Ooops w/2.6.7-rc2 & XFS

Ooops w/2.6.7-rc2 & XFS

[Phy Prabab]

Hello,

Ran into this problem today.  The machine in question
is a dual Xeon 2.4GHz w/a 1.5T 3Ware sata volume
carved out  vi LVM2 and formated with XFS:

nfsd: non-standard errno: -990
Filesystem "dm-1": XFS internal error
xfs_btree_check_sblock at line 342 of file
fs/xfs/xfs_btree.c.  Caller 0xc025c414
 [<c0240098>] xfs_btree_check_sblock+0x65/0xed
 [<c025c414>] xfs_inobt_lookup+0x19f/0x3cf
 [<c025c414>] xfs_inobt_lookup+0x19f/0x3cf
 [<c0259832>] xfs_dialloc+0x316/0xb74
 [<c027ef70>] xfs_vget+0xcf/0xe8
 [<c016f836>] iput+0x3c/0x76
 [<c026c8e2>] xlog_grant_log_space+0x260/0x408
 [<c0261736>] xfs_ialloc+0x7a/0x4bd
 [<c027c595>] xfs_dir_ialloc+0x9b/0x323
 [<c027923b>] xfs_trans_reserve+0x8b/0x1f9
 [<c0283988>] xfs_mkdir+0x2d9/0x77d
 [<c021d664>] xfs_acl_vhasacl_default+0x3b/0x49
 [<c028ecdb>] linvfs_mknod+0x44c/0x451
 [<c037cec9>] tcp_transmit_skb+0x40b/0x693
 [<c011864d>] __wake_up_common+0x3f/0x5e
 [<c028ed17>] linvfs_mkdir+0x2a/0x2e
 [<c0164eaf>] vfs_mkdir+0x9f/0x118
 [<c01e7b3f>] nfsd_create+0x443/0x48a
 [<c01eed79>] nfsd3_proc_mkdir+0xd3/0x11b
 [<c01e32a4>] nfsd_dispatch+0x15e/0x20e
 [<c03a9d7d>] svc_process+0x406/0x61d
 [<c01e2f97>] nfsd+0x1ed/0x39c
 [<c01e2daa>] nfsd+0x0/0x39c
 [<c01022b5>] kernel_thread_helper+0x5/0xb


I unmounted and repaired which allowed the unit to run
for another couple of hours only to get hosed again
with the above trace.

Any suggestions?

Thank you for your time.
Phy


	
		
__________________________________
Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/ 

Re: Ooops w/2.6.7-rc2 & XFS

[Chris Wedgwood]

On Sun, May 30, 2004 at 07:32:25PM -0700, Phy Prabab wrote:

> nfsd: non-standard errno: -990

XFS has detected corruption.  Umount and run xfs_repair.


  --cw

Re: Ooops w/2.6.7-rc2 & XFS

[Olaf Frączyk]

On Mon, 2004-05-31 at 05:38, Chris Wedgwood wrote:
> On Sun, May 30, 2004 at 07:32:25PM -0700, Phy Prabab wrote:
> 
> > nfsd: non-standard errno: -990
> 
> XFS has detected corruption.  Umount and run xfs_repair.
> 
Hi,

Read his email to the end. He did it, but the error ocurred again.

Regards,

Olaf

Re: Ooops w/2.6.7-rc2 & XFS

[Chris Wedgwood]

On Sun, May 30, 2004 at 07:32:25PM -0700, Phy Prabab wrote:

> I unmounted and repaired which allowed the unit to run
> for another couple of hours only to get hosed again
> with the above trace.

repaired == xfs_repair?  what did that say?

> Any suggestions?

this machine has ECC/whatever and memtest86 passes cleanly right?
also, when the error occured again was it exactly the same both times?


thanks,

  --cw

Re: Ooops w/2.6.7-rc2 & XFS

[Phy Prabab]

> repaired == xfs_repair?  what did that say?
ran xfs_repair, no errors reported.

> this machine has ECC/whatever and memtest86 passes
> cleanly right?
ECC 1.5G.  I have already swapped out memeory just to
be on the safe side.

I ran memtest with no errors

> also, when the error occured again was it exactly
> the same both times?

It happened 4 times and every single time, it had the
identical trace.

I brought on line a new volume with new drives to see
if I can reproduce this (same machine and same
kernel).

Any other suggetions/questions?

Thank you for the assistance.

Phy





	
		
__________________________________
Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/