Xen users - Apr 2007 - XFS problem in domU with Xen 3.0.4 - 2.6.16.33-xen

Hello,

i''ve a dom0 running many domUs using ext3 filesystems, and only one 
using XFS (only / partition in all cases)

The kernel is the official 2.6.16.33-xen from xen 3.0.4 (from xensource)

Whe are using a PERC5/i (Dell PowerEdge 1900) hardware RAID controller 
with SAS disks in RAID5 mode, which excludes hard disk hardware failures

from times to times the XFS domU freezes :

XFS internal error XFS_WANT_CORRUPTED_GOTO at line 1563 of file 
fs/xfs/xfs_alloc.c. Caller 0xf490175c
[<f4900435>] xfs_free_ag_extent+0x325/0x700 [xfs]
[<f490175c>] xfs_free_extent+0xdc/0x110 [xfs]
[<f490175c>] xfs_free_extent+0xdc/0x110 [xfs]
[<f495e9be>] kmem_zone_alloc+0x3e/0xc0 [xfs]
[<f495ea66>] kmem_zone_zalloc+0x26/0x50 [xfs]
[<f4913e23>] xfs_bmap_finish+0x113/0x1a0 [xfs]
[<f493bc49>] xfs_itruncate_finish+0x299/0x3b0 [xfs]
[<f495350b>] xfs_trans_ijoin+0x2b/0x80 [xfs]
[<f49596f1>] xfs_inactive+0x421/0x510 [xfs]
[<c01958df>] inotify_inode_is_dead+0xf/0x90
[<f4968a7f>] linvfs_clear_inode+0x9f/0xb0 [xfs]
[<c019d0e0>] dquot_drop+0x0/0x90
[<c0186448>] clear_inode+0xc8/0x160
[<c0151f87>] truncate_inode_pages+0x17/0x20
[<c01872f6>] generic_delete_inode+0x106/0x160
[<c0187523>] iput+0x53/0x70
[<c017d397>] do_unlinkat+0xc7/0x120
[<c011b31f>] do_sched_setscheduler+0x5f/0xe0
[<c0105461>] syscall_call+0x7/0xb
xfs_force_shutdown(sda1,0x8) called from line 4200 of file 
fs/xfs/xfs_bmap.c. Return address = 0xf496969c
Filesystem "sda1": Corruption of in-memory data detected. Shutting
down
filesystem: sda1

The freeze happens during the logrotate ( it seems to be during the 
syslog rotation since we loose the syslog from the day before )

The first freeze was about after a couple of weeks uptime, and this time 
the domU was up only for  a couple of days

does anyone have any idea and/or the same problem ?

Thanks in advance

@+
Jérôme


_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users