For the last 7 months, I've been getting the following error in
/var/log/messages every night during the cron.daily execution. I've finally
tracked it down to happening during my tripwire run, and I suspect
(based on linear time into the run, and sizes of files) the problem file
lies somwhere in /usr/lib.
The error message has been identical for months, so I assume I have a bad
spot that is not spreading. I'd like to find the affected file, rename it
and ignore the problem for a while longer.
If I know the sector and lbasector, can I determine the inode and/or
the actual file affected?
The error message is:
Jan 23 04:24:34 helium kernel: hdc: dma_intr: status=0x51 { DriveReady SeekComplete Error }
Jan 23 04:24:34 helium kernel: hdc: dma_intr: error=0x40 { UncorrectableError }, LBAsect=4200315, sector=4200248
Jan 23 04:24:34 helium kernel: end_request: I/O error, dev 16:01 (hdc), sector 4200248
as I said before, the sector number has never changed in months.
thanks,
-chuck
--
ACCEL Services, Inc.| Specialists in Gravity, Magnetics | 1(713)993-0671 ph.
1980 Post Oak Blvd. | and Integrated Interpretation | 1(713)960-1157 fax
Suite 2050 | |
Houston, TX, 77056 | Chuck Campbell | [email protected]
| President & Senior Geoscientist |
"Integration means more than having all the maps at the same scale!"
> If I know the sector and lbasector, can I determine the inode and/or
> the actual file affected?
probably makes more sense to just "e2fsck -c".
On Wed, 23 Jan 2002, Chuck Campbell wrote:
> For the last 7 months, I've been getting the following error in
> /var/log/messages every night during the cron.daily execution. I've finally
> tracked it down to happening during my tripwire run, and I suspect
> (based on linear time into the run, and sizes of files) the problem file
> lies somwhere in /usr/lib.
>
> The error message has been identical for months, so I assume I have a bad
> spot that is not spreading. I'd like to find the affected file, rename it
> and ignore the problem for a while longer.
>
> If I know the sector and lbasector, can I determine the inode and/or
> the actual file affected?
find /usr/lib -type f|sed -e 's!.*!cat & >/dev/null || echo &!'|sh
On Jan 23, 2002 12:00 -0600, Chuck Campbell wrote:
> If I know the sector and lbasector, can I determine the inode and/or
> the actual file affected?
>
> The error message is:
>
> Jan 23 04:24:34 helium kernel: hdc: dma_intr: status=0x51 { DriveReady SeekComplete Error }
> Jan 23 04:24:34 helium kernel: hdc: dma_intr: error=0x40 { UncorrectableError }, LBAsect=4200315, sector=4200248
> Jan 23 04:24:34 helium kernel: end_request: I/O error, dev 16:01 (hdc), sector 4200248
>
> as I said before, the sector number has never changed in months.
If you run 'badblocks /dev/hdc1' it will do a full (read-only by default)
surface scan of the disk and report the bad blocks. This still doesn't
tell you the filename though.
You can use "debugfs /dev/hdc1" and then "icheck 525031" (assuming
you have a 4kB block ext2/ext3 filesystem on this drive) and then
"ncheck <inum>" for the inode number returned by icheck to find the
filename.
As someone else reported, running "e2fsck -c" will add this block to
the bad blocks list, and re-assign another block for the file in question.
It runs 'badblocks' in the background with the correct parameters (read
only check, correct blocksize for the filesystem).
Cheers, Andreas
--
Andreas Dilger
http://sourceforge.net/projects/ext2resize/
http://www-mddsp.enel.ucalgary.ca/People/adilger/