On Feb 02, 2007  13:16 -0600, Kevin L. Buterbaugh wrote:
> Sorry, meant to include that.  Here's the relevant information from the 
> client (scnode01):
> 
> Feb  2 12:48:15 scnode01 kernel: LustreError: 
> 16536:0:(client.c:576:ptlrpc_check_status()) @@@ type == 
> PTL_RPC_MSG_ERR, err == -19 [EMAIL PROTECTED] x13/t0 
> o38->[EMAIL PROTECTED]@tcp:12 lens 240/272 ref 1 fl Rpc:R/0/0 rc 0/-19
> Feb  2 12:48:15 scnode01 kernel: LustreError: mdc_dev: The configuration 
> 'client' could not be read from the MDS 'mds-test'.  This may be the 
> result of communication errors between the client and the MDS, or if the 
> MDS is not running.

Client couldn't connect to the MDS.  -19 = -ENODEV

> And from the MDS (lustrem):
> 
> 3894:0:(client.c:940:ptlrpc_expire_one_request()) @@@ timeout (sent at 
> 1170442057, 5s ago) [EMAIL PROTECTED] x1/t0 
> o8->[EMAIL PROTECTED]:6 lens 240/272 ref 1 fl Rpc:/0/0 rc 0/0
> Feb  2 12:48:07 lustrem kernel: LustreError: 
> 3894:0:(client.c:940:ptlrpc_expire_one_request()) @@@ timeout (sent at 
> 1170442082, 5s ago) [EMAIL PROTECTED] x4/t0 
> o8->[EMAIL PROTECTED]:6 lens 240/272 ref 1 fl Rpc:/0/0 rc 0/0
> Feb  2 12:48:07 lustrem kernel: LustreError: 

These messages indicate failure to connect to the OSTs (op 8 = OST_CONNECT).
What is in the OST syslog?  Are you positive that /dev/sda1 and /dev/sdb1
on the two nodes are set up the same way, so that e.g. lustre1+sda1 isn't
talking to the same disk as lustre2+sdb1?

Also minor nit - you don't need to have a partition table, it can hurt
performance on some RAID setups because of the 512-byte offset of IOs
due to the DOS partition table.


Cheers, Andreas
--
Andreas Dilger
Principal Software Engineer
Cluster File Systems, Inc.

_______________________________________________
Lustre-discuss mailing list
[email protected]
https://mail.clusterfs.com/mailman/listinfo/lustre-discuss

Reply via email to