk group; lor example. if your largest disk is 4 UB. then create a ()-UB
volume.
Note: If you arc working in the Virtual Academy (VA) lab environment, your
largest disk will have a size 01'2 GB. In this environment. you can use a 3GB
volume size.
4 If you are using VEA. view the commands executed by VEA to create the
namevo12 volume during this section of the lab.
5 View the volumes and compare the layouts.
6 Remove all of the volumes in the namedgdisk group.
Copyright ~ 200(; Swnanter Corporation All rights reserved

Using Ordered Allocation While Creating Volumes
You can complete this exercise using either the VEA or the CLI interface.
Solutions are provided for both.
Nute: In order to perform the tasks in this exercise, you should have at least tour
Create a 20-I1B. two-column striped volume with a mirror in the namedg disk
gruup. Name the volume namevoll.
2 Display the volume layout. How are the disks allocated in the volume'} Which
disk devices are used'?
3 Remove the volume you just made. and re-create it by specifying the four disks
in an order different from the original layout. Use the command line to create
the volume in this step.
4 Display the volume layout. How arc the disks allocated this time')
5 Remove all of the volumes in the namedg disk group.
CopynqN ,~,2006 Svmantec corooreuoo All nobts reserved

Optional Lab: Creating Volumes with User Defaults: CLI
I
This optional guided practice illustrates how to use the files:
• /ete/default/vxassist
• /ete/default/alt_vxassist
to create volumes with defaults specified by the user. Note that some of the default
values may nut apply to VEA because VEA uses explicit values for number of
columns. stripe unit size. and number of mirrors while creating striped and
mirrored volumes.
Create two files in lete/default:
a lh~ V.1. c'dilOL cTc:lil' ,I iik c'a!b! ''::':cIS,::LE;t 11;,;[ inc'ludl" lill'
Ji)!ldw!ng:
# when mirroring create three mirrors
nmirror=3
bUsing llie vi cduor.crc.uc,; tile called ,,1 v xa s si s r. Ih:1I iurludc- the
loll(llllg:
# use 256K as the default stripe unit size for
# regular volumes
stripeunit:256k
2 Use these files when creating the following volumes:
a Create a IOO-MB volume called namevoll using Layout em.i r r o r:
b Create a IOO-M8. two-column stripe volume called namevol2 using - d
al t _ vxassist so that Volume Manager uses the default file:
3 View the layout of these volumes using VEA or by using vxprint -g
namedg -htr. What do you notice?
4 Remove any vxassist default files that you created in this optional lab
section, The presence of these tiles can impact subsequent labs where default
behavior is assumed.
Copyright © 2006 Sv-nantec Corporation All rights reserved

Copynght -: 200fi Svmaotec Corpoeauon All fights reserved

I
'symant('(
Lab 5
This lab provides practice in making basic configuration changes. In this lab. you
add mirrors and logs to existing volumes. and change the volume read policy. You
also resize volumes. rename disk groups. and move data between systems.
Tile: Lab Solutiu1ls torthr- lab arc I()c<tkd PII lil,' !"II,),lIlg I'<lg,':
"L;111:, Solutions: lakinf!. fl,,,i,, Cuni"iguFlti(lli Ckln;!cs." I"lg(' [--+7
Prerequisite Setup
To perform this Jab.you need a lab system with Storage Foundation pre-installed.
configured and licensed, In addition to this. you also need fuur external disks to be
used during the labs.
At the beginning ofthis lab. you should have a disk group called namedg that has
four external disks and no volumes in it.
A-29Lab 5: Making Basic Configuration Changes
Copyngh! © 2006 Symanter; Corporation All rights reservecl

Host name trainl
sharing disks with my
system
(my partner system)
My Data Disks: Solaris: clt#dO -
clt#d5
HP-UX: c4tOdO -
c4tOd5
AIX:hdisk21-
hdisk26
Linux: sda - sdf
2nd Internal Disk: Soluris: cOt2dO
HP-UX: c3t15dO
AIX: hdiskl
Linux: hdb
Location of Lab Scripts /student/labs/sf/
(if any): sf50
object names
A-3D VERITAS Storage Foundation 5.0 for UNIX: Fundamentals
ClJpynghl'; 2006 Svmantec Corporation. All nqt-ts reserved

I
You can complete this exercise using either the YEA or the CLI interface.
Note: In order to perform the tasks in this exercise, you should have at least four
1 Ensure that you have a disk group called namedgwith four disks in it. Ir not.
create the disk group using four disks.
Note: If you have completed the previous lab steps you should already have
the namedgdisk group with four disks and no volumes.
2 Create a 50-MB, two-column striped volume called namevol1 in namedg.
3 Display the volume layout. "ow are the disks allocated in the volume'! Note
the disk devices used for the volume.
4 Add a mirror to namevol1.and display the volume layout. What is the layout
of the second plcx? Which disks are used for the second plex?
5 Add a dirty region log to namevol1and specify the disk to use lor the DRL.
Display the volume layout.
6 Add a second dirty region log to namevol1and specify another disk to use
for the DRL. Display the volume layout.
7 Remove the first dirty region log that you added to the volume. Display the
volume layout. Can you control which log was removed?
8 Find out what the current volume read policy for namevol1 is. Change the
volume read policy to round robin, and display the volume layout.
9 Remove the original mirror (namevol1-01) from namevoll. and display
the volume layout.
10 Remove namevoll.
Resizing a Volume
You can complete this exercise using either the YEA or the ell interface.
Lab 5: Making Basic Configuration Changes A-31
Copynght ~ 2006 Svmamec Corporation AlIlIgtlls reserved

If you have not already done so. remove the volumes created in the previous
lab in namedg.
2 Create a 20-MB concatenated mirrored volume called namevoll in namedg.
Create a Veritas tile system on the volume and mount it to / namel. Make sure
that the file system is not added to the tile system table.
3 View the layout of the volume and display the size of the tile system.
4 Add data to the volume by creating a file in the tile system and verify that the
fi le has been added.
5 Expand the tile system and volume to 100 MB. Observe the volume layout to
see the change in size. Display f lc system size.
Resizing a File System Only: ell
Note: This exercise should be performed using the command line interface
because the VEA docs not allow you to create a tile system smaller in size than the
underlying volume. You also cannot change the size of the volume and the tile
system separately using the GUI.
1 Create a 50-MB concatenated volume named namevo12 in the namedg disk
group.
2 Create a Veritas tile system on the volume by using the mkfs command.
Specify the tile system size as 40 MB.
3 Create a mount point / name2 on which to the mount the tile system. if it does
not already exist.
4 Mount the newly created tile system on the mount point /name2.
5 Verify disk space using the df command (or the bdf command on HP-L!X).
Observe that the total size of the tile system is smaller than the size of the
volume.
6 Expand the tile system to the full size of the underlying volume using the
f sadm -b news i ze option.
7 Verify disk space using the df command (or the bdf command on HP-UX).
COPYflgh! ;!O06 Svmuun-c Corporauoo All "Uf'l~ reserved

8 Make a lile on the file system mounted at /name2. so that the tree space is less
than 50 percent of the total file system size.
I
9 Shrink the file system to 50 percent of its current size. What happens"
1 Try to rename the namedg disk group to namedgl while the /namel and
/ name2 file systems are still mounted. Can you do it'!
2 Observe the contents of the / dev /vx/ rdsk and / dev /vx/ dsk directories
and their subdirectories. What do you see')
3 Unmount all the mounted file systems in namedg disk group.
4 Rename the namedg disk group to namedgl. Do not forget to start the
volumes in the disk group after the renaming if you are using the command
line interface.
5 Observe the contents of the / dev /vx/ rdsk and / dev /vx/ dsk directories
and their subdirectories. What has changed?
6 Observe the disk media names. Is there any change?
7 Mount the / namel and / name2 file systems. and observe their contents.
Note: If you arc sharing a disk array. each participant should make sure that the
prefix used for object names is unique.
Copy new data to the / namel and / name2 file systems. For example. copy
the jete/hosts file to /namel and the jete/group file to /name2.
2 View all the disk devices on the system.
Copyright I,,' 2006 Syrnantec Corporation All rigl11sreserved

3 Unmount all file systems in the namedgl disk group and deport the disk
group. Do not give it a new owner. View all the disk devices on the system.
4 Identify the name of the system that is sharing access to the same disks as your
system. If you are not sure. check with your instructor. Note the name of the
partner system here.
Partner system hostnamc:
5 Using the command line interface, perform the following steps on your partner
system:
Note: If you are working on a standalone system, skip step a in the following
and use your own system as the partner system.
a Remote login to the partner system.
b Import the namedgl disk group on the partner system, start the volumes in
the imported disk group. and view all the disk devices on the system.
c While still logged in to the partner system, mount the /namel and
/ name2tile systems. Note that you will need to create the mount
directories on the partner system before mounting the file systems. Observe
the data in the fi Ie systems.
d Unmount the file systems on your partner system.
e On your partner system, deport namedg1 and assign your own machine
name, for example. trainS. as the New host.
Exit from the partner system.
6 On your own system import the disk group and change its name back to
namedg. View all the disk devices on the system.
7 Deport the disk group namedg by assigning the ownership to anotherhost.
View all the disk devices un the system. Why would you do this?
8 From the command line display detailed information about one of the disks in
the disk group using the vxdi sk 1 is t device_ tag command. Note the
hos t id field in the output.
9 Import namedg. Were you successful?
A-34 VERITAS Storage Foundation 5.0 for UNIX.-Fundamentals
Copynght ~~ LOOf) Symantec Ccrporauon All nqhts reserved

10 Now import namedg and overwrite the disk group lock. What did you have to
do to import it and why?
I
11 From the command line display detailed information about the same disk in the
disk group as you did in step X using the vxdi sk 1 i st device_tag
command. Note the change in the hostid field in the output.
Copyright i; 2006 svmemec Corporation. All rigl1l5 reserved

Preparation for Defragmenting a Veritas File System Lab
A lab exercise in the next lesson requires that you run a script that sets up files with
different size extents. Because the script can take a long time to run, you may want
to begin running the script now, so that the necessary environment is created by the
next lab time.
Identify the device tag for the second internal disk on your lab system. If you
do not have a secondinternal disk or if you cannot use the second internal disk,
use one of the external disks allocated to you.
Second internal disk (or the external disk used in this lab): _
2 Initialize the second internal disk (or the external disk used in this lab) using a
non-CDS disk format.
3 Create a non-cds disk group called testdg using the disk you initialized in
step 2.
4 In the tes tdg disk group create a I-CiB conca ten atcd volume called
testvol initializing the volume space with zeros using the .i n i t.e z e r o
option to vxassist.
5 Create a VxFS file system on testvol and mount it on Ifs_test.
6 Ask your instructor for the location of the exten t s . sh script. Run the
extents. sh script.
Note: This script can take about 15 minutes to run.
7 Verify that the VRTSsptsoftware is already installed on your system. If not.
ask your instructor for the location of the software and install it.
Note: Before Storage Foundation 5.0. the VRTSsptsoftware was provided as
a separate support utility that needed to be installed by the user. With 5.0. this
software is installed as part of the product installation.
8 Ensure that the directory where the vxbench command is located is included
in your PATH definition.
Copyngh!r~:200[; SvmantecCornorauon AlIlIgh<;reserved

,S)l1mme(.
I
Lab 6
and administering the file change log.
In this lab. you practice file system administration. including dcfragmcntauon and
administering the file change log.
The Lah Solurion« fpr Ihis lab <lr,' Ipc·<lIUI Oil the li)llowing pagc·
"Lab (, Solurions: :rillliniskrlllg Fill' SV'iCJllS." I'ag.: B·h7
Setup
configured and licensed. In addition to this, you also need four external disks and
the second internal disk to be used during the labs. If you do not have a second
internal disk ur if you cannot use the second internal disk, you need five external
disks to complete the labs.
At the beginning of this lab. you should have a disk group called namedg that has
four external disks and no volumes in it. The second internal disk should be empty
and unused.
Note: If you are working in a North American Mobile Academy lab environment.
you cannot use the second internal disk during the labs. If thai is the case. select
one of the external disks 10 complete the lab steps.
Copyright (f 200G Symantec Corporation All rights reserved
A-37

environment. For your reference. you may record the information here, or refer
My Data Disks: Soluris: clt#dO -
clt#d5
HP-UX: c4tOdO -
c4tOd5
AIX: hdisk21-
hdisk26
l.inux: sda - sdf
2nd lnternal Disk: Solaris: cOt2dO
HP-UX: c3t15dO
AIX: hdiskl
Linux: hdb
Location of Lah Scripts /student/labs/sf/
(if any): sf50
object names
COl'yrll.jhl ,,;: 2006 Svmantec Corporation An fights mservec

I
Note: If you have already performed these steps at the end of the last lab, then you
can skip this section and proceed with Defragmenting a Veritas File System
section.
do not have a second internal disk or if you cannot use the second internal disk,
non-CDS disk format
3 Create a non-CDS disk group called testdg using the disk you initialized in
step 2.
4 In the testdg disk group create a I-GB concatenated volume called
testvol initializing the volume space with zeros using the ini t=zero
option to vxassist.
5 Create a VxFS file system on testvol and mount It on Ifs test.
6 Ask your instructor for the location of the extents. sh script. Run the
extents. sh script.
7 Verify that the VRTSspt software is already installed on your system. If not.
Note: Before Storage Foundation 5.0, the VRTSspt software was provided as
Lab 6: Administering File Systems A-39
Copyright <h.2006 Svmantec Corporation. All rights reserved

Defragmenting a Veritas File System
The purpose of this section is to examine the structure of a fragmented and an
untragmcntcd tile system and compare the file system's throughput in each case,
The general steps in this exercise arc:
Make and mount a file system
Examine the structure of the new file system for extents allocated
Then examine a fragmented tile system and report the degree of fragmentation
in the file system
Use a support utility called vxbench to measure throughput to specific files
within the fragmented tile system
De-fragment the tile system. reporting the degree offragmentation
Repeat executing the vxbench utility using identical parameters to measure
throughput tu the same tiles within a relatively un fragmented tile system
Compare the total throughput before and after the dcfragmcntauon process
In the namedg disk group create a I-GB concatenated volume called
namevoll.
2 Create a VxFS tile system on namevol1 and mount it on / namel.
3 Run a fragmentation report on / namel to analyze directory and extent
fragmentation. Is a newly created, empty tile system considered fragmented?
In the report. what percentages indicate a tile system's fragmentation'?
4 What is a fragmented tile system"
5 If you were shown the following extent fragmentation report about a tile
system. what would you conclude?
Directory Fragmentation Report
Dirs Total Immed Immeds Dirs to Blocks to
total
Searched Blocks Dirs to Add Reduce
199185 85482 115118 5407 5473
Reduce
5655
6 Unmount / namel and remove namevoll in the namedg disk group,
Note: The following steps will use the Ifs_test tile system to analyze the
impact of fragmentation on the tile system performance, Verify that the
extents. sh script has completed before you continue with the rest of this lab,
A--40 VERITAS Storage Foundation 5,0 for UNIX: Fundamentals
Copynqht '.;.2006 Syrnantcc Corporation All rights reserved

7 Run a fragmentation report on Ifs_test to analyze directory and extent
fragmentation. Is If s t est fragmented? Why or why nor' What should be
done"
I8 Use the Is -Te command to display the extent attributes of the tiles in the
Ifs test file system. Note that on the Solaris platform you need to use the
1s command provided by the VxFS file system software to be able to use the
-e option.
9 Measure the sequential read throughput to a particular file. for example. an
8MB file on an 8K extent (for example. Ifs_test/test48). in a
fragmented file system using the vxbench utility and record the results. Use
an 8K sequential I/O size.
Notes:
You need to use the vxbenchutility that is appropriate for the platform
you arc working on. for example vxbench_9 on Solaris 9. To identify the
appropriate vxbenchcommand. use the 1s -1 lopt/VRTSspt/FSI
VxBenchcommand. If this path is not in your PAT" environment
variable. use the fullpath of the command while running the corresponding
vxbenchutility.
Remount the file system before running each 110 test.
10 Repeat the same test for an 8Mb file on an 8Mb extent (ILX example. using the
Ifs_test/test58 file). Note that the file system must be remounted
between the tests. Can you explain why"
11 Defragment I fs _test and gather sumlllary statistics after each pass through
the file system. Aller the defragmentation completes, determine if I f s_test
is fragmented' Why or why not?
Note: The defragmentation can take about 5 minutes to complete.
12 Measure the throughput of the unfragmcnted file system using the vxbench
utility on the same files as you did in steps <) and 10. Is there any change in
throughput'.'
Notes:
You need to use the vxbenchutility that is appropriate for the platform
you are working on. for example vxbench_9 on Solaris 9. To idcnti ty the
appropriate vxbenchcommand. use the Is -1 Iopt/VRTSsptl Fsi
VxBenchcommand. If this path is not in your PATH environment
variable. use the fullpaih of the command while running the corresponding
vxbench utility.
The file system must be remounted before each test to clear the read
buffers.
If you have used external shared disks on a disk array used by other
systems for this lab. the performance results may be impacted by the disk
Copyright ([ 2()06 svmamec Corporation All rights reserved

array cache and may not provide a valid comparison between a fragmented
and defragmented file system.
13 What is the difference between an unfragmented and a fragmented file system')
14 Is anyone environment more prone to needing dcfragmcntatiou than another'!
Reading the File Change Log (FCL)
In the namedgdisk group create a new IO-MB volume called namevoll.
Create a YxFS tile system on namevoll and mount II on /fcl test
2 Turn the FCL on for /fcl test. and ensure that it is on.
3 Go to the directory that contains the FCL.
4 Display the superblock for /fcl_test.
5 How do you know that there have been no changes in the file system yet'?
6 Add some files to /fcl_test. Then remove one of the tiles you just added.
8 How do you know that changes have bcen made to the file system')
9 Print the number of the FCL.
10 Which files are identified by the inode numbers that are listed in the Create
type'!
11 Unmount the fcl_ test tile system and remove namevoll.
12 The next two lab sections are optional labs on analyzing and dcfragmenting
fragmented tile systems. If you are not planning to carry out the optional labs.
unmount / f s_tes t file system and destroy the tes tdg disk group;
otherwise. skip this step.
CcpYflgrl't 2006 Symaotec Corporation All Jlgl'1ts reserved

Optional Lab Exercises
The next set of lab exercises is optional and may be performed if YOll have time.
These exercises provide additional practice in defragmenting a file system and
monitoring fragmentation.
Optional Lab: Defragmenting a Veritas File System IThis section uses the If s_t est file system to analyze the impact of
fragmentation on the performance ota variety otl/O types on files using small and
large extent sizes.
Recreate the fragmented Ifs_test file system using the following steps:
a Unmount the Ifs test file system.
b Recreate a vxfs file system in the testvol in testdg.
c Mount the file system to Ifs_test.
d Ask your instructor for the location of the extents. sh script. Run the
extent s .sh script.
2 Run a series of performance tests for a variety ofI/O types using the vxbench
utility to compare the performance of the tiles with the RKextent size
(/fs_test/test48) and the ROOOKextent size (/fs_test/test58) by
performing the following steps.
Complete the following table when doing the performance tests.
Copyright II'; 2006 Symantec Corporation Ail rights reserved.

Test Type Time (seconds) Throughput (KB/second)
Before After Defrag Before After Defrag
llefl'ag Defrag
Sequential 2.709 .526 2953.22 15202.10
reads. XK
extent
Sequential .547 .549 14634.57 14576.20
reads. HOOOK
extent
Random 8.268 6.267 967.54 1276.53
reads. XK
extent
Random 6.541 6.468 1223.02 1236.91
reads. 80()tJK
extent
Note: Results can vary depending on the nature of the data and the model of
array used. No performance guarantees are implied by this lab.
3 Ensure that the directory where the vxbench utility is located is included in
your PATH definition.
4 Sequential 1/0 Test
Note: You must unmount and remount the file system Ifs_test before each
step to dear and initialize the butter cache.
To test the 8K extent size:
5 Random I/O Test
6 Defragment the If s_test tile system. The dcfragmentation process takes
some time.
7 Repeat the vxbench performance tests and complete the table with these
performance results.
8 Compare the results of the dcfragmcntcd tile system with the fragmented tile
system.
9 When finished comparing the results in the previous step. un mount the
Ifs_test tile system and destroy the testdg disk group.
COPYligN l' 200!,) Svmanlec Corporation All nqhts rr-served

Optional Lab: Additional Defragmenting Practice
I
In this exercise, you monitor and defragment a file system by using the fsadm
command.
Create a new 2-GB striped volume called narnevoll in narnedgdisk group.
Create a VxFS file system on narnevoll and mount it on I fs_ test.
2 Repeatedly copy a small existing tile system to If s_test using a new target
directory name each time until the target file system is approximately R5
percent full. For example, on the Solaris platform:
for i in 1 2 3
> do
> cp -r lopt Ifs test/opt$l
> done
Note: Monitor the file system size using df -k on the Solaris platform and
bdf on the I1P-UX platform, and CTRL-C out of the for loop when the file
system becomes approximately R5 percent full.
3 Delete all files in the Ifs_test file system over 10 MB in size.
4 Check the level of fragmentation in the Ifs_test tile system.
5 Repeat steps 2 and .j using values 4 5 for i in the loop. Fragmentation of
both free space and directories will result.
6 Repeat step 2 using values 6 7 for i. Then delete all files that are smaller
than 64K to release a reasonable amount of space.
7 Defragment the file system and display the results. Run fragmentation reports
both before and after the detragmentation and display summary statistics after
each pass. Compare the f sadmreport from step 4 with the final report from
the last pass in this step.
8 Unmount the If s test file system and remove the narnevoll volume used
in this lab.
Copyright Sl 2006 Syrnantec Corporation All right:'> reserved

A-46 VERITAS Storage Foundation 5.0 for UNIX Fundamentals
Copvnqet ~ 2006 Svmautec Corporation All (l{JI'I1S reserved

I
'SY111anttx
Lab 7
In this lab. you practice recovering from a variety of hardware failure scenarios.
resulting in disabled disk groups and failed disks. First you recover a temporarily
disabled disk group and then you use a set of interactive lab scripts to investigate
and practice recovery techniques. Each interactive lab script:
Sets up the required volumes
Simulates and describes a failure scenario
Prompts you to fix the problem
Finally. a set of optional labs are provided to enable you to investigate disk failures
further and to understand the behavior of spare disks and hot relocation.
The l.ab Solutio11' ror 1111'0 Jdb arc Incat,'" ,11 IlL' 1<"lpIing page.
"Lab "7 SOillli,)!lS: j(,',ol, in~ I Lm!;ll,' I'rohlcmv." p;ti!' B-~)
Prerequisite Setup
configured and licensed. In addition to this. you also need four external disks to be
Copyright It, 2006 Symantec Corporation. All rights reservec
A-47

In preparation for this lab, you will need the following information about your lah
environment. For your reference, you may record the intormation here, or refer
f--c0bject Sample Value You,' Value
lIy Data Disks: Solaris: c 1t #dO -
clt#d5
I-II'-UX: c4tOdO -
c4tOd5
AIX: hdisk21-
hdisk26
l.inux: sda - sdf
Location of Lab Scripts: /student/labs!sf/
sf50
object names
Copyrl(.lht";:· L006 Symantec Corporation All rights reserved

Recovering a Temporarily Disabled Disk Group
Remme all disks except lor one (namedgOl) from the namedg disk group.
I2 Create a Ig volume called namevoll in namedg disk group.
3 Create a file system on namevoll and mount it to / namel.
4 Copy the contents of / etc/ defaul t directory to / name1 and display the
contents of the file system.
5 Ask your instructor for the location ofthe faildg_temp script. and note the
location here:
Script location: _
6 Start writing to a file in the /namel file system at the background using the
following command:
dd if=/dev/zero of=/namel/testfile bs=1024
count=500000 &
7 In one terminal change to the directory containing the script and before the I/O
completes, execute faildg_temp namedg command.
Notes:
The faildg_temp script disables the single path to the disk in the disk
group to simulate a hardware failure. This is just a simulation and not a real
failure; therefore, the operating system will still be able to seethe disk after
the failure. The script waits until you arc ready with analyzing the failure,
to re-enable the path to the disk in the disk group.
If the 110 you started in step 6 completes before you can simulate the
failure, you can start it again to observe the I/O failure.
8 Wait lor the (/0 to fail and in another terminal observe the error displayed in
the system log.
9 Use the vxdisk -0 alldgs list and vxdg list commands to
determine the status of the disk group, and the disk.
10 What happened to the file system')
11 When you are done with analyzing the impact of the failure, change to the
terminal where the faildg_temp script is waiting and enter "e" to correct
the temporary failure.
Lab 7: Resolving Hardware Problems A-49
Copyrighte: 2006 Svmentec Corporation All rights reserved

Note: In a real failure scenario. after the hardware recovery, you would need to
first verify that the operating system can see the disks and then verify that
Volume Manager has detected the change in status. If not, you can force
VxVM to scan the disk by executing the vxdctl enable command. This
will not be necessary for this lab.
12 Assuming that the failure was due to a temporary fiber disconnection and that
the data is still intact. recover the disk group and start the volume. Verify the
disk and disk group status using the vxdisk -0 alldgs 1 is t and vxdg
1 i st commands.
13 Remount the tile system and verify that the contents are still there. Note that
you will need to perform a tile system check before you mount the tile system.
14 Unmount the file system and remove namevoll. At the end of this section
you should be left with a namedg disk group with a single disk and three
initialized disks that arc free to be used in a new disk group.
Preparation for Disk Failure Labs
Overview
The following sections use an interactive script to simulate a variety of disk failure
scenarios. Your goal is to recover from the problem as described in each scenario.
Use your knowledge ofVxVM administration. in addition to the VxVM recovery
tools and concepts described in the lesson. to determine which steps to take to
ensure recovery. After you recover the test volumes, the script verities your
solution and provides you with the result. You succeed when you recover the
volumes without corrupting the data.
For most of the recovery problems, you can use any of the VxVM interfaces: the
command line interface, the VERITAS Enterprise Administrator (VEA) graphical
user interface, or the vxdiskadm menu interface. Lab solutions are provided for
only one method. If you have questions about recovery using interfaces not
covered in the solutions, see your instructor.
Setup
Due to the way in which the lab scripts work, it is important to set up your
environment as described in this setup section:
If your system is set to use enclosure-based naming, then you must turn 011'
enclosure-based naming before running thelab scripts.
2 Create a disk group named tes tdg and add three disks (preferably of the
same size) to the disk group. Assign the following disk media names to the
disks: testdgOl, testdg02, and testdg03.

