The Myth of Disaster Recovery Testing
I had lunch the other day with a friend who manages a fairly large IT department. We were discussing disaster recovery testing and came to the conclusion that it is a myth. Companies require DR testing, IT directors talk about it and data protection software vendors promote it, but in practice, disaster recovery testing happens rarely.
Now, when I call disaster recovery testing a myth, I am using hyperbole, but barely. The Disaster Recovery Preparedness Council 2014 annual report says 73% of survey participants worldwide are failing in terms of disaster readiness. More than 60% do not have a fully documented DR plan and just one third of survey participants actually test their plan. Remarkably, 65% that try do not pass their own test.
On a personal level, I have the opportunity to speak to groups of systems administrators and I always ask the question, “who here performs regular disaster recovery tests?” Typically, one or two out of 30 will raise their hand.
Why this epidemic of lack of preparedness for a disaster? The answers I usually hear fall into two categories – time and resources. I would add a third reason that most will not confess to – they do not want to be in the 65% that report back to their management that they failed their own DR test.
Time is a problem because it often takes weeks or months to build out the disaster recovery plan, days or weeks to prepare for a test and days to perform the disaster recovery exercise. For many shorthanded IT staffs, this amount of time is simply not available.
My friend falls into the resources category. He has all of his data replicated to a second site, but does not have the compute resources at that location to actually test anything. His DR data center is simply a data repository which guarantees no recovery SLAs will be met. He has no choice, he is budget limited.
So, in about 300 words, we have painted a pretty bleak picture of the current state of recovery readiness for businesses. We have also laid out a pretty good argument for cloud backup and DR services. Solutions where resources can be accessed as needed without a large capital expense, a solution that is secure and reliable and a solution that can guarantee recovery SLAs.
The key to all of this is recovery assurance. Once you have a plan written and you know your recovery point and recovery time objectives and resources are in place that allows testing, can you easily perform the test? At Unitrends we use our own product, ReliableDR to automate recovery assurance testing for VMware and Hyper-V virtual machines and Windows physical servers. We schedule DR tests that report back the recovery point and recovery time actuals and if they were compliant with the objectives. If you use the Unitrends Cloud and our disaster recovery services, you receive recovery assurance.