Drp For Menora

DRP BCP for Menora Pini Cohen EVP & Senior Analyst Architecture & Infrastructure Strategies [email_address]

Market Status and Trends DRP BCP The second Lebanon war has raised the BCP\DRP issue. Organizations realize that all of Israel might be under missile threat This has two implications: Better shelters are considered for Data Centers Out of Israel BCP\DRP solutions are considered DRP is more and more related to Regulations . Israel Standard for Risk Management #5300 is starting to effect the DRP\BCP market. We expect more vertical regulations to follow Interesting project – EDS at FIBI with HDS – DRP outside Israel with 10MB/SEC line and journaling

Agenda Current BCP\DRP situation Definitions Categories Pitfalls Justification BCP\DRP service levels Related Technologies DRP related technologies

Gaps in Best Practices <40% (less then!) of BC plans address resumption Moving from backup site to primary site <50% (less then!) of BC plans address Personnel transportation Telecommunication and network outages Only 66% of organizations regularly validate data backups Only 65% of BC plans include policies on speaking to the media One-third of plans do not include recovery “sequences” for business functions 75% of recovery configurations are “ out of sync ” with their production configurations Significant lack of senior-executive participation in planning Only 25% of organizations include BC training for all key personnel

DR\BCP Facts As Datacenters consume more electricity, the chance for fire increases Primary Reason for DR/BC 33% Regulatory Compliance 34% Stakeholder Protection 16% Past Business Interruption 7% Public Relations 10% Other Causes of Data Loss Hardware System 44% Human Error 32% Software 14% Virus 7% Natural Disaster 3%

Organizations Without Documented Plans Source: Cutter Emergency Preparedness: The Need for Enterprise-Wide Efforts

What do you consider to be your organization’s biggest threat with regard to being able to continue business? Source: Cutter Emergency Preparedness: The Need for Enterprise-Wide Efforts

Approximately what percentage of your organization’s total IT budget is dedicated to emergency preparedness, business continuity, and disaster recovery? Source: Cutter Emergency Preparedness: The Need for Enterprise-Wide Efforts

Agenda Current BCP\DRP situation Definitions Categories Pitfalls Justification BCP\DRP service levels DRP related technologies

Definitions BCM (Business Continuity Management) The development, maintenance and implementation of strategies; plans and actions to ensure the continued availability of critical business processes and services Disaster Recovery Planning ( DRP ) is a subset of Business Continuity Planning (BCP) that focuses on the processes, plans, and procedures to recover an organization’s technology infrastructure to support critical business processes and functions within a predetermined time after a disaster or unplanned disruption in operations. www.tenders.tas.gov.au/domino/dtf/dtf.nsf/0/1ed10d1fa4fcdb39ca256f1d00060555/$FILE/BCM%20Seminar%20220306.ppt

Example of BCM activities \ considerations Evacuation Setting up an information centre, to register employees and make an inventory of missing or wounded people Care for employees; families and victims; community Setting up communication and IT networks Creating alternative office space Reallocation - Transportation & hotels near the alternative office space Managing / Recovering day to day business Security (physical security) Reputation Management Source: www.tenders.tas.gov.au/domino/dtf/dtf.nsf/0/1ed10d1fa4fcdb39ca256f1d00060555/$FILE/BCM%20Seminar%20220306.ppt stki modifications

The Ongoing Trend: Eliminating Data Loss on Recovery IT must deliver current-state recovery capability by application or LOB For each application, business leaders have traditionally defined: RTOs: Recovery time objectives RPOs: Recovery point objectives LBO: Lost business opportunity Required recovery classes Financial impact of outages (e.g., hour, day, week, month) IT should provide a categorization framework for the evaluation Crisis Classification Recovery From RPO State RTO Lost Business Opportunity Recovery Point Failure Point Lost Data RPO Time Restart Point

