David,
It looks like your system and syslog are very similar to mine.
Are you using two aic7xxx controllers for 8 * 9Gb disks?
And is it SMP?
Mine is PII-500 2CPU, with 9 * 18GB on two aic7xxx U2W.
I've got very same scsi errors with you.
Due to the double disk failure, I tried rebuilding with the latest
linux-2.2.12 and 2.2.11 raid patch.(Previously, I used 2.2.10)
After that, one of the scsi channel hung about an hour of resync process.
I tried three times, and got same failure.
After that, I checked with aic7xxx mailing list, and found a message
which mentions about SMP issue. As I am a nuewbie, only thing I can do is
trial-and-error. I recompiled the kernel with no SMP, and it runs
five hours and 48 minutes until now. Mkraid took more than three
hours to resync my 8*18 Gb array.
Then I ran mke2fs with "-c" flag. I am not sure if it's necessary or
not. I did it just to make the SCSI systems busy and to see if it
crahes again or not. This took about an hour.
Now I am writing many 2GB files to fill up the file system.
After that, I will delete the files, and repeat the writing.
This is to make my file system busy and busy to see if I can see
any SCSI compalints.
--
sysuh
In article <[EMAIL PROTECTED]> you write:
>Another strange message, apparently after a disk failure:
>
>Sep 9 01:32:25 zorn kernel: scsi0: MEDIUM ERROR on channel 0, id 1, lun 0, CDB:
> Read (10) 00 00 89 a9 bf 00 00 80 00
>Sep 9 01:32:25 zorn kernel: Info fld=0x89aa2b, Current sd08:01: sense key
>Medium Error
...