Note: If you do not have enough disks. you can destroy disk groups created in
other labs (fur example. namedg)in order to create the testdg disk group.
I
3 Before running the automated lab scripts. set the DGenvironment variable in
your root profile to the name of the test disk group that you are using:
Rerun your profile by logging out and logging back on. or manually running it.
4 Ask your instructor fur the location of the lab scripts.
Note: This lab can only be performed on Solaris. HP-UX. and Linux.
Recovering from Temporary Disk Failure
In this lab exercise. a temporary disk failure is simulated. Your goal is to recover
all of the redundant and nonredundant volumes that were on the failed drive, The
lab script run _di sks sets up the test volume configuration. simulates a disk
failure. and validates your solution for recovering the volumes. Ask your instructor
for the location of the run_disks script.
Before You Begin: Ensure that the environment variable DGis set to the name of
the testdg disk group. For example:
DG="testdg"
export DG
1 From the directory that contains the lab scripts. run the script run_disks. and
select option I. "Turned on' drive (temporary failure)":
This script sets up two volumes:
testl with a mirrored layout
test2 with a concatenated layout
Note: II'you receive an error messageabout the / image file system becoming
full during volume setup. ignore the error message. This error will not have
any impact on further lab steps or lab results.
2 Read the instructions in the lab script window. The script simulates a disk
power-off by saving and overwriting the private region on the drive that is used
by both volumes. Then. when you are ready to power the disk back on. the
script restores the private region as it was before the failure.
3 Assume that the failure was temporary. In a second terminal window. attempt
to recover the volumes.
4 After you recover the volumes, type e in the lab script window. The script
veri fics whether your solution is correct.
Cop~rlght 'E: 2006 Symanter; Corporauon All rights reserved

Recovering from Permanent Disk Failure
In this lab exercise, a permanent disk failure is simulated. Your goal is to replace
the failed drive and recover the volumes as needed. The lab script run_disks
sets lip the test volume configuration. simulates a disk failure, and validates your
solution for recovering the volumes. Ask your instructor for the location of the
run_disks script.
Before You Begin: Check to ensure that the environment variable DG is set to the
name otthe testdg disk group:
If DG is not set, set it before you continue:
DG="testdg"
export DG
From the directory that contains the lab scripts, run the script run_disks, and
select option 2, "Power failed drive (permanent failure)":
This script scts up two volumes:
Note: If you receive an error message about the / imagefile system becoming
powcr-ortby saving and overwriting the private region on the drive that is used
by both volumes. The disk is detached by VxVM.
3 In a second terminal window, replace the permanently failed drive with either a
new disk at the same SCSI location or by another disk at another SCSI
location. Then. recover the volumes.
verifies whether your solution is correct.
5 When you have completed this exercise. if the disk device that was originally
used during disk failure simulation is in onl ine inval id state, rcinitialize
the disk to prepare for later labs.
A--52 VERITAS Storage Foundation 5.0 for UNIX: Fundamentals
Copynyll ~ 2l1Uh Svmantec Corporation. All fights r.:,,>crvcd

Recovering from Intermittent Disk Failure (1)
IIn this lab exercise. intermittent disk failures are simulated. but the system is still
OK. Your goal is to move data from the failing drive and remove the failing disk.
The lab script run_disks sets up the test volume configuration and validates
your solution for resolving the problem. Ask your instructor for the location of the
run_disks script.
BeforeYou Begin:Check to ensure that the environment variable DG is set to the
name of the testdg disk group:
If it is not set. set it before you continue:
DG="testdg"
export DG
From the directory that contains the lab scripts. run the script run_disks. and
select option 3. "Intermittent Failures (system still ok)":
Note: If you receive an error messageabout the / image tile system becoming
2 Read the instructions in the lab script window. You arc informed that the disk
drive used by both volumes is experiencing intermittent failures that must be
addressed.
3 In a second terminal window. move the data on the failing disk to another disk.
and remove the failing disk.
4 After you resolve the problem. type e in the lab script window. The script
verities whether your solution is correct.
5 When you have completed this exercise. add the disk you removed from the
disk group back to the testdg disk group so that you can use it in later labs.
Copyriglll '0 2006 Svmantec Corporation. All rigtHS reserved

The next set of lab exercises is optional and may he performed if you have time.
These exercises provide additional recovery scenarios, as well as practice in
replacing physical drives and working with spare disks. A final activity explores
how to use the Support website, which is an excellent troubleshooting resource.
Optional Lab: Recovering from Intermittent Disk Failure (2)
In this optional lab exercise, intermittent disk failures are simulated, and the
system has slowed down significantly, so that it is not possible to evacuate data
from the failing disk. The lab script run_disks sets up the test volume
configuration and validates your solution for resolving the problem. Ask your
instructor for the location of the run _di sks script.
IfDG is not set, set it before you continue:
DG="testdg"
export DG
select option 4. "Intcrmirtcnt Failures (system too slow)":
tes t 1 with a mirrored layout
tes t2 with a concatenated layout
Note: If you receive an error message about the / image tile system becoming
full during olume setup. ignore the error message. This error will not have
2 Read the instructions in the lab script window. You are informed that:
The disk drive used by both volumes is experiencing intermittent failures
that need to be addressed immediately.
The system has slowed down significantly, so it is not possible to evacuate
the disk before removing it.
3 In a second terminal window. perform the necessaryactions to resolve the
problem.
Copyright ~,20()£ Syr'lilntec CnrpOfiltloP All nqnts reserved

Optional Lab: Recovering from Temporary Disk Failure - Layered
Volume
IIn this optional lab exercise. a temporary disk failure is simulated. Your goal is to
recover all of the volumes that were on the failed drive. The lab script run_disks
sets up the test volume configuration and validates your solution for resolving the
problem. Ask your instructor for the location of the run_disks script.
BeforeYou Begin:Check to ensure that the environment variable DGis set to the
If DG is not set. set it before you continue:
DG="testdg"
export DG
select option 5. "Turned off drive with layered volume":
tes t 1 with a concat-mirror layout
power-offby saving and overwriting the private region on the drive that is used
by both volumes. Then. when you arc ready to power the disk back on. the
3 Assume that the failure was temporary. In a second terminal window, attempt
4 Arter you recover the volumes. type e in the lab script window. The script
veri lies whether your solution is correct.
Copyright K 2006 Svmantec Corporancn All rights reserved

Optional Lab: Recovering from Permanent Disk Failure> Layered
Volume
In this optional lab exercise, a permanent disk failure is simulated. Your goal is to
replace the failed drive and recover the volumes as needed. The lab script
run_disks sets up the test volume configuration and validates your solution for
resolving the problem. Ask your instructor for the location of the run disks
script.
Before Yuu Begin: Check to ensure that the environment variable DG is set to the
If DG is not set, set it before you continue:
DG="testdg"
export DG
From the directory that contains the lab scripts, run the script run_di sk s, and
select option 6. "Power failed drive with layered volume":
testl with a concat-mirror layout
Note: If you receive an error message about the / image tile system becoming
full during volume setup, ignore the error message. This error will not have
by both volumes. The disk is detached by YxYM.
The rest of this lab exercise includes optional lab instructions where you perform a
variety of basic recovery operations.
VERITAS Storage Foundation 5.0 for UNIX: FundamentalsA-56

Destroy the testdg disk group and add the three disks back to the namedg
disk group, At this point you should have one disk group called namedgwith
four empty disks in it. There should be no volumes in the namedgdisk group,
II'you had destroyed the namedgdisk group in previous lab sections. re-create
it.
I
Optional Lab: Removing a Disk from VxVM Control
2 In the namedgdisk group create a IOO-MB.mirrored volume named
namevoll. Create a Veritas file system on namevol1 and mount it to
/ name1directory,
3 Display the properties of the volume, In the table. record the device and disk
media name of the disks used in this volume,
4 Remove one of the disks that is being used by the volume for replacement.
5 Confirm that the disk was removed,
6 From the command line. check that the state of one of the plexcs is DISABLED
and REMOVED,
7 [I' you are not already logged in YEA. start YEA and connect to your local
system. Check the status of the disk that has been removed.
8 Replace the disk back into the namedgdisk group.
9 Check the status of the disks. What is the status of the replaced disk')
10 Display volume information. What is the state of the plexes of namevol 1';
11 In YEA. what is the status of the replaced disk') What is the status of the
volume')
12 From the command line. recover the volume. During and after recovery. check
the status of the plex in another command window and in YEA,
Lab 7: Resotving Hardware Problems A-57
Copyright t: 2006 Symaotec Corporanon All rights reserved

Optional Lab: Replacing Physical Drives (Without Hot Relocation)
Note: If you have skipped the previous optional lab section called Removing a
Disk From VxVM Control. you may need to destroy testdg and add the three
disks back to the narnedgdisk group before you start this section. If you had
destroyed the narnedgdisk group in previous lab sections, re-create it.
Ensure that the narnedgdisk group has a mirrored volume called narnevoll
with a Veritas file system mounted on / narne1.I f not, create a 100-IIB
mirrored volume called narnevoll in the narnedgdisk group, add a VxFS
file system to the volume, and mount the tile system at the mount point
/narne1.
2 If the vxrelocd daemon is running, stop it using ps and kill, in order to
stop hot relocation from taking place. Verify that the vxrelocd processes are
killed before you continue.
Notes:
If you have executed the run_di sks script in the previous lab sections,
the vxrelocd daemon may already be killed.
There are two vxre locd processes on the Solaris platform. You must kill
both of them at the same time.
3 Next, simulate disk failure by writing over the private region using the
overwritepr script followed by vxdctl disable and vxdctl
enable commands. Ask your instructor for the location of the script.
While using the script, substitute the appropriate disk device name for one of
the disks in use by narnevoll, lor example on Linux use sbd, on Solaris and
HP-UX use cltBdO.
4 When the error occurs, view the status of the disks from the command line.
5 View the status of the volume from the command line.
6 In V EA, what is the status of the disks and volume?
Note: 011 the HP-UX platform, the vxdctl di sable command may cause
the StorageAgent used by the VEA GUI to hang. If this happens, the VEA (iUI
docs not detect the changes. Use the following command to restart the agent:
/opt/VRTSobc/pa133/bin/vxpalctrl -a StorageAgent
-c restart
7 Rcscan lor all attached disks:
A-58 VERITAS Storage Foundation 5.0 for UNIX Fundamentals
Copyngt'l~' ;W06SvrnamecCorporation All fights reserved

8 Recover the disk by replacing the private and public regions on the disk. In the
command. substitute the appropriate disk device name. lor example on Linux,
use sbd:
Note: This step is only necessary when you replace the failed disk with a brand
new one. If it were a temporary failure. this step would not be necessary.
I9 Bring the disk back under VxVM control:
10 Check the status of the disks and the volume,
11 From the command line. recover the volume.
12 Check the status of the disks and the volume to ensure that the disk and volume
are fully recovered.
13 Unmount the / namel tile system and remove the namevoll volume.
Optional Lab: Exploring Spare Disk Behavior
Note: If you have not already done so. destroy testdg and add the three disks
back to the namedg disk group before you start this section.
You should have four disks (namedgO 1 through namedg04) in the disk group
namedg. Set all disks to have the spare flag on.
2 Create a IOO-MB mirrored volume called sparevol.
Is the volume successfully created') Why or why not?
3 Attempt to create the same volume again. but this time specify two disks to
use. Do not clear any spare flags on the disks.
4 Remove the sparevol volume.
5 Verify that the relocation daemon (vxrelocd) is running. Ifnot. start it as
follows:
6 Remove the spare flags from three of the four disks.
7 Create a I OO-MB concatenated mirrored volume called spare2vol.
8 Save the output ofvxprint -g namedg -thr to a file.
Copyright,£ 2006 Svmantec Corporation All rights reserved

9 Display the properties of the spare2vol volume. In the table, record the
device and disk media name of the disks used in this volume. You are going to
simulate disk failure on one of the disks. Decide which disk you are going to
fail.
overwri tepr script followed by vxdctl disable and vxdctl
enable commands. Ask your instructor for the location of the script.
While using the script, substitute the appropriate disk device name for one of
the disks in use by spare2voL for example on Linux use sbd, on Solaris
and HP-UX use clt8dO.
11 Run vxprint -g namedg - rth and compare the output to the vxprint
output that you saved earlier. What has occurred'?
Notc: You may need to wait a minute or two for the hot relocation to complete.
12 In YEA, view the disks. Notice that the disk is in the disconnected state.
Notc: On the HP-UX platform. the vxdct 1 disable command may cause
the StorageAgent used by the YEA GUI to hang. If this happens, the YEA GUI
does not detect the changes. Use the following command to restart the agent:
-c restart
13 Run vxdisk -0 alldgs list. What do you notice'?
14 Rcscan for all attached disks.
15 In YEA. view the status of the disks and the volume.
command, substitute the appropriate disk device name, for example on Linux,
use sbd:
17 Bring the disk back under Yx YM control and into the disk group.
18 In YEA, undo hot relocation for the disk.
19 Wait until the volume is fully recovered before continuing. Check to ensure
that the disk and the volume are fully recovered.
20 Remove the spare2vol volume.
A-60 VERITAS Storage Foundation 5.0 for UNIX.' Fundamentals
Copvnqht «. 2006 svroeotec Corporation All fights reserved

Optional Lab: Using the Support Web Site
Access the latest information on VERITAS Storage Foundation.
Note: If you are working in the Virtual Academy lab environment. you may
not be able to access the Veritas Technical Support web site, because the DNS
configuration was changed during software installation by the prepare_ns
script. To restore the original DNS configuration. change to the directory
containing the lab scripts, execute the restore_ns script and try to access
the web site again.
I
2 What is the VERlTAS Support mission statement"
Hint: It is in the Support Handbook (page 3).
3 How many on-site support visits are included in a Extended Support contract"
l low about with a Business Critical Support"
Hint: In the Support l lundbook. see table on page 4 and explanation on page 5.
4 Which AIX platform is supported for Storage Foundation 5,()"
5 Access a recent Hardware Compatibility List for Storage Foundation. Which
Brocade switches are supported by VERITAS Storage Foundation and lligh
Availability Solutions 5.0 on Solaris?
6 Where would you locate the Patch with Maintenance Pack I for VERITAS
Storage Solutions and Cluster File Solutions 4.0 IlH Solaris
7 Perform this step only if you are working in the Virtual Academy lab
environment. If you have executed the restore_ns script to restore the
name resolution configuration at the beginning of this lab section in step I,
change to the directory containing the lab scripts and execute the
prepare_ns script before you continue.
If necessary:
cd /script_location
,/prepare ns
Lab 7: Resotving Hardware Problems A-51
Copyright,S: 2006 Svmantec Corporation All rights rescrvec

CGpynght 'f: 2(JI)6 Svmar-tec Cnrporanon All fights reserved

, symantee_
I
Lab 1
For Lab Exercises, see Append~
For Lab Solutions, see Append~--J
Lab 1 Solutions: Introducing the Lab Environment
In this lab. you are introduced to the lab environment. the systems. and disks thai
you will use throughout this course. You II ill also record some prerequisite
information that will prepare you for the installation ofVERITAS Storage
Foundation and the labs that follow throughout this course.
The Lab Exercises for this lab are located on the following page:
Lab 1 Solutions: Introducing the Lab Environment 8-3
Copyright ~")2006 Svnantoc Corporation All rights reserved

Lab Environment Introduction
The instructor will describe the classroom environment, review the configuration
and layout of the systems, and assign disks for you to use. The content olthis
activity depends on the type of classroom, hardware, and the operating system:s)
deployed.
Lab Prerequisites
Record the following information to be provided by your instructor:
Ohject Sample Value Your Value
Host name trainl
Domain name classrooml.int
(FQHI) .int
system
(my partner system)
'Iy Root I)isk: Solaris: cOtOdO
HP-UX: clt15dO
AIX:hdiskO
Linux: hda
~~ternal Disk: Solaris: cOt2dO
HP-UX: c3t 15dO
AIX: hdiskl
Linux: hdb
'Iy Data Disks: Solaris: c 1t #dO -
clt#d5
HJ>-UX: c4tOdO -
c4tOd5
AIX:hdisk21-
hdisk26
l.inux: sda - sdf
Cccynqnt (:. 2U06 S),Hl31l1CC Corporation All Pyt1tS roservec

Software:
sf50
Location of the fp /student/labs/sf/
program: sf50/bin
Location or V F:RITAS /student/
license keys: sf50 -
entr
-lic.txt
I
Lab 1 Solutions: Introducing the Lab Environment B-5
Copyright ~ 2006 Symantec Corporation All rights reserved

Instructor Classroom Setup
Perform the following steps to enable zoning configurations for the Storage
Foundation 5-day course (not required for High Availability Fundamentals):
1 Use course_setup script: Select Classroom.
(Setup scripts are all included in Classroom SAN configuration Version 2).
Select Function To Perform:
Select Zoning by Zone Name
2 - Select Zoning and Hostgroup Configuration by Course Name
3 - Select/Check Hostgroup Configuration
2 Select option 3 - Select/Check Hostgroup Configuration.
Select BostGroup Configuration to be Configured:
1 - Standard Mode: 2 or 4 node sharing, No DMP
2 DMP Mode: 2 node sh~ring, switch~ble between 1 path and 2
pa th access
3 - Check active HDS Hostgroup Configuration
3 Select option 2 - DMI' Mode. Wait and do not respond to prompts.
4 Exit to first level menu.
5 Select option I - Select Zoning by Zonc Name,
Select Zoning Configuration Required:
1 - Mode 1: 6 sets of 2 Systems sharing 12 LUNs, no Tape Library
available (HDS DMP Mode - 6 x 2 Systems - Single Path to 12
LUNs)
2 - Mode 2: 3 sets of 4 Systems sharing 24 LUNs, no Tape L.ibrary
available (HDS DMP Mode 6 x 2 Systems - Dual Paths to 12 LUNs)
6 Select option I - Mode I (single path to 12 LUNs).
7 Select option 4 - Solaris as the OS.
S Exit out ofthe course_setup script.
9 Reboot each system using reboot - - - 1".
B-6 VERITAS Storage Foundation 5.0 for UNIX: Fundamentals
Copyn!-)ht ~ 20()6 Svrnantec copo-auoo. AlillyhlS reserved

, symanter
I
Lab 2
command-line interface.
Fo.! Lab Solutions, see Appendix B.
Lab 2 Solutions: Installation and Interfaces
In this exercise. you install V[RITAS Storage Foundation 5.0 on your lab system.
You also explore the VxVM user interfaces. including the V[RITAS Enterprise
Administrator interface. the vxdiskadm menu interface. and the command line
interface.
Prerequisite Setup
To perform this lab. you need a lab system with the appropriate operating system
and patch sets pre-installed. At this point there should be no Storage Foundation
software installed on the lab system. The lab steps assume that the system has
access to the Storage Foundation 5.0 software and that you have a Storage
Foundation 5.0 Enterprise demo license key that can be used during installation.
Lab 2 Solutions Installation and Interfaces B-7
COp)'rigt11 r,;:.2006 Syrnaruec Corporation. All righls reserved

Host name trainl
Domain name classrooml,int
Fully qualified hostname trainl,classrooml
(FQIII) ,int
1y Boot Disk: Solaris: cOtOdO
HP-UX: clt15dO
AIX: hdiskO
Linux: hda
Foundation 5,0 soft ware/ sf / sf50
Software:
Location ofVERITAS /student/
license keys: sf 50
-entr -
lic,txt
Location of Lah Scripts: /student/labs/sf/
sf50
B-8 VERITAS Storage Foundation 5.0 for UNIX Fundamentals
Copyngt11G 2006 Svroaruec COfpOr(lIW'" Ill! "llnts reserved

Preinstallation
Determine if there are any VRTS or SYMC packages currently installed on your
system.
Solaris pkginfo I grep -i VRTS
pkginfo I grep -i SYMC
HP-lJX swlist -1 product I grep VRTS
swlist -1 product I grep SYMC
Note: If you have chosen to install the VxVM bundle that
comes with the IIiv2 operating system software. you will see
versions 3.) ofVERITAS Volume Manager software including
the VEA packages.
f--
AIX lslpp -1 'VRTS* '
lslpp -1 'SYMC*'
Linux rpm -qa I grep VRTS
rpm -qa I grep SYMC
2 Before installing Storage Foundation. save the following important system
files into hackup files named with a '". preVM" extension. Also. save your
hoot disk information to a file for later use (do not store the file in /tmp). You
may need the boot disk information when you bring the boot disk under VxVM
control in a later lab.
Solaris ep fete/system /ete/system.preVM
ep /ete/vfstab /ete/vfstab.preVM
prtvtoe /dev/rdsk/boot -
disk device -
name >
/ete/bootdisk.preVM
IW-lJX ep /stand/system /stand/system.preVM
ep /ete/fstab /ete/fstab.preVM
AIX ep /ete/filesystems /ete/filesystems.preVM
ep /ete/vfs /ete/vfs.preVM
Linux ep /ete/grub.eonf /ete/grub.eonf.preVM
ep /ete/modules.eonf /ete/modules.eonf.preVM
3 Are any VERlTAS license keys installed on your system'? Check for installed
licenses.
vxlierep
Note: The vxlicrep utility may not be available on your system at this
point.
Lab 2 Solutions: Installation and Interfaces
Copyligl1t r£; 2006 Symantec Corporation All rigtlls reserved
I
8-9

4 To test if DNS is configured in your environment, check if nslookup
resolves the hostnamc to a fully qualified hostname by typing nslookup
hostname. If there is no DNS or if the host name cannot be resolved to a fully
qualified hostname. carry out the following steps:
a Ensure that the fully qualified hostname is listed in the
jete/hosts tile. For example,
cat jete/hosts
192.168.xxx.yyy train#.domain train#
where domain is the domain name used in the classroom, such as
elassrooml. into
If the fully qualified hosinamc is not in the jete/hosts tile. add it as an
alias to hostname.
b Change to the directory containing lab scripts and execute the
prepare _ns script. This script ensures that your lab system only uses
local files for name resolution.
./prepare ns
Copynghl ,,~.2006 Symantec Corporation All nqhb reserved

Installing VERITAS Storage Foundation
Navigate to the directory containing the Vcritas Storage Foundation software.
Ask your instructor for the location of the installer script. Using the
installer script. run a precheck to determine if your system meets all
preinstallation requirements. If any requirements (other than the license
software not being installed) are not met. follow the instructions to take any
required actions before you continue, Note that you can look into the log file
created to see the details of the checks the script performs.
cd /software_location
./installer -precheck system
where system is the hostname of your lab system.
Select the option number for Vcrltas Storage Foundation when prompted.
I
2 Navigate to the directory containing the Veriras Storage Foundation software.
Install and perform initial configuration of Storage Foundation (VxVM and
VxFS) using the following steps:
a Start the installer script.
./installer
b Select I for Install/Upgrade a Product option.
c Select the Veritas Storage Foundation software to install.
On the I IP-UX platform. confirm that you wish to continue the
installation of this version.
d Enter the name of your system when prompted.
e Obtain a license key from your instructor and record it here. Type the
license key when prompted.
License Key:
Enter n when you are asked if you want to enter another license key.
Select to install All Veritas Storage Foundation packages when prompted.
9 Press Return to scroll through the list of packages.
Lab 2 Solutions: Installation and Interfaces 8-11
Copynqhtt'. 2006 Symantec Cnrporatioe All rights reserved

h Accept the default ory to configure SF.
HP-liX On the HP-UX plauonn. the installer script starts the software
installation without asking any configuration questions. When the
softwure installation is complete. it prompts you to reboot your
system. Continue with the eonliguration using. / installer
- con figure utter the system is rebooted.
cd /
shutdown -ry now
Alter reboot:
./installer -configure system
Do not set up enclosure-based naming for Volume Manager.
Do not set up a default disk group.
kif an error message is displayed that the tully-qualified host name could not
be queried, press return to continue.
Obtain the domain name from your instructor and type the fully qualified
host name of your system when prompted. For example:
train5.classrooml.int
m Do not enable Storage Foundation Management Server Management. The
system will be a standalone host.
n Select y to start Storage Foundation processes.
o Wait while the installation proceeds and processes arc started.
p When the installation script completes, you will be asked to reboot your
system. Perform the next lab step (lab step 3) to modify the root profile
before rebooting your system.
q This step is only for North American Mobile Academy lab environment. If
you arc working ill a different lab environment, skip this step.
If you are working in a North American Mobile Academy lab environment
with iSCSI disk devices, change to the directory containing the lab scripts
and execute the iSCSl setup lab script. TIllS SCript disables DMP
support fur iSCSI disks so that they can be recognized correctly by Volume
Manager.
Only if you arc working in a North American Mobile Academy lab
environment:
COPYlIgtlt .~ 2006 Symanrec Corporation All nqnts reserved

./iscsi setup
3 Check in /. profile to ensure that the following paths are present.
Note: Your lab systems may already be configured with these environment
variable settings. l lowever, in a real-li fe environment you would need to carry
out this step yourself
Solaris. PATH=$PATH:/usr/lib/vxvm/bin:/opt/VRTSob/bin:
AIX /usr/sbin:/opt/VRTS/bin:/opt/VRTSvxfs/sbin
MANPATH=$MANPATH:/opt/VRTS/man
export PATH MANPATH
Linux PATH=$PATH:/usr/lib/vxvm/bin:/opt/VRTSob/bin:
/usr/sbin:/opt/VRTS/bin:/opt/VRTSvxfs/sbin
MANPATH=$MANPATH:/opt/VRTS/man
export PATH MANPATH
MANSECT=$MANSECT:lm ; export MANSECT
IW·UX PATH=/usr/lib/vxvm/bin:/opt/VRTSob/bin:
/opt/VRTS/bin:/usr/sbin:
/usr/lbin/fs/vxfs4.1:$PATH
export PATH
I
4 Reboot your system.
Solaris shutdown -y -i6 -gO
HI'·LJX No needto reboot your system becauseit hasalready been rebooted
after the package installation.
Lab 2 Solutions Installation and Interfaces 8-13
C0pyright ih 2006 svrneotec Corporation AI! rights reserved

Setting Up VERITAS Enterprise Administrator
Is the YEA server running'! lfnot, start it.
Solar-is, vxsvc -m (to confirm that the server is running)
HI'-UX, vxsvc (if the SenTI" is not already running)
AIX
Linux vxsvcctrl status (to confirm that the server is running)
vxsvcctrl start (if the server is not already running)
2 Start the YEA graphical user interface.
vea &
Note: On some systems, you may need to configure the system to use the
appropriate display. For example, if the display is pel: O.before you run
YEA, typc:
DISPLAY=pcl:0
export DISPLAY
It is also important that the display itself is configured to accept connections
from your client. If you receive permission errors when you try to start YEA,
in a terminal window on the display system, type:
xhost system or xhost +
where system is thc hostname of the client on which you are running the
vea command.
3 In the Select Profile window. click Manage Profiles button and configure YEA
to always start with the Default profile.
Set "Start VEA using profile" option to Default and click Close, then click
OK to continue.
4 Click "Connect to a Host or Domain link" and connect to your system as
root. Your instructor provides you with the password.
Hostname: (For example, train13)
lJsername: root
Password: (Your instructor provides the password.)
5 On the left pane (object tree) view. drill down the system and observe the
various categories of VxYM objects.
6 Select the Assistant perspective on the quick access bar and view tasks for
systemname/StorageAgent.
7 Using the System perspective find out what disks are available to the OS.
Copynght l;; 200fi Svmantec Corporanon All rights reserved

In the System perspective object tree, expand your host and the
StorageAgent, then selectthe Disks node. Examine the Device column in
the grid.
8 Execute the Disk Scan command and observe the messages on the console
view. Click on a message to see the details.
In the VEA System perspective object tree, select your host. Select
Actions->Rescan.
I9 What commands were executed by the Disk Scan task')
Navigate to the Log perspective. Select the Task Log tab in the right pane
and double-click the "Scan for new disks" task.
10 Exit the VEA graphical interface.
In the VEA main window, select Filc->Exit. Confirm when prompted.
Lab 2 Solutions: Installation and Interfaces 8-15
Copyright'i" 2006 Symantec Corporation All lighTs resorveo

