In case you're interested in the details, during our update last week, several indexes in the table that contains all 3.2 million posts or so became corrupted. Attempts to rollback the change didn't succeed in fixing the problem. The end result was that retreiving topics and posts took much more CPU than they should, and that meant that all of you would get timeout errors, because everything slowed way down. Thankfully, with the help of the Ideal Science guys, we were able to fix the problem. We'll continue to monitor it, but so far, CPU utilization and request speed are back to normal.