We had a string of performance issues, punctuated by about thirty minutes of outright downtime, tonight. We think it was caused by a one-off maintenance process that was taking up too much of the master database's resources. Of course, that shouldn't bring the site down, but a few other issues we have with our web processes timing out and leaving their TCP connections to the database open conspired to cause a more severe issue. I'll be working on site stability in the next few weeks, and hopefully we can keep this from happening again (though the maintenance process that appears to have been the cause of tonight's outage was a one-off task, so we should be okay in the meantime).
jephn:
I was trying to log on. I thought the internet was telling me to go read a book.