Disaster recovery is designed to safeguard a company’s data, but sometimes getting the system to work properly can result in catastrophe.
CBL Data Recovery Technologies Inc. has published a top 10-style list of storage blunders,
most of which are the result of human error. One story refers to a “”killer”” Unix network system that was regularly backed up, but there was nothing attached to the back end so the data disappeared into the ether. In another instance, an organization froze its disks after its office flooded before sending them off to a data recovery lab. (Read the complete list here.)
“”It shows people are naive,”” said president of Armonk, N.Y.-based CBL Bill Margeson. “”In IT there are brilliant people that are great at building . . . or selling you what you need. But when it comes to an unusual sequence of events, they are without experience, without tools, and they do make some silly decisions.””
The problems stem from junior IT staff who don’t fully understand the systems they’re working with, or a poor choice of solution in the first place, he said. Senior IT personnel aren’t always available to oversee every aspect of data backup and recovery, so sometimes steps get missed or procedure isn’t properly followed.
If you publish a list of foul-ups you’re bound to get worst case scenarios, noted Giga Information Group Inc. analyst Anders Lofgren, but disaster is one of the thornier aspects of IT management. “”It is getting much more difficult as data capacities continue to grow. There’s no doubt that it’s becoming much trickier and I’m sure there’s mistakes along the way,”” he said. Those mistakes are often aggravated by complex open systems environments, he added, which use multiple flavours of Unix and competing operating systems.
Mistakes are easily made, said Dave McCartney, systems administrator with Statistics Canada. He used to deal with local area network administration, using back-up solutions running on 8-mm tape and a combination of other storage tools. “”If you don’t rotate the tapes, then chances are you’re going to need to recover something and it’s going to be on something you don’t have backed up. That’s just Murphy’s Law,”” he said. “”As long as you change the tapes, you’re fine. But the problem is the manual intervention part of it. It gets you.””
Margeson may be taking his cue from David Letterman, but disaster recovery is no laughing matter. The aforementioned case of the flooded office and frozen disks happened in Venezuela two years ago. “”The data they were recovering in that case was civil engineering data they needed to find where their streets were. Sometimes its not something to fool with,”” he said.
But he defends the use of humour to highlight a serious problem. “”Part of it is psychological,”” he said. “”We’re trying to not point fingers, we’re trying to make people aware and humour is pretty good (at that). It’s in their vested interests. You can’t always criticize, so we hope that humour gives it a little tweak.””
Sometimes a crazy idea like freezing disks might just work, though. McCartney pointed out an article he read on Techrepublic.com which recommends freezing overheated hard drives. The trick is to recover the data while the disk is still cold enough to be read. But like most crazy ideas, it’s a one-shot deal. The disk is useless when it thaws out.