Trouble with Adaptec AIC7902 Ultra320 SCSI

[Andy Farkas]

On 23 Apr 2003, Michael Bruening wrote:

> I ran into SCSI controller errors during a large file transfer from a
> FreeBSD4.8-RELEASE system. Logs showed these errors had appeared once
> before: ahd card state is dumped. Details below include (1) dmesg, (2)
> logs generated when large recursive scp was interrupted, (3) logs from
> a similar occurrence from couple days earlier around 3am (when system
> was probably quite idle), and (4) output of "pciconf -lv". System is
> dual Xeon Micron with two SCSI drives and kernel compiled with
> hyperthreading and SMP (see dmesg below for more detail on the system).

Can't help with your specific problem, but (3) above is when the periodic
daily script usually kicks in and rattles the disks quite a bit.

Quick question: do you get the "<<< dump card state ends >>>" messages
spammed to all users logged in?  Just wondering if anyone can confirm I'm
not going crazy... (I posted about this a week ago..)


> Apr 21 18:18:11 dpsfbd01 /kernel: <<<<<<<<<<<<<<<< Dump Card State Ends >>>>>>>>>>>>>>>>>>

> Apr 21 18:18:11 dpsfbd01 /kernel: <<<<<<<<<<<<<<<< Dump Card State Ends >>>>>>>>>>>>>>>>>>

> Apr 21 18:18:11 dpsfbd01 /kernel: <<<<<<<<<<<<<<<< Dump Card State Ends >>>>>>>>>>>>>>>>>>

> Apr 21 18:21:11 dpsfbd01 /kernel: <<<<<<<<<<<<<<<< Dump Card State Ends >>>>>>>>>>>>>>>>>>

> Apr 21 18:21:12 dpsfbd01 /kernel: <<<<<<<<<<<<<<<< Dump Card State Ends >>>>>>>>>>>>>>>>>>

> Apr 21 18:21:12 dpsfbd01 /kernel: <<<<<<<<<<<<<<<< Dump Card State Ends >>>>>>>>>>>>>>>>>>

> Apr 16 03:02:48 dpsfbd01 /kernel: <<<<<<<<<<<<<<<< Dump Card State Ends >>>>>>>>>>>>>>>>>>


It would be good if you could run syslogd with -vv flag to confirm that
the above messages are logged at <kern.emerg> level.

--

 :{ andyf at speednet.com.au

        Andy Farkas
    System Administrator
   Speednet Communications
 http://www.speednet.com.au/