We have a number of Debian woody servers, mostly Intel kit. One of our
P4 boxen decided to reboot itself twice over the weekend. last -x:
runlevel (to lvl 2) 2.4.20 Sat Jun 12 08:04 - 10:12 (1+02:07)
reboot system boot 2.4.20 Sat Jun 12 08:04 (1+02:07)
shutdown system down 2.4.20 Sat Jun 12 07:50 - 10:12 (1+02:22)
runlevel (to lvl 6) 2.4.20 Sat Jun 12 07:49 - 07:50 (00:00)
No previous logins within hours of this. syslog shows eth0 ran out of
resources, followed by a regular cronjob (very memory intensive),
followed by out of memory errors, then a switch to runlevel 6.