11 Create a root equivalent administrative account named adminl for use of
YEA.
Solaris,
Linux
ur-ux
AIX
I. Create a new administrative account named adminl:
useradd adminl
passwd adminl
2. Type a password for adminl.
3 Modify the jete/group file to add the vrtsadm group and
specify the root and adminl user, by using the vi editor:
vi /ete/group
4. In the tile. move to the location where you want to insert the
vrt sadmentry. change to insert mode by typing .i , then add the line:
vrtsadm::99:root,adminl
5. When you are finished editing. press [Esc] to leave insert mode.
6. Then. save the tile and quit:
:wq
I. Create a newadministrative account named adminl by using
SAM or command line utilities:
useradd adminl
passwd adminl
2. Type a password lor adminl.
3. Add the vrtsadm group and specify the root and adminl users
as members. Use SAM or modify the jete/group file by using the
v i editor:
vi /ete/group
4. In the file. move to the location where you want to insert tile
vrtsadm entry. change to insert mode by typing i, then add the
line:
vrtsadm::99:root,adminl
5. When you are finished editing. press [Esc] to leave insert mode.
11. Then, save the tile and quit:
:wq!
mkgroup -A vrtsadm
useradd -m -G vrtsadm adminl
passwd adminl
(Type the password.)
12 Test the new account. After you have tested the new account, exit YEA.
a Launch VEA:
vea
b Select "Connect tu a Host or Domain", and specify the host name:
Hustname: (For example, train13)

c Select the "Connect using a different user account" option and click
Connect.
d Enter the username and password for the new user:
User: adminl
Password: (Type the password that you created for adminl.)
e After confirming the account, select File->Exit.
Exploring vxdiskadm IFrom the command line. invoke the text-based Vx YM menu interface.
vxdiskadm
2 Display information about the menu or about specific commands.
Type? at any of the prompts within the interface.
3 What disks are available to the OS?
Type list at the main menu, and then type all.
4 Exit the vxdiskadm interface.
Type q at the prompts until you exit vxdiskadm.
Lab 2 Sotutions: tnstallation and Interfaces 8-17
Cnnvnqtu t. 2006 Svrnantec Corporation All riqt1f;; reservort

Optional Lab: Accessing CLI Commands
Note: This exercise introduces several commonly used VxVM commands. These
commands and associated concepts are explained in detail throughout this course.
If you have used Volume Manager before. you may already be familiar with these
comma lids. If you are new to Volume Manager, this exercise aims to show you the
amount of information you can gct from the manual pages. Note that you do not
need to read all of the manual pages for this exercise.
From the command line. invoke the VxVM manual pages and read about the
vxassist command.
man vxassist
2 What vxassist command parameter creates a VxVM volume?
The make parameter is used in creating a volume.
3 From the command line. invoke thc VxVM manual pages and read about the
vxdisk command.
man vxdisk
4 What disks are available to Vx VM"
All the available disks are displayed in the list.
5 From the command line. invoke thc VxVM manual pages and read about the
vxdg command.
man vxdg
6 How do you list locally imported disk groups'?
vxdg list
7 From the command line. invoke the VxVM manual pages and read about the
vxp ri nt; command.
man vxprint
B-18 VERITAS Storage Foundation 5,0 for UNIX Fundamentals
Copyright 'f 20011Svmantec Corporation All nqhts reserved

Optional Lab: More Installation Exploration
When docs the Vx VM license expire"
vx1icrep I more
2 What is the version and revision number of the installed version ofVx v'M?
Solaris pkginfo -1 VRTSvxvm
In the output. look at the Version field.
--
ur-ux sw1ist I grep -i vxvm
The version is in the second column ofthe output.
AIX lslpp -1 VRTSvxvm
In the output. look under the column named Level.
Linux rpm -qi VRTSvxvm
3 Which daemons are running after the system boots under Vx VM control"
Solaris ps -eflgrep -i vx
vxconfigd, vxrelocd, vxnotify, vxcached,
vxesd, vxconfigbackupd, vxsvc, vxpal,
vxsmf.bin
HI'-lJX ps -ef I grep -i vx
vxconfigd, vxrelocd. vxnotify, vxcached,
vxesd, vxconfigbackupd, vxsvc, vxfsd, vxiod,
vxpal, vxsmf.bin
Lab 2 Solutions: Installation and Interfaces B-19
Copvuqht 'f: 2006 Symantec Corporation All rights roservoc
I

COPYright 2D06 Swnantec Corporauoo All fI~hlS reserved

'S}111illitCC.
Lab 3
interface.
I
Lab 3 Solutions: Creating a Volume and File System
In this lab. you create new disk groups. simple volumes. and file systems. mount
and unmount the file systems. and observe the volume and disk properties. The
first exercise uses the VEA interface. The second exercise uses the command line
interface.
If you use object names other than the ones provided. substitute the names
accordingly in the commands.
Caution: In this lab. do not include the boot disk in any of the tasks.
Prerequisite Setup
Note: Although you should not have to perform disk labeling. here are some tips
that may help if your disks are not properly formatted:
On Solaris, use the format command to place a label on any disks that are not
properly labeled for use under Solaris. Ask the instructor for details.
On Linux, if you have problems initializing a disk. you may need to run this
command: fdi sk / dev / di sk.
Use options - 0 and -w to write a new DOS partition table. (The disk may have
previously been used with Solaris.)
Copyngh!"!;) 2006 Symantec Corporauon All rights reservoc
8-21Lab 3 Solutions: Creating a Volume and File System

In preparation for this lab. you will need the following information about your lab
~ '-{)
, "
Host name trainl
-'l,,~ Data Disks: Solaris: c 1t #dO -
clt#d5
HP-UX: c4tOdO -
c4tOd5
AIX:hdisk21-
hdisk26
Linux: sda - sdf
object names
Copyright .; 2006 Svmar-tec Comoranon AU nqt-ts resorvoo

Creating a Volume and File System: VEA
Run and log on to the YEA interface asthe root user.
vea &
2 View all thedisk deviceson thesystem.What is thestatusolthc disks assigned
to you for the labs"
a lising the system perspective (StorageAgent view), drill down the
object tree, and select the Disks node.
b View the disks in the grid.
Normally the disks should be in Not I nitialized state.
I3 Selectan uninitialized disk and initialize it using the YEA, Observethechange
in the Statuscolumn. What is the statusof the disk now
a Select the disk in the grid, and select Actions->Initialize Disk.
b Verify the selecteddisk in the Initialize Disk view and click OK.
The status of the disk should change to Free.
4 Createa new disk group using thedisk you initialized in the previous step.
Name the new disk group namedg1. Observethe changein the disk status.
Note: If you are sharinga disk array.makesurethat the prefix you are using
for the disk group namesis unique.
a Select the newly initialized disk in the grid, and select Actions->New
Disk Group.
b In the New Disk Group wizard, click Next to skip the Welcome page.
c Type the name of the disk group. Ensure that Enable Cross-platform
Data Sharing (CDS) remains checked. If necessary,make changes to
the selecteddisks, and click Next.
d Confirm the disk selection.
e Do not select a disk group organization principle when prompted.
Click Finish.
The status ofthe disk should change to Imported and the disk media name
and the disk group name should be visible in the disk grid.
5 Using YEA createa new volume of size Ig in namedg 1. Name the new
volume namevol1. Createa file systemon it and makesurethat the file
systemis mountedat bout time to / name1 directory.
a Select the Volumes node in the object tree and select Actions->New
Volume.
b In the New Volume Wizard, click Next on the welcome page.
c Select the disk group name and click Next.
Lab 3 Solutions: Creating a Volume and File System 8-23
Copyright:s 2006 Svmantec corpcrauon. All lights reservec.

d Let volume manager decide what disks to use for this volume, and
click Next to continue.
e Enter volume name and size, and leave the other options at their
default values. Click Next to continue.
Leave the "Create as a Snapshot Cache Volume" option unchecked
and click Next.
9 On the Create File System page, select to create a VxFS me system.
Enter the mount point called / narnel and verifv that Add to file
system table and (for Solarts) Mount at buot options are checked.
Click Next.
h Verify the summary information, and click Finish.
6 Check if the tile system is mounted and verify that there is an entry for this file
system in the tile system table,
Select the File Systems node in the object tree and observe the list of
mounted me systems in the right pane view. The / narnel me system
should be listed here. Note the "Mounted" and "In File System Table"
columns.
You can also use the command line to verify the changes as follows:
Solaris mount
cat /etc/vfstab
HP-lIX, mount
The / narnel file system should show as mounted and there should be a
line in the me system table to ensure that it is mounted at boot time.
7 View the properties of the disk in the namedgl disk group and note the
Capacity and the Unallocated space fields,
Select Disks in the object tree, right click the disk in the narnedgl disk
group, and select Properties.
8 Try to create a second volume. namevo12.in the namedgl and specify a
size slightly larger than the unallocaicd space on the existing disk in the disk
group. for example 4g in the standard Symantec classroom systems, Do not
create a tile system on the volume, What happens?
a Select the Volumes node in the object tree and select Actions->New
Volume.
c Select the disk group name and click Next.
d Let volume manager decide what disks to use for this volume, click
Next to continue.
CO[Jyugt11E 2006 Svmantec Corporauo».All nqhts r-sorveo

default values, click Next to continue.
Leave the Create as a Snapshot Cache Volume option unchecked. and
click Next.
9 On the Create File System page leave the "No tile system" option
checked and click Next.
You should receive an error indicating that Volume Manager cannot
allocate the requested spacefor the volume, and the volume is not created.
9 Add a disk to the namedgl disk group.
a Select the disk to be addcd to the disk group.
b Select Actions->Add Disk to Disk Group.
e Click Next on the Welcome page, verifv that the namedgl disk group
is selectedand that the disk is listed under Selecteddisks, and click
Next.
d Confirm "hen prompted.
e Verify the summary information and click Finish.
10 Create the same volume. namevo12.in the namedgl disk group using the
same size as in step X. Do not create a tile system.
a Select the Volumes node in the object tree and select Actions->Ncw
Volume.
e Select the disk group name and click Next.
d Let volume manager decide what disks to use for this volume, click
Next to continue.
default values, click Next to continue.
Leave the Create as a Snapshot Cache Volume option unchecked. and
click Next.
9 On the Create File System page leave the "No tile system" option
checked and click Next.
This time the volume creation should complete successfully.
11 Observe the volumes by selecting the Volumes object in the object tree. Can
you tell which volume has a mounted file system')
Select Volumes node in the object tree.
In the right pane view you should notice that the tile system and mount
point columns have tile system information for namevoll and not for
namevo12.
Copyriqht it; 2006 Symantec Corporation All rights reserved
I
8-25

12 Create a VxFS file system on namevo12and mount it to / name2 directory.
Ensure that the file system is not mounted at boot time. Check if the / name2
tile system is currently mounted and verify that it has not been added to the file
system table.
a Select the namevol2 volume and select Actions->File System->
New File System.
b Verity that the file system type is vxfs, enter the mount point, uncheck
the "Add to tile system table" option, and click OK.
c Select the File Systems node in the object tree and observe the list of
mounted file systems in the right pane view. The / name2 file system
should be listed here. Note the Mounted and In File System Table
columns.
You can also use the command line to verify the changes as follows:
Solaris mount
cat /etc/vfstab
BP-lJX, mount
The / name2 file system should show as mounted but there should be no
change in the file system table.
13 Observe the commands that were executed by VEA during this section ofthe
lab.
a Select the Logs perspective in the quick accessbar.
b Click Task Lug in the right pane view.
c Observe the commands executed by VEA during this section of the lab
by double clicking the individual tasks and observing the Task Details
view.
Creating a Volume and File System: CLI
View all the disk devices on the system. What is the status of the disks assigned
to you for the labs')
If you completed the first section of this lab, you should have two disks in
namedgl in online status. The rest of the disks assigned to you should
be in online invalid status. If you have a disk in error status,
contact your instructor.
B-26 VERITAS Storage Foundation 5.0 for UNIX. Fundamentals
Copynqht '<', 2006 Symautec corooreuoo. All nqhls reserved

2 Select an uninitialized disk and initialize it using the CLI. Observe the change
in the Status column. What is the status of the disk now'!
vxdisksetup -i device_tag
where device_ tag is c#t#d# for Solaris and IIP-UX, hdisk# for AIX
and sd# for Linux platforms.
The status of the disk should change to online but the DISK and
CROUP columns should still be empty.
Name the new disk group namedg2. Observe the change in the disk status.
Note: If you are sharing a disk array. make sure that the prefix you are using
vxdg init namedg2 namedg20I=device_tag
where devi ce_ tag is c#t #d# for Solaris and HP-UX, hdisk# for AIX
The status of the disk is still online but the [)ISK and CROUP columns
now show the new disk media name and the disk group name respectively.
4 Using the vxassist command. create a new volume of size Ig ill namedg2.
Name the new volume namevo13.
vxassist -g namedg2 make namevo13 Ig
5 Create a Vcritas file system on the namevo13 volume. mount the file system
to the / name3 directory.
mkfs -F vxfs /dev/vx/rdsk/namedg2/namevo13
Note: On LiIllIX, use mkf s - t.
mkdir /name3
mount -F vxfs /dev/vx/dsk/namedg2/namevo13 /name3
Note: On Llnux, use mount -to
Copyright I[) 2006 Symantec. Corporation All nqhts. reserved
I
B-27

Make sure that the tile system is mounted at boot time.
Solaris vi /etc/vfstab
...
/dev/vx/dsk/namedg2/namevo13
/dev/vx/rdsk/namedg2/namevo13
/name3 vxfs o yes -
UP-LJX vi /etc/fstab
...
/dev/vx/dsk/namedg2/namevo13 /name3 vxfs
rw,largefiles,delaylog 0 2
6 Unmount the / name3 tile system, verify the un mount, and remount using the
mount -a command to mount all tile systems ill the file system table.
umount /name3
mount
mount -a
mount
7 Identity the amount of free space in the namedg2 disk group. TIY to create a
volume in this disk group named namevo14 with a size slightly larger than
the avai lablc free space. for example 5g on standard Symantcc classroom
systems. What happens'?
Note: The disk sizes in Symanicc Virtual Academy lab environments are
slightly less than 2g. Ensure that you use the correct value suitable to your
environment instead of the 5g example used here.
vxdg -g namedg2 free
The free space is displayed in sectors in the LENGTH column.
vxassist -g namedg2 make namevo14 5g
You should receive an error indicating that Volume Manager cannot
allocate the requested space for the volume, and the volume is not created.
8 Initialize a new disk and add it to the namedg2 disk group. Observe the
change in free space.
vxdg -g namedg2 adddisk namedg202=device_tag
where device tag is c#t#d# for Solaris and HP-UX,
hdisk# for AIX and sd# for Linux platforms.
vxdg -g namedg2 free
9 Create the same volume. namevo14, in the namedg2 disk group using the
same size as in step 7.
vxassist -g namedg2 make namevo14 5g
VERITAS Storage Foundation 5,0 for UNIX: Fundamentals8-28
Copyright _~'200G S)"IWlO>tE'C CorpOf(:l!iuli All fights reserved

Note: The 5g volume size is usedas an example here. You may need to use
a value more suitable to your lab environment if you arc not working in a
standard Symantec classroom.
This time the volume creation should complete successfully.
10 Display volume information for namedg2 disk group using the
vxprint -g namedg2 -htr command. Can you identify which disks arc
used for which volumes?
vxprint -g namedg2 -htr
I11 List the disk groups on your system using the vxdg list command.
vxdg list
If you have followed the labs so far, you should have two disk groups listed
namedgl and namedg2.
Lab 3 Solutions: Creating a Volume and File System B-29
Copyright IE 2006 Svmamec Corporation All rights reserved

Removing Volumes, Disks and Disk Groups: ell
Unmount the / name3file system and remove it from the file system table.
Solaris umount /name3
vi /etc/vfstab
Navigate to the line with the entry corresponding to the /name3 lile
system and type dd 10 delete the line.
Type :wq 10 saveand close the tile.
HP-UX, umount /name3
Linux vi /etc/fstab
Navigate 10 the line with the entry corresponding to the /name3 file
system and type dd 10 delete the line.
Type :wq 10 save and close the file.
2 Remove the namevo14volume in the namedg2disk group. Observe the disk
group configuration information using the vxprint -g namedg2 -htr
command.
vxassist -g namedg2 remove volume namevol4
vxprint -g namedg2 -htr
There should be only namevol3 volume, and the second disk,
namedg202, should be unused.
3 Remove the second disk (namedg202)from the namedg2disk group.
Observe the change in its status.
vxdg -g namedg2 rmdisk namedg202
Note that the disk is still ill online state; it is initialized.
4 Destroy the namedg2disk group.
vxdg destroy namedg2
5 Observe the status of the disk devices on the system.
CopynytH~ 2006 SvmantecCon.orauon Alillyn~ reserved

Removing Volumes, Disks and Disk Groups: VEA
Unmount both / namel and / name2 file systems using YEA. Accept to
remove the file systems from the file system table ifprompted. Check if the file
systems are unmounted and veri fy that any corresponding entries have been
removed from the file system table.
a Select the File Systems node in the object tree and select / namel file
system.
c Confirm the unmount and select Yes when prompted to remove it
from the file system table.
d Selectthe / name2 tile system. SelectActions->Unmount File System.
Confirm the unmount.
Both tile systemsshould disappear from the file system list in VEA. You
can usethe command line to verify the changesas follows:
Solaris mount
cat /etc/vfstab
U1'-l'X. mount
The / namel and / name2 file systemsshould not be among the mounted
tile systems,and the tile system tahle should not contain any entries
corresponding to thesetile systems.
2 Remove the namevo12 volume in the namedgl disk group.
a Select the Volumes node in the object tree and select namevol2
volume.
b Select Actions->Delete Volume. Cnnflrm when prompted.
3 Select the Disk Groups node in the object tree and observe the disks in the
namedgl disk group. Can you identify which disk is empty')
The %Used column should show 0% for the unused disk which is the
seconddisk in the disk group (namedgl02).
4 Remove the disk you identified as empty from the namedgl disk group.
Select the empty disk and select Actions->Remove Disk From Disk
Group.
5 Observe all the disks on the system. What is the status of the disk you removed
from the disk group?
pane view.
Copyrigh10 2006 swoeotec Corporation. All !ighls reserved
I
8-31

The disk removed in step 4 should be in Free state.
6 Destroy the namedgl disk group.
a Select the Disk Croups node in the object tree and the namedgl disk
group in the right pane view.
b Select Actions->Destroy Disk Croup. Confirm when prompted.
7 Observe all the disks on the system. What is the status of the disks?
pane view.
If you have followed all the lab steps, you should have 4 disks in Free state;
they are already initialized but not in a disk group.
Copyngtll e. 20U6 SYIl1<1l1le{; Corpor auon A,II flgtlt<: reservert

'symanl<x.
I
Lab 4
volumes.
Lab 4 Solutions: Selecting Volume Layouts
In this lab. you create simple concatenated volumes. striped volumes. and mirrored
volumes. You also practice creating a a layered volume and using ordered
allocation while creating volumes.
Prerequisite Setup
To perform this lab. you need a lab system with Storage Foundation pre-installed,
Copvnqtn b 2006 Symarrter; Cmpor3MI1 All righlS reserved
8-33

environment. For your reference. you may record the information here, or refer
Oh.jeet Sample Value Your Value
Host name trainl
My Data Disks: Sularis: c It #dO -
clt#d5
IIP-UX: c4 t OdD -
c4tOd5
AIX:hdisk21-
hdisk26
l.inux: sda - sdf
object flames
Ccpynqnt ';' I'U[}6 Svn.antec Corporauon All lights rcscrveo

Creating Volumes with Different layouts: Cli
Add four initialized disks to a disk group called namedg. Verify your action
using vxdisk -0 alldgs list.
Note: If you arc sharing a disk array. make sure that the prefix you are using
for the disk group name is unique.
a If you have completed the Creating a Volume and File System lab (lab
3), you should already have four initialized disks. If not, initialize four
disks for use in Volume Manager:
where devi ce_ tag is c#t#d# for Solaris and HP-UX, hdisk# for
AIX and sd# for Linux platforms.
(Do the above command for anv disks that have not beeninitialized for
Volume Manager useand that will be used in this lab.)
b Create a new disk group and add disks:
vxdg init namedg namedgOl=devicel_tag
namedg02=device2_tag namedg03=device3_tag
namedg04=dev~ce4 tag
Alternatively, you cau also create the disk ~roup using a single disk device
and then add each additional disk as follows:
vxdg -g namedg adddisk namedg##=device_tag
vxassist -g namedg make namevoll SOm
3 Display the volume layout. What names have been assigned to the plex and
subdisks?
To view the assigned names, view the volume using:
vxprint -g namedg -thr I more
vxassist -g namedg remove volume namevoll
5 Create a 50-MB striped volume on two disks in namedg and specify which
vxassist -g namedg make namevol2 SOm layout=stripe
namedgOl namedg02
What names have been assigned to the plex and subdisks?
To view the assigned names,view the volume using:
vxprint -g nallledg -thr I more
Copyright'C 2006 Syrnaruec Corporation, A.IIri9nt>. reserved
I
8-35

6 Create a 20-MB, two-column striped volume with a mirror in namedg. Set the
vxassist -g namedg make namevol3 20m
layout=mirror-stripe neol=2 stripeunit=256k
What do you notice about the plexes?
View the volume using vxprint -g namedg -thr I more.
Notice that you now have a second plex.
7 Create a 20-MB. two-column striped volume with a mirror. Set the stripe unit
size to 128K. Select at least one disk that you should not use. Name the volume
namevo14.
vxassist -g namedg make namevo14 20m
layout=mirror-stripe neol=2 stripeunit=128k
lnamedg03
This operation should fail because there are not enough disks available in
the disk group. A two-column striped mirror requires at least four disks.
8 Create a 20-MB 3-column striped volume with a mirror. Specify three disks to
be used during VOIUlllC creation. Name thc volume namevo14.
vxassist -g namedg -b make namevol4 20m
layout=mirror-stripe neol=3 namedgOl namedg02
namedg03
Again, this operation should fail because there are not enough disks
available in the disk group. At least six disks are required for this type of
volume configuration.
9 Create the same volume specified in the previous step, but without the mirror.
vxassist -g namedg -b make namevol4 20m layout=stripe
neol=3 namedgOl namedg02 namedg03
What names have been assigned to the plcx and subdisk s?
To view the assigned names, view the volume using:
vxprint -g namedg -thr I more
10 Remove the volumes created in this exercise.
vxassist -g namedg remove volume namevol2
vxassist -g namedg remove volume namevo14
11 Remove the disk group that was used in this exercise.
vxdg destroy namedg
Copyngllt t; 2006 Svmantec Corporatmn AlIl!ghts reserveo

Creating Volumes with Different Layouts: VEA
If you had exited out orVEA. start it and connect back to your system.
vea&
2 Add four initialized disks to a disk group called namedg. Verify your action in
the main window.
a In the System perspective, drill down to the Disks node in the object
tree.
b Select a disk, and select Actions->New Disk Group.
e In the New Disk Group wizard, skip the welcome page,specify the disk
group name, selectthe disks you want to usefrom the Available Disks
list, and click Add.
d Click Next, confirm your selection, do not select any Organization
Principle, and click Finish.
a Select the namedg disk group, and select Actions->New Volume.
b In the New Volume wizard, let VxVI1 determine which disks to use.
e Type the name of the volume, and specify a size of 50 MB. Verify that
the Concatenated layout is selectedin the Layout region.
d Complete the wizard by accepting all remaining defaults to create the
volume.
4 Display the volume layout. Notice the naming convention ofthe plcx and
subdisk.
a Select the volume in the object tree, and select Actions->Volume
View.
b In the Volume View window, click the Expand button. Compare the
information in the Volume View window to the information under the
Mirrors, Logs, and Subdisks tabs in the right pane of the main
window.
a Select the volume, and select Actions->Dclete Volume.
b In the Delete Volumc dialog box, click Yes.
6 Create a 50-MB striped volume on two disks in namedg. and specify which
b In the New Volume wizard, select"Manually selectdisks to usefor this
volume." Move two disks into the Included box, and then click Next.
Copyright © 2006 Svrnantec Corporation. All nqhts reserved
I
B-37

e Type the name of the volume, and specify a size of 50 MB.
d Select the Striped option in the Layout region. Verify that the number
of columns is 2.
e Complete the wizard by accepting all remaining defaults to create the
volume,
View the volume.
a Select the volume, and select Actions->Volume View.
b Close the Volume View window when you arc satisfied.
7 Create a 20-MI3. two-column striped volume with a mirror in namedg. Set the
b In the New Volume wizard, let VxVM determine which disks to use.
of columns is 2. Set the Stripe unit size to 256K (512 sectors on Solaris,
AIX, and Linux, 256 sectors on HP-lJX).
e Mark the Mirrored check box in the Mirror Info region.
Complete the wizard by accepting all remaining defaults to create the
volume,
View the volume. Notice that you now have a second plcx.
a Select the volume, and select Actions->Volume View.
b Close the Volume View window when you arc satisfled.
8 Create a 20-MI3. two-column striped volume with a mirror. Set the stripe unit
size to 128K. Select at least one disk you should nor use. Name the volume
namevo14.
b In the New Volume wizard, select "Manually select disks to use for this
volume." Muve one disk into the Excluded box, and then click Next.
e Tvpe the name of the volume, and specify a size of 20 MB.
of columns is 2. Set the Stripe unit size to 256 (sectors), or 12HK.
e Mark the l1irrored check box in the Mirror Info region.
volume.
This operation should fail, because there arc not enough disks available in
the disk group. A two-column striped mirror requires at least four disks.
Ccpvnqtu ~ L006 Symantec Corpcreuon All ngtlts reserved

9 Createa 20-MB 3-column striped volume with a mirror. Specify threedisks to
be usedduring volume creation. Name the volume namevo14.
b In the New Volume wizard, let 'xVI1 determine which disks to use.
d Select the Striped option in the Layout region. Change the number of
columns to 3.
e Mark the Mirrored check box in the Mirror Info region. Click Next.
You receive an error and are not able to complete the wizard.
Wasthe volume created?
Again, this operation should fail, becausethere arc not enough disks
available in the disk group. At least six disks arc required for this type of
volume configuration.
I
10 Createthe samevolume specified in step9. but without the mirror.
Note: If you did not cancelout of the previous step.thenjust unchcck the
mirrored option and continue the wizard.
d Select the Striped option in the Layout region. Change the number of
columns to 3.
volume.
Wasthe volume created"
Yes, the volume is created this time.
11 Deleteall volumes in the namedg disk group.
a Select the namedg disk group, then select the Volumes tab.
b Highlight all volumes in the window.
e Select Actions->Delete Volume.
d Click Yes To All.
12 View the commandsexecutedby YEA during this sectionor the lab.
a Select the Logs perspective in the quick accessbar.
b Click Task Log in the right pane view.
e Double-click the individual tasks and observe the Task Details view.
Lab 4 Solutions: Selecting Volume Layouts 8-39
Copyright if: 2006 Symantcc Corporation. All rigt!ts reserved

