Back To School – Disaster Recovery Lessons

The past week and a half have brought us the trifecta of business continuity / disaster recovery (BC/DR) stories: Hurricane Irene, the San Diego blackout, and the 10th anniversary of 9/11. I’ll try not to belabor the point – stuff happens, and business need to be prepared – and instead highlight some of the more important lessons.

It’s unfortunate that our industry refers to “disaster recovery”. The word disaster sounds like something that happens rarely. In our normal lives, that’s true. But in technology, “disaster” has come to mean anything that causes the loss of a server (or more), or some data. And this happens all the time. Aberdeen reported a couple of years ago that more than 60% of business report having 1 to 5 “disasters” every year.

A simple breakdown of the causes of business continuity disruptions:

• Natural events (eg: hurricane, earthquake)
• Equipment failure
• Man made (error or intentional)

When you stop and think about all these potential causes, it’s not hard to see why disasters happen all the time.

For the same reason that so many people fail to buy life insurance or drive without auto insurance, so too do so many companies operate without the “insurance” of a BC/DR plan. It costs money for something you hope you never need. It hurts to write that check! But you have to do it. To do otherwise is to accept what is really an unacceptable risk. Every year, after another list of disasters, more companies finally get protection.. But there are still too many businesses that haven’t taken this very important precaution.

The thing about paying for BC/DR “insurance” is, it’s gotten a lot cheaper in the last few years. The emergence of virtualization and shared resource cloud environments means that Cloud DR providers are now able to offer DR for a fraction of what it used to cost, or costs to do it yourself. Like an insurance carrier, a Cloud DR vendor can do “actuarial” analysis and provide DR capabilities to many customers at once, at a fraction of what each would have to pay to “self-insure”.

Example: nScaled had one customer declare a disaster as a result of Irene. For the time they were failed over to nScaled’s data centers, if they had instead been completely down, they would have lost close to $1 million in revenues. This customer pays on the order of $50k a year to nScaled for its BC/DR protection, so even if they declare a disaster only once every 20 years, it’s worth it!

Class dismissed.