Debian Lenny Random SCSI Errors

jmoschetti45 · 12-04-2009, 05:46 PM

What happens: A SCSI drive randomly seems to corrupt, but a reboot fixes it.

Parts of log files relevant.

Code:

/var/log/debug:
sd 0:0:2:0: [sdc] Mode Sense: f3 00 10 08

Code:

/var/log/dmesg
scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 7.0
<Adaptec 2902/04/10/15/20C/30C SCSI adapter>
aic7850: Single Channel A, SCSI Id=7, 3/253 SCBs

scsi 0:0:2:0: Direct-Access     IBM      DXHS18Y          0430 PQ: 0 ANSI: 3
scsi0:A:2:0: Tagged Queuing enabled.  Depth 8
scsi target0:0:2: Beginning Domain Validation
scsi target0:0:2: FAST-10 SCSI 10.0 MB/s ST (100 ns, offset 15)
scsi target0:0:2: Domain Validation skipping write tests
scsi target0:0:2: Ending Domain Validation

sd 0:0:2:0: [sdc] 35566480 512-byte hardware sectors: (18.2 GB/16.9 GiB

sd 0:0:2:0: [sdc] Write Protect is off
sd 0:0:2:0: [sdc] Mode Sense: f3 00 10 08
sd 0:0:2:0: [sdc] Write cache: enabled, read cache: enabled, supports DPO and FUA

sd 0:0:2:0: [sdc] Attached SCSI disk

sd 0:0:2:0: Attached scsi generic sg2 type 0

Code:

/var/log/kern.log (this is just the main part, it goes on for a long time)
sd 0:0:2:0: [sdc] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK
end_request: I/O error, dev sdc, sector 3832287
Buffer I/O error on device sdc1, logical block 479028
lost page write due to I/O error on sdc1
Buffer I/O error on device sdc1, logical block 479029
lost page write due to I/O error on sdc1
sd 0:0:2:0: [sdc] Unhandled error code
sd 0:0:2:0: [sdc] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK
end_request: I/O error, dev sdc, sector 3486735
sd 0:0:2:0: [sdc] Unhandled error code
sd 0:0:2:0: [sdc] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK
end_request: I/O error, dev sdc, sector 3832295
Buffer I/O error on device sdc1, logical block 479029
lost page write due to I/O error on sdc1
sd 0:0:2:0: [sdc] Unhandled error code
sd 0:0:2:0: [sdc] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK
end_request: I/O error, dev sdc, sector 3486735
..............
end_request: I/O error, dev sdc, sector 17776464
end_request: I/O error, dev sdc, sector 17776464
I/O error in filesystem ("sdc1") meta-data dev sdc1 block 0x10f3f11       ("xlog_iodone") error 5 buf count 3584
xfs_force_shutdown(sdc1,0x2) called from line 1043 of file /build/buildd/linux-2.6-2.6.30/debian/build/source_i386_none/fs/xfs/xfs_log.c.  Return address = 0xe051c7f5
Filesystem "sdc1": Log I/O Error Detected.  Shutting down filesystem: sdc1
Filesystem "sdc1": xfs_log_force: error 5 returned.
Please umount the filesystem, and rectify the problem(s)
Filesystem "sdc1": xfs_log_force: error 5 returned.
Filesystem "sdc1": xfs_log_force: error 5 returned.
............

Unmounting it is impossible, even with force and lazy option. It appears unmounted, but trying to run xfs_check or xfs_repair says it's still in use. Remounting it doesn't do any good, the file system is unusable.

code933k · 12-04-2009, 06:11 PM

That is a well know bug in many distributions (SCSI Linux kernel driver seems to be the problem)
I don't experiment this in HDDs but CD-rom instead.
My kernel is currently 2.6.31 ... And still no clue.

However, try adding hdc=noprobe (or corresponding) to your boot loader. See if that works around the problem for you.

jmoschetti45 · 12-12-2009, 10:47 AM

Doesn't seem to help any. Problem has more info now though:

Code:

Shelly:~# xfs_check /dev/sdb1
xfs_check: unexpected XFS SB magic number 0x15594019
/usr/sbin/xfs_check: line 28:  7035 Floating point exceptionxfs_db$DBOPTS -i -p xfs_check -c "check$OPTS" $1
Shelly:~# xfs_repair /dev/sdb1
xfs_repair: warning - cannot set blocksize on block device /dev/sdb1: Device or resource busy
Phase 1 - find and verify superblock...
bad primary superblock - bad magic number !!!

attempting to find secondary superblock...
..................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................found candidate secondary superblock...
unable to verify superblock, continuing...
.Sorry, could not find valid secondary superblock
Exiting now.
Shelly:~# fdisk -l /dev/sdb1

Disk /dev/sdb1: 1612 MB, 1612122624 bytes
255 heads, 63 sectors/track, 195 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x1e993d6a

Disk /dev/sdb1 doesn't contain a valid partition table

code933k · 12-13-2009, 08:43 PM

It seems you caught one or both bugs together... I had to move from XFS because of a bug in a 2.6.2(8|9) kernel