Oh, you wanted to recover those backups?

Incident: Suppose your company's mission was to inform others about the business of technology. Surely you'd want your own IT operations to be above the cut, if only to maintain a modicum of credibility with your readers.

In one of the more amazing not-practicing-what-you-preach stories of recent history, the editorial and design staff of tech-centric Business 2.0 somehow managed to lose its entire June 2007 issue to a hard-drive failure on its editorial server sometime during the night of April 23, 2007.

As wrenching as it must have felt for its editorial staff to realize that a month's worth of work was gone, that sinking feeling in the pit of the stomach must have been compounded by its status as a standard-bearer for sound data-protection practices, including frequent and comprehensive backups.

Well, at least the Business 2.0 IT staff did the backup part. What they never seemed to do was the restore part, as a test to make sure the backups were readable and that they worked. And that's exactly what failed when the admins tried to recover from the catastrophic crash.

Fallout: As for the editorial content of the issue, it turned out that the lawyers had copies of every article in a final, copyedited state. But the layout and design of the magazine had to be re-created from scratch. A heroic effort from Business 2.0's art department saved the day and got the magazine produced (reproduced?) in record time, on deadline for press.

Moral: As any halfway sensible IT person will tell you, it's not enough just to back up the data; you also have to test those backups periodically to make sure they actually work.