What Disaster Are You Planning For?

When the subject of disaster recovery comes up many IT professionals' minds immediately flash to an epic event like a fire, hurricane, tornado or earthquake. While this is fine for a point of reference, what about planning for the more mundane disaster? These simple disasters can often cost you as much in revenue and brand reputation than their larger alternatives.I have nothing against planning for a disaster that assumes the loss of the primary data center and moving operations to an alternate site. Clearly this is something you should plan for. The problem that I have seen is that when planning for these once in a lifetime disasters people often loose site of the risk involved in the once a month mini-disasters. Mini-disasters are situations that occur and impact a small section of your data center. It can be a double drive failure on a RAID array, application data getting corrupted or the server/virtual machine that the application runs on crashing for some reason.

As we will discuss in our upcoming webcast "What's Missing From Your DR Plan for 2011?" mini-disasters tend to get left out of most disaster plans and application rollout projects. Mini-disasters don't capture headlines, users have no idea in many cases why their application isn't available, they just start calling IT and asking when it will be fixed. Then they wait and there goes productivity. Lost user productivity can delay production which will impact revenue. The situation is worse when customers have no idea why they can no longer place an order or use a particular service. Customers don't wait for you to fix the problem, they just go somewhere else. These mini-disasters also send the IT staff into a wasteful fire-drill mode and put friction in the relationship between IT and the rest of the organization.

For these mini-disasters most IT pros count on the backup process to bring things back to life. Probably for many applications that is a fair expectation but even if all the back data is actually recoverable, there is a gap in how often that data has been protected and there is a time delay in how long it will take to restore that data back into place, especially if it needs to be copied across an Ethernet network. The net impact is that you should count on a minimum of four hours of downtime when recovering from a backup system. Server virtualization and virtualization specific backup applications can help, as can application availability applications. All of which we will cover in our upcoming entries.

Track us on Twitter: http://twitter.com/storageswiss

Subscribe to our RSS feed.

George Crump is lead analyst of Storage Switzerland, an IT analyst firm focused on the storage and virtualization segments. Find Storage Switzerland's disclosure statement here.