You can complete this exercise using either the VEA or the CLl interface.
First, ensure that any volumes created in the previous labs are removed from
the namedgdisk group.
VEA
a Select the namedg disk group and click the Volumes tab in the right
pane view.
b To remove a volume, highlight the volume in the window, and select
Actions=c-Dclete Volume.
eLl
vxprint -g namedg -htr I more
For each volume in the namedg disk group:
vxassist -g namedg remove volume volume_name
2 Create a I OO-MB Striped Mirrored volume with no logging. Name the volume
namevoll.
VEA
a Select the namedg disk group, and select Actions=c-New Volume.
b In the New Volume wizard, let 'xVI1 determine which disks to use.
e Type the volume name, specify a volume size of 100 IIB, and select a
Striped Mirrored layout.
d Ensure that the Columns and the Total mirrors fields arc both set to
the default value of 2.
volume.
CLI
vxassist -g namedg make namevoll 100m
layout=stripe-mirror nmirror=2 ncol=2
3 If you are using VEA, view the commands executed by VEA to create the
namevoll volume during this section of the lab.
COPYright ',' 2006 Svmanu-c Corporauoo All rt(.lhts rc sorveo

a Select the Logs perspective in the quiek accessbar.
e Double-click the specific task and observe the Task Details view.
4 Create a Concatenated Mirrored volume with no logging called namevo12.
The size of the volume should be greater than the size of the largest disk in the
disk group; for example. if your largest disk is 4 GB. then create a 6-GB
volume.
Note: If you are working in the Virtual Academy (VA) lab environment, your
largest disk will have a size 01'2 GB. In this environment. you can use a 3GB
volume size.
VEA Ia Selectthe namedg disk group, and select Actions->New Volume.
e Type the volume name, an appropriate volume size, and select a
Concatenated Mirrored layout.
d Ensure that the Total mirrors field is set to the default value of 2.
volume.
CLI
vxassist -g namedg -b make namevo12 6g
layout=concat-mirror nmirror=2
5 II'you arc using VEA. view the commands executed by VEA to create the
namevo12 volume during this section of the lab.
a Select the Logs perspective in the quick aeeessbar.
e Double-click the specific task and observe the Task Details view.
6 View the volumes and compare the layouts.
VEA
a Uighlight the namedg disk group and select Actions->Volumc View.
b Click the Expand button in the Volumes window.
You can also highlight each volume in the object tree and view
information in the tabs in the right pane.
CLJ
Copvnqn 'c 2006 Symantec Corporation A11li'lllt<; reserved

vxprint -g namedg -htr I more
7 Remove all of the volumes in the namedgdisk group.
VEA
a Select the namedg disk group, and click the Volumes tab in the right
pane view.
e Select Actiuns->Delete Volume.
d Click Yes To All.
eLl
Using Ordered Allocation While Creating Volumes
disks in the disk group that you arc using.
Createa 20-MB. two-column striped volume with a mirror in the namedgdisk
group. Name the volume namevoll.
VEA
e Type the name of the volume, and specify a size of 20 M B.
of columns is 2.
e Mark the Mirrored check box in the Mirror Info region.
volume.
CLI
Copyrlght'~ 20il6 Syrnantec Corporation All ngtlls reserved

vxassist -g namedg make namevoll 20m
layout=mirror-stripe ncol=2
2 Display the volume layout. l low are the disks allocated in the volume') Which
disk devices are used':'.
VEA
a Selectthe Volumes node in the object tree and namevoll in the right
pane view.
b Select Actions->Layout View. Note the plex number and the column
number for each subdisk on each disk.
IeLl
vxprint -g namedg -htr
Notice which two disks are allocated to the first piex and which two disks
are allocated to the second piex and record your observation.
3 Remove the volume you just made. and re-create it by specifying the four disks
in an order different trorn the original layout. Use the command line to create
the volume in this step.
eLl
vxassist -g namedg -0 ordered make namevoll 20m
layout=mirror-stripe ncol=2 namedg04 namedg03
namedg02 namedgOl
4 Display the volume layout. I low are the disks allocated this time".
VEA
a Select the Volumes node in the object tree and namevoll in the right
pane view.
b Select Actions->Layout View. Note the plcx number and the column
number for each subdisk on each disk.
eLl
The plexes are now allocated in the order specified on the command line.
Copyright~;; 2006 Symaruec Corporation All rights teservec

VEA
a Select the namedg disk group and click the Volumes tab in the right
pane vicw.
c Select Actions->Delctc Volume.
d Click Yes To All.
eLl
Optional Lab: Creating Volumes with User Defaults: CLI
This optional guided practice illustrates how to use the files:
• lete/default/vxassist
• /ete/default/alt vxassist
to create volumes with defaults specified by the user. Note that some of the default
values may not apply to VEA because VEA uses explicit values for number of
columns, stripe unit size. and number of mirrors while creating striped and
mirrored volumes.
Create two tiles in jete/default:
cd fete/default
a Using the vi editor, create a tile called vxassist that includes the
following:
# when mirroring create three mirrors
nmirror=3
b Using the vi editor, create a tile called alt_vxassist that includes the
following:
# use 256K as the default stripe unit size for
# regular volumes
stripeunit=256k
Copynght 'j" ';;O[)fiSvmantec Corporation All rights reserved

2 Use these files when creating the following volumes:
a Create a IOO-MB volume called namevoll using Layout erni r r or:
vxassist -g namedg make namevo11 100m
1ayout=mirror
b Create a IO()-MB, two-column stripe volume called namevo12 using -d
alt_vxassist so that Volume Manager uses the default file:
vxassist -g namedg -d a1t vxassist make namevo12
100m 1ayout=stripe
I3 View the layout of these volumes using VEA or by using vxprint -g
namedg -htr. What do you notice?
The first volume should show three plexes rather than the standard
two.
The second volume should show a stripe sizeof 256K instead of the
standard 64K.
4 Remove any vxassist default Illes that you created in this optional lab
section. The presence of these files can impact subsequent labs where default
behavior is assumed.
rm /etc/defau1t/vxassist
rm /etc/defau1t/a1t vxassist
Copyri~t1t (t 2006 Svmantec Corporation All rigtlts reserved.

COPYright .~ ;:'006 Svmantec Corporation. All fights reserved

I
, symnnter.
Lab 5
For Lab ExerCise. s, see Appendix A. ]
Lab 5 Solutions: Making Basic Configuration Changes
This lab provides practice in making basic configuration changes. In this lab, you
add mirrors and logs to existing volumes. and change the volume read policy. You
also rcsize volumes. rename disk groups, and move data between systems.
The Lab Exercises for this lab arc located on the following page:
Prerequisite Setup
configured and licensed. In addition to this. you also need four external disks to he
At the beginning of this lab, you should have a disk group called namedg that has
Lab 5 Solutions: Making Basic Configuration Changes
Copyright c. 2006 Symantec Cnrporauon. All fights reserved
B-47

In preparation for this lab. you will need the following information about your lab
back to the first lab where you initially documented this informal ion.
Ohject Sample Value Your Value
Host name trainl
system
(my partner system)
1y Data Disks: Solaris: c It #dO -
clt#d5
HP-UX: c4tOdO -
c4tOd5
AIX:hdisk21-
hdisk26
l.inux: sda - sdf
HP-UX: c3t15dO
AIX: hdiskl
Linux: hdb
l.ocation of Lah Scripts /student/labs/sf/
(if any): sf50
Prefix to he used with name
ubject names
Copyrlght'.[; 2006 Svmantec Corporation All notusreserved

You can complete this exercise using either the VEA or the CLl interlace.
Solutions arc provided lor both.
Ensure that you have a disk group called namedg with four disks in it. I f not.
create the disk group using four disks.
Note: I I'you have completed the previous lab steps you should already have
the namedg disk group with four disks and no volumes.
YEA ISelect the Disk Groups node in the ohject tree and select namedg.
err
2 Create a 50-M8. two-column striped volume called namevoll in namedg.
VEA
b In the New Volume wizard, let Vx'11 determine which disks to use.
e Type the volume name, specify a volume size of 50 MR, and select a
Striped layout.
d Complete the wizard by accepting all remaining defaults to create the
volume.
CLI
vxassist -g namedg make namevoll SOm
layout=stripe ncol=2
3 Display the volume layout. l low are the disks allocated ill the volume'! Note
the disk devices used for the volume.
YEA
a Select the Volumes node in the object tree and namevoll in the right
pane view.
b Select Actlons=c-Lavour View.
Note the disk devices used for the first plex.
Lab S Solutions: Making Basic Configuration Changes B-49
Copynght 'Z' 2006 Symantec Corporation All rights resorveo

eLl
Notice which two disks are allocated to the first plex and record your
observation.
4 Add a mirror to namevo 11. and display the volume layout. What is the layout
of the second plex? Which disks arc used for the second plcx?
VEA
a Highlight the volume to be mirrored, and select Actions->Mirror->
Add.
b Accept the defaults in the Add Mirror dialog box and click OK.
c Select namevoll and select Aetions->Layout View.
Note the disk devices used for the second plex. Note that the default layout
used for the second plex is the same as the flrst plex,
CLl
vxassist -g namedg mirror Ilamevoll
Note the disk devices used for the second plex, Note that the default layout
used for the second plex is the same as the first plex.
5 Add a dirty region log to namevo11 and specify the disk to use tor the DRL.
Display the volume layout.
VEA
a Highlight the namevoll volume, and select Actions->Log->Add.
b In the Add Log dialog box, select Manually assign destination disks.
c Select one of the disks and click Add to add it to the Selected disks list.
d Click OK to complete.
Note: If you receive an error messageindicating that YEA could not
allocate enough space for the log, ignore the message.
e Highlight the volume under the Volumes node in the object tree and
click the Logs tab.
CLI
vxassist -g namedg addlog namevoll logtype=drl
nallledgOl
vxprint -g namedg -rth
Ccpynqht ,t 2006 Syruantec Corporauon All nqhts reserved

6 Add a second dirty region log to namevoll and specify another disk to use
for the DRL. Display the volume layout.
VEA
a Highlight the namevoll volume. and select Actions->log->Add.
b In the Add Log dialog box. select Manually assign destination disks.
c Selectone of the disks and click Add to add it to the Selecteddisks list.
d Click OK to complete.
Note: If you receive an error messageindicating that VEA could not
allocate enough spacefor the log. ignore the message.
e Highlight the volume under the Volumes node in the object tree and
click the Logs tab. IeLl
vxassist -g namedg addlog namevoll logtype=drl
namedg02
7 Remove the first dirty region log that you added to the volume. Display the
volume layout. Can you control which log was removed?
VEA
a Highlight the namevoll volume. and select Actions-->Log->
Remove.
b In the Remove Log dialog box. selectthe log you want to remove and
click Add to add it to the Selectedlogs list.
c Click OK and confirm when prompted to complete.
d Highlight the volume under the Volumes node in the object tree and
click the Logs tab.
eLl
vxassist -g namedg remove log namevoll !namedgOl
8 Find out what the current volume read policy for namevoll is. Change the
volume read policy to round robin. and display the volume layout.
VEA
a Right click the namevoll volume in the right pane view, and select
Properties. Observe the existing value of the Read policy lield. It
should indicate the default value of Basedon layouts.
Lab 5 Solutions: Making Basic Configuration Changes B-51
Copyright .~ 2006 Swnaruec Corporation AlIl1gtlls reserved.

b Highlight the namevoll volume, and select Actions->Set Volume
Usage.
c Select the Round robin option and click OK.
d Right click the namevoll volume in the right pane view, and select
Properties. Observe the existing value of the Read policy field. It
should have changed to Round robin.
CLl
You should observe that the read policy shows as SELECT which is the
value used for selected based on layouts.
vxvol -g namedg rdpol round Ilamevoll
The value of the attribute will change to ROUND.
9 Remove the original mirror (namevoll- 01) from namevoll, and display
the volume layout.
VEA
a Highlight the namevoll volume in the object tree, and click the
Mirror~ tab in the right pane.
b Right-click a plex, and select Acrlons=-c-Removc Mirror.
c III the Remove Mirror dialog box, click Yes.
d Highlight the namevoll volume, and select Actions->Layout View.
Note that the DRL log is not removed automatically when you remove the
mirror by specifying the plex name.
CLl
vxassist -g namedg remove mirror namevoll
! disk_used_by_original_mirror
Note that the DRL log will also he removed automatically with this
command becausethe volume is IIU longer mirrored.
10 Rcmovcnamevoll.
VEA
Highlight the namevoll volume, and select Actions=c-Delete Volume.
C0Pyrlgh! ~ 20M Svmantec Corporation All rights reserved

eLl
Resizing a Volume
I
You can complete this exercise using either the VEA or the ell interface.
If you have not already done so. remove the volumes created in the previous
lab in namedg.
lEA
For each volume in your disk group, highlight the volume, and select
Actions=-c-Dclctc Volume.
eLl
vxassist -g namedg remove volume volume_name
2 Create a 20-MB concatenated mirrored volume called namevoll in namedg.
Create a Veritas tile system on the volume and mount it to !namel. Make sure
that the file system is not added to the tile system table.
lEA
a Highlight the namedg disk group, and select Actions->New Volume.
b Specify a volume name, the size, a concatenated layout, and select
mirrored.
c Ensure that "Enable logging" is not checked.
d Add a VxFS tile system and set the mount point. Uncheck the Add to
tile system table option.
e Complete the wizard.
eLl
vxassist -g namedg make namevoll 20m layout=mirror
mkfs -F vxfs /dev/vx/rdsk/namedg/namevoll
Note: On Linux, usemkf s - t.
mkdir / namel (if necessary)
mount -F vxfs /dev/vx/dsk/namedg/namevoll /namel
Note: On Linux, usemoun t - t.
Copyright 'i- 2006 Symantec Corporation. All rights reserved

3 View the layout of the volume and display the size of the file system.
VEA
Highlight the volume in the object tree and click each of the tabs in the
right pane to display information about Mirrors, Logs, and Subdisks.
You can also select Actions->Volume View, click the Expand button, and
compare the information to thc main window.
To view the tile system size, select the File Systems node in the object tree
and observe the Size column for the / namel file system in the right pane
view.
eLl
Solaris, vxprint -g namedg -rth
Linux, df -k /namel
AIX
IIP-UX vxprint -g llamedg -rth
bdf /namel
4 Add data to the volume by creating a file in the file system and verity that the
tile has been added.
echo "hello name" > /namel/hello
5 Expand the tile system and volume to 100 MS. Observe the volume layout to
see the change in size. Display tile system size.
VEA
a Highlight the volume and select Actions->Resize Volume.
b In the Resize Volumc dialog box, spcclty 100 M8 in the "New volume
size" field, and click OK.
c Right click the volume and select Properties to observe the change in
size.
d For thc file system size, select the File Systems node in the object tree
and observe the Size column for the / namel file system.
CDpynght"; 2006 Symantec Couiorauor, AlIlIgt1ls ro servec

eLl
Solaris, vxresize -g namedg namevoll 100m
Linux, vxprint -g namedg -rth
AIX
df -k /namel
HP-I.IX vxresize -g namedg namevoll 100m
bdf /namel
IResizing a File System Only: ell
Note: This exercise should be performed using the command line interface
because the VEA docs not allow you to create a file system smaller in size than the
underlying volume. You also cannot change the size of the volume and the file
system separately using the GUI.
1 Create a 50-MB concatenated volume named narnevo12 in the narnedg disk
group.
vxassist -g namedg make namevo12 SOm
2 Create a Veritas file system on the volume by using the mkf s command.
Specify the file system size as 40 MB.
mkfs -F vxfs /dev/vx/rdsk/namedg/namevo12 40m
Note: On Linux, usemkfs -to
3 Create a mount point /narne2 on which to the mount the file system. ifit does
not already exist.
mkdir / name2 (if necessary)
4 Mount the newly created file system on the mount point /narne2.
mount -F vxfs /dev/vx/dsk/namedg/namevo12 /name2
Note: On Limn, usemount - t.
Copyrigll1 & 2006 Syrrmntcc Corporation. All rights reserved

Observe that the total size of the file system is smaller than the size of the
volume.
Solaris, df -k
Linux,
AIX
B!'-UX bdf
6 Expand the file system to the full size of the underlying volume using the
fsadm -b newsize option.
fsadm -b 50m -r /dev/vx/rdsk/namedg/namevol2 /name2
Solaris, df -k
Linux,
AIX
UI'-UX bdf
8 Make a tile on the tile system mounted at / name2, so that the free space is less
than 50 percent of the total tile system size.
dd if=/dev/zero of=/name2/25_mb bs=1024k count=25
9 Shrink the tile system to 50 percent of its current size. What happens'?
fsadm -b 25m -r /dev/vx/rdsk/namedg/namevol2 /name2
The command fails. You cannot shrink the tile system because blocks are
currently in use.
1 Try to rename the namedgdisk group to namedgl while the / namel and
/ name2 tile systems are still mounted. Can you do it'?
VEA
a Highlight the namedgdisk group, and select Actions->Rename Disk
Group.
b Type in the new name and click OK.
You receive an error messageindicating that the volumes in the disk group
are in use.
8-56 VERITAS Storage Foundation 5.0 for UNIX Fundamentals
Copyright if 20U6 Svmantec Corporauor- Allnghl':' rcservec

eLl
vxdg -n namedgl deport namedg
You receive an error messageindicating that the volumes in the disk group
are in USI'.
2 Observe the contents of the / dev/vx/ rdsk and / dev/vx/ dsk directories
and their subdirectories. What do you see'!
ls -lR /dev/vx/rdsk
This directory contains a subdirectory for each imported disk group,
which contains the character devices for the volumes in that disk group.
ls -lR /dev/vx/dsk
This directory contains a subdirectory for each imported disk group,
which contains the block devices for the volumes in that disk group.
I
3 Unmount all the mounted fi lc systems in narnedg disk group.
YEA
a Select the File Systems node in the object tree and highlight the file
systemsyou want to unmount in the right pane view.
c Confirm when prompted.
eLl
umount /namel
umount /name2
4 Rename the narnedg disk group to narnedgl. Do not forget to start the
volumes in the disk group after the renaming if you are using the command
line interface.
YEA
a Highlight the namedg disk group, and select Actions->Rename Disk
Group.
b Type in the new name and click OK.
eLl
vxdg -n namedgl deport namedg
vxdg import namedgl
Lab 5 Solutions Making Basic Configuration Changes B-57
Copyright 'Z: 2006 Symantec Corporation. All righl~ rcservoc

vxvol -g namedgl startall
5 Observe the contents of the / dev/vx/ rdsk and / dev/vx/ dsk directories
and their subdirectories. What has changed'?
Is -lR /dev/vx/rdsk
Is -lR /dev/vx/dsk
The device subdirectories are rebuilt with the new name of the disk group.
6 Observe the disk media names. Is there any change'!
VEA
Select namedgl in the object tree and click the Disks tab.
Observe the Internal name column. There should be no change in disk
media names.
CLl
vxdisk -0 al1dgs list
vxprint -g namedgl -htr
There should be no change in disk media names.
7 Mount the / narnel and / narne2 file systems, and observe their contents.
VEA
For each volume:
a Highlight the volume, and select Actions->File System->Mount File
System.
b Type the Mount point and unselect the Add to file system table option.
c Click OK to complete.
CLl
mount -F vxfs /dev/vx/dsk/namedgl/namevol1 /namel
mount -F vxfs /dev/vx/dsk/namedgl/namevo12 /name2
Note: On Linux, usemount - t.
Is -1 /namel
Is -1 /llame2
Copynghl,: 20[10 S,mantec Corpor auor- All flghls reserved

You can complete this exercise using either the VEA or the CLI interlace.
Solutions arc provided for both.
Note: If you are sharing a disk array. each participant should make sure that the
prefix used for object names is unique.
Copy new data to the /namel and /name2file systems. For example. copy
the jete/hosts file to /namel and the jete/group file to /name2.
ep tete/hosts /namel
ep jete/group /name2
I2 View all the disk devices on the system.
VEA
pane. Notc thc Status column.
CLI
3 Unmount all file systems in the namedgldisk group and deport the disk
group. Do not give it a new owner. View all the disk devices on the system.
VEA
a Select the File Systems node in the object tree and highlight the file
systems you want to unmount.
d Select the disk group and select Actions->Deport Disk Group. Click
OK.
e Confirm your request when prompted in the Deport Disk Group
dialog box.
Select the Disks node in the object tree and observe the disks in the
right pane. Note the change in the Status column.
CLI
umount /namel
umount /name2
vxdg deport namedgl
Copynght (f; 2006 Symantec Corporal Ion All rights reserved

4 Identify the name ofthe system that is sharing access to the same disks as your
system. If you are not sure, check with your instructor. Note the name of the
partner system here.
Partner system hostname: _
5 Using the command line interface. pcrfonn the following steps on your partner
system:
Note: If you arc working on a standalone system. skip step a in the following
and use your own system as the partner system.
a Remote login to the partner system,
rlogin partner_system_hostname
b Import the narnedgl disk group on the partner system. start the volumes in
the imported disk group, and view all the disk devices on the system.
On the partner system:
vxdg import namedgl
vxvo1 -g namedgl starta11
vxdisk -0 al1dgs list
c While still logged in to the partner system, mount the / narnel and
/ name2 file systems. Note that you will need to create the mount
directories on the partner system before mounting the file systems. Observe
the data in the tile systems.
mkdir /namel
mkdir /name2
mount -F vxfs /dev/vx/dsk/namedgl/namevo11 /namel
mount -F vxfs /dev/vx/dsk/namedgl/namevo12 /name2
Note: On Linux, use moun t - t.
Is -1 /namel
Is -1 /name2
The data should be the same as it was on your own system.
d Unmount thc file systems on your partner system.
umount /namel
umount /name2
e On your partner system. deport narnedgl and assign your own machine
name, tor example, trainS, as the New host.
Cop/ughl G:2006 Symantoc Corporation. All nqtus reserved

vxdg -h your_system_name deport namedgl
Exit from the partner system.
Type Exit.
6 On your own system imparl the disk group and change its name back to
namedg. View all the disk devices on the system.
VEA
Ia Select the disk ~roup under the Disk Groups node and select
Actions->Import Disk Group.
b In the Import Disk Group dialog box, type namedg in the New name
field, verify that the "Start all volumes" option is checked, and click
OK.
c Select the Disks node in the object tree and observe the disks in the
right pane. The status should change to Imported.
CLI
vxdg -n namedg import namedgl
vxvol -g namedg startall
7 Deport the disk group namedg by assigning the ownership to anotherhost.
View all the disk devices on the system. Why would you do this')
V£A
a Select the disk group under the Disk Groups node and select
Actions->Deport Disk Group.
b In the Deport Disk Group dialog box, check Deport options, and type
anotherhost in the New host field.
c Click OK and confirm when prompted.
In the list of disks, the status of the disks in the deported disk group is
displayed as Foreign.
You would do this to ensure that the disks are not imported accidentally
by any system other than the one whose name you assignedto the disks.
CLI
vxdg -h anotherhost deport namedg
Copyright © 2006 Svmantec Corporation All rights reserved.

8 From the command line display detailed information about one of the disks in
the disk group using the vxdisk list device_tag command. Note the
hostid field in the output.
vxdisk list device tag
where device_tagis c#t#d# for Solaris and HP-UX, hdisk# for AIX
9 Import namedg. Were you successful'?
VEA
a Selectthe disk group and selectActions->Import Disk Group.
b In the Import Disk Group dialog box, click OK.
This operation should fail, becausenamedg belongsto another host.
CLI
vxdg import namedg
This operation should fail, becausenamedg belongsto another host.
10 Now import namedg and overwrite the disk group lock. What did you have to
do to import it and why"
VEA
a Selectthe disk group and selectActions->Import Disk Group.
b In the Import Disk Group dialog box, mark the Clear host ID check
box, verify that the "Start all volumes" option is checked, and click
OK.
CLI
vxdg -c import namedg
11 From the command line display detailed information about the same disk in the
disk group as you did in step 8 using the vxdi sk 1i st devi ce_ tag
command. Note the change in the hostid field in the output.
vxdisk list device tag
where device_ tag is c#t#d# for Solaris and HP-UX, hdisk# for AIX
Copvnqnt '& 2006 Svmantec Corporauon. All rights reserved

VEA
a Selectthe namedg disk group and click the Volumestab in the right
pane view.
b Highlight all volumesin the window.
c SelectActions->Dclcte Volume.
d Click YesTo All.
eLl
Ivxassist -g namedg remove volume namevoll
vxassist -g namedg remove volume namevo12
Copyright If: 2006 Swnantec Corrorauon. All rights reserved

A lab exercise in the next lesson requires that you run a script that sets up tiles with
different size extents. Because the script can take a long time to run, you may want
to begin running the script now, so that the necessary environment is created by the
ncx t lab time.
non-CDS disk formal.
Solaris,
Linux,
AIX
vxdisksetup -i device_tag format=sliced
where device_tag is c#t#d# for Solaris.
II P-LJX Note: Check the status of the second internal disk using the vxdisk
1i st command. If the disk is displayed as an LVM disk. ensure that
it is not used by any active LVM volume groups and take it out of
LVM control using the pvremove command. Ir the pvremove
command Jails due to an exported volume group information left on
the disk. re-create an LVM header using the force option
(pvcreate -f /dev/rdsk/device_name) before using the
pvremove command to remove it.
vxdisk list
If necessary:
vgdisplay -v /dev/vgOO
pvcreate -f /dev/rdsk/device_tag
pvremove /dev/rdsk/device tag
where dev i ce_ tag is the device name of the second internal
disk in the format c#t#d#.
vxdctl enable
vxdisk list
vxdisksetup -i device tag format=hpdisk
where device_ tag is c#t#d# for HP-UX.
3 Create a non-cds disk group called testdg using the disk you initialized in
step 2.
vxdg init testdg testdgOl=device_tag cds=off
where device_ tag is c#t#d# fur Sularis and HP-UX.
8-64
CoPVrtghl : 2000 Svmanlec Corporauoo All nghts reserved

4 In the testdg disk group create a 1-(i8 concatenated volume called
testvol initializing the volume space with zeros using the Ln i t e zero
option to vxassist.
vxassist -g testdg make testvol 19 init=zero
5 Create a VxFS file system on testvol and mount it on /fs_test.
mkfs -F vxfs /dev/vx/rdsk/testdg/testvo1
mkdir /fs test
mount -F vxfs /dev/vx/dsk/testdg/testvol /fs_test
Note: On Llnux, use mount -to
I6 Ask your instructor for the location of the extents. sh script. Run the
extents. sh script.
/student/labs/sf/sf50/extents.sh
7 Verify that the VRTSspt software is already installed on your system. Ifnot,
Note: Before Storage Foundation 5.0. the VRTSspt software was provided as
Sularis, pkginfo I grep VRTSspt
Linux,
AIX
HP-liX sw1ist -1 product I grep VRTSspt
echo $PATH I grep -i vxbench
If necessary:
export PATH=$PATH:/opt/VRTSspt/FS/VxBench
Copyright?:, 2006 Syrnantec Corporation. All IIghl<; reserved

Coovnqht:'; 2006 Syrw}i'-'C Corporatron. All rights roverveo

,S}lmmh'(.
I
Lab 6
and administering the file change log,
For Lab Exercises. see Appendix A.
Lab 6 Solutions: Administering File Systems
In this lab. YOIl practice file system administration. including defrngmentation and
administering the file change log.
Setup
configured and licensed. In addition to this. you also need four external disks and
the second internal disk to be used during the labs. If you do not have a second
internal disk or if you cannot use the second internal disk. you need live external
disks to complete the labs.
At the beginning otthis lab. you should have a disk group called narnedg that has
four external disks and no volumes in it. The second internal disk should be empty
and unused.
Note: If you are working in a North American Mobile Academy lab environment.
you cannot use the second internal disk during the labs. II'that is the case. select
one or the external disks 10 complete the lab steps.
Copyriglll "'; 2006 Symantec Corporation. All rights teservcrt
8-67

My Data Disks: Solaris: clt#dO -
clt#d5
HP-UX: c4tOdO -
c4tOd5
IIIX: hdisk21-
hdisk26
l.inux: sda - sdf
HP-UX: c3t15dO
IIIX: hdiskl
Linux: hdb
Location of Lab Scripts Istudent/labs/sf/
(if any): sf50
object names
Copynght:;: 2006 Symantec Corporauon All fights reserved

