need help with elusive hardware problem

[Mark Vollmers]

Hi, all. I have an issue that I'm sure is hardware related, but one that I havn't been able to pin down yet. We are running a D-class server with 10.20. We have a RAID drive mounted as /home and an external tape drive. When I go to run backup, it starts fine and then stops on me and gives me a bunch of SCSI errors (syslog is attached. This is representative of what I'm seeing). I have been screwing around with the controller speed, since there were a bunch of /home file system errors (async, system reading errors, etc.) I think that I might have that nailed down, but I still have the SCSI errors. The SCSI adaptor card was replaced last winter, and the controller was just replaced a few weeks ago (old one fried, but there were scsi errors a few weeks before that, along with async ones and the like). The cable was just changed. We have been having SCSI problems for a while. The problem really only appears with backup (I assume due to volume of files being moved or size of files).
My question is: what do I persue? are the errors that I am seeing more likely to be RAID controller or SCSI card? I've been chasing this bugger for about a month now, and would really like to be able to get a successful backup run. Any thoughts? Thanks!

Mark

[Mark Vollmers]

whoops. Forgot the syslog. Sorry

mark

[Vincenzo Restuccia]

Verify SCSI termination,BCC and controller.

[Patrick Wallek · Accepted solution]

Mark,

Do you have the latest SCSI cumulative patch installed? The latest one is PHKL_23259 for 10.20.

Here is the link describing the patch:
http://us-support.external.hp.com/cki/bin/doc.pl/sid=9932c38e0385ce88b1/screen=ckiDisplayDocument?docId=200000054200660

It mentions the error you are getting as fixed in one of the previous iterations of this patch.

[Mark Vollmers]

Vincenzo, I've checked the termination a while ago, back when it first started to happen.
Patrick- I don't have that patch installed. I'll get that one and plop it in. Could I really be lucky enough to have the patch fix it, though? Thanks.

Mark

[Tracey]

Several years ago my D210 was having similar problems when the system load wash high. It took months, and many crashes before we pinned it down to an internal cable for the interanl drives. We didn't have a array, but we an an external storage box that was daisy chained into the main controller.

[Alan Edwards]

I had a similar problem (on a D box as well) and after HP changed everything; tape drive (twice) cables, and motherboard, it turned out to be the CD ROM drive that was causing the problem. Try disconnecting the CD ROM drive and running your backup.

Alan