In an hour or two, we will have to take the site down for a few minutes as we reboot our master DB with a new kernel.
There've been a few other things going on behind the scenes in the last day or two.
Two of our free web slaves had crashed, but have since been fixed. Luckily, the Linux kernel has showed signs of not sucking anymore, and we'll see how those all hold up with a kernel upgrade.
The bugs we had hit with InnoDB have been fixed by the developers, and we are converting a slave to InnoDB fully right now. Then we will test replication and other goodies further. This will mean all sorts of good things when it finally works.
Update: Done. The site will still probably be slow while caches continue to fill. There's this one problem we've been chasing for more than a month now, and it refuses to be found. It resurfaced today. I'm going to harass the MySQL devs again with a very detailed report to see if they can't figure it out this time.