A machine I managed crashed today. The console (25 lines) indicated a
kernel fault in an interrupt handler. After reboot, two pre-crash
kernel oopses were in the log (see attached). The commonality between
the two seems to be a spin_lock being already locked in mm/slab.c.
The oops times were 10:43 and 11:01; the machine crashed somewhere
between 11:01 and 11:10 judging by /var/log/cron (how can I get a more
accurate time of crash?).
This machine is running FC2 kernel 2.6.8-1.521 #1 Mon Aug 16 09:01:18
EDT 2004 i686 i686 i386. The machine had been up for about two weeks
(when it was first put in service with this OS installation), and there
was no apparent unusual activity on it today. It's an NFS fileserver.
Any ideas? Is this likely a problem in the kernel, or with my hardware?
Thanks,
David
Show replies by date