Market status and Trends IBM BCRS selected clients: Menora, Dexia, Ayalon, Mekorot, Novolog, Yashir Insurance, New Kopel, Saifan, Poalim Sahar, Given imaging, Postil, Danel, Psagot, Albat, Harel Insurance, IEC, Amidar, Technion, Keren Makafet, Tambor, Mehish, Malam IDBI Med1 – Hosting selected clients: Teva, Leumit Health Services, Asuta, Migdal, Brom, Broadcom, TI, Tadiran Kesher, Clal Insurance, Cafronet Med1 – DRP selected clients: Leumi Card, Bank Israel, YES , ZIM, AudioCodes, Partner (transmission pop) , Meshek and Calcala Company , Netivey Gas, Smarteam, AIG, Shva, Bituach Haklai HP is hosting (DRP – selected clients) - Tel Aviv Stock Exchange , Oradio, BNP Bank, Mashbir Machsanei Ofna, Bank Yahav, Elihu Insurance, HSBC Bank, Mashbir Haklai All Rights Reserved @STKI Moshav Bnei Zion, Israel +972 9 74 444 74 www.stki.info

BCM - A Viewpoint BEING PROPERLY PREPARED IS A COMPLEX SCIENCE Marsh www.tenders.tas.gov.au/domino/dtf/dtf.nsf/0/1ed10d1fa4fcdb39ca256f1d00060555/$FILE/BCM%20Seminar%20220306.ppt

Agenda Current BCP\DRP situation Definitions Pitfalls Justification BCP\DRP service levels DRP related technologies

Pitfalls and mistakes – BCP \ BCM IT is in charge BCP vs. DRP - Lack of integration with DR (e.g., data center plans not providing voice/data network, systems, and personnel support for work-area recovery ) Lack of education and awareness of explicit roles and responsibilities Lack of maintenance/updates Business executives believe backed-up information can be recovered quickly Users test the ability to recovery systems separately but do not check consistency of systems and data

Business Case for BCP & DRP What is the impact to the company if an unplanned outage caused the company to delay delivering a critical product release to a customer ? What is the impact to the company if an unplanned outage to not be able to provide Customer Care support? What is the impact to the company if an unplanned outage caused the company to delay processing Cash Receipts ? What is the impact to the company if an unplanned outage caused the company to delay processing and printing Customer Bills ? What is the impact to the company if an unplanned outage caused the company to delay Financial Reporting ? What is the impact to the company if an unplanned outage caused the company to delay Payroll for employees?

Recovery Objectives and Costs Mission-critical: Platinum RTO & RPO: Zero Cost: 6x-8x can be as high as 20x Business-critical: Gold RTO: <8 hours RPO: <15 minutes Cost: 4x-5x Basic: Silver RTO: <72 hours for critical RPO: <24 hours Cost: 1.6x Standalone: Bronze Best-effort recovery RTO: >1 week RPO: 1 week Cost: x Establish a realistic and measurable framework What’s in a “Nine”? 99. 999 99 9 0 = = = = 5 min 58 min 8.8 hr 3.6 days

BC & DRP Service Level Model continued

Related Technologies Related Technologies (DRP, High Availability, Backup) Clustering Replication – Synchronous, Semi-Synchronous, A-Synchronous Oracle RAC – Real Applications Cluster Geo-Clustering Boot from SAN CDP – Continuous Data Protection Server Virtualization and P2V Backup to disk – VTL – De-Duplication Interesting new concept by Israeli Startup – Continuity Software

Site-Level Clustering Addresses unplanned downtime (not planned) Campus environment (<5 miles) RTO, RPO — minutes Potential failover performance impact DR issue with single database Complex implementation and scripts Example solutions: HP MC Service Guard, IBM HACMP, Sun (True) Cluster, Microsoft Cluster, Veritas Cluster, Oracle RAC, Sybase HA, Parallel Sysplex Clustering Options Database A Failover for B Database B Failover for A Cluster Shareable Disk Group Failover for A, B, and C Cluster Shareable Disk Group Database A Master Node Database B Master Node Database C Master Node Failover N + 1 Failover

