From: Andy Malato (andym@oak.njit.edu)
Date: Thu Apr 28 2005 - 08:19:57 EDT
Thanks to all who answered my question :
I just noticed that one of our Tru64 systems had rebooted itself a
couple of days ago. I found the following the /var/adm/messages :
Apr 21 21:12:20 cams1 vmunix:
Apr 21 21:12:20 cams1 vmunix: trap: invalid memory read access
from kernel
mode
Apr 21 21:12:20 cams1 vmunix:
Apr 21 21:12:20 cams1 vmunix: faulting virtual address:
0x0000000000000000
Apr 21 21:12:20 cams1 vmunix: pc of faulting instruction:
0xfffffc0000645ad4
Apr 21 21:12:20 cams1 vmunix: ra contents at time of fault:
0xfffffc00005c2f14
Apr 21 21:12:20 cams1 vmunix: sp contents at time of fault:
0xfffffe0544e575f0
Apr 21 21:12:20 cams1 vmunix:
Apr 21 21:12:20 cams1 vmunix: panic (cpu 0): kernel memory fault
Apr 21 21:12:20 cams1 vmunix: syncing disks... done
Does anyone know what the cause of the above is? This system has
been running fine for years without problems. Does this indicate a
hardware problem, such as with the system's memory or CPU ?
I'd like to especially thank Jason Orendorf for helpling me identify the
cause of the problem which turned out to be related to the arla client
which uses XFS and the panic was occurring after certain calls to the XFS
code. We are running arla on several machines, including an identical
system with the same OS level, without problems.
For now I've turned off the arla client and have switched to the TransARC
AFS client. The reason for using Arla was that the TransARC client seems
to have problems where its cache manager drops one's token at random
times.
Thanks also goes to Dr Thomas Blinn, David Gutierrez and Whitney Latta who
also contributed advise on diagnosing the aforementioned error messages.
Regards,
---Andy
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:50:18 EDT