RAID1 and kernel panic

Sascha Ziemann (szi@khs-ag.de)
Wed, 15 Jul 1998 18:32:50 +0200


Hi,

I did a simple RAID1 test: I created a RAID1 stack with two partitions
on two different disks on the same SCSI bus (Adaptec 2940AU, Seagate ST
34555N). After removing the power for one disk the RAID driver wrote to
syslog, that there is a disk error and that from now on only one disk
will be used:

Jul 15 13:25:06 gate kernel: scsi0 channel 0 : resetting for second half of retries.
Jul 15 13:25:06 gate kernel: SCSI bus is being reset for host 0 channel 0.
Jul 15 13:25:06 gate kernel: SCSI disk error : host 0 channel 0 id 3 lun 0 return code = 80000
Jul 15 13:25:06 gate kernel: scsidisk I/O error: dev 08:22, sector 7274688, absolute sector 7547793

But after that I got continously SCSI errors for the switched off
device:

Jul 15 13:26:03 gate kernel: SCSI disk error : host 0 channel 0 id 3 lun 0 return code = 26030000
Jul 15 13:26:03 gate kernel: scsidisk I/O error: dev 08:22, sector 7367978, absolute sector 7641083
Jul 15 13:26:03 gate kernel: (scsi0:0:0:0) Synchronous at 10.0MHz, offset 15.
Jul 15 13:26:06 gate kernel: SCSI disk error : host 0 channel 0 id 3 lun 0 return code = 26030000
Jul 15 13:26:06 gate kernel: scsidisk I/O error: dev 08:22, sector 7367492, absolute sector 7640597
Jul 15 13:26:07 gate kernel: SCSI disk error : host 0 channel 0 id 3 lun 0 return code = 26030000
Jul 15 13:26:07 gate kernel: scsidisk I/O error: dev 08:22, sector 7367738, absolute sector 7640843

And in the end a Kernel panic: "Interrupt routine called
reentrantly".

I think this is not a problem of the RAID driver, but the SCSI driver.
Someone gave me a mystic hint to remove the CD-ROM, but I would like
to know what the error is before I start such strange actions. Anybody
here who has an idea, what is going wrong? And why Linux has a problem
with the interrupt handler, because of a not existing disk? I think
this must be a bug.

-- 
Help a swallow land at Capistrano.

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html