On 2010-02-20 at 16:28:15 [+0100], Oliver Tappe <zooey@xxxxxxxxxxxxxxx> wrote: > [ ... ] > > So basically I don't know what's been going on, but I will now continue with > checking the logs on the VM itself. Maybe something interesting can be found > there. Well, yes and no: it looks like there was a system-time inconsistency at around 11:40 CET / 12:04 CET, since the apache transfer log contains intermixed entries for both times. Soon after that, mysql went down, which brought some PHP-watchdog into action, which sent a lot of complaints about 'mysql having gone away' into the logs. Mysqld then got restarted (probably by the watchdog), which didn't seem to have rectified the problem, as soon afterwards the http_prefork process got visited by the OOM killer (twice). After that, the logs got quiet. I don't know which process actually continued to hog the CPUs and why, probably apache and/or mysql - but what really makes me wonder is what peculiarity might have caused the system time inconsistency? cheers, Oliver