xfsdump, xfsrestore segmentation faults

[Christopher Jones <cjones@lifesci.ucsb.edu>]

Hi everyone,

I'm experiencing some problems with XFS on a Redhat 7.2 installation 
using the SGI supplied installer, and was hoping that someone might be 
able to provide some advice.

I've had the system up for 2 months now, but in the last 2 weeks, have 
fould that both xfsdump and xfsrestore will crash the system.  I had 
consistently been doing backups just fine, until around January 28th, 
during a dump, the system load peaked up in the 40's as reported by top 
(although cpu and memory were fine).

My typical dump would look like (sans the escapes):

/sbin/xfsdump -F -o -l 9 -L session091 \
               -M tue-2002-02-08 \
               -f /dev/nst0 /home

and a restore:

xfsrestore -if /dev/nst0 /home

I have two LVM volumes (home and /scratch01):

Filesystem            Size  Used Avail Use% Mounted on
/dev/sda1             1.9G  368M  1.5G  19% /
/dev/sda6              97M  4.8M   92M   5% /boot
/dev/sda7             9.8G  4.9G  4.9G  50% /usr
/dev/sda8             3.9G  243M  3.6G   7% /var
/dev/sda10             55G   18G   36G  33% /d01
none                  753M     0  753M   0% /dev/shm
/dev/vg03/lv03         98G   64G   33G  66% /home
/dev/vg00/lv00         78G   70G  8.7G  89% /scratch01

Sytem software:

Linux 2.4.9-13SGI_XFS_1.0.2smp i686

lvm_1.0.1-rc4
xfsdump-1.1.12-0
xfsprogs-1.3.16-0
xfsprogs-devel-1.3.16-0

After the first time the dump seg faulted, I did an xfs_repair, which 
found many orphaned inodes, and placed them in lost+found (on the order 
of 200).

Here's the output from one of the failed dumps:

[root@mistral log]# /sbin/xfsdump -F -o -l 0 -L session087 -M 
tue-2002-02-08 -f
/dev/nst0 /
/sbin/xfsdump: using scsi tape (drive_scsitape) strategy
/sbin/xfsdump: version 3.0 - Running single-threaded
/sbin/xfsdump: level 0 dump of mistral:/
/sbin/xfsdump: dump date: Fri Feb  8 03:35:18 2002
/sbin/xfsdump: session id: e04d9485-3601-4d5b-86f0-1a946c5f117f
/sbin/xfsdump: session label: "session087"
/sbin/xfsdump: ino map phase 1: skipping (no subtrees specified)
/sbin/xfsdump: ino map phase 2: constructing initial dump list
/sbin/xfsdump: ino map phase 3: skipping (no pruning necessary)
/sbin/xfsdump: ino map phase 4: skipping (size estimated in phase 2)
/sbin/xfsdump: ino map phase 5: skipping (only one dump stream)
/sbin/xfsdump: ino map construction complete
/sbin/xfsdump: estimated dump size: 382444928 bytes
/sbin/xfsdump: preparing drive
/sbin/xfsdump: WARNING: media may contain data. Overwrite option specified
Segmentation fault

I've tried the -v trace, but don't get any different output.

Is anyone else having problems with these cominations?

Can anyone suggest a next step in troubleshooting?

Thanks very much in advance,

Chris