We have over the last few months experienced a lot of PVFS client stability
problems with both pvfs-2.6.2 and pvfs-2.6.3. The syslog would typically
show the following message just before a crash:
Jul 30 20:42:06 shc kernel: Unable to handle kernel NULL pointer dereference
at 0000000000000000 RIP:
Jul 30 20:42:06 shc kernel: <ffffffff8839c2a3>{:pvfs2:wait_for_a_slot+126}
Jul 30 20:42:06 shc kernel: PGD 1c3f3a067 PUD 1f2a1e067 PMD 0
Jul 30 20:42:06 shc kernel: Oops: 0000 [2] SMP
Jul 30 20:42:06 shc kernel: CPU 3
We have been using kernel 2.6.16.48 and 2.6.16.52 on x86_64 architecture.
Typically users have been busy copying files into the PVFS file system,
using cp or rsync, just before the crash.
Any insight or suggestions for how to debug this, is much appreciated.
Regards,
Jan Lindheim
_______________________________________________
Pvfs2-users mailing list
[email protected]
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-users