[Lustre-discuss] Panic problem

[Andreas Dilger]

On 2010-07-08, at 14:32, Marco Aurelio L Gomes wrote:
> We have here a lustre 1.8.1 filesystem running on a CentOS 5.4 system using kernel 2.6.18-128.1.14.el5_lustre.1.8.1. Since 3 days ago, we're having problems with kernel panic in our mds machines (our setup  has 2 mds and 2 oss), and when panic occurs, the other machine mount lustre mdt filesystem and became principal mds, but after the recover, it panic also. I attached the dump from kernel panic and would like to know if someone has this kind of problem, and if someone can help me.

I think the first error is fixed in 1.8.3 already, please search in bugzilla:

> lustre_shrink_reply_v2()) ASSERTION(msg->lm_bufcount > segment) failed


The next bug:

> Kernel BUG at ...s/root/usr/local/src/aufs.wcvs/aufs/fs/aufs/f_op.c:706

is not in the lustre code.  I don't recall what "aufs" is, but presumably some kind of filesystem.

Cheers, Andreas
--
Andreas Dilger
Lustre Technical Lead
Oracle Corporation Canada Inc.