Note: If you have already performed these steps at the end ofthe last lab, then you
can skip this section and proceed with Dctragmcnting a Veritas File System
section.
Identify the device tag lor the second internal disk on your lab system. If you
Second internal disk (or the external disk used in this lab):~ _
I2 Initialize the second internal disk (or the external disk used in this lab) using a
non-CIJS disk format
Solaris vxdisksetup -i device_tag format=sliced
where device_ tag is c #t#d# for Solaris.
HP-UX Note: Check the status of the second internal disk using the vxdi sk
list command. lfthe disk is displayed as an LYM disk. ensure that
it is not used by any active LYM volume groups and take it out or
LYM control using the pvremove command. II' the pvremove
command fails due to an exported volume group information left on
the disk. re-create an LYM header using the torce option
(pvcreate - f / dev /rdsk/ device_name) before using the
pvremove command to remove it.
vxdisk list
If necessary:
vgdisplay -v /dev/vgOO
pvcreate -f /dev/rdsk/device_tag
pvremove /dev/rdsk/device_tag
where device_ tag is the device name of the second internal
disk in the format c#t#d#.
vxdctl enable
vxdisk list
vxdisksetup -i device_tag format=hpdisk
where devi ce_tag is c#t#d# for .IP-UX.
3 Create a non-CDS disk group called testdg using the disk y()U initialized in
step 2.
vxdg init testdg testdgOl=device_tag cds=off
where device_ tag is c#t#d# for Solaris and HP-UX.
4 In the testdg disk group create a I-GB concatenated volume called
testvol initializing the volume space with zeros using the init=zero
option to vxassist.
Copynght ~ 2006 Symantec Corporation All rights reserved
8-69

vxassist -g testdg make testvo1 19 init=zero
5 Create a VxFS tile system on testvol and mount II on /fs test.
mkfs -F vxfs /dev/vx/rdsk/testdg/testvo1
Note: On Linux, usemkfs - t.
mkdir /fs test
mount -F vxfs /dev/vx/dsk/testdg/testvo1 /£s test
6 Ask your instructor for the location of the extents. sh script. Run the
exten ts . sh script.
/student/1abs/sf/sf50/extents.sh
7 Verify that the VRTSspt software is already installed on your system. Ifnot,
Note: Before Storage Foundation S.O,the VRTSspt software was provided as
pkginfo I grep VRTSspt
sw1ist -1 product I grep VRTSspt
echo $PATH I grep -i vxbench
If necessary:
Copynght C' 2006 Svmar-tec Corporation All rights reserved

Defragmenting a Veritas File System
The purpose of this section is to examine the structure of a fragmented and an
unfragmented file system and compare the file system's throughput in each case.
The general steps in this exercise are:
Make and mount a file system
Examine the structure of the new file system for extents allocated
Then examine a fragmented file system and report the degree of fragmentation
in the file system
Use a support utility called vxbench to measure throughput to specific files
within the fragmented file system
De-fragment the file system, reporting the degree or fragmentation
Repeat executing the vxbench utility using identical parameters to measure
throughput to the same files within a relatively unfragmented file system
Compare the total throughput before and after the defragmentation process
In the namedg disk group create a I-GB concatenated volume called
namevoll.
vxassist -g namedg make namevoll 19
2 Create a VxFS file system on namevoll and mount it on / namel.
Note: On Linux, use mkf s - t..
Note: On Linux, use moun t - t.
3 Run a fragmentation report on / namel to analyze directory and extent
fragmentation. Is a newly created, empty file system considered fragmented"
In the report. what percentages indicate a file system's fragmentation?
fsadm -D -E /namel
Searched Blocks Dirs to Add Reduce Reduce
total 2 0 2 0 0 0
Extent Fragmentation Report
Total Average Average Total
Files File Blks # Extents Free Blks
0 0 0 1030827
blocks used for indirects: 0
Lab 6 Solutions: Administering File Systems 8-71
Copyrtqht © 2006 Symantec Corporation. All rights reserved
I

Free Extents By Size
1: 1 2 : 1 4: 2
8: 2 16: 1 32: 2
64: 1 128: 2 256: 1
512: 2 1024: 1 2048: 0
4096: 1 8192: 1 16384: 0
32768: 1 65536: 1 131072: 1
262144: 1 524288: 1 1048576: 0
2097152: 0 4194304: 0 8388608: 0
16777216 : 0 33554432: 0 67108864: 0
134217728: 0 268435456: 0 536870912: 0
1073741824: 0 2147483648: 0
A newly created file system with no files or directories cannot be
fragmented.
The following table displays the percentages you should be observing in
the output of the fragmentation report to determine if a tile system with
tiles and directories is fragmented.
Percentage Unfragmented Badly
Fragmented
'Yo of Free hlocks in extents smaller < 5'Yo >50%
than 64 blocks
% of Free blocks in extents smaller < If}O > 5'Y.
than 8 blocks
% blks allocated to extents 64 blks or > 5'Yo <5%
larger
4 What is a fragmented file system?
A fragmented tile system is a tile system where the free space is in
relatively small extents scattered throughout different allocation units
within the tile system.
5 If you were shown the following extent fragmentation report about a file
system. what would you conclude"
total
Searched Blocks Dirs to Add Reduce
199185 85482 115118 5407 5473
Reduce
5655
COPYrlghl''- 2006 Svmantec Corporation All fight:. reseo.eo

A high total in the Dirs to Reduce column indicates that the directories are
not optimized. This file system's directories should be optimized by
directory defragmentation.
6 Unmount Inamel and remove namevoll in the namedg disk group.
umount Iname1
Note: The following steps will use the Ifs_test tile system to analyze the
impact of fragmentation on the tile system performance. Verily that the
extents. shscript has completed before you continue with the rest of this lab.
7 Run afragmentation report on Ifs test to analyze directory andextent
fragmentation. Is Ifs_test fragmented? Why or why not'! What should be
done')
fsadm -D -E Ifs test
Searched Blocks Dirs to Add Reduce Reduce
total 2 o o2 1 o
Total Average
Files File Blks
55 5102
Average
# Extents
Total
641
Free Blks
750037
1: 16891 2 : 11505 4: 25446
8 : 10868 16: 1384 32: 2
64: 0 128: 0 256: 10
512: 0 1024: 1 2048: 0
4096: 1 8192: 0 16384: 0
32768: 1 65536: 1 131072 : 1
262144: 1 524288: 0 1048576: 0
2097152: 0 4194304: 0 8388608: 0
16777216: 0 33554432: 0 67108864: 0
134217728: 0 268435456: 0 536870912: 0
1073741824: 0 2147483648 : 0
Dirs to Reduce column is O.Therefore, the directories do not needto he
optimized. But the extents needto be optimized. Because:
Copyright r{; 2006 SYll13nt0C Corpo'8110II All rights reserved
I
8-73

'YoFree blocks in extents smaller than 64 blks: 33.44 «50%) - OK
% Free blocks in extents smaller than 8 blks: 18.89 (>5°1.,)- Not OK
%, blks allocated to extents 64 blks or larger: 42.07 (>5%) - OK
Therefore, the file system's extents should be defragmented.
8 Lise the Is -Te command to display the extent attributes of the files in the
Ifs_test tile system. Note that on the Solaris platform you need to use the
Is command provided by the YxFS file system software to be able to use the
-e option.
/usr/lib/fs/vxfs/bin/ls -Ie /fs test
Is -Ie /fs_test
-rw-r--r-- 1 root other 2048000 Jul 14 17:57 test42 ores 0
ext 2
ext 4
ext 6
-rw-r--r-- 1 root other 8192000 Jul 14 17:57 test48 :res 0
ext 8
-rw-r--r-- 1 root other 8192000 Jul 14 17:57 test 50
ext 2
ext 4
ext 6
-rw-r--r-- 1 root other 8192000 Jul 14 17:57 test58 :res 0
ext 8000
Two files that will be used in performance tests have been highlighted in
the sample output provided here.
9 Measure the sequential read throughput to a particular tile. for example, an
RMB tile on an RK extent (for example, /fs_test/test48), in a
fragmented tile system using the vxbench utility and record the results. Use
an RK sequential I/O size.
Notes:
You need to use the vxbench utility that is appropriate for the platform
you are working on, for example vxbench_9 on Solaris 9. To identify the
appropriate vxbench command, use the Is -1 Iopt/VRTSspt /FS /
VxBench command. If this path is not in your PATH environment
variable, use the fullparh of the command while running the corresponding
vxbench utility.
Copynqh! r£ 2006 Symautec Corporation All rights reserved

Remount the file system before running each I/O test.
Solaris mount -F vxfs -0 remount
/dev/vx/dsk/testdg/testvol /fs -
test
/opt/VRTSspt/FS/VxBench/vxbench_9 -w read -i
iosize=8k,iocount=1000 /fs - test/test48
HI'-lJX mount -F vxfs -0 remount
/dev/vx/dsk/testdg/testvol /fs - test
/opt/VRTSspt/FS/VxBench/vxbench_ll.23_pa64
-w read -i iosize=8k,iocount=1000
/fs .-
test/test48
A sample output is provided here as an example:
total: 7.147 sec 1119.40 KB/s cpu: 0.12 sys 0.00 user
I10 Repeat the same test for an 8Mb file on an 8Mb extent (for example. using the
Ifs_test/test58 file). Note that the lile system must be remounted
between the tests. Can you explain why?
The file system must be remounted to clear the read buffers.
lopt/VRTSspt/FS/VxBench/vxbench -
9 -w read -i
iosize=8k,iocount=1000 Ifs - test/test58
HI'·LJX mount -F vxfs -0 remount
/dev/vx/dsk/testdg/testvol Ifs - test
10pt/VRTSspt/FS/VxBench/vxbench_ll. 23 _pa64
-w read -i iosize=8k,iocount=1000
Ifs - test/test58
total: 0.206 sec 38911.83 KB/s cpu: 0.17 sys 0.01 user
11 Defragment If s _ t es t and gather summary statistics after each pass through
the file system, After the dcfragrnentation completes. determine if I f s_ test
is fragmented? Why or why not?
Note: The dctragmcntation can take about 5 minutes to complete.
fsadm -e -E -s Ifs test
Total Average Average Total
Files File Blks # Extents Free Blks
55 5102 641 750037
Lab 6 Sotutions: Administering File Systems 8-75
Copyright ~ 2006 Symantoc COrporation All rights reserved

1: 16891 2: 11505 4: 25446
8: 10868 16: 1384 32: 2
64: 0 128: 0 256: 10
512: 0 1024: 1 2048: 0
4096: 1 8192: 0 16384: 0
32768: 1 65536: 1 131072: 1
262144: 1 524288: 0 1048576: 0
2097152: 0 4194304: 0 8388608: 0
16777216: 0 33554432: 0 67108864: 0
134217728 : 0 268435456: 0 536870912: 0
1073741824: 0 2147483648: 0
Pass 1 Statistics
Extents Reallocations Ioctls Errors
Searched Attempted Issued FileBusy NoSpace Total
total 35210 16151 45 0 0 0
Pass 2 Statistics
Extents Reallocations Ioctls Errors
Searched Attempted Issued FileBusy NoSpace Total
total 18296 8643 33 33 0 33
Total Average
Files File Blks
55
Average
# Extents
333
Total
Free Blks
7446052833
% Free blocks in extents smaller than 64 b1ks: 8.89
1: 2173 2:
8: 1122 16:
128:
1024:
8192:
65536:
524288:
4194304:
33554432:
268435456:
64: 994
512: 5
4096: 0
32768: 0
262144: 1
2097152 : 0
16777216: 0
134217728: 0
38 4: 1161
1104 32: 1021
989 256: 605
3 2048: 0
0 16384: 0
1 131072: 0
0 1048576: 0
0 8388608: 0
0 67108864 : 0
0 536870912: 0
C()P~flgtH ~_-200n Svmantec Corporauon All fights wspr'/<:'d

1073741824: 2147483648: oo
The file system no longer needsto he defragmentcd, because:
%, Free blocks in extents smaller than 64 blks: 8.89 «50'10) - OK
(much better than before)
'Yt,Free blocks in extents smaller than 8 blks: 0.93 «I %) - OK (milch
better than before)
01., blks allocated to extents 64 blks or larger: 46.94 (>5%) - OK
(slightl~' better than before)
12 Measure the throughput of the untragmented file system using the vxbench
utility on the same files as you did in steps 9 and 10, Is there any change in
throughput"
Notes:
You need to use the vxbench utility that is appropriate for the platform
you are working on. for example vxbench_9 on Solaris 9, To identify the
appropriate vxbench command. use the 1 s -1 / opt /VRTSs pt / FS/
VxBench command. If this path is not in your PATH environment
variable. use the tullpath of the command while running the corresponding
vxbench utility.
The file system must be remounted before each test to clear the read
buffers.
If you have used external shared disks on a disk array used by other
systems for this lab. the performance results may be impacted by the disk
array cache and may not provide a valid comparison between a fragmented
and defragmented file system.
/opt/VRTSspt/FS/VxBeneh/vxbeneh -
9 -w read -i
iosize=8k.ioeount=1000 /fs -
test/test48
--
HP-UX mount -F vxfs -0 remount
/opt/VRTSspt/FS/VxBeneh/vxbeneh_11.23_pa64
-w read -i iosize=8k.ioeount=1000
/fs -
test/test48
total: 0.241 see 33187.31 KB/s epu: 0.13 sys 0.01 user
Copwiqbt D 2U06 Symantec Corporation. Alillghl5 reserved
I

/dev/vx/dsk/testdg/testvol /fs test-
/opt/VRTSspt/FS/VxBeneh/vxbeneh 9 -w read -i -
iosize=8k,ioeount=1000 /fs --test/test58
UP-UX mount -F vxfs -0 remount
/Opt/VRTSspt/FS/VxBeneh/vxbeneh -
11.23 _pa64
-w read -i iosize=8k,ioeount=1000
/fs -
test/test58
total: 0.202 see 39650.48 KB/s epu: 0.18 sys 0.00 user
There is an improvement in throughput for both casesbut the
improvement is highest for the file using small extent sizes (that is for
Ifs test/test48).
13 What is the difference between an unfragmented and a fragmented file system'?
A fragmented tile system has free space scattered throughout the file
system in relatively small extents whereas an unfragmented tile system has
free space in just a few relatively large extents.
14 Is any onc environment more prone to needing dcfragmcntation than another'?
Yes,volatile environments wherein files are grown, shrunk, erased,
moved, with ownership changes, and so on are prone to fragmentation.
Stable environments, such as Oracle databases and logs, have very little
impact on the supporting tile system so require infrequent
dcfragmentation.
Reading the File Change Log (FCL)
In the namedg disk group create a new 10-1113 volume called namevoll.
Create a YxFS tile system on namevoll and mount it on /fel_test.
vxassist -g namedg make namevol1 10m
Note: On Llnux, usemkfs - t.
mkdir /fel test
mount -F vxfs /dev/vx/dsk/namedg/namevoll /fel test
2 Turn the FCL on for /fel test. and ensure that it is on.

feladm on Ifel test
feladm state Ifel test
ON
3 (10 to the directory that contains the FCL.
ed Ifel test/lost+found
Is
ehangelog
I4 Display the superblock for /fcl_test.
feladm print 0 Ifel_test
5 l low do you know that there have been no changes in the file system yet?
The superblock (fo££) and the end of the FCL file (loff) are the same
number.
6 Add some tiles to /fcl test. Then remove one of the tiles youjust added.
ed Ifel test
toueh a b e
rmb
feladm print 0 Ifel test
8 Ilow do you know that changes have been made to the file system"
The superblock (foff) and the end of the FCL file (10££) are different
numbers.
9 Print the number of the FCL.
feladm print 1024 Ifel_test
The fields are Change Type, Inode Number, Inode Generation, and
Timestamp
The Unlink and Rename types list the name of the file on the following
line, preceded by the parent's inode number.
10 Which tiles are identified by the inode numbers that are listed in the Create
type?
vxlsino inode_number Ifel test
11 Unmount the fel_test tile system and remove namevoll.
cd I
umount Ifcl_test
Copyrighl.:t 2006 Svmanter; Corporation All rights reserved

vxassist -g nallledg remove volume namevoll
12 The next two lab sections are optional labs on analyzing and defragmeming
fragmented file systems. If you are not planning to carry out the optional labs,
unmount Ifs_test tile system ami destroy the testdg disk group:
otherwise. skip this step.
umount Ifs test
vxdg destroy testdg
Copyright if· 2006 Symanrac Corporation All nqbts reserved

The next set of lab exercises is optional and may be performed iI'you have time.
These exercises provide additional practice in defragmenting a file system and
monitoring fragmentation.
Optional Lab: Defragmenting a Veritas File System
I
This section uses the Ifs_test file system to analyze the impact of
fragmentation on the performance of a variety of 110types on files using small and
large extent sizes.
Recreate the fragmented Ifs_test file system using the following steps:
a Unmount the Ifs test file system
UllIount Ifs test
b Recreate a vx ts file system in the testvol in testdg.
mkfs -F vxfs /dev/vx/rdsk/testdg/testvol
Note: On Linux, usemkf s - t.
c Mount the file system to /fs_test.
mount -F vxfs /dev/vx/dsk/testdg/testvol /fs test
Note: On Liuux, usemount -to
d Ask your instructor for the location of the extents. sh script. Run the
extent s. sh script.
Note: This script can take about I:; minutes to run.
/student/labs/sf/sf50/extents.sh
2 Run a series of performance tests for a variety of 1i0 types using the vxbench
utility to compare the performance of the Illes with the XK extent size
(/fs_test/test48) and the ROOOKextent size (/fs_test/test58) by
performing the following steps.
Complete the following table when doing the performance tests.
Copyright (i; 2006 Symantec Corporation All rights reservec

Test Type Time (seconds) Throughput (KB/second)
Before After Defrag Before After Defrag
Defrag Defrag
Sequential 2.709 .526 2953.22 15202.10
reads.XK
extent
Sequential .547 .549 14634.57 14576.20
reads.8000K
extent
Random 8.268 6.267 967.54 1276.53
reads.RK
extent
Random 6.541 6.468 1223.02 1236.91
reads.8UOUK
extent
Note: Results can vary depending on the nature of the data and the model of
array used. No performance guarantees arc implied by this lab.
3 Ensure that the directory where the vxbench utility is located is included in
your PATH definition.
4 Sequential 1/0 Test
Note: You must unmount and remount the file system Ifs_test before each
step to clear and initialize the buffer cache.
mount -F vxfs -0 remount /dev/vx/dsk/testdg/testvol
/fs test
Note: On Linux, use mount -to
vxbench_platform -w read -i iosize=8k,iocount=lOOO
/fs test/test48
/fs test
Note: On Linux, use mount - t.
vxbench_platform -w read -i iosize=8k,iocount=lOOO
/fs test/test58
5 Random 1/0 Test
Copynyht 20011Svrnantec Corporaurm AI! fl9h!S rese-veo

/fstest
Note: On Linux, usemount -to
vxbench_platform -w rand_read -i
iosize=8k,iocount=lOOO,maxfilesize=8000
/fs test/test48
To test the ROOOK extent size:
/fs test
Note: On Linux, usemount -to
vxbench-platform -w rand_read -i
iosize=8k,iocount=lOOO,maxfilesize=8000
/fs test/test58
I6 Defragment the If s_tes t file system. The defragmenration process takes
some time.
Solaris. /opt/VRTSvxfs/sbin/fsadm -e -E -d -D -s /fs -
test
Linux,
AIX
BI'-LJX /usr/lbin/fs/vxfs5.0/fsadm -e -E -d -D -s /£s -
test
7 Repeat the vxbench performance tests and complete the table with these
performance results.
8 Compare the results of the defragmenrcd tile system with the fragmented file
system.
9 When finished comparing the results in the previous step. unmount the
Ifs_test file system and destroy the testdg disk group.
umount /fs_test
vxdg destroy testdg
Optional Lab: Additional Defragmenting Practice
In this exercise. you monitor and defragment a file system by using the f sadm
command.
Create a new 2-GB striped volume called namevoll in namedgdisk group.
Create a VxFS file system on namevoll and mount it on Ifs_test.
vxassist -g namedg make namevoll 2g layout=stripe
mkfs -F vxfs Idev/vx/rdsk/namedg/namevoll
Lab 6 Solutions: Administering File Systems B-83
Copyright © 2006 Symantec Coepo.auon All dqtlls recerven

Note: On Linux, use mkfs - t.
mkdir / f s _ tes t (if the directory docs not already exist)
mount -F vxfs /dev/vx/dsk/namedg/narnevoll Ifs test
Note: On Linux, use mount -to
2 Repeatedly copy a small existing file system to If s_test using a new target
directory name each time until the target tile system is approximately 85
percent full. Fur example, 011 the Solaris platform:
for i in 1 2 3
> do
> cp -r lopt Ifs test/opt$i
> done
Note: Monitor the tile system size using df - k on the Solaris platform and
bdf on the HP-UX platform, and CTRL-C out of the for loop when the file
system becomes approximately 85 percent full.
3 Delete all files in the Ifs_test tile system over 10 MB ill size.
find Ifs test -size +20480b -exec rm {} ;
find Ifs test -size +20480 -exec rm {} ;
4 Check the level offragmentation ill the Ifs_test file system.
fsadm -D -E /fs test
5 Repeat steps 2 and .3 using values 4 5 fur i ill the loop. Fragmentation of
both free space and directories will result.
6 Repeat step :2 using values 6 7 for i. Then delete all tiles that are smaller
than 6.:1K to release a reasonable amount of space.
find Ifs test -type f -size -64k -exec rm {} ;
find Ifs test -type f -size -128 -exec rm {} ;
7 Defragment the file system and display the results. Run fragmentation reports
both before and after the defragmentation and display summary statistics after
each pass. Compare the f sadrn report from step 4 with thc final report from
the last pass in this step.
fsadm -e -E -d -D -s /fs test
8 Unmount the If s_test file system and remove the namevoll volume used
ill this lab.
umount Ifs test
Copvnqht 'i:. ~006 Symantec CorPOf?'1, '

'S)11U1I1rCC
Lab 7
For J::ab Solutions, see Appendix B.
I
Lab 7 Solutions: Resolving Hardware Problems
In this lab. you practice recovering from a variety of hardware failure scenarios.
resulting in disabled disk groups and railed disks. First you recover a temporarily
disabled disk group and then you use a set of interactive lab scripts to investigate
and practice recovery techniques. Each interactive lab script:
Sets up the required volumes
Simulates and describes a failure scenario
Prompts you to fix the problem
Finally. a set of uptiona I labs are provided to enable you to investigate disk failures
further and to understand the behavior of spare disks and hot relocation.
Prerequisite Setup
configured and licensed. In addition to this. you also need four external disks to be
Copyrtght if; 2006 Symantec Corporation. All rights reserved
8-85Lab 7 Solutions: Resolving Hardware Problems

:ly Data Disks: Sularis: c It #dO -
clt#d5
HP-UX: c4tOdO -
c4tOd5
AIX:hdisk21-
hdisk26
Linux: sda - sdf
Location of Lab Scripts: !student!labs!sf!
sf50
object names
Copvnqht G;2006 Symaotec Corporation All rights reserved

Recovering a Temporarily Disabled Disk Group
Remove all disks except for one (namedgOl) from the namedg dis~ g;P.
vxdg -g namedg rmdisk namedg04 . l ~.
vxdg -g namedg rmdisk namedg03 C~~:
vxdg -g namedg rmdisk namedg02 .J
I
2 Create a Ig volume called namevoll in namedg disk group.
vxassist -g namedg make namevoll 19
3 Create a file system on namevoll and mount it to / namel.
mkdir /namel
4 Copy the contents of / etc/ defaul t directory to / namel and display the
contents of the file system.
ep -r jete/default /namel
Is -IR /~n~a~n~le~ll- ----------
-...~
5 Ask your instructor for the location of the fa i 1dg_ temp script, and note the
location here:
Script location: . _
6 Start writing to a file in the / namel file system at the background using the
following command:
dd if=/dev/zero of=/namel/testfile bs=1024
count=500000 &
7 In one terminal change to the directory containing the script and before the I/O
completes. execute faildg_temp namedg command.
Notes:
The faildg_temp script disables the single path to the disk in the disk
group to simulate a hardware failure. This is just a simulation and not a real
failure: therefore. the operating system will still be able to see the disk after
the failure. The script waits until you are ready with analyzing the failure.
to re-enablc the path to the disk in the disk group.
II'the 110 you started in step (, completes before you can simulate the
failure. you can start it again to observe the 110 failure.
ed /script_location
./faildg temp namedg
Disabling device_tag
Lab 7 Solutions: Resolving Hardware Problems 8-87
Copyright It, 2006 Symantec Corporation. All tights reserved

Enter e when you are ready for the disks to be re-
enabled:
8 Wait for the 110to fail and in another terminal observe the error displayed in
the system lug.
Solar-is, tail -f /var/adm/messages
Linux,
AIX
HP-lIX tail -f /var/adm/syslog/syslog.log
9 Use the vxdisk -0 alldgs list and vxdg list commands to
determine the statusof the disk group, and the disk.
vxdg list
The disk group should show asdisabled and the disk status should change
to online dgdisabled.
10 What happenedto the file system'!
The till:' system is also disabled.
11 When you are done with analyzing the impact of the failure, changeto the
terminal where the fa i ldg_ temp script is waiting and enter "e" to correct
the temporary failure.
Note: In a real failure scenario.after the hardware recovery, you would needto
first verify that the operating system can seethe disks and then verity that
Volume Manager hasdetected the change in status. If nut, you can force
VxVM to scan the disk by executing the vxdctl enable command. This
will not be necessaryfor this lab.
On the terminal where the faildg_ temp script is waiting:
Enter e when you are ready for the disks to be re-
enabled: e
12 Assuming that the failure was due to a temporary fiber disconnection and that
the data is still intact. recover the disk group and start the volume. Verity the
disk and disk group statususing the vxdi sk - 0 a Ildgs 1i stand vxdg
1i s t commands.
umount /namel
vxdg deport namedg
vxdg import namedg
vxdg list
Copynyht'~ 200n Syn:ilIHC',: Corporation All rights reserved

The disk group should now be enabled and the disk status should change
back to online.
13 Remount the file system and verify that the contents arc still there. Note that
you will need to perform a tile system check before you mount the tile system.
fsck -F vxfs /dev/vx/rdsk/namedg/namevoll
mount -F vxfs /dev/vx/dsk/namedg/namevoll /narnel
Is -IR /namel
I14 Unmount the tile system and remove namevoll. At the end of this section
you should be left with a namedg disk group with a single disk and three
initialized disks that are free to be used in a new disk group.
umount /namel
Preparation for Disk Failure Labs
Overview
The following sections use an interactive script to simulate a variety of disk failure
scenarios. Your goal is to recover lrom the problem as described in each scenario.
Use your knowledge of VxVM administration, in addition to the VxVM recovery
tools and concepts described in the lesson. to determine which steps to take to
ensure recovery. Aller you recover the test volumes, the script verifies your
solution and provides you with the result. You succeed when you recover the
volumes without corrupting the data.
For most of the recovery problems, you can use any ofthc VxVM interfaces: the
command line interface, the VERITAS Enterprise Administrator (VEA) graphical
user interface, or the vxdi skadm menu interface. Lab solutions are provided for
only one method. II'you have questions about recovery using interfaces not
covered in the solutions. seeyour instructor.
Setup
Due to the way in which the lab scripts work. it is important to set up your
environment as described in this setup section:
II'your system is set to use enclosure-based naming, then you must turn off
enclosure-based naming before running the lab scripts.
2 Create a disk group named testdg and add three disks (preferably of the
same size) to the disk group. Assign the following disk media names to the
disks: testdgOl. testdg02. and testdg03.
vxdisksetup -i device_tag(if"eccssar~')
vxdg init testdg testdgOl=device_tagl
testdg02=device_tag2 testdg03=device_tag3
Copyright!£: 2006 Symantec Corporation All right;; reserved

