Virtual Server Recovery (VSR) Tom Humphreys IBM Technical Solution Manager
Business Drivers Clients need faster recovery and increased server recovery reliability Need for faster recovery and increase reliability of recovery in the event of Challenges of an IT disruption recovery on physical infrastructure: High cost as associated with maintaining redundant in-house servers High cost and complexity associated with need for travel to physical infrastructure at time of disaster Risk of failure to recovery on unlike hardware
IT Drivers There are many challenges associated with disaster recovery on a physical infrastructure Identical hardware required for recovery Site A Site B • Costs are doubled, complexity is increased • Accurate configuration information is critical Idle hardware • Standby hardware is needed for faster recovery time objective • Servers that need repurposing at time of disaster are labor intensive and time consuming Slow, complex recovery process • Different solutions for different availability tiers • Separate procedures for system and data disks Configure Install Start Install Configure Hardware OS “Single-step backup operating • Separate procedures for desktop PCs automatic and system (OS) recovery” restore • Multi-step recovery process agent
Context: Real impact of business downtime Average annual revenue loss, productivity loss and downtime hours by industry 100,000 Financial Services Revenue loss per downtime our 3.57 10,000 Retail log scale ($K) 7.75 1,000 Manufacturing 8.01 100 Healthcare 21.7 10 Public Sector 9.41 0.5 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Productivity loss per downtime hour Source:IDC’s Business Value Research, 2009-2011
VSR Cloud Recovery Delivers both performance and economy Warm or Shared Hot Site Economy Days Tape recovery Speed of recovery Hours Performance Hot Site Clustering Disk Recovery Minutes R S V Clustering & Replication Seconds Active-Active with Data Mirroring $ $$ $$$$$$ Business Impact
VSR Cloud Recovery …and overcomes the challenge of Recovery Variability If one of your critical servers goes down: � • In the very best case, how long will it take to restore an equivalent server? • In the very worst case, how long will it be? • Where in between in the most likely case? Likely case: Likely case: 48 hrs +/- 10 hrs 3k mins +/- 600 mins Best case Worst case Best case Worst case 20 40 48 60 80 1k 2k 3k 4k 5k Hours Minutes
Cloud Virtualised Server Recovery 1 4 VSR Storage Login to IBM Cloud Client VSR Portal IBM Cloud VSR Portal 2 3 IBM Recovery Location Client Servers Internet
VSR value proposition …in a nutshell • Dramatically improved RTO and RPO • Extremely reliable protection from server interruptions to full DR for physical and/or virtual servers • Immediate deployment: avoid CAPEX, design, planning, installation • Easy to use • Hypervisor agnostic • Access to recovery infrastructure remotely, so client need not travel to recovery centre • Client avoids the hassles of recovery on non-like hardware • Client may more easily and frequently test the failover and recovery • Our ‘cross-server consistency’ feature ensures client applications run seamlessly when their servers failover to our cloud recovery infrastructure • Our Always Available Virtual Machine service element can be archive, test and production overflow, besides recovery
IBM provide support …to an international integrated credit insurance company Business challenge • Customer received tape based IT DR services • Challenge to provide significantly improved recovery time and recovery point Solution • IBM performed a Business Impact Analysis to understand resilience • Deployed Virtualised Server Recovery (VSR) as a managed service to support business critical services and core applications Benefits • Core application Recovery Time Objective (RTO) reduced, giving the ability to restart all applications quicker • Managed service means a reduced involvement of Customer staff • Testing facilities improved
Recommend
More recommend
