Last friday I attended a session from Bob van der Werf, Sr. Systems Engineer at VMware about Disaster Recovery. He spoke about some D/R Concepts and the challenges that are faced when implementing a disaster recovery plan and what solution VMware provides to help us out. The product called VMware Site Recovery Manager will provide an automated recovery to a Recovery Site. The BETA of this new product is expected to be released in february 2008, and the RTM is due mid-2008.
Continuity Management/Contingency Planning/Disaster Recovery Planning
A sound disaster recovery / business continuity plan is essential to protect the well being of an organization. This cannot really be over emphasized… yet many enterprises still side step the issue or hold plans which are clearly out of date or inadequate.
Part of the reason for this is the complexity of the task. This is not helped by some vendors selling planning products which are themselves extremely difficult to master.
When Bob asked the attendees who has a D/R-plan in place about half of the crowd responded by raising their hand. The reaction to the question who actually tested the plan recently was even more disapointing, about half of the respondees of the first question raised their hand.
Reasons why you should have a D/R-plan and should test it:
- Rule Compliancy (SOX or National Financial Rules)
- High SLA due to competitiveness in the market
- Increasing security threats
When creating a D/R-plan certain aspects are imported to identify two important aspects are:
RPO – Recovery Point Objective
The point to which data must be restored
RTO – Recovery Time Objective
The time by which data must be restored
Recovery is expensive so you have to make choices what to do keeping in mind the RPO and RTO, also keep in mind that has to be a recovery site available.
Vmware’s vision for Disaster Recovery Solutions:
- Fast
- Reliability
- Managebility
- Affordability
VMware Site Recovery Manager(SRM) is totally integrated with Virtual Center (a plug-in for the VI-client).
Mike Laverick from RTFM Education wrote an technical article about SRM based on what he had heared at VMworld 2007 in LA.
The replication method used to replicate the data which resides on the storage array’s to the Recovery site depends on the array-vendor. There will be seperate HCL regarding SRM and Storage-type, the HCL will list the supported types of storage that have a supported replication method used by SRM.
Inside SRM you can define plans which outline in which order the vm’s should get available in the recovery site.
Keep in mind that there will be no recovery without downtime.
So in short: SRM gives you a large button you can press whenever you have to do a disaster recovery.