Basically, around 4:30 - 4:45 am US EDT (-0400), window.gnome.org ran out of memory for unknown reasons. Ross called me, about an hour later, and I called Matt Galgoci, but we were unable to perform a remote reboot through the Dell Remote Access Card. (we had the same problem with button.gnome.org last week, problem needs to be investigated.) After some delays, we got a colo technician there about 11:00 am and rebooted the system; came up fine. No clues into the syslog as to what caused the EOM situation. Future remediation: - Fix the DRAC configuration - Debug why we are running out of memory (from the memory usage logs from mrtg, we seem to be gradually running out of memory in some cases, though we also seem to be less gradually running out of memory in others) - If we can find out what parts of the system are triggering the EOM, look at limiting them via ulimit (?) Regards, Owen
Attachment:
signature.asc
Description: This is a digitally signed message part