xfs
[Top] [All Lists]

re[2]: Lockup with LVM snapshot

To: Eric Sandeen <sandeen@xxxxxxx>
Subject: re[2]: Lockup with LVM snapshot
From: Greg Freemyer <freemyer@xxxxxxxxxxxxxxxxx>
Date: Thu, 31 Oct 2002 13:34:29 -0500
Cc: xfs mailing list <linux-xfs@xxxxxxxxxxx>
Organization: The NorcrossGroup
Sender: linux-xfs-bounce@xxxxxxxxxxx
Thanks Eric,

Current CVS kernel does indeed allow my backup to run.

It is obvious that you guys have made a number of improvements in XFS/LVS 
integration in the last several months.

I'm looking forward to the real 1.2 release.

One question: do I still have an unusual data/meta-data problem on my LV that 
this kernel handles better  Or did this kernel fix the on disk data itself and 
now I could go back to the old kernel?

Greg
 >>  Greg - 

 >>  Flashing lights means you're in kdb... but if you're in X, you're stuck
 >>  there.  Blindly typing "go<enter>" may get you going again.

 >>  I put a fix in recently that prevents XFS from generating a transaction
 >>  on a read-only snapshot, I'm guessing this is the same thing you're
 >>  hitting.

 >>  (if, for some reason (post-crash?) there are files with blocks past eof,
 >>  xfs would try to clean those up, even if the underlying device is
 >>  read-only, which will fail).

 >>  You might try the very latest CVS code, or grab the next 1.2 prerelease
 >>  that's available.

 >>  Alternatively, if you tar up the "live" filesystem once, it will
 >>  probably fix up the state of any files that may be causing this problem.

 >>  -Eric

 >>  On Thu, 2002-10-31 at 10:38, Greg Freemyer wrote:
 >>  > I have a test machine which has started locking up when I run my backup
 >>  script.
 >>  > 
 >>  > The script creates a LVM snapshot for each of 3 LVs in turn and then
 >>  uses xfsdump to back them up.
 >>  > 
 >>  > After the lockup, nothing seems to work, but the keyboard lights are
 >>  flashing.
 >>  > 
 >>  > I'm running a pure CVS kernel from Sept. 6.
 >>  > 
 >>  > My warn.log file has:
 >>  > 
 >>  > Oct 20 01:15:22 TruStore1000 kernel: XFS mounting filesystem lvm(58,4)
 >>  > Oct 20 01:15:28 TruStore1000 kernel: lvm - lvm_map: ll_rw_blk write for
 >>  readonly LV /dev/TruStore-Data/config_snap
 >>  > Oct 20 01:15:28 TruStore1000 kernel: I/O error in filesystem
 >>  ("lvm(58,4)") meta-data dev 0x3a04 block 0x1a5c8
 >>  > Oct 20 01:15:28 TruStore1000 kernel:     ("xlog_iodone") error 5 buf
 >>  count 2560
 >>  > Oct 20 01:15:28 TruStore1000 kernel: xfs_force_shutdown(lvm(58,4),0x2)
 >>  called from line 949 of file xfs_log.c.  Return address = 0xc022b827
 >>  > Oct 20 01:15:28 TruStore1000 kernel: Log I/O Error Detected.  Shutting
 >>  down filesystem: lvm(58,4)
 >>  > Oct 20 01:15:28 TruStore1000 kernel: Please umount the filesystem, and
 >>  rectify the problem(s)
 >>  > Oct 20 01:15:29 TruStore1000 kernel: XFS mounting filesystem lvm(58,3)
 >>  > 
 >>  > The filesystem config_snap has already been deleted.
 >>  > 
 >>  > I have not tried to reproduce this, but I suspect I can.  It has
 >>  happened twice so far.
 >>  > 
 >>  > Is it likely a updated CVS kernel would fix this?
 >>  > 
 >>  > Is there more data I can provide?
 >>  > 
 >>  > Thanks,
 >>  > Greg
 >>  > =============
 >>  > Greg Freemyer
 >>  > Internet Engineer
 >>  > Deployment and Integration Specialist
 >>  > Compaq ASE - Tru64 v4, v5
 >>  > Compaq Master ASE - SAN Architect
 >>  > The Norcross Group
 >>  > www.NorcrossGroup.com
 >>  > 
 >>  -- 
 >>  Eric Sandeen      XFS for Linux     http://oss.sgi.com/projects/xfs
 >>  sandeen@xxxxxxx   SGI, Inc.         651-683-3102






Greg Freemyer
Internet Engineer
Deployment and Integration Specialist
Compaq ASE - Tru64 v4, v5
Compaq Master ASE - SAN Architect
The Norcross Group
www.NorcrossGroup.com


<Prev in Thread] Current Thread [Next in Thread>