(This is a conversation I’ve had several times but figured it merited a blog post given it came up again with a customer this week.)
How much is an untested Disaster Recovery (DR) plan worth? Similarly, how much is an infrequently tested DR plan worth? (say once a year)
Pardon my being blunt….but somewhere between nothing and not much. In the event of a disaster (i.e. one of the more stressful times of your career), you have to execute a long series of steps perfectly or at least close to it….and we’re not even considering equipment failure issues from equipment sitting on the shelf for months to years. How likely is that to work well?
This isn’t necessarily anyone’s fault though — DR testing takes real time and organizational commitment. Often, the DR testing process is disruptive to the existing environment….or at best, a successful DR test requires a serious amount of time and often a “manual reset” (and you won’t know if the “reset” was done correctly until your next DR test — rather the catch-22).
When you have a situation like this, the answer is not something along the lines of “beatings will continue until morale improves”. Rather, you need to examine the technical and procedural underpinnings to find architectural level solutions.
Ever hear of intelligent storage and server virtualization? Now imagine that you could non-disruptively test your DR plan on a quarterly basis without significant time involved.
Right now VMware has the best (if not only) answer to this challenge for most customers in their Site Recovery Manager product. There’s a good video here which covers the product quite well (a bit dated and mildly cheesy at times but quite good overall).
Varrow’s own Jason Nash has a good 2 part video on SRM as well.
EMC even has some good SRM videos (as they offer arrays that integrate very well with SRM). EMC is unique in offering easy failback capabilities as well.
Is all this cheap? Well, no….but would you actually want a “cheap” DR plan? It’s not cost prohibitive however (as evidenced by the number of Varrow customers currently using Site Recovery Manager) which is rather amazing given that something like this wasn’t even possible just a few years ago.
Viva la progress!