[1139] in linux-scsi channel archive
Re: medium error, what next?
daemon@ATHENA.MIT.EDU (Gerard Roudier)
Fri Dec 20 16:55:25 1996
Date: Fri, 20 Dec 1996 22:09:26 +0000 (GMT)
From: Gerard Roudier <groudier@club-internet.fr>
To: "Stephen R. van den Berg" <srb@cuci.nl>
cc: linux-scsi@vger.rutgers.edu, ncr53c810@colorado.edu
In-Reply-To: <199612201118.MAA06113@hera.cuci.nl>
Use scsi-config as suggested by Leonard to change the Read-Write Recovery
Page.
Set ARRE to 1 as suggested by Stefan.
If EER bit change is supported by your drive set it to 1
If DCR bit ................................. set it to 0
If Read retry count ................................. set it to max value
Save the changes as follow:
1 - Use try change button.
2 - If all seems ok, you can save the current page, and try to access the
drive. The page will be restored to saved one on next reset.
3 - If all seems ok, you can read the saved page, make changes and save it.
The page is now saved permanently for next reboot.
Be carefull. Bad parameters value can make your drive unusable.
Use "dd" with a suitable value of "skip" and "count" parameters in order to
only try to read the wrong sector. You will spare time.
Retry the dd severall times if necessary.
Using a bootable disquette instead with "dd" in ram disk or a mounted
read only root file system would allow you to spare time if the system
crashes at each "dd" command.
If that did not work, :(
Set AWRE to 1.
Use a suitable "dd" command that just tries to write the wrong sector
with some zero values.
Retry this command severall times if necessary.
If that did not work ... stop :(
If you're not too much infortunate, you just will lose a file and fsck
will probably give you information in order to retreive and restore
the file.
Gerard.
On Fri, 20 Dec 1996, Stephen R. van den Berg wrote:
>
> This is what I found in the logs (I had deliberately done a
> dd bs=1024 if=/dev/sdd of=/dev/null):
>
> Dec 20 12:05:41 dionysus kernel: scsi0: MEDIUM ERROR on channel 0, id 3, lun 0, CDB: 0x08 07 f8 78 78 00
> Dec 20 12:05:41 dionysus kernel: Current error sd08:30: sns = f0 3
> Dec 20 12:05:41 dionysus kernel: ASC=11 ASCQ= 0
> Dec 20 12:05:41 dionysus kernel: Raw sense data:0xf0 0x00 0x03 0x00 0x07 0xf8 0x7f 0x18 0x00 0x00 0x00 0x00 0x11 0x00 0x00 0x80
> Dec 20 12:05:41 dionysus kernel: scsidisk I/O error: dev 08:30, sector 522366
>
> This is on a 2.0.27 machine, under ELF, using the BSD-ncr driver.
>
> What should I do at this point (beside panic :-)?
> How do I tell the drive to remap this particular sector? Or, alternately,
> maybe I should make Linux skip it?
>
> Anyone who can spare some words of wisdom here?
> --
> Sincerely, srb@cuci.nl
> Stephen R. van den Berg (AKA BuGless).
> Our four weapons are: fear, surprise, a ruthless efficiency, an almost
> fanatical devotion to the Pope... and nice red uniforms.
>
>
