I'll attempt to answer my own question here.

It appears that because the OSSes were started before the MDS that these 
nodes are still trying to obtain MGS data from that host but even though 
it came back, they still can't.  This appears to be bug # 13715.

A few weeks from now I'll be able to restart the lustre OSSes and see if 
this corrects the problem.

Bill

Bill Wichser wrote:
> I'm seeing kernel messages from my three OSS machines and don't know 
> what they mean.  These messages started after a orderly shutdown of all 
> my machines over the weekend for a new UPS install.  All the OSSes are 
> logging the same error.
> 
> kernel: LustreError: 7025:0:(client.c:519:ptlrpc_import_delay_req()) @@@ 
> IMP_INVALID  [EMAIL PROTECTED] x56653/t0 
> o101->[EMAIL PROTECTED]@o2ib_0:26 lens 232/240 ref 1 fl Rpc:/0/0 rc 0/0
> 
> Is there something that I need to do to eliminate these?  The filesystem 
> appears to be completely functional otherwise.
> 
> The run environment is lustre 1.6.4.2 on SLES 10.
> 
> Thanks,
> Bill
> _______________________________________________
> Lustre-discuss mailing list
> [email protected]
> http://lists.lustre.org/mailman/listinfo/lustre-discuss
_______________________________________________
Lustre-discuss mailing list
[email protected]
http://lists.lustre.org/mailman/listinfo/lustre-discuss

Reply via email to