Fixing disk errors

[Michael Mansour]

Hi,

> I've been on holidays and one of my production
> systems
> has produced disk errors on a linux RAID 0 stripe
> (FC1). After correcting hundreds of errors using
> fsck
> in single-user mode, I continue to get the following
> 2
> errors on each fsck run:
> 
> # fsck -f /dev/md6
> fsck 1.34 (25-Jul-2003)
> e2fsck 1.34 (25-Jul-2003)
> Pass 1: Checking inodes, blocks, and sizes
> Error reading block 6689 (Attempt to read block from
> filesystem resulted in short read) while doing inode
> scan.  Ignore error<y>? yes
> 
> Force rewrite<y>? yes
> 
> Error reading block 7714 (Attempt to read block from
> filesystem resulted in short read) while doing inode
> scan.  Ignore error<y>? yes
> 
> Force rewrite<y>? yes
> 
> I can mount the filesystem without dramas, but am
> wondering how I can fix the above issues
> permanently?
> 
> Thanks.
> 
> Michael.

As a supplement to the above information, when I run
an:

# e2fsck -f -y /dev/md6

on that filesystem (not mounting it at boot time), the
following is also produced in the /var/log/messages
file:

Jul 17 17:26:31 server kernel: end_request: I/O error,
dev 21:09 (hde), sector 30876
Jul 17 17:26:31 server kernel: hde: 0 bytes in FIFO
Jul 17 17:26:31 server kernel: hde: timeout waiting
for DMA
Jul 17 17:26:31 server kernel: hde: dma_intr:
status=0x51 { DriveReady SeekComplete Error }
Jul 17 17:26:31 server kernel: hde: dma_intr:
error=0x40 { UncorrectableError }, LBAsect=69544196,
high=4, low=2435332, sector=3087
8
Jul 17 17:26:31 server kernel: end_request: I/O error,
dev 21:09 (hde), sector 30878
Jul 17 17:26:31 server kernel: hde: 0 bytes in FIFO
Jul 17 17:26:31 server kernel: hde: timeout waiting
for DMA
Jul 17 17:26:31 server kernel: hde: dma_intr:
status=0x51 { DriveReady SeekComplete Error }
Jul 17 17:26:31 server kernel: hde: dma_intr:
error=0x40 { UncorrectableError }, LBAsect=69544198,
high=4, low=2435334, sector=3088
0
Jul 17 17:26:31 server kernel: end_request: I/O error,
dev 21:09 (hde), sector 30880
Jul 17 17:26:31 server kernel: hde: 0 bytes in FIFO
Jul 17 17:26:31 server kernel: hde: timeout waiting
for DMA
Jul 17 17:26:31 server kernel: hde: dma_intr:
status=0x51 { DriveReady SeekComplete Error }
Jul 17 17:26:31 server kernel: hde: dma_intr:
error=0x40 { UncorrectableError }, LBAsect=69544200,
high=4, low=2435336, sector=3088
2
Jul 17 17:26:31 server kernel: end_request: I/O error,
dev 21:09 (hde), sector 30882
Jul 17 17:26:31 server kernel: hde: 0 bytes in FIFO
Jul 17 17:26:31 server kernel: hde: timeout waiting
for DMA
Jul 17 17:26:31 server kernel: hde: dma_intr:
status=0x51 { DriveReady SeekComplete Error }
Jul 17 17:26:31 server kernel: hde: dma_intr:
error=0x40 { UncorrectableError }, LBAsect=69544202,
high=4, low=2435338, sector=3088
4

Any ideas how I can fix this?

It should be noted that /dev/md6 is an 80Gb stripe on
two 40Gb partitions on two 80Gb IDE disks. The rest of
the partitions on the same drives all function fine
and are in mirrored sets.

Thanks.

Michael.


Find local movie times and trailers on Yahoo! Movies.
http://au.movies.yahoo.com