** Description changed:

  Symptoms:
  
- These symptoms were also present in a Hardy installation.  Upgraded to
- Intrepid in the hopes of resolving the issue.
+ The system is extremely unstable during high loads of the NFS service.
+ Symptoms include:
+ 
+    1. XFS filesystem becoming unavailable
+    2. System freezing completely
+    3. Various kern.log messages
+ 
+ Common message:
+ 
+ XFS internal error XFS_WANT_CORRUPTED_RETURN at line 281 of file 
/build/buildd/linux-2.6.24/fs/xfs/xfs_alloc.c. Caller 0xffffffff882bd4c6 
+ Filesystem "dm-2": XFS internal error xfs_trans_cancel at line 1163 of file 
/build/buildd/linux-2.6.24/fs/xfs/xfs_trans.c. Caller 0xffffffff8830b9b7 
+ Filesystem "dm-2": Corruption of in-memory data detected. Shutting down 
filesystem: dm-2 
+ Please umount the filesystem, and rectify the problem(s)
+ 
+ These symptoms were present in a Hardy installation.  Upgraded to
+ Intrepid in the hopes of resolving the issue but no improvement is
+ noticeable.  All log messages in this bug report are from 8.04.
  
  Hardware:
  
  Dell PowerEdge 2950 (system) specs: http://is.gd/4Eohj
  Dell Powervault MD1000 (storage) specs: http://is.gd/4EqcR
  
  Ubuntu release:
  
      8.10 64bit server
      Linux miceserver 2.6.27-15-server #1 SMP Tue Oct 20 07:30:41 UTC 2009 
x86_64 GNU/Linux
  
  Overview of setup:
  
  * System *
  Internal drive /dev/sda
  - sda1 as /boot (ext3)
  - sda5 as LVM physical volume with volume group miceserver2
  
  Logical volumes:
  
  miceserver2-root as /root (ext3)
  miceserver2-swap_1 as swap
  
  * Storage *
  Each pair of enclosures is RAID50 and is connected to a PERC6/e controller.
  
  First pair of enclosures - virtual volume - /dev/sdb
  Second pair of enclosures - virtual volume - /dev/sdc
  
  sdb and sdc are made RAID0 via LVM volume group "DataVG"
  
  That group contains logical volume "micedata" which has 39TB XFS
  filesystem exported via NFS.
  
  The disks.txt attachment details the entire disk subsystem including
  complete LVM configuration.
  
  Notes:
  
  a) Physical memory was swapped out (no change).

** Description changed:

  Symptoms:
  
- The system is extremely unstable during high loads of the NFS service.
- Symptoms include:
+ System is extremely unstable during high loads of NFS service exporting
+ 39TB of storage.  Symptoms include:
  
-    1. XFS filesystem becoming unavailable
-    2. System freezing completely
-    3. Various kern.log messages
+    1. XFS filesystem becoming unavailable
+    2. System freezing completely
+    3. Various kern.log messages
  
  Common message:
  
- XFS internal error XFS_WANT_CORRUPTED_RETURN at line 281 of file 
/build/buildd/linux-2.6.24/fs/xfs/xfs_alloc.c. Caller 0xffffffff882bd4c6 
- Filesystem "dm-2": XFS internal error xfs_trans_cancel at line 1163 of file 
/build/buildd/linux-2.6.24/fs/xfs/xfs_trans.c. Caller 0xffffffff8830b9b7 
- Filesystem "dm-2": Corruption of in-memory data detected. Shutting down 
filesystem: dm-2 
+ XFS internal error XFS_WANT_CORRUPTED_RETURN at line 281 of file 
/build/buildd/linux-2.6.24/fs/xfs/xfs_alloc.c. Caller 0xffffffff882bd4c6
+ Filesystem "dm-2": XFS internal error xfs_trans_cancel at line 1163 of file 
/build/buildd/linux-2.6.24/fs/xfs/xfs_trans.c. Caller 0xffffffff8830b9b7
+ Filesystem "dm-2": Corruption of in-memory data detected. Shutting down 
filesystem: dm-2
  Please umount the filesystem, and rectify the problem(s)
  
  These symptoms were present in a Hardy installation.  Upgraded to
  Intrepid in the hopes of resolving the issue but no improvement is
  noticeable.  All log messages in this bug report are from 8.04.
  
  Hardware:
  
  Dell PowerEdge 2950 (system) specs: http://is.gd/4Eohj
  Dell Powervault MD1000 (storage) specs: http://is.gd/4EqcR
  
  Ubuntu release:
  
      8.10 64bit server
      Linux miceserver 2.6.27-15-server #1 SMP Tue Oct 20 07:30:41 UTC 2009 
x86_64 GNU/Linux
  
  Overview of setup:
  
  * System *
  Internal drive /dev/sda
  - sda1 as /boot (ext3)
  - sda5 as LVM physical volume with volume group miceserver2
  
  Logical volumes:
  
  miceserver2-root as /root (ext3)
  miceserver2-swap_1 as swap
  
  * Storage *
  Each pair of enclosures is RAID50 and is connected to a PERC6/e controller.
  
  First pair of enclosures - virtual volume - /dev/sdb
  Second pair of enclosures - virtual volume - /dev/sdc
  
  sdb and sdc are made RAID0 via LVM volume group "DataVG"
  
  That group contains logical volume "micedata" which has 39TB XFS
  filesystem exported via NFS.
  
  The disks.txt attachment details the entire disk subsystem including
  complete LVM configuration.
  
  Notes:
  
- a) Physical memory was swapped out (no change).
+ a) Physical memory was swapped out (no change)
+ b) Memory was originally 8 GB and now has 32 GB

-- 
Large XFS filesystem crashing / unmounting
https://bugs.launchpad.net/bugs/481080
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to