[OpenAFS] openafs fileserver crashes.

Matthew Andrews mnandrews@lbl.gov
Fri, 09 Nov 2001 14:12:31 -0800


Hello,
I bave an openafs fileserver which has been crashing fairly regularly
under moderate load. the machine is an i586 running redhat 7.1 with
the 2.4.2-2 kernel, and openafs-1.2.2 from the rpms


the symptoms are usually that at least one of the fileserver processes
die, leaving the rest owned by init. salvage then starts running, and
hangs. sometimes, I am unable to kill all of the remaining fileserver
processes, and if that happens, the only way to bring the fileserver
back is to reboot the machine.

the most recent time the server hung, the machine actually hung, and
had to be power cycled.

I don't have logfiles for the previous failures, but I have turned on
debugging in the Fileserver, and have seen no suspicious messages on the
FileLog.old after a crash.

any ideas on where to proceed from here? I'm going to make sure that the
core limit is not set before starting bos. is there anything else I need
to do to ensure that a fileserver will leave a core file when it dies?

-Matthew Andrews