kernel: aacraid: Host adapter reset request. SCSI hang ?

Hello,
 
We recently purchased two Dell PowerEdge 2650 servers with PERC3/Di
controllers. Both servers are executing RedHat Linux 9.0. On both servers we
are encountering the following error:
 
<<< Portion of server message log >>>
May 31 16:14:07 server1 kernel: aacraid: Host adapter reset request. SCSI
hang ?
May 31 16:14:17 server1 kernel: scsi: device set offline - command error
recover failed: host 0 channel 0 id 0 lun 0
May 31 16:14:17 server1 kernel: SCSI disk error : host 0 channel 0 id 0 lun
0 return code = 6000000
May 31 16:14:17 server1 kernel:  I/O error: dev 08:03, sector 83200
May 31 16:14:17 server1 kernel:  I/O error: dev 08:03, sector 13568
May 31 16:14:17 server1 kernel:  I/O error: dev 08:03, sector 13616
May 31 16:14:17 server1 kernel:  I/O error: dev 08:03, sector 83200
May 31 16:14:17 server1 kernel:  I/O error: dev 08:03, sector 22030904
May 31 16:14:17 server1 kernel:  I/O error: dev 08:03, sector 88348712
May 31 16:14:17 server1 kernel:  I/O error: dev 08:03, sector 72976
May 31 16:14:17 server1 kernel:  I/O error: dev 08:03, sector 13624
May 31 16:14:17 server1 kernel:  I/O error: dev 08:03, sector 13752
May 31 16:14:17 server1 kernel:  I/O error: dev 08:03, sector 13768
May 31 16:14:17 server1 kernel:  I/O error: dev 08:03, sector 72976
<<< I/O error messages continue until the server is rebooted >>>
 
 
Here are a few notes regarding the error and operating environment:
 
- The error occurs with RedHat's kernel RPMs kernel-smp-2.4.20-9 and
kernel-smp-2.4.20-13.9. As of today, we are testing kernel-2.4.20-9 to
determine if the problem occurs under a non-smp environment.
- The time between failures varies from several hours to several days.
- The failures occur both during light and heavy system loads.
- PowerEdge 2650 BIOS is at 1.10 A10
- Backplane firmware is at 1.01
- PERC3/Di BIOS is at V2.7-1 (build 3170)
- A full system diagnostics has been successfully executed on both servers.
- The RAID media has been successfully 'verified' on both servers.
 
Thank you in advance for your assistance in helping to get this problem
resolved.
 
Javier
 
 
JLR Consulting, PO Box 638, Bernville, PA 19506-0638
mailto:jlr-slrnY02Ltykg9ELcbhaJlAC/[email protected]