On Fri, 2009-03-20 at 10:15 -0400, Thomas Wakefield wrote:

> 
> Mar 20 09:51:59 disk0 kernel: device-mapper: snapshots: Invalidating
> snapshot: Unable to allocate exception.

This is an LVM (device-mapper in fact) error.  Nothing to do with
Lustre.

> Mar 20 09:51:59 disk0 kernel: Buffer I/O error on device dm-1, logical
> block 22676124
> Mar 20 09:51:59 disk0 kernel: lost page write due to I/O error on dm-1
> Mar 20 09:51:59 disk0 kernel: Buffer I/O error on device dm-1, logical
> block 22676125
> Mar 20 09:51:59 disk0 kernel: lost page write due to I/O error on dm-1

These are all generic kernel I/O messages, probably related to the prior
LVM error.

> Mar 20 09:51:59 disk0 kernel: LDISKFS-fs error (device dm-1):
> ldiskfs_get_inode_loc: unable to read inode block - inode=21347250,
> block=21333880

This looks like ldiskfs reporting an error likely caused by the previous
errors.

> Then after a reboot here is a DMESG bug that concerned me:

Probably a result of the problems with your disk.  I would suggest
getting your disk problems fixed, completely, first, then doing an fsck
on the MDT to see what damage the disk problems caused the filesystem

b.


Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
Lustre-discuss mailing list
[email protected]
http://lists.lustre.org/mailman/listinfo/lustre-discuss

Reply via email to