Hi,
The university I work at has 4 Blade HS20 servers connected to a IBM
DS400 storage via Fibre Channel.
The fibre channel module is:
06:01.0 Fibre Channel: QLogic Corp. QLA2312 Fibre Channel Adapter (rev 02)
06:01.1 Fibre Channel: QLogic Corp. QLA2312 Fibre Channel Adapter (rev 02)
# lsmod
Module Size Used by
...
qla2xxx 143584 1
# l /lib/firmware
total 244
lrwxrwxrwx 1 root root 21 2007-05-05 22:02 ql2300_fw.bin -> ql2300_fw.bin.3.03.18
With recent kernels filesystem corruption occurs under high disk I/O.
I have just tried 2.6.19 and I got the same, but kernel 2.6.18 and
before looks safe.
The test I use is the following:
I first checkout the kernel's git tree:
# git-clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git linux-2.6.git
Then
# for i in {100..400}; do; rsync -a linux-2.6.git/ c2_$i& done
.... waiting for finish
# for i in {100..400}; do; rsync -a linux-2.6.git/ c3_$i& done
.... waiting for finish
# for i in {100..400}; do; rsync -a linux-2.6.git/ c4_$i& done
Some error messages:
ReiserFS: warning: is_tree_node: node level 32010 does not match to the expected one 1
ReiserFS: sdb1: warning: vs-5150: search_by_key: invalid format found in block 8093723. Fsck?
ReiserFS: sdb1: warning: vs-13070: reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [961831 1064427 0x0 SD]
I am able to reproduce this errors.
With other filesystems I get similar errors.
I suspect a bug in qla2xxx, but I am not sure.
Can you help me how to track down this problem?
Thanks in advance.
PS: please cc to me.
--
CSILLAG Tamas (cstamas) - http://digitus.itk.ppke.hu/~cstamas
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