Note: If you do not have enough disks, you can destroy disk groups created in
other labs (for example, namedg) in order (0 create the testdg disk group.
3 Before running the automated lab scripts. set the DG environment variable in
your root profile to the name of the test disk group that you are using:
Solaris, vi / .profile
HP-liX DG=testdg; export DG
Linux vi /root/.bashrc
DG=testdg; export DG
Rerun your profile by logging out and logging back on, or manually running it.
4 Ask your instructor for the location of the lab scripts.
Note: This lab can only be performed on Solaris, HP-UX, and Linux.
Recovering from Temporary Disk Failure
In this lab exercise, a temporary disk failure is simulated. Your goal is to recover
all of the redundant and nonredundant volumes that were on thc failed drive. The
lab script run _di sks sets up the test volume configuration, simulates a disk
failure, and validates your solution for recovering the volumes. Ask your instructor
for the location of the run_disks script.
Before You Begin: Ensure that the environment variable DGis set to the name of
the testdg disk group. For example:
DG="testdg"
export DG
select option I. "Turned off drive (temporary failure)":
./run_disks
1) Lab 1 - Turned off drive (temporary failure)
2) Lab 2 - Power failed drive (permanent failure)
3) Lab 3 - Intermittent Failures (system still ok)
4) Optional Lab 4 - Intermittent Failures (system too
slow)
5) Optional Lab 5 - Turned off drive with layered
volume
6) Optional Lab 6 - Power failed drive with layered
volume
x) Exit
Your Choice? 1
Copyright;';' 2006 svroaotec Corporation All nqhts reserved

tes t 1 with a mirrored layout
Note: If you receive an error messageabout the / image file system becoming
by both volumes. Then. when you arc ready to power the disk back on. the
3 Assume that the failure was temporary. In a second terminal window. attempt
Assume that the drive that was turned off and then back on was cl t2dO
for a Solaris or HP-UX system or sdb for a Linux system (actual device
name will vary by system).
Notc: When performing recovery procedures, run vxprint and vxdisk
list oftcn to seewhat is changing after issuing recovery commands:
vxprint -g testdg -htr
To recover from the temporary failure:
a Ensure that the opcrating system recognizes the device:
Solaris devfsadm
Note: Becauseyou have not changedthe SCSI location of the
drive. running devfsadm may not be necessary.However,
running this command verifies the existence and validity of the
disk label. Prior to Solaris 7. you can usedrvconf ig and
disks.
HI'-LJX ioscan -c disk
insf -e
Linux partprobe Idev/sdb
b Verify that thc operating system recognizes the device:
Solaris prtvtoc Idev/rdsk/c1t2dOs2
HP-LJX ioscan -fnC disk
(Verify that the disk is ill CLAIMED statc.)
Linux fdisk -1 Idev/sdb
Lab 7 Solutions: Resolving Hardware Problems
Copyright f' 2006 Symantec Corporation. All rights reserved
I
B-91

c Force the VxVM configuration daemun tu reread all of the drives in
the system:
vxdctl enable
d Reattach the device tu the disk media record:
vxreattach
e Recover the volumes:
vxrecover
Start the nun redundant volume:
vxvol -g testdg -£ start test2
4 After YOIl recover the volumes, type e in the lab script window. The script
Recovering from Permanent Disk Failure
In this lab exercise, a permanent disk failure is simulated. Your goal is to replace
the tailed drive and recover the VOIIlIllCS as needed. The lab script run_disks
sets up the test volume configuration, simulates a disk failure, and validates your
solution for recovering the volumes. Ask your instructor for the location of the
run_disks script.
echo $DG
IfDG is not set, set it before you continue:
DG="testdg"
export DG
select option 2, "Power tailed drive (permanent failure r":
./run_disks
slow)
volume
Copytlghl~' 20(;6 Svmantec Corporaucn. All rights rescrvoo

volume
x) Exit
Your Choice? 2
test 1 with a mirrored layout
Note: If you receive an error messageabout the / image file system becoming
I2 Read the instructions in the lab script window. The script simulates a disk
by both volumes. The disk is detached by YxYM.
Assume that the failed disk is testdg02 (clt2dO for a Solaris or HP-lJX
system or sdb for a Linus system) and the new disk used to replace it is
cl t3dO for a Solaris or H P-lJX system or sdd for a Linux system (actual
device name will vary by system), which is originally uninitialized.
Note: When performing recovery procedures, run vxprint and vxdisk
list often to seewhat is changing after issuing recovery commands:
To recover from the permanent failure:
a Initialize the new drive:
Solarts, vxdisksetup -i clt3dO
Hl'-UX
Linux vxdisksetup -i sdd
b Attach the disk media name (testdg02) to the new drive:
Solaris. vxdg -g testdg -k adddisk testdg02=clt3dO
HP-UX
Linux vxdg -g testdg -k adddisk testdg02=sdd
c Recover the volumes:
Copyright ~ 2006 Symanrer. Corporation. All nqt.ts reserved

vxrecover
d Start the nonrcdundant volume:
vxvol -g testdg -f start test2
Alternatively, you can use the vxdiskadm menu interface:
a Invoke vxdiskadm:
vxdiskadm
b From the vxdiskadm main menu, select the option, "Replace a failed
or removed disk." When prompted, select clt3dO for a Solaris or
HP-lJX system or sdd for a Linux system to initialize and replace
testdg02.
Note: If you receive an error while using vxdiskadm about a
vxprint operation requiring a disk group, ignure the errur.
e Start the nunredundant volume:
5 When you have completed this exercise, if the disk device that was originally
used during disk failure simulation is in online invalid state, rcinuialize
the disk to prepare tor later labs.
For example:
Recovering from Intermittent Disk Failure (1)
In this lab exercise. intermittent disk failures arc simulated. but the system is still
OK. Your goal is to move data from the failing drive and remove the failing disk.
The lab script run_disks sets up the test volume configuration and validates
your solution for resolving the problem. Ask your instructor for the location of the
run_disks script.
Before You Begin: Check to ensure that the environment variable DGis set to the
echo $DG
If it is not set, set it before you continue:
DG="testdg"
export DG
Copyright '';' 2000 Svmautec Corporation. All nqhts reserved

select option 3. "Intermittent Failures (system still ok I":
./run_disks
slow)
volume
volume
x) Exit
Your Choice? 3
I
test1 with a mirrored layout
Note: If you receive an error message about the / image file system becoming
lull during volume setup. ignore the error message. This error will not have
2 Read the instructions in the lab script window. You are in limned that the disk
drive used by both volumes is experiencing intermittent failures that must be
addressed.
3 In a second terminal window. move the data on the failing disk to another disk.
and remove the failing disk.
Assume that testdg02 (clt2dO for a Solaris or HP-UX system or sdb
for a Linux system, and with plex tes tl- 01 from the mirrored volume
testl) is the drive experiencing intermittent problems (actual device
list often to see what is changing after issuing recovery commands:
To recover:
a Set the read policy to read from a preferred plex that is not on the
failing drive before evacuating the disk. This technique prevents
'xVM from accessing the failing drive during a read, if possible:
vxvol -g testdg rdpol prefer testl testl-02
Copyright '9 2006 Symantec Corporation All fights reserveo

b Evacuate data from the failing drive to one or more other drives by
using the vxdiskadm menu interface. Invoke vxdiskadm:
vxdiskadm
c From the vxdiskadm main menu, select the option, "Move volumes
from a disk." Evacuate the volumcs on testdg02 to another disk in
the disk group, such as testdg03.
d Remove the failing disk by using the vxdiskadm menu interface.
From the vxdiskadm main menu, select the option, "Remove a disk."
Remove the disk testdg02.
e Set the volume read policy back to the original read policy:
vxvol -g testdg rdpol select testl
Note: In this exercise, you still succeed even if you do not change the read
policy or you do not remove the failing disk after evacuation.
Warning: If the lab is repeated and a disk that has been used as a
replacement disk in a previous lab is now used as a new disk to replace the
failing disk without moving the volumes, the test results may succeed
although they should fail. If this happens, remove the volume called image
in the testdg disk group and re-run the lab.
4 Atier you resolve the problem. type e in the lab script window. The script
5 When you have completed this exercise, add the disk you removed from the
disk group back to the testdg disk group so that you can use it in later labs.
For example:
vxdg -g testdg adddisk testdg02=device_tag
COpYfI<,Ihl',~20(}6 Svmaruec Corporation All rig!'>!!'reserved

The next set of lab exercises is optional and may be performed iI'you have time.
These exercises provide additional recovery scenarios, as well as practice in
replacing physical drives and working with spare disks. A final activity explores
how to use the Support website, which is an excellent troubleshooting resource.
Optional Lab: Recovering from Intermittent Disk Failure (2)
In this optional lab exercise, intermittent disk failures are simulated, and the
system has slowed down significantly, so that it is not possible to evacuate data
from the failing disk. The lab script run_disks sets up the test volume
configuration and validates your solution for resolving the problem. Ask your
instructor for the location of the run _ di sks script.
echo $DG
I
DG="testdg"
export DG
select option 4. "Intermittent Failures (system too slow)":
./run_disks
slow)
volume
volume
xl Exit
Your Choice? 4
2 Read the instructions in the lab script window. You are informed that:
Lab 7 Solutions: Resotving Hardware Probtems 8-97
Copyright I£: 2006 Symantec Corporation All rights reserved

The disk drive used by both volumes is experiencing intermittent failures
that need to be addressed immediately.
The system has slowed down significantly, so it is not possible to evacuate
the disk before removing it.
3 In a second terminal window, perform the necessary actions to resolve the
problem.
Assume that testdg02 (c1t2dO for a Solaris or HP-UX system or sdb
for a Linux system and with plex test1- 01 from the mirrored volume
test1) is the drfve experiencing intermittent problems (actual device
vxdisk -0 a11dgs list
To recover:
a Remove the failing disk for replacement by using the vxdiskadm
menu interface. Invoke vxdiskadm:
vxdiskadm
b From the vxdiskadm main menu, select the option, "Remove a disk
for replacement". Remove the disk testdg02. 00 not use a
replacement disk yet.
vxprint operation requiring a disk group, ignore the error.
c To ensure that you have an uninitialized new disk to use as the
replacement disk, you may need to copy zeros to the beginning of the
failing disk and then uninitialize it before using it as the replacement
disk.
To carry out this task, you can use the / script_l ocat ion/bin/
cleandisk device_tag command where script_location is
the home directory from which you are running the automated lab
scripts. For example:
Solarls, /script - location/bin/cleandisk clt2dO
HP-UX
Linux /script locat~on/bin/cleandisk sdb-
d Replace the failed disk with a new disk by using the vxdiskadm menu
interface. From the vxdiskadm main menu, select the option,
"Replace a failed or removed disk." Select an uninitialized disk to
replace testdg02.
COPYrighl '~' 2006 S~rT'<lnlf'r Corporation All, .:',.,~, "('Iwd

e Start the nonredundant volume.
In this optional lab exercise. a temporary disk failure is simulated. Your goal is to
recover all of the volumes that were on the failed drive. The lab script run _di sks
sets up the test volume configuration and validates your solution for resolving the
problem. Ask your instructor for the location of the run _ d i sks script.
echo $DG
DG="testdg"
export DG
IOptional Lab: Recovering from Temporary Disk Failure· Layered
Volume
select option 5. "Turned off drive with layered volume":
./run_disks
slow)
volume
volume
x) Exit
Your Choice? 5
testl with a concat-mirror layout
Note: II'you receive an error message about the / image file system becoming
Copynghl rt: 2006 Svmantec Corporation. All riqhts reserved

power-off by saving and overwriting the private region on the drive that is used
by both volumes, Then. when you are ready to power the disk back on, the
3 Assume that the failure was temporary. In a second terminal window, attempt
Assume that the drive that was turned off and thcn back on was cl t2dO
for a Solaris or HP-UX system or sdb for a Linux system (actual dcvicc
namc will vary by system).
Notc: When pcrforming recovery procedures, run vxprint and vxdisk
To recover from thc temporary failure:
a Ensure that the opcrating system recognizcs thc device:
Solaris devfsadm
1'I0te: Because you have not changed the SCSI location of the
drive. running devf sadmmay not be necessary. However.
running this command vcruics the existence and validity of the
disk label. Prior to Solaris 7. you can use drvconf ig and
disks.
HP-UX ioscan -c disk
insf -e
Linux partprobe /dev/sdb
b Vcrify that thc operating system rccognizes the device:
Solaris prtvtoc /dev/rdsk/clt2dOs2
HP-UX ioscan -fne disk
(Verify that the disk is in CLAIMED state.)
Linux fdisk -1 /dev/sdb
c Forcc the VxVI1 conliguration daemon to reread all ofthe drives in
the system:
vxdctl enable
d Reattach the device to the disk media record:
vxreattach
CupYrlyhl'~ 20;')6Symantec Corporauon AUuulns reserved

e Recover the volumes:
vxrecover
Start the non redundant volume:
4 Alter you recover the volumes. type e in the lab script window. The script
IOptional Lab: Recovering from Permanent Disk Failure· Layered
Volume
In this optional lab exercise. a permanent disk failure is simulated. Your goal is to
replace the failed drive and recover the volumes as needed. The lab script
run_disks sets up the test volume configuration and validates your solution for
resolving the problem. Ask your instructor for the location of the run _ di sks
script.
echo $DG
DG="testdg"
export DG
From the directory that contains the lab scripts, run the script run_disks. and
select option 6, "Power failed drive with layered volume":
./run_disks
slow)
volume
volume
x) Exit
Your Choice? 6
Copyright i~2006 Symantec Corporation All nqhts reserved

testl with aconcur-mirror layout
Note: Ifyou receive an error message about the / image file system becoming
power-uffby saving and overwriting the private region on the drive that is used
by buth volumes, The disk is detached by YxYM.
3 In a second terminal window, replace the permanently jailed drive with either a
location. Then, recover the volumes.
Assume that the failed disk is testdg02 (el t2dO for a Solaris or HP-UX
system or sdb for a Linux system) and the new disk used to replace it is
elt3dO fur a Solaris HP-LJX system ur sdd for a Linux system, which is
originally uninitialized (actual device names will vary by system).
list uften to seewhat is changing after issuing recovery commands:
Tu recover from the permanent failure:
a Initialize the new drive:
Solaris. vxdisksetup -i clt3dO
1U'-liX
Linux vxdisksetup -i sdd
b Attach the disk media name (testdg02) tu the new drive:
Solaris, vxdg -g testdg -k adddisk testdg02=clt3dO
HP-UX
Linux vxdg -g testdg -k adddisk testdg02=sdd
c Recover the volumes:
vxrecover
d Start the nonredundant volume:
Alternatively, you can use the vxdiskadm menu interface:
8-102 VERITAS Storage Foondation 5.0 for UNIX: Fundamentals
COPYright! 2006 Symaruec Corporation All nghts reserved

a Invoke vxdiskadm:
vxdiskadm
b From the vxdiskadm main menu, select the option, "Replace a failed
or removed disk." When prompted, select c1t3dO for a Solaris or
HP-liX system or sdd for a Linux system to initialize and replace
testdg02.
e Start the nonredundant volume:
I4 Alter you recover the volumes, type e in the lab script window. The script
5 When you have completed this exercise, if the disk device that was originally
used during disk failure simulation is in onl ine inval id state, reinitialize
the disk to prepare for later labs.
For example:
The rest of this lab exercise includes optional lab instructions where you perform a
variety of basic recovery operations.
Optional Lab: Removing a Disk from VxVM Control
Destroy the testdg disk group and add the three disks back to the namedg
disk group. At this point you should have one disk group called namedgwith
four empty disks in it. There should be no volumes in the namedgdisk group.
If you had destroyed the namedgdisk group in previous lab sections, re-create
it.
vxdg destroy testdg
vxdg ini t namedg namedg01=devi ce_ tagl (if the disk group docs
not exist)
vxdg -g namedg adddisk namedg02=device_tag2
namedg03=device_tag3 namedg04=device_tag4
2 In the namedgdisk group create a IOO-MB. mirrored volume named
namevoll. Create a Veritas file system on namevoll and mount it to
/ namel directory.
vxassist -g namedg make namevol1 100m layout=mirror
mkfs -F vxfs /dev/vx/rdsk/namedg/namevol1
Note: On LiIlUX, usemkfs - t.
Copyright '~2006 Symantec Corporation. All nqhts reserved

Note: 011 Linux, usemount -to
3 Display the properties of the volume. In the table, record the device and disk
media name of the disks used in this volume.
vxprint -g namedg -thr
For example, the volume namevoll usesnamedg02 and namedg04:
Device Tag Disk Media Name
Solaris, Disk I clt2dO namedg02
HP-UX
Disk 2 clt3dO namedg04
Device Tag Disk ;Iedia Name
Linux Disk I sde namedg02
Disk 2 sdf namedg04
4 Remove one of the disks that is being used by the volume for replacement.
vxdg -g namedg -k rmdisk namedg02
5 Confirm that the disk was removed.
6 From the command line. check that the state of one of the plexes is DISABLED
and REMOVED.
7 If you are not already logged in VEA, start VEA and connect to your local
system. Check the status of the disk that has been removed.
III VEA, the disk is shown as disconnected, because the disk has been
removed for replacement.
8 Replace the disk back into the namedg disk group.
vxdg -g namedg -k adddisk namedg02=device_tag
where devi ce_ tag is c#t#d# for Solaris and HP-lJX, hdisk# for AIX
9 Check the status of the disks. What is the status of the replaced disk?
The status of the disk is ONLINE.
10 Display volume information. What is the state of the plexes ofnamevoll'!
Copynqbt ~,2000:.Symantec Corporauon. All ngh:s reserved

The plex using the disk you removed and replaced is marked RECOVER.
11 In YEA. what is the statusof the replaceddisk') What is the statusof the
volume')
The disk is reconnected; its status shows Imported as normal. Select the
volume in the left pane, and click the Mirrors tab in the right pane. The
plex is marked recoverable.
vxrecover
In YEA, the status of the plex changesto Recovering, and eventually to
Attached.
With vxprin t, thc status of the plcx changesto STALE and eventually to
ACTIVE.
I12 From the command line. recoverthe volume. During andalter recovery.check
the statusof the plex in anothercommandwindow and in VEA.
Optional Lab: Replacing Physical Drives (Without Hot Relocation)
Note: If you haveskipped the previous optional lab section called Removing a
Disk From YxYM Control. you may needto destroy testdg and add the three
disks back to the namedg disk group before you start this section. If you had
destroyedthe namedg disk group in previous lab sections.re-createit.
If necessary:
vxdg destroy testdg
vxdg init namedg namedgOl=device_tagl(ifthediskgroupdoesnot
exist)
namedg03=device_tag3 namedg04=device_tag4
Ensurethat the namedg disk group hasa mirrored volume called namevoll
with a Yeritas tile systemmountedon /namel. If not. createa IO()-MB
mirrored volume called namevoll in the namedg disk group. add a YxFS
file systemto the volume. and mount the file systemat the mount point
/namel.
[fnecessary:
vxassist -g namedg make namevoll 100m layout=mirror
Copyright£:: 2006 svmaotec Corporation All rights reservec

Note: On Linux, use mount - t.
2 If the vx r e I ocd daemon is running. stop it using psand ki 11,in order to
stop hot relocation from taking place. Verify that the vxrelocd processes are
killed before you continue.
Notes:
Ifyou have executed the run_di sks script in the previous lab sections,
the vxrelocd daemon may already be killed.
There are two v x r e Loc d processes on the Solaris platform. You must kill
both of them at the same time.
ps -ef I grep vxrelocd
kill -9 pidl [pid21
ps -ef I grep vxrelocd
enable commands. Ask your instructor for the location of thc script.
While using the script. substitute the appropriate disk device name for one of
the disks in useby namevoll. lor example on Linux use sbd,on Solaris and
HP-UX use clt8dO.
cd /script_location
./overwritepr device_tag
vxdctl disable
vxdctl enable
4 When the CITor occurs. view the status of the disks from the command line.
The physical device is no longer associated with the disk media name and
the disk group.
5 View the status of the volume from thc command line.
The plex displays a status of DISABLED NODEVICE.
6 In VEA, what is the status of the disks and volume')
Note: On the HP-UX platform, the vxdctl disable command may cause
the StorageAgcnt used by the VEA GUI to hang. If this happens, the VEA GUI
/opt/VRTSobc/pa133/bin/vxpa1ctr1 -a StorageAgent
-c restart
The status of the disk is Disconnected, and the volume has a Recoverable
status for the plex,
7 Rescan tor all attached disks:
Copynght (~2006 Svmanrec coroc.aucn All rights reserved

vxdctl enable
command. substitute the appropriate disk device name. for example on l.inux.
use sbd:
Note: This step is only necessary when you replace the failed disk with a brand
new one. If it were a temporary failure. this step would not be necessary.
I9 Bring the disk back under YxYM control:
vxdg -g namedg -k adddisk disk_name=devlce tag
where disk nameis the disk media name of the failed disk and
devi ce_ tag is the device name of the disk device used to replace the failed
one.
10 Check the status of the disks and the volume.
vxprint -thf
11 From the command line. recover the volume.
vxrecover
12 Check the status of the disks and the volume to ensure that the disk and volume
are fully recovered.
13 Unmount the / namel file system and remove the namevoll volume.
umount /namel
Optional Lab: Exploring Spare Disk Behavior
Note: If you have not already done so. destroy testdg and add the three disks
back to the namedg disk group before you start this section.
If necessary:
vxdg destroy testdg
namedg03=device_tagJ namedg04=device_tag4
You should have four disks (namedgO1 through namedg04) in the disk group
namedg. Set all disks to have the spare !lag on.
Copynghl··~ 2006 Syrnantec Corporation All rights reserved

vxedit -g namedg set spare=on namedgOl
vxedit -g namedg set spare=on namedg02
2 Createa IOO-MB mirroredvolumecalledsparevol.
vxassist -g namedg make sparevol 100m layout=mirror
Is thevolumesuccessfullycreated?Why or why not?
No, the volume is not created, and you receivethe error:
... Cannot allocate space for size block volume
The volume is not created becauseall disks arc setasspares,and
vxassist or VEA do not tind enoughfree spaceto create the volume.
3 Attempt to createthesamevolumeagain.but this time specifytwo disksto
use.Do not clearany spare flagson thedisks.
vxassist -g namedg make sparevol 100m
layout=mirror namedg03 namedg04
Notice that VxVM overrides its default and appliesthe two sparedisks to
the volume becausethe two disks were specilied by the administrator.
4 Removethe sparevol volume.
vxassist -g namedg remove volume sparevol
5 Verify thattherelocationdaemon(vxrelocd) is running.11'1101, startit as
follows:
vxrelocd root &
6 Removethe spare flagsfrom threeof thefour disks.
vxedit -g namedg set spare=off namedgOl
vxedit -g namedg set spare=off namedg02
vxedit -g namedg set spare=off namedg03
7 Createa IOO-MB concatenatedmirrored volumecalledspare2vol.
vxassist -g namedg make spare2vol 100m layout=mirror
8 Savetheoutputofvxprint -g namedg -thrtoafile.
vxprint -g namedg -thr > /tmp/savedvxprint
9 Displaythepropertiesof thespare2vol volume.In thetable,recordthe
deviceanddisk medianameof thedisksusedin this volume.Youarcgoing to
simulatedisk failure on oneof thedisks.Decidewhich disk Y0U aregoingto
fail.

For example, the volume spare2vol uses namedgOl and namedg02:
Device Tag I>isk Media Name
Disk I clt2dO namedgOl
I>isk 2 clt3dO namedg02
10 Next, simulate disk failure hy writing over the private region using the
enabl e commands.Ask your instructor for the location of the script.
While using the script. substitute the appropriate disk device name for one of
the disks in use by spare2voL for example on Linux use sbd, on Solaris
and HP-UX use clt8dO.
cd /scr~pt l ocst i on
./overwritepr device_tag
vxdctl disable
vxdctl enable
11 Run vxprint - 9 namedg - rth and compare the output to the vxprint
output that you saved earlier. What has occurred')
Note: You may need to wait a minute or two for the hot relocation to complete.
Hot relocation has taken place. The failed disk has a status of NODEVICE.
VxVM has relocated the mirror of the failed disk onto the designated
spare disk.
12 In VEA, view the disks. Notice that the disk is in the disconnected state.
Note: On the IIP-UX platform, the vxdctl disable commandmay cause
the StorageAgent used by the VEA GUI to hang. If this happens, the VEA GUI
-c restart
13 Run vxdisk -0 alldgs list. What do you notice')
This disk is displayed as a failed disk.
14 Rcscan for all attached disks.
vxdctl enable
15 In VEA, view the status of the disks and the volume.
Uighlight the volume and click each of the tabs ill the right pane. You call
also select Actions->Volume View and Actlons=c-Dtsk View to view
status information.
Copyright 'E 2006 Symantec Corporation. All rigt15 reserved
I

command, substitute the appropriate disk device name, for example on Linux,
use sbd:
17 Bring the disk back under VxVM control and into the disk group.
vxdg -g namedg -k adddisk namedg##=device_tag
18 In VEA, undo hot relocation for the disk.
Right-click the disk group and select Undo Hot Relocation. In the dialog
box, select the disk for which you want to undo hot relocation and click
OK. After the task completes, the alert on the disk group should be
removed.
Alternatively, from the command line, run:
vxunreloc -g namedg namedg##
where namedg## is the disk media name of the failed and replaced disk.
19 Wait until the volume is fully recovered before continuing. Check to ensure
that the disk and the volume arc fully recovered.
Note: The vxprint command shows the subdisk with the UR tag.
20 Remove the spare2vol volume.
vxassist -g namedg remove volume spare2vol
Optional Lab: Using the Support Web Site
Access the latest information on VERITAS StorageFoundation.
Note: If you are working in the Virtual Academy lab environment, you may
not be able to accessthe Veritas Technical Support web site, becausethe DNS
configurauon was changedduring software installation by the prepare_Ds
script. To restore the original DNS configuration, changeto the directory
containing the lab scripts. execute the re store _DSscript and try to access
the web site again.
If necessary:
cd /scr~pt locat~on
./restore ns
Go to the VERIT AS Technical Support Web site at
http://support.veritas.com.
From the Select Product Family menu, select Storage Foundation. From
the Select Product menu, select Storage Foundation for UNIX.

