We've been experiencing frequent system hangs with our Altix 3700: it
stops in a state where cannot go anywhere but the kbd prompt. For a
long time SGI support have been unable to get a clue on the problem,
so I'm asking around here for any pointer - what can be done to
troubleshoot this? The logs don't seem to indicate anything useful,
and SGI's standard investigating procedures were no help either...

Earlier we suspected that perhaps some overload makes the system
unstable, but these days it barely keeps going for a day even without
much applications running. The system has SLES 10 now (
kernel), but we've had the same trouble earlier with SLES 9.3, too.

Thanks for any info -- Zoli in the wilderness