It’s amazing that hard drives don’t fail more often; a metal disk rotating at close to 200 miles per hour, with a read-write head floating above it by less than the thickness of a human hair. It’s like flying a 747 an inch off the ground, for years.

Well, on Saturday afternoon our metaphorical 747 had an … er … incident. Of course, Murphy dictates that these kinds of disasters can only happen on weekends. Saturday afternoon is Murphy’s usual preferred time for hard drive and/or plumbing failures.

I want to thank the people at Hosting Matters for working overtime to get the drive replaced and all our data restored. Restoring the data was the trickiest part, because as the old drive was failing, and valiantly still trying to serve data to our readers, it was not properly closing database tables. In its death throes, it left a bit of a mess.

But I’m happy to report that all of our important data is completely intact. We lost the access log data for the past month, but that’s no big deal. And in addition to our new drive, we now have another two gigabytes of RAM in the database server. (Query caching, here we come.)

I’ve been testing as much as possible this morning, and all systems appear to be functioning. The only thing I can’t test is what happens when the lizard army starts making demands on the facilities, so here’s an open thread.

I think I need more coffee.