On the next window, click the "documents published in the last 30 days"
link.
This will show you any information that has been published for Storage
Foundation in the last 30 days.
2 What is the VERITAS Support mission statement'?
Hint: It is in the Support Handbook (page 3).
Select the Support Handbook link at the bottom of the page. On page 3:
"'VI' will provide world-class technical expertise acting as the customer
advocate to maximize their investment in VERIT AS solutions."
I3 How many on-site support visits are included in a Extended Support contract?
How about with a Business Critical Support?
Hint: In the Support Handbook, see table on page 4 and explanation on page S.
Extended Support: No on-site support visits are included.
Business Critical Support: Six on-site support visits are included.
4 Which AIX platform is supported for Storage Foundation s.o'?
Select Compatibility & Reference link under Support Resources title.
Set Show Document Types drop-down list to Manuals and
Documentation. Set Show Results For drop down list to 5.0(AIX).
Select the VERIT AS Storage Foundation (tm) 5.0 - Release Notes (AIX).
See the Supported software section with the following information:
- AIX 5.2 ML6 (legacy)
- AIX 5.3 TL4 with SJl4
Veritas 5.0 products also operate on AIX 5.3 with SP3, but you must
install an AIX interim fix. See the following TechNote for information
on downloads, service pack availability, and other important issues
related to this release.
http://support.veritas.com/docs/282024
5 Access a recent Hardware Compatibility List for Storage Foundation. Which
Brocade switches are supported by VERITAS Storage Foundation and Iligh
Availability Solutions 5.0 on Solaris?
Select the Compatibility & Reference tab.
Click on the appropriate MCL link.
6 Where would you locate the Patch with Maintenance Pack I for VERITAS
Storage Solutions and Cluster File Solutions 4.0 fix Solaris?
Select the Software Updates & Downloads link on the left navigation bar
and locate the patch.
7 Perform this step only if you are working in the Virtual Academy lab
environment. If you have executed the restore_ns script to restore the
Copyright <12006 Symantcc Corporation. All rights reserveu

name resolution configuration at the beginning of this lab section in step I,
change to the directory containing the lab scripts and execute the
prepare _ns script before you continue,
If necessary:
cd /script locat~on
./prepare_ns

Glossary
A
access control list (ACL) A list of
users or groups whu have access privileges
to a specified file. A file may have its own
ACL or may share an ACL with other tiles.
ACLs allow detailed access permissions
lor multiple users and groups.
active/active disk arrays This type of
muliipathed disk array enables you to
access a disk in the disk array through all
the paths to the disk simultaneously.
active/passive disk arrays This type
of multipathed disk array allows one path
to a disk to be designated as primary and
used to access the disk at any time.
agent A process that manages predefined
YERITAS Cluster Server (YCS) resource
types. Agents bring resources online. take
resources offline. and monitor resources to
report any state changes to YCS. When an
agent is started. it obtains configuration
information from yeS and periodically
monitors the resources and updates yeS
with the resource status.
AIX coexistence label Data on disk
that identifies the disk to the AIX volume
manager (LYM) as being controlled by
VxYM. The contents have no relation to
YxYM lD Blocks.
alert An indication that an error or failure
has occurred on an object on the system.
When an object fails or experiences an
error. an alert icon appears.
alert icon An icon that indicates that an
error or failure has occurred on an object
on the system. Alert icons usually appear
in the status area of the YEA main window
and on the affected object's group icon.
allocation unit A basic structural
component of YxFS. The YxFS Version 4
and later file system layout divides the
entire file system space into fixed size
allocation units. The first allocation unit
starts at block zero. and all allocation units
are a fixed length of 32K blocks.
application volume A volume created
by the intelligent storage provisioning
(lSP) feature of YERITAS Volume
Manager (YxYM).
associate The process of establishing a
relationship between YxYM objects; for
example. a subdisk that has been created
and defined as having a starting point
within a plcx is referred to as being
associated with that plex.
associated plex A plcx associated with
a volume.
associated subdisk A subdisk
associated with a plex.
asynchronous writes A delayed write
in which the data is written to a page in the
system's page cache. but is not written to
disk before the write returns to the caller.
This improves performance. but carries the
risk of data loss if the system crashes
before the data is flushed to disk.
atomic operation An operation that
either succeeds completely or Iai Is and
leaves everything as it was before the
operation was started. If the operation
succeeds. all aspects of the operation take
effect at once and the intermediate states of
change are invisible. If any aspect of the
operation fails, then the operation aborts
without leaving partial changes.
attached A state in which a YxYM
object is both associated with another
object and enabled lor use,
Copyright if: 2006 Svrnantec Corporation. All rigtm. reserved.
Gtossary-1

attribute Allows the properties of a LUN
to be defined in an arbitrary conceptual
space. such asa manufacturer or location.
B
back-rev disk group A disk group
created using a version of Yx YM released
prior to the release of CDS. Adding CDS
functionality rolls over to the latest disk
group version number.
block The minimum unit of data transfer
to or from a disk or array.
Block-Level Incremental Backup (BLI
Backup) A YERITAS backup capability
that docs not store and retrieve entire files.
Instead. only the data blocks that have
changed since the previous backup are
backed up.
boot disk A disk used for booting
purposes. This disk may be under YxYM
control for some operating systems.
boot disk group A disk group that
contains the disks from which the system
may be booted.
bootdg A reserved disk group name that
is an alias for the name of the boot disk
group.
browse dialog box A dialog box that is
used to view and/or select existing objects
on the system. Most browse dialog boxes
consist of a tree and grid.
buffered 1/0 During a read or write
operation. data usually goes through an
intermediate tile system butter before
being copied between the user buffer and
disk. If the same data is repeatedly read or
written, this file system buffer acts as a
cache. which can improve performance.
See direct 1'0 and nnbuffcrcd//0.
button A window control that the user
clicks to initiate a task or display another
object (such as a window or menu).
c
capability A feature that is provided by a
volume. For example, a volume may
exhibit capabilities, such as performance
and reliability to various degrees. Applies
to the ISP feature ofYxYM.
CDS disk A disk whose contents and
attributes are such that the disk can be used
for CDS as part of a CDS disk group. In
contrast, a non-CDS disk can neither be
used for CDS nor be part of a CDS disk
group.
CDS disk group A YxYM disk group
whose contents and attributes are such that
the disk group can be used to provide tor
cross-platform data sharing. In contrast, a
non-CDS disk group (that is. a back-rev
disk group or a current-rev disk group)
cannot be used lor cross-platform data
sharing. A CDS disk group can only
contain CDS disks.
CFS YERITAS Cluster File System.
check box A control button used to
select optional settings. A check mark
usually indicates that a check box is
selected.
children Objects that belong to an object
group.
clean node shutdown The ability ofa
node to leave the cluster gracefully when
all access to shared volumes has ceased.
clone pool A storage pool that contains
one or more full-sized instant volume
snapshots of volumes within a data pool.
Applies to the ISP feature ofYxYM.
cluster A set of host machines (nodes)
that share a set of disks.
Glossary-2 VERITAS Storage Foundation 5.0 for UNIX: Fundamentals

cluster file system A Y xFS fi lc system
mounted on a selected volume in cluster
(shared) mode.
cluster manager An externally-
provided daemon that runs on each node in
a cluster. The cluster managers on each
node communicate with each other and
inform YxYM of changes in cluster
membership.
cluster mounted file system A shared
file system that enables multiple hosts to
mount and perform file operations on the
same tile. A cluster mount requires a
shared storage device that can be accessed
by other cluster mounts of the same file
system. Writes to the shared device can be
performed concurrently from any host on
which the cluster file system is mounted.
To be a cluster mount, a file system must
be mounted using the mount -0
cl uster option. See local monnted fi!e
.11'.111.'111.
cluster-shareable disk group A disk
group in which the disks are shared by
multiple hosts (also referred to as a shared
disk group).
column A set of one or more subdisks
within a striped plex. Striping is achieved
by allocating data alternately and evenly
across the columns within a plex.
command log A log file that contains a
history of YEA tasks performed in the
current session and previous sessions.
Each task is listed with the task originator,
the start/finish times. the task status. and
the low-level commands used to perform
the task.
concatenation A layout style
characterized by subdisks that arc arranged
sequentially and contiguously.
configuration copy A single copy of a
configuration database.
configuration database A set of
records containing detailed infonnation on
existing YxYM objects (such as disk and
volume attributes). A single copy of a
configuration database is called a
con figuration copy.
contiguous file A file in which data
blocks are physically adjacent on the
underlying media.
Cross-platform Data Sharing (CDS)
Sharing data between heterogeneous
systems (such as SUN and HP). where
each system has direct access to the
physical devices used to hold the data and
understands the data on the physical
device.
current-rev disk group A disk group
created using a version of YxYM
providing CDS functionality; however, the
CDS attribute is not set. If the CDS
attribute is set for the disk group, the disk
group is called a CDS disk group.
CVM The cluster functionality of
YERITAS YxYM.
D
data blocks Blocks that contain the
actual data belonging to files and
directories.
data change object (DCO) A YxYM
object that is used to manage information
about the FastResync maps in the DCO log
volume. Both a DCO object and a DCO
log volume must be associated with a
volume to implement Persistent
FastResync on that volume.
data pool The first storage pool that is
created within a disk group. Applies to the
ISP feature ofYxYM.
data stripe This represents the usable
data portion of a stripe and is equal to the
stripe minus the parity region.
Copynght ~.2006 Symanter Corporation. All fights reserved
Glossary-3

data synchronous writes A form of
synchronous 1/0 that writes the tile data to
disk before the write returns, but only
marks the inode for later update. If the tile
size changes, the inode will be written
before the write returns. In this mode, the
tile data is guaranteed tu be on the disk
before the write returns, but the inodc
modification times may be lust if the
system crashes,
Deo log volume A special volume that
is used tu huld Persistent FastResync
change maps.
defragmentation Reorganizing data on
disk to keep tile data blocks physically
adjacent so as to reduce access times.
detached A state in which a VxVM
object is associated with another object,
but not enabled for use.
Device Discovery Layer (DOL) A
facility of VxVM for discovering disk
attributes needed for VxVM DMP
operation.
device name The device name or
address used to access a physical disk.
SUch as cOtOdOs2 on Solaris, cOtOdO
on HP-UX, hdiskl on AIX, and hda on
Linux. In a SAN environment, it is more
convenient to use cnclosure-bascd nannng,
which forms the device name by
concatenating the name of the enclosure
(such as encO) with the disk's number
within the enclosure, separated by an
underscore (for example, encO _ 2). The
term disk access name can also be used to
refer to a device name.
dialog box A window in which the user
submits information tu VxVM. Dialog
boxes can contain selectable buttons and
fields that accept information.
direct extent An extent that is
referenced directly by an inodc.
direct I/O An unbuffered form of 1/0
that bypasses the tile system's buffering of
data. With direct 1/0, the tile system
transfers data directly between the disk and
the user-supplied butler. See buffered 110
and IIl1hll/t('I'{'t/ I/O.
dirty region logging The procedure by
which the Vx VM monitors and logs
modi fications to a plex. A biunap of
changed regions is kept in an associated
subdisk called a /og subdisk,
disabled path A path to a disk that is
not available for 110. A path can be
disabled duc to real hardware failures or if
the user has used the vxdmpadm
disable command on that controller.
discovered direct I/O Discovered
Direct 1;0 behavior is similar to direct 1/0
and has the same alignment constraints,
except writes that allocate storage or
extend the tile size do not require writing
the inodc changes before returning to the
application.
disk A collection of read/write data
blocks that are indexed and can be
accessed fairly quickly. Each disk has a
universally unique identifier.
disk access name The name used to
access a physical disk. The c#t#d#s#
syntax identities the controller, target
address, disk, and partition. The term
device 1I<1111e can also be used to refer to the
disk access name.
disk access records Configuration
records used to specify the accesspath to
particular disks. Each disk access record
contains a name, a type, and possibly some
type-specific information. which is used by
the VxVM in deciding huw to access and
manipulate the disk that is defined by the
disk access record.
Glossary-4 VERITAS Storage Foundation 5.0 for UNIX.' Fundamentals
COPyllght ., 2()Ob Svmaruec ccroorauc» AI! fights rcservec

disk array A collection of disks
logically arranged into an object. Arrays
tend to provide benefits, such as
redundancy or improved performance.
disk array serial number This is the
serial number of the disk array.: it is
usually printed on the disk array cabinet or
can be obtained by issuing a vendor-
specific SCSI command to the disks on the
disk array. This number is used by the
[)MP subsystem to uniquely identify a disk
array.
disk controller The controller (lIBA)
connected to thc host or the disk array that
is represented as the parent node of the
disk by the operating system; it is called
the disk controller by the multipathing
subsystem ofVxVM.
For example, if a disk is represented by the
device name:
/devices/sbus@lf,O/
QLGC,isp@2,lOOOO/sd@8,O:c
then the disk controller for the disk
sd@8 r °:cis:
QLGC,isp@2,lOOOO
This controller (HBA) is connected to the
host.
disk enclosure An intelligent disk array
that usually has a backplane with a built-in
Fibre Channel loop, and which permits
hot-swapping of disks.
disk group A collection of disks that arc
under VxVM control and share a common
configuration. A disk group configuration
is a set of records containing: detailed
information on existing VxVM objects
(such as disk and volume attributes) and
their relationships. Each disk group has an
administrator-assigned name and an
internally defined unique !D.
disk group ID A unique identifier used
to identify a disk group.
disk ID A universally unique identifier
that is given to each disk and can be used
to identify the disk, even if it is moved.
disk media name A logical or
administrative name chosen for the disk,
such as disk03. The term disk name is
also used to refer to the disk media name.
disk media record A configuration
record that idcnti fics a particular disk, by
disk ID, and gives that disk a logical (or
administrati ve) name.
disk name A logical or administrative
name chosen for a disk that is under the
control ofVxVM, such as disk03. The
term disk I1I1!dia 17(1l11eis also used to refer
to a disk name.
dissociate The process by which any
link that exists between two Vx VM objects
is removed. For example, dissociating a
subdisk trom a plcx removes the subdisk
from the plex and adds the subdisk to the
tree space pool.
dissociated plex A plex dissociated
from a volume.
dissociated subdisk A subdisk
dissociated from a plex.
distributed lock manager A lock
manager that runs on different systems and
ensures consistent access to distributed
resources.
dock To separate or attach the main
window and a subwindow,
Dynamic Multipathing (DMP) The
method that VxVM uses to manage two or
more hardware paths directing 1/0 to a
single drive.
Copyright t,[, 20G6 Symantec Corporation. All nqhts rcsetvco
Glossary-5

E
enabled path A path to a disk that is
available for 110,
encapsulation A process that converts
existing partitions on a specified disk to
volumes, Ifany partitions contain tile
systems, thc tile system table entries are
modi tied so that the file systems are
mounted on volumes instead,
Encapsulation is not applicable on some
systems,
enclosure A disk array,
enclosure-based naming An
alternative disk naming method, beneficial
in a SAN environment. which forms the
device name by concatenating the name of
the enclosure (such as encO) with the
disk's number within the enclosure,
separated by an underscore (for example,
encO 2),
extent A group of contiguous file system
data blocks that arc treated as a unit. An
extent is detined by a starting block and a
length,
extent attributes The extent allocation
policies associated with a tile,
external quotas file A quotas tile
(named quotas) must exist in the root
directory of a tile system for quota-related
commands to work, See internal quotas
fil« and quotasfile,
F
fabric mode disk A disk device that is
accessible on a Storage Area Network
(SAN) through a Fibre Channel switch,
FastResync A fast resynchronization
feature that is used to perform quick and
efficient rcsynchronizution of stale mirrors,
and to increase the efficiency of the
snapshot mechanism.
Fibre Channel A collective name for the
fiber optic technology that is commonly
used to set up a Storage Area Network
(SAN),
file system A collection of files
organized together into a structure, The
UNIX tile system is a hierarchical
structure consisting of directories and files,
file system block The fundamental
minimum size of allocation in a tile
system, This is equivalent to the uf s
fragment size,
file system snapshot An exact copy
ota mounted tile system at a specific point
in time, Used to perform online backups,
fileset A collection of tiles within a tile
system,
fixed extent size An extent attribute
associated with overriding the default
allocation policy of the tile system,
fragmentation The ongoing process on
an active tile system in which the tile
system is spread further and further along
the disk, leaving unused gaps or fragments
between areas that are in use, This leads to
degraded performance because the tile
system has fewer options when assigning a
tile to an extent.
free disk pool Disks that are under
Yx VM control, but that do not belong to a
disk group,
free space An area of a disk under
YxYM control that is not allocated to any
subdisk or reserved for use by any other
YxYM object.
free subdisk A subdisk that is not
associated with any plcx and has an empty
putil [0] field,
Glossary-6 VERITAS Storage Foundation 5,0 for UNIX: Fundamentals
Copyuqht ~ 2006 Svmaruoc Corporauon All nghts reserved

G
gap A disk region that does not contain
VxVM objects (subdisks l.
GB Gigabyte (230
bytes or 1024
megabytes ).
graphical view A window that displays
a graphical view of objects. In VEA. the
graphical views include the Object View
window and the Volume Layout Details
window.
grid A tabular display of objects and
their properties. The grid lists YxVM
objects. disks. controllers. or file systems.
The grid displays objects that belong to the
group icon that is currently selected in the
object tree. The grid is dynamic and
constantly updates its contents to reflect
changes to objects.
group icon The icon that represents a
specific object group.
GUI Graphical User Interface.
H
hard limit The hard limit is an absolute
limit on system resources lor individual
users for lile and data block usage on a file
system. See quotas.
host A machine or system.
hostid A string that identities a
host to the VxVM. The hostid for a host
is stored in its volboot file. and is used
in defining ownership of disks and disk
groups.
hot relocation A technique of
automatically restoring redundancy and
access to mirrored and RAID-5 volumes
when a disk fails. This is done by
relocating the affected subdisks to disks
designated as spares and/or tree space in
the same disk group.
hot swap Refers to devices that can be
removed from. or inserted into. a system
without lirst turning oil the pmer supply
to the system.
HP-UX coexistence label Data on disk
that identi lies the disk to the HP volume
manager (LYM) as being controlled by
VxVM. The contents of this label are
identical to the contents of the VxYM ID
block.
I/O clustering The grouping of multiple
110 operations to achieve better
performance.
indirect address extent An extent that
contains references to other extents. as
opposed to tile data itself. i single indirect
address extent references indirect data
extents, A double indirect address extent
references single indirect address extents.
indirect data extent An extent that
contains file data and is referenced through
all indirect address extent.
initiating node The node on which the
system administrator is running a uti lity
that requests a change to VxVM objects.
This node initiates a volume
rcconfiguration.
inode A unique identifier lor each file
within a file system. which also contains
metadaia associated with that file.
inode allocation unit A group of
consecutive blocks that contain inode
allocation information for a given filesct.
This information is in the form of a
resource summary and a tree inode map.
Intelligent Storage Provisioning
(ISP) ISP enables you to organize and
manage your physical storage hy creating
application volumes. ISP creates volumes
lrorn available storage with the required
Copyright ((: 2006 Svmantec Corporation. All rights reserved
Gtossary-7

capabilities that you specify. To achieve
this, ISP selects storage by referring to the
templates for creating volumes.
intent The intent of an ISP application
volume is a conceptualization of its
purpose as defined by its characteristics
and implemented by a template.
intent logging A logging scheme that
records pending changes to the tile system
structure. These changes are recorded in a
circular intent log tile.
internal quotas file VxFS maintains an
internal quotas file for its internal usage.
The internal quotas tile maintains counts of
blocks and inodes used by each user. See
extcrnal quotusfilc and quotas.
J
JBOe The common name for an
unintelligent disk array which may, or may
not, support the hot-swapping of disks. The
name is derived from "just a bunch of
disks."
K
K Kilobyte (2111
bytes or 1024 bytes).
L
large file A tile larger than 2 gigabytes.
VxFS supports files up to two terabytes in
size.
large file system A file system more
than 2 gigabytes in size. VxFS supports
tile systems up to 32 terabytes in size.
latency For tile systems, this typically
refers to the amount of time it takes a given
tile system operation to return to the user.
launch To start a task or open a window.
local mounted file system A file
system mounted on a single host. The
single host mediates all file system writes
to storage from other clients. To be a local
mount. a tile system cannot be mounted
using the mount -0 cluster option.
See cluster mountedfile svsteni.
log plex A plex used to store a RAID-5
log. The term log plex may also be used to
refer to a dirty region logging plex.
log subdisk A subdisk that is used to
store a dirty region log.
LUN Logical Unit Number. Each disk in
an array has a LUN. Disk partitions may
also be assigned a LUN.
M
main window The main VEA window.
This window contains a tree and grid that
display volumes, disks. and other objects
on the system. The main window also has a
menu bar and a toolbar.
master node A node that is designated
by the software as the "master" node. Any
node is capable of being the master node.
The master node coordinates certain
VxVM operations.
mastering node The node to which a
disk is attached. This is also known as a
disk (JI'I/E'I'.
MB Megabyte (220 bytes or 1024
kilobytes).
menu A list of options or tasks. A menu
item is selected by pointing to the item and
clicking the mouse.
menu bar A bar that contains a set of
menus for the current window. The menu
bar is typically placed across the top of a
window.
metadata Structural data describing the
attributes of files on a disk.
Cor."lllt,j 2f}116 Symaotec Corporation All fights reserveo.

mirror A duplicate copy of a volume and
the data therein (in the form of an ordered
collection of subdisks). Each mirror is one
copy of the volume with which the mirror
is associated. The terms mirror and plcx
can be used synonymously.
mirroring A layout technique that
mirrors the contents of a volume onto
multiple plexes. Each plex duplicates the
data stored on the volume. but the plcxes
themselves may have different layouts.
multipathing Where there are multiple
physical access paths to a disk connected
to a system. the disk is called multipathcd.
Any software residing on the host (for
example. the DMP driver) that hides this
fact from the user is said to provide
multipathing functionality.
multivolume file system A single file
system that has been created over multiple
volumes. with each volume having its own
properties.
N
node In an object tree. a node is an
element attached to the tree. In a cluster
environment. a node is a host machine in a
cluster.
node abort A situation where a node
leaves a cluster (on an emergency basis)
without attempting to stop ongoing
operations.
node join The process through which a
node joins a cluster and gains access to
shared disks.
nonpersistent FastResync A form of
FastResync that cannot preserve its maps
across reboots of the system because it
stores its change map in memory.
o
object An entity that is defined to and
recognized internally by YxYM. The
YxYM objects arc: volume. plex, subdisk,
disk. and disk group. There arc actually
two types of disk objects- -one type for the
physical aspect of the disk and the other for
the logical aspect.
object group A group of objects of the
same type. Each object group has a group
icon and a group name. In YxYM. object
groups include disk groups. disks.
volumes. controllers. free disk pool disks.
uninitializcd disks. and file systems.
object location table (OLT) The
information needed to locate important file
system structural elements. The OLT is
written to a fixed location on the
underlying media (or disk).
object location table replica A copy
of the OLT in case ofdata corruption. The
OLT replica is written to a fixed location
on the underlying media (or disk).
object tree A dynamic hierarchical
display ofYxYM objects and other objects
on the system. Each node in the tree
represents a group of objects of the same
type.
Object View Window A window that
displays a graphical view of the volumes.
disks. and other objects in a particular disk
group. The objects displayed in this
window are automatically updated when
object properties change. This window can
display detailed or basic information about
volumes and disks.
p
page file A fixed-size block of virtual
address space that can he mapped onto any
of the physical addresses available on a
system.
Ccpvriqbt ,t 2006 Svmantec Corporation All nqtrts reserved
Gtossary-9

parity A calculated value that can be
used to reconstruct data after a failure.
While data is being written to a RAID-5
volume. parity is also calculated by
performing an exclusive OR (XOR)
procedure on data. The resulting parity is
then written to the volume. Ifa portion ofa
RAID-5 volume fails, the data that was on
that portion of the failed volume can be re-
created from the remaining data and the
parity.
parity stripe unit A RAID-5 volume
storage region that contains parity
information. The data contained in the
parity stripe unit can be used to help
reconstruct regions of a RAID-5 volume
that are missing because of 110 or disk
failures.
partition The standard division of a
physical disk device. as supported directly
by the operating system and disk drives.
path When a disk is connected to a host,
the path to the disk consists of the Host
Bus Adapter (HBA) on the host. the SCSI
or fiber cable connector and the controller
on the disk or disk array. These
components constitute a path to a disk. A
failure on any of these results in DMP
trying to shift all l/Os for that disk onto the
remaining (alternate) paths.
pathgroup In case of disks that arc not
multipathed by vxdmp, VxVM will see
each path as a disk. In such cases, all paths
to the disk can be grouped. This way only
one of the paths from the group is made
visible to VxVM.
persistent FastResync A form of
FastResync that can preserve its maps
across reboots of the system by storing its
change map in a DCO log volume on disk.
persistent state logging A logging
type that ensures that only active mirrors
are used for recovery purposes and
Glossary-10
prevents failed mirrors from being selected
for recovery. This is also known as kernel
logging,
physical disk The underlying storage
device, which mayor may not be under
VxVM control.
platform block Data placed in sector 0,
which contains OS-specitic data for a
variety of platforms that require its
presence for proper interaction with each
of those platforms. The platform block
allows a disk to masquerade as if it was
initialized by each of the specific
platforms.
plex A duplicate copy of a volume and
the data therein (in the form of an ordered
collection of subdisks). Each plex is one
copy of the volume with which the plex is
associated.
preallocation The preallocation of
space for a file so that disk blocks will
physically be part of a file before they are
needed. Enabling an application to
preallocate space for a tile guarantees that
a specified amount of space will be
available for that tile. even if the file
system is otherwise out of space.
primary fileset A fileset that contains
the tiles that are visible and accessible to
users.
primary path In active/passive type disk
arrays, a disk can be bound to one
particular controller on the disk array or
owned by a controller. The disk can then
be accessed using the path through this
particular controller.
private disk group A disk group in
which the disks are accessed by only one
specific host.
Copvr, ;1'1" 200f) Symantec Corporation All fight" rcserveo

private region A region of a physical
disk used to store private, structured
VxVM information. The private region
contains a disk header. a table of contents,
and a configuration database.
The table of contents maps the contents of
the disk. The disk header contains a disk
!D. All data in the private region is
duplicated for extra reliability.
properties window A window that
displays detailed information about a
selected object.
public region A region of a physical
disk managed by VxVM that contains
available space and is used tor allocating
subdisks.
Q
Quick 110 file A regular VxFS file that is
accessed using the : : cdev : vxf s :
extension.
Quick I/O for Databases Quick 1'0 is a
VERITAS File System feature that
improves database performance by
minimizing read/write locking and
eliminating double buffering of data. This
allows online transactions to be processed
at speeds equivalent to that of using raw
disk devices. while keeping the
administrative benefits of file systems.
QuickLog VERITAS QuickLog is a high
performance mechanism for receiving and
storing intent log information for VxFS file
systems. Quickl.og increases performance
by exporting intent log information to a
separate physical volume.
quotas Quota limits on system resources
for individual users for file and data block
usage on a fi le system. See hard limit and
soli limit.
quotas file The quotas commands read
and write the external quotas file to get or
change usage limits. When quotas are
turned on. the quota limits are copied from
the external quotas file to the internal
quotas file. See external quotasfile
ill/emu/ quotastile. and quotas,
R
radio buttons A set ol'buuons used to
select optional settings. Only one radio
button in the set can be selected at any
given time. These buttons toggle on or ofT.
RAID A Redundant Array of Independent
Disks (RAID) is a disk array set up with
part of the combined storage capacity used
for storing duplicate information about the
data stored in that array. This makes it
possible to regenerate the data if a disk
failure occurs.
read-writeback mode A recovery
mode in which each read operation
recovers plex consistency for the region
covered by the read. Plex consistency is
recovered by reading data from blocks of
one plex and writing the data to all other
writable plcxcs.
reservation An extent attribute
associated with preallocating space for a
file.
root disk The disk containing the root
file system. This disk may be under VxVM
control.
root disk group In versions ofVxVM
prior to 4.0. a special private disk group
had to exist on the system. The root disk
group was always named rootdg. This
requirement does not apply to VxVM 4..v
and higher.
root file system The initial file system
mounted aspart of the UNIX kernel startup
sequence.
Copyright'L 2006 Symantcr- Corporanon All rights re<;p.rvcri
Glossary-11

root partition The disk region on which
thc root tile system resides.
root volume The VxVM volume that
contains the root tile system, if such a
volume is designated by the system
configuration.
rootability The ability to place the root
tile system and the swap device under
VxVM control. The resulting volumes can
then be mirrored to provide redundancy
and allow recovery in the event of disk
fai lure.
rule A statementwritten in the VERITAS
ISI' language that specifies how a volume
is to be created.
5
scroll bar A sliding control that is used
to display different portions of the contents
ora window.
Search window The VEA search tool.
The Search window provides a set of
search options that can be used to search
for objects on the system,
secondary path In active/passive type
disk arrays, the paths to a disk other than
the primary path are called secondary
paths. A disk is supposed to be accessed
only through the primary path until it fails,
after which ownership of the disk is
transferred to one of the secondary paths.
sector A unit of size, which can vary
between systems. A sector is commonly
512 bytes.
sector size Sector size is an attribute of
a disk drive (or SCSI LUN till' an array-
type device) that is set when the drive is
formatted. Sectors are the smallest
addressable unit of storage on the drive,
and arc the units in which the device
performs I/O.
Glossary-12
shared disk group A disk group in
which the disks are shared by multiple
hosts (also referred to as a cluster-
shareable disk group).
shared VM disk A VxVM disk that
belongs to a shared disk group.
shared volume A volume that belongs
to a shared disk group and is open on more
than one node at the same time.
Shortcut menu A context-sensitive
menu that only appears when you click a
specific object or area.
slave node A node that is not
designated as a master node.
slice The standard division of a logical
disk device. The terms partition and slice
arc sometimes used synonymously.
snapshot A point-in-time copy of a
volume (volume snapshot) or a tile system
(tile system snapshot).
snapped file system A file system
whose exact image has been used to create
a snapshot ti Ie system.
snapshot file system An exact copy of
a mounted file system at a specific point in
time. Used to do online backups. Scefi!e
SI'stCII/ snapshot,
soft limit The soft limit is lower than a
hard limit. The soli limit can be exceeded
for a limited time. There are separate time
limits lor tiles and blocks. See hard fill/it
and quota.
spanning A layout technique that
permits a volume (and its tile system or
database) that is too large to tit on a single
disk to span across multiple physical disks.
sparse plex A plex that is not as long as
the volume or that has holes (regions of the
plcx that do not have a backing subdisk ).
Copvnqm " 20G() SyrT1cll't.;( Corporauon All JI<JhIS rcserveu