Remote Logging Capture log data and transmit to remote site Pros Minimal database data loss at remote site (RPO) Can be synchronized with other applications Cons Complex and long recovery times at remote site (RTO) Complex to implement and maintain Increased processing at both sites No provision for non-database data Log Propagation Logging solutions are complex and time consuming IMS DB2 Oracle CICSlogs Logs Applied to Database

Synchronous, Storage-Based, Data Replication Data writes committed at both locations Sites <50 miles apart Database/application restart required RTO = <4 hours, RPO = last committed transaction Leveraged infrastructure and no quiesce required Vendor-specific solutions Many times the applications are started manually Synchronous Replication Examples: EMC SRDF, IBM PPRC, HDS HRC, Compaq DRM, HP Continuous Access XP Controller Send Less Than 50 Miles Wait Until Complete

Geographic cluster Hardware Veritas Cluster Server Hardware Shared Storage Primary Site Node A Secondary Site Node C Hardware Primary Site Node B DR Agent VCS VCS DR Agent Source: eval.symantec.com/mktginfo/enterprise/other_resources/ent-vcs_for_vmware_11-2006.en-us.ppt with modifications Local Cluster Heartbeat Global Cluster Heartbeat Hardware-Based Data Replication

Continuous data protection (CDP) “ Any PIT” recovery Some CDP flavors: Replicate all writes (complete copy of Server) Replicated only application related writes – for easier “bare metal restore” Considerations Introduces a new process into your backup strategy Typically requires as much as 1.5x your current capacity Primary storage Secondary storage CDP appliance/server Server Control data Data Backup/copy data Source: www.forrester.com/Events/Content/0,5180,-1277,00.ppt with modification

Many to one CDP solutions + “bare metal” restore High Availability (Target) Server Off-Site Disaster Recovery (DR) Server Optional Centralized Tape Backup Server Consolidation Data Migration Snapshot Production (Source) Servers Direct offsite connection or two-stage connection through local HA server Source: www.synegi.com/docs/nsi-synegi-presentation.ppt

Virtualization and P2V recovery Scenario – Phase 1 : Initial P2V Your Servers are : Protected Ready for Patches / Upgrade tests

How De-Duplication Works: Compares latest backup data to stored data None, only pointers to previous data Source (with STKI modification) : Sepaton http://www.go-dpa.com/media/TSM_Webinar.ppt Current Data New Data Stored Previously Stored Data Eliminates duplicate data within Exchange, Oracle ® and Microsoft ® SQL databases Eliinates storage of unchanged portions of files Eliminates storage of duplicate files (single Instance)

VDI – Overview Thin Client or PC Virtual Infrastructure 3 Server Local users are very excited about this technology Selected VMWARE VDI installations: Siemens All Rights Reserved @STKI Moshav Bnei Zion, Israel +972 9 74 444 74 www.stki.info

Microsoft Application Virtualization (formerly SoftGrid) allows applications to be deployed in real-time to any client from a virtual application server It removes the need for local installation of the applications. Softgrid applications are also sandboxed from each other. Softgrid selected clients: Education (Bar Ilan, Afek collage, Hebrew University), IDF Market status and Trends Source: http://www.virtuall.nl/articles/ProductInformation/PQR%20- %20Reality%20of%20Virtualization%20Solutions%20v1.1.ppt All Rights Reserved @STKI Moshav Bnei Zion, Israel +972 9 74 444 74 www.stki.info

Summery The basic issues (and pitfalls) of DRP\BCP are here to stay (management, rolls, processes) New technologies enable better options Some of these technologies need to mature The essence of DRP is process not technology

Thank you Pini Cohen EVP & Senior Analyst Architecture & Infrastructure Strategies [email_address]

Drp For Menora

More Related Content

What's hot (20)

Similar to Drp For Menora (20)

More from Pini Cohen (20)

Recently uploaded (20)

Drp For Menora