How to check Adaptec AIC7902B hostraid status / AIC7902B channel B question
 

A few months ago we purchased a Supermicro 2U server. I installed WBEL respin 1 on it, updated it but stayed with the the stock kernel, this is a RHEL3 clone.

This machine contains the infamous Adaptec AIC7902B.

IMHO the documentation about this controller could be a lot better, the only thing that's remotely useful is the file README.aic79xx !!

If I understand correctly the device contains 2 channels which can be configured separately

I have configured Channel A with hostraid, 2 RAID1 volumes exist (72 GB (seagate, this one bootable) and 36 GB (ibm))
I have configured Channel B without hostraid so it will function as an 'ordinary' scsi controller

On "channel B" I attached a Sony DDS4 tape drive
Somewhere you can set a master "Channel" or so, this is set to channel A

The kernel (2.4.21-4) acknowledges these settings, both a320raid and aic79xx are loaded as expected,
the two raid volumes are recognized as /dev/sda and /dev/sdb as they probably should. It also recognizes my tape drive.

Linux then starts correctly without any errors.

Finally now my questions
1) how do I check the status of both RAID1 arrays?
cat /proc/scsi/a320raid/1 just gives a version/model/build date information
cat /proc/scsi/aic79xx/0 gives more info, but of course only of channel B


2) is it allowed to attach a 'slow' SCSI tape device to Channel B? I read somewhere this can cause bus resets and stuff, and is not received well by raid controllers.
My hardware supplier insists both channels behave separately and I can add the device safely. But what about the master setting ?

3) A few days ago I really hit some problems:
kernel: scsi: device set offline - not ready or command retry failed after bus reset: host 1 channel 0 id 3 lun 0
(we're hosting a vmware VM on it with 8 files of 2GB each)

the next logic thing were of course filesystem errors (EXT3-fs error read failures etc...)
While restarting the system, the controller said the 2nd drive of the 2nd array (36 GB) was degraded.

I rebuilt the array, checked the file system, did some stuff on it, then verified the array with the bios utility: no problems, both drives were optimal.
Do you guys think this is an hardware problem or can it be caused by the tape device (maybe the first kernel device suggests this??)

If've disconnected the tape device as a precaution.

Perhaps there's some good soul out here who has useful information ?

Sincerely,
Guy

You can check the status with StorMan (closed source)