splitter ; bar that separates two panes of
a window (such as the object tree and the
grid). A splitter can be used to adjust the
sizes of the panes.
status area An area of the main window
that displays an alert icon when an object
fails or experiences some other error.
Storage Area Network (SAN) ;
networking paradigm that provides easily
rcconfigurable connectivity between any
subset of computers, disk storage and
interconnecting hardware such as switches.
hubs and bridges.
storage checkpoint ; facility that
provides a consistent and stable view of a
file system or database image and keeps
track of modi fied data blocks since the last
checkpoint.
storage pool ; policy-based container
within a disk group in VxVM. for use by
ISP. that contains LUNs and volumes.
storage pool definition ; grouping of
template sets that defines the
characteristics of a storage pool. Applies to
the ISP feature ofVxVM.
storage pool policy Defines how a
storage pool behaves when more storage is
required, and when you try to create
volumes whose capabilities are not
permitted by the current templates. Applies
to the ISP feature ofVxVM.
storage pool set ; bundled definition
of the capabilities ofa data pool and its
clone pools. Applies to the ISP feature of
VxVM.
stripe A set of stripe units that occupy
the same positions across a series of
columns.
stripe size The SUIll of the stripe unit
sizes that compose a single stripe across all
columns being striped.
stripe unit Equally sized areas that arc
allocated alternatelv on the subdisks
(within columns) 0'1' each striped plcx. In
an array. this is a set oflogically
contiguous blocks that exist on each disk
before allocations are made from the next
disk in the array. ; stripe IIl1i! may also be
referred to as a stripe element.
stripe unit size The size or each stripe
unit. The default stripe unit size is 32
sectors ( I()K). / stripe unit si:e has also
historically been referred to as a stripe
width.
striping / layout technique that spreads
data across several physical disks using
stripes. The data is allocated alternately to
the stripes within the suhdisks of each
plex.
structural fileset A special filesct that
stores the structural elements of a VxFS
file system in the form of structural files.
These files define the structure or the file
system and arc visible only when using
utilities such as the file system debugger.
subdisk ; consecutive set of contiguous
disk blocks that form a logical disk
segment. Subdisks can be associated with
plcxes to form volumes.
super-block A block containing critical
information about the file system such as
the file system type, layout, and size. The
VxFS super-block is always located 8192
bytes from the beginning of the file system
and is 8192 bytes long.
swap area A disk region used to hold
copies of memory pages swapped out by
the system pager process.
swap volume ; VxVM volume that is
configured for use as a swap area.
synchronous writes A form or
synchronous I/O that writes the lile data to
disk. updates the inode times, and writes
Copyright {. 2006 Svmantec Corpoeauon All rights reserved
Glossary-t S

the updated inode to disk. When the write
returns to the caller, both the data and the
inode have been written to disk.
T
task properties window A window
that displays detailed information about a
task listed in the Task Request Monitor
window.
Task Request Monitor A window that
displays a history of tasks performed in the
current VEA session. Each task is listed
with the task originator. the task status. and
the start/ finish times for the task.
T8 Terabyte (2~o bytes or 1024
gigabytes ).
template A meaningful collection oflS!'
rules that provide a capability for a
volume. Also known as a volume template.
template set consists of related
capabilities and templates that have been
collected together for convenience to
create IS!' volumes.
throughput For tile systems, this
typically refers to the number of 1/0
operations in a given unit of time,
toolbar A set otbuuons used to access
VEA windows. These include another
main window. a task request monitor. an
alert monitor, a search window, and a
customize window.
transaction A set of configuration
changes that succeed or fail as a group.
rather than individually. Transactions arc
used internally to maintain consistent
configurations.
tree A dynamic hierarchical display of
objects on the system. Each node in the
tree represents a group of objects of the
same type.
Glossary-14
u
ufs The UNIX tile system type. Used as
parameter in some commands.
UFS The UNIX tile system; derived from
the 4.2 Berkeley Fast File System.
unbuffered 1/0 JlO that bypasses the tile
system cache to increase I/O performance.
This is similar to direct l/O, except when a
tile is extended. For direct 1/0, the inode is
wriuen to disk synchronously; for
unbuffered 1/0. the inode update is
delayed. See buffered 110and direct I/O.
uninitialized disks Disks that are not
under Vx YM control.
user template Consists of related
capabilities and templates that have been
collected together till' convenience for
creating IS!' application olumcs.
v
VCS VERITAS Cluster Server.
VEA VERITAS Enterprise Administrator
graphical user interface.
VM disk A disk that is both under VxVM
control and assigned to a disk group. VM
disks are sometimes referred to as Volume
Mal/agel' disks or simply disks. In the
graphical user interface, VM disks are
represented iconically as cylinders labeled
D.
VMSA Volume Manager Storage
Administrator. an earlier version of the
VxVM C1UI used prior to VxVM version
3.5.
volboot file A small tile that is used
to store the host ID of the system on which
VxVM is installed and the values of
bootdg and defaultdg.
CUl>ynght _ 2001) Svmautcc Corporuuon All nqhts reserved

volume A virtual disk or entity that is
made up of portions of one or more
physical disks. i volume represents an
addressable range of disk blocks used by
applications such as file systems or
databases. A volume is a collection offrom
one to 32 plexes.
volume configuration device The
volume configuration device (/ dev/vx/
conf ig) is the interface through which all
configuration changes to the volume
device driver are performed.
volume device driver The driver that
forms the virtual disk drive between the
application and the physical device driver
level. The volume device driver is
accessed through a virtual disk device
node whose character device nodes appear
in /dev/vx/rdsk, and whose block
device nodes appear in / dev/vx/ dsk.
volume event log The volume event
log device (/ dev/ vx/ event) is the
interface through which volume driver
events are reported to the utilities.
Volume Layout Window A window
that displays a graphical view of a volume
and its components. The objects displayed
in this window are not automatically
updated when the volume's properties
change.
volume set A volume set allows several
volumes to be treated as a single object
with one logical 110 interface, Applies to
the ISP feature ofVxVM.
volume template i meaningful
collection of ISP rules that provide a
capability for a volume. Also known as a
template.
Volume to Disk Mapping Window A
window that displays a tabular view of
volumes and their underlying disks. This
window can also display details such asthe
subdisks and gaps on each disk.
vxconfigd The VxVM
configuration daemon, which is
responsible for making changes to the
VxVM configuration. This daemon must
be running before VxVM operations can
be performed.
vxfs The VERITAS File System type.
Used as a parameter in some commands.
VxFS VERITAS File System.
VxVM V[RITAS Volume Manager.
VxVM ID block Dat:l on disk that
indicates the disk is under VxVM control.
The VxVM ID Block provides dynamic
VxVM private region location, GUID, and
other information.
Copyright oi.: 2006 Symantec Corporation. All rights reserved
Glossary-15

C()pyrlght; 2006 Svmnun-c Corl'or.:lllull AIlllqlns reservco

Index
Files and Directories
/dev/vx/dsk 3·7
/dev/vx/rdsk 3-7
/etcdefauh/ts 6-6
'ctc/default/vxassist a-tu
/ctc/filcsystems 3·17,3·19.3·20
iele/fs/vxf,6·5
/erc/tstuh 3-17, 3·20
ICle/rc.dirc:2.d/S02vxvm-recover 7·23
lelc/rc2.d/S50isisd 2-22
/etc/system 2·7
/etc/vfs 6·6
/ctc/vfsrab 3·17, 3·20
/etc/vxclm 2·6
InpliVRTS/bin 6·5
iopllVRTS/inslall 'logs 2-11
lopliVRTSlman 2·19
iopl/VRTSob/bin 2-21
10plIVRTSvxfsisbin 6·5
/sbin 6·5
Isbin 'fs 6·5
'usr/lib/fs.vxfs 6·5
/v ar/vx/isis/vx isis.log 2·23
A
address-length pair 6·4
AIX disk 1-4
AIX physical volume 1·4
array 1·9
B
backing lip the VxVM configuration 7·11
Bad Block Relocation Area 1·4
BBRA 1·4
block clustering 6·3
block device file 3-18
block-based allocation 6·3
bootdg 3-7, 3·33
c
cos i-n
CDS disk 1·12
CDS disk groups
converting disk groups 5·25
CDS disk layout 1·11
cfgmgr 7·14
chts 3·20
(,L1 2-16
cu commands in VEA 2·18
cluster 2-8
Cluster File System 2·9
cluster group 3·10
cluster management 3·3
column 4·4
command line interface 2·16. 2-19
concatenated 3·14, 4·9
Concatenated Mirrored 3·15. 4·9, 4·23
concatenated volume 1-16, 4·3
creating 4·10
concarcnution 1·16
advantages 4·7
disadvantages 4·7
configuration backup and restoration 7·11
configuration database 7·6
controller 1·7
creating a layered volume 4-18
creating a volume 3-12
CLI 3-12
errs 3·19
cron 6·12
cross-platform data sharing 1·11
convening disk groups 5·25
requirements for CDS disk groups 5-24
D
data change object 3·22
data redundancy 1·16
databases on file systems 2-9
delaultdg 2·12, 3·7
Index-l
Copyright G2006 Svmautec Corporation All nqtns reserved

defrugmentution
scheduling 6-12
dcfragmcnting a lile system 6-11
deponing a disk group
and renaming 5-17
10 new host 5-17
VEA 5-18
destroying a disk group 3-33
lLl 3-33
VEA 3-33
devtsadrn 7-14
device path 3-25
dcvicetug 3-25
directory fragmental ion 6-9
dirty region logging 5-7
disaster recovery Inlro-10
disk
adding new 7-14
adding 10 a disk group in VEA 3-11
AIX 1-4
configuring for VxVM 3-4
displaying summary infonnation 3-25
failing 7-4
forced removal 7-21
HP-UX 1-4
Linux 1-5
naming 1-7
recognizing by operating system 7-14
removing in VEA 3-32
replacing tailed in vxdiskadm 7-15
replacing in CLI 7-15
replacing in VEA 7-15
uninirializing 3-32
unrelocaring 7-26
viewing in ell 3-24
viewing informal ion about 3-26
disk accessname 3-5
disk access record 1-14. 3-5
disk array 1-9
multiparhed 1-9
disk enclosure 2-12
disk failure 7-4
parlial7-22
permanent 7-7
resolving intermittent failure 7·20
temporary 7-7
disk failure handling 7-4
disk group
adding a disk in VEA 3-11
clearing host locks 5-19
creating 3-8
creating in VEA 3-10
creating in vxdiskadm 3-9
definition 1-13
deporting 5-17
destroying 3-33
destroying in CLI 3-33
destroying in VEA 3-33
displaying deponed 3-28
displaying tree space in 3-28
displaying properties for 3-28
forcing an import 5-19
high availability 3-6
importing 5-19
importing and renaming 5-19
importing as temporary in ell 5-20
purpose 1-13, 3-6
renaming in VEA 5-22
reserved names 3-7
shared 3-10
disk group contigurauou 1-13
disk group I D 3-25
disk group properties
viewing 3-29
Disk Group Properties window 3-29
disk group versions 5-23
disk ID 3-25
disk initialization 3-4
disk layout 1-11
changing 3-4
disk media name 1-13. 3-5, 3-8, 7·6
dclaull1-13
disk media record 7-6
disk name 3-25
disk naming 3-8
AIX 1-8
HP-UX 1-7
Linux 1-8
Solaris 1-7
disk properties 3-27
disk replacement 7-13
disk spanning 1-15
disk status
Deponed 3-26
Disconnected 3-26
External 3·26
Index-2 VERITAS Storage Foundation 5,0 for UNIX: Fundamentals

Free 3-26
Imported 3-26
Not Initialized 3-26
online 3-24
online invalid 3-24
disk structure 1-3
Disk View window 4-17
disks
adding to a disk group 3-8
displaying detailed information 3-25
evacuating data 3-31
renarnmg 5-21
uninitialized 3-4
dynamic lUN
resizing 5-15
dynamic mulripathing 2-12.3-3
E
ENABLED stare 7-19
encapsulation 3-4
enclosure 2-12
enclosure-based naming 2-12
benefits 3-3
error disk status 7-6
evacuating a disk 3-31
exclusive OR 4-6
EXT2 6-5
EXT36-5
Extended File System 6-5
extended partition 1·5
extent 6·4
extent fragmemat ion 6-9
extent-based allocation 6·3. 6·4
extents
defragmenting 6·11
F
rAilED disks 7·4
rAILING disks 7·4
FastResync 3·15
rct s-zo
fdisk 1·5
Fibre Channel 2·12
file change log 6·20
compared to intent log 6·20
file system
adding to a volume 3·16. 3-18
adding to a volume in ell 3·18
consistency checking 6-16
defragmcnring 6·11
lragmcntat ion 6·9
fragmentation rep(1I1S6·10
fragmentation types 6·9
intent log 6·15
intcnt log resizing 6·17
logging and performance 6·-19
logging options 6·18
mounting at boot 3·20
resizing 5-14
resizing in VFA 5·12
resizing methods 5-11
file system free space
identifying 6·8
file system type 3·16. 6·8
FlashSnap 2·8
torced removal of a disk 7·21
fragmentation 6·9
directory 6·9
extent 6·9
free space pool 3·5
fsadrn 5·14. 6·9. 6·10
fsck 6·15. 6·16
tsck pass 3·17
G
group name 3·25
H
HI'S 6·5
l lierarchical File System 6·5
high availability 2·8. 5·16
hOSIlocks
clearing 5..19
hostid 3·25
hot rclocat ion
definition 7·22
failure detection 7·23
notification 7·23
VERITAS Storage Foundation 5_0 for UNIX: Fundamentals
Copylighl© 2006 Symantec Corporation. AJIrighls reserved
lndex-S

process 7·23
recovery 7-23
selecting space 7·24
unrclocating a disk 7·26
HP·UX disk 1-4
I/O failure
identifying 7-4
importing a disk group
and renaming 5·19
torcing 5-19
VEA 5-20
initialize zero 3-15
inodc 6-4
insf 7·14
installation log tile 2·11
installation menu 2·10
installer 2-11
installlS 2-11
installing V"VM 2-10
package space requirements 2·7
verifying on AIX 2-15
verifying on HP·UX 2·14
veri lying on Linux 2·15
veri lying on Solaris 2·14
verifying package installation 2·14
installp 2·11
installs!' 2·11
installvm 2·11
Intelligent Storage Provisioning 3-10. 3·22
intent log
resizing 6-17
intent logging 6·15
interfaces 2-16
command line inter/ace 2·16
VERIT AS Enterprise Administrator 2·16
vxdiskadm 2-16
intermittent disk failure
resolving 7-20
ioscan 7·14
iosize 6-13
J
JFS 1-5. 6-5
JFS26·5
Joumalcd File System 6·5
journaling 6-15
K
kernel issues
and VxFS 2-7
L
layered volume 1·16. 4-18
advantages 4-19
creating 4·18
creating in ell 4·23
disadvantages 4-19
preventing creation 3·15
viewing in ell 4-24
viewing in VEA 4-24
layered volume layouts 4-22
licensing 2·5
generating a license key 2·6
obtaining a license key 2·5
vl.icense 2-6
Linux disk 1-5
listing installed packages 2-14
load balancing 4-7
location code 1-8
logging 3·15. 5·7
and VxFS performance 6-19
for mirrored volumes 5·7
logging options
for a file system 6-18
logical unit number 1-7
Logical Volume Manager 14
logtype 4-12
lsdcv 7-14
IsIS 3·20
Islpp 2·15
LUN 1·7
and resizing V"VM structures 5-15
LVM 1-4
M
man 2-19
lndex-d VERITAS Storage Foundation 5,0 for UNIX: Fundamentals
COPlIfIgt1t 2(106Svrnantec Ccrporanon All nghts reservco

manual pages 2·19
maxfilcsize 6·13
metadara6·3
mirror
adding 5·5
adding in CLI 5·6
adding in VEA 5·6
removing 5·5
mirror-concat 4·22
mirrored volume 1·16, 4·5
creating 4·12
mirroring 1·16
advantages 4·8
disadvantages 4-8
enhanced 4-18
mirroring a volume 3-15, 4-9
mirrors
adding 4·12
mirror-stripe layout 4-20
mkdir 3-18
rnkfs 3·18
mkfs options 6·7
mrnap 6-14
mount 3-18, 6·18
mount at boot 3-17
CLI 3-20
mount opt ions
delaylog 6-18
log 6·18
tmplog 6-18
mount poi nt 3·17
moving a disk
vxdiskadm 5·15
multipathed disk array 1·9
N
naming disks
defaults 3-8
ncol 4-11
New Volume wizard 3-13
newfs 3·18
nlog 4-12
nmirror 4-12
node 2·8
NODEVICE state 7·21
nodg 3·7
nostripe 4·10
o
Object Data Manager 1·8
off-host processing 2·8
online disk status 7·6
online invalid status 3·24
online status 3·24
operating system versions 2·3
ordered allocat ion 4·27
order of columns 4·28
order of mirrors 4·28
organization principle 3·10
p
packages
listing 2·14
space requirements 2·7
parity 1-16,4·6
partial disk iailure 7-22
partition 1·7
PAl116·5
permanent disk failure 7·7
physical disk
naming 1·7
Physical Volume Reserved Area 1·4
pkgadd 2·11
pkginfo 2·14
plex 1·14.4·5
definition 1·14
naming 1-14
plcx name
default 1·14
Preferences window 2·17
primary partition 1·5
pri vate region 1·11, 3·4. 7-4
private region size 1·11
AIX 1·11
HP·UX 1·11
Linux 1·11
Solaris 1·11
projection 4·17
prtvtoc 7·14
Copyright~: 2006 Svmantec Como-anon All rights reserved
lndex-S

public region 1-11. 1-13. 7-4
PVRA 1-4
Q
resizing a dynamic LUN 5-15
rcsizing a file system 5·14
rcsizing a volume 5-10
VEA 5-12
with vxnssist 5-14
with vxresize 5-13
resiziug a volume and file system 5-11
rcsizing a volume with a file system 5-10
response file 2·11
rliuk 3-22
rpm 2-11. 2-15
Quick 1/0 2-9
R
RAID 1-15
RAID array
benefits with VxVM Intro-10
RAID levels 1-15
RAID-5 column 4-6
RAID-5 volume 1-16.4-6
random read 6-14
random write 6-14
raw device tile 3-18
read policy 5·8
changing in CLI 5-9
changing in VEA 5-9
preferred plcx 5·8
round robin 5-8
selected pic" 5·8
recovering a volume
VEA 7-16
s
recovering volumes
and volume states 7-19
redundancy 1-16
relocating subdisks 7-24
REMOVED state 7-21
removing a disk
forced 7-21
VEA 3-32
S95vx'Jll-recover 7-23
SAN 2-12
SAN management 3-3
selected plex read policy 5-8
sequential read 6-14
sequential write 6-14
size of a volume 3-14
slice 1-7
sliced disk 1-12
snap object 3-22
software packages 2·7
space requirements 2-7
spare disks
managing 7-25
STALE state 7-19
removing a volume 3·30
renaming a disk 5-21
renaming a disk group 5·22
replacing a disk 7-13
CLI 7·15
VEA 7·15
replacing a failed disk
vxdiskadm 7·15
replicated volume group 3·22
Rescun option 7-14
resilience 1-16
resilient volume 1-16
storage
allocating lor volumes 4-25
storage area network 2-12
storage attributes
specifying lor volumes 4-25
specifying in VEA 4-26
storage cache 3-22
stripe unit 4-4. 4-6
default si/e 3-14. 4-9
striped 3-14. 4-9
Striped Mirrored 3-15. 4-9. 4-23
striped volume 1-16,4-4
creating 4-11
stripe-mirror 4-22
stripe-mirror layout 4-21
stripeuuit 4-11
lndex-G VERITAS Storage Foundation 5,0 for UNIX: Fundamentals
Copl/flght G 2006 SYl'1311tCC Corporation All fights reserved

striping 1·16
advantages 4·7
disadvantages 4·8
subdisk 1·14
definition 1·14
subdisk name
default 1·14
subvolume 4·18
summary fi le 2·11
support for Vx V M 2·4
swinstall z-t t
swlist 2·14
T
target 1·7
Task History window 2·18
tasks
clearing history 2·18
technical support for VxVM 2·4
temporary disk failure 7·7
true mirror 4·5
true mirroring 1·16
type 3·25
u
UFS 6·5
allocation 6·3
uninitialized disks 3-4
UNIX File System 6·5
unrelocating a disk 7·26
upgrading a disk group version 5·23
user interfaces 2·16
v
VEA 2·16
adding a disk to a disk group 3·11
adding a mirror 5·6
changing volume read policy 5·9
clearing task history 2·18
creating a disk group 3·10
creating a layered volume 4·23
creating a volume 3·13
deporting a disk group 5·18
destroying a disk group 3·33
disk properties 3-27
Disk View window 4·17
displaying volumes 3·23
importing a disk group 5·20
installing 2·21
installing the serv er and client 2·21
monitoring events and tasks 2·23
multiple views of objects 2·17
recovering a volume 7·16
remote administration 2·17
removing a disk 3-32
replacing a disk 7·15
resizing a volume 5·12
scanning disks 7·14
security 2·17
setting preferences 2·17
starting 2·22, 2·23
lask History window 2·18
viewing a layered volume 4·24
viewing CLI commands 2·18
viewing disk group properties 3·29
viewing disk information 3·26
Volume Layout window 4·14
Volume to Disk Mapping window 4·15
Volume View window 4·16
VrRITAS Cluster Fik System 2·9
VERITAS Cluster Server 2·8
VLRITAS Enterprise Administrator 2·16.2·17
VLRITAS hie System 6·5
VERITAS Quick 1'0 for Databases 2·9
VFRITAS Volume Manager 2-9
VrRIT AS Volume Replicator lntro-f O, 3·22
versrorung
and disk groups 5·23
V(JDA 1·5
VCiRA 1·4
virtual storage objects 1·10
vl.icensc 2·6
vol subdisk num 1·14
volboot 3·7
volume 1·10. 3·5
accessing 1·10
adding a tile system 3-16
adding a Ilk system in CLl 3·18
adding a minor 5·5
adding a mirror in VEA 5·6
creating 3·12
lndex-?VERITAS Storage Foundation 5.0 for UNIX' Fundamentals
Copyrighllf', 2006 Svmantec Corporation All rights reserved

creating a layered volume 4-18
creating in Cl.l 3-12
creating layered in CLI 4-23
creating layered in VEA 4-23
creating mirrored and logged 4-12
creating on specific disks 4-26
definition 1-10, 1-14
disk requirements 3-12
estimating size 4-13
expanding the size 5-10
layered layouts 4-22
logging 3-15
mirroring 3-15, 4-9
recovering in VEA 7-16
reducing the size 5-10
removing 3-30
removing a mirror 5-5
rcsizing 5-10
resizing in VEA 5-12
rcsizing methods 5-11
resizing with vxassist 5-14
resizing with vxrcsizc 5-13
specifying ordered allocation 4-27
specifying storage attributes in VEA 4-26
starting manually 5-20
viewing layered in CLI 4-24
viewing layered in VEA 4-24
volume attributes 3-13
Volume Group Descriptor Area 1-5
Volume Group Reserved Area 1-4
volume layout 1-15
concatenated 1-16
displaying in Cl.I 3-21
layered 1-16
mirrored 1-16
RAID-S 1-16
selecting 4-3
striped 1-16
Volume Layout window 4-14
Volume Manager control 1-11
Volume Manager disk 1-13
naming 1-13
Volume Manager Support Operations 2-16,2-
20
volume name
default 1-14
volume read policy 5-8
changing in Cl.I 5-9
changing in VEA 5-9
volume recovery 7-13
Volume Replicator 2-8
volume size 3-14
volume states
utter attaching disk media 7-18
after recovering volumes 7-19
alter running vxreauuch 7-12
alter temporary disk failure 7-12
after volume recovery 7-19
Volume to Disk Mapping window 4-15
Volume View window 4-16
volumes
allocating storage for 4-25
vrtsadm 2-22
VRTSap 2-7
VRTStj'pm 2-7
VRTShdoc 2-7
VRTSfspro 2-7
VRTSmuob 2-7
,RTSob 2-7
VRTSobadmin 2-7
VRTSobgui 2-7
VRTStep 2-7
'RTSvmdoc 2-7
Vk TSvmman 2-7
'RTSv mpro 2-7
VRTSvxts 2-7
v xassisr 3-12, 5-11, 5-14
vxassist growby 5-14
vxusxis: growro 5-14
vxussist shrinkby 5-14
vxussist shrinkto 5-14
vxbcnch
options 6-14
vxcunfigbackup 7 -11
vxconligbackupd 7-11
vxconfigrestore 7-11
vxdctl enable 3-5, 3-11, 7-6, 7-14
vxdg destroy 3-33
vxdisk list 3-9, 3-24, 3-25, 7-6, 7-14
vxdisk rcsize 5-15
vxdiskadm 2-16, 2-20, 3-4
creating a disk group 3-9
replacing a Jailed disk 7-15
Index-8 VERITAS Storage Foundation 5.0 for UNIX: Fundamentals

starting 2·20
vxdiskunsetup 3·32
VxFS 6·5
allocation 6·3, 6·4
and logging 6·19
command locations 6·5
command syntax 6·6
dcfragmenting 6·11
file system switchout mechanisms 6·6
file system type 6·8
Iragmentarionreports 6·10
fragmentation types 6·9
idenii rying free space 6·8
intcnt log 6·15
intent log resizing 6·17
logging options 6·18
maintaining consistency 6·16
resizing 5·14
resizing in VEA 5·12
using by default 6·6
vxinstall 2·11
vxmake 4·19
vxprint 3·21, 4·24
vxreal1ach 7·16
vxrecover 7·16
vxrelocd 7·23
vxresize 5·11, 5·13
vxunreloc 7-26
VxVM
configuration backup 7·11
user interfaces 2-16
VxVM and RAID arrays lntrc-f O
VxVM configuration daemon 3-5
vxvol rdpol prefer 5·9
vxvol rdpol round 5-9
vxvol rdpol select 5-9
vxvol stopall 5-18
x
XOR 1·16,4-6
VERITAS Storage Foundation 5.0 for UNIX. Fundamentals Index-9
Copyright~' 2006 Symantec Corporation. All fignts reserved

Index-10 VERITAS Storage Foundation 5.0 for UNIX: Fundamentals
Cupyrlgtlt 2006 Sym,H11CC Corporation All fights reserved

Veritas storage foundation_5.0_for_unix_-_fundamentals

More Related Content

What's hot (20)

Similar to Veritas storage foundation_5.0_for_unix_-_fundamentals (20)

Recently uploaded (20)

Veritas storage foundation_5.0_for_unix_-_fundamentals