Hi,
The Lustre volume stops responding every few weeks. The
log is attached. The Lustre installation has a single OSS and serves a
big scratch volume
and one home volume to 21 hpc client nodes. The home volumes have disk
quota enforced. I had to restart the cluster due to the Lustre problem.
The version of Lustre
is 1.6.0.1 and OS is CentOS 4. Once restarted, the Lustre volume works
fine for another few weeks. Any comments and help are much appreciated.
Thanks.
Regards
Balagopal
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: lustre-crash.txt
Url:
http://lists.lustre.org/pipermail/lustre-discuss/attachments/20071009/855a25dd/attachment-0002.txt