SUMMARY: RZ28 Disk being killed by new SCSI controller

From: Chris Dale <oucd_at_alinga.newcastle.edu.au>
Date: Fri, 14 Jul 1995 12:25:03 +1000 (EST)

My original Posting...

> This one has my local field service engineer stumped.
>
> I have just had a second SCSI controller installed on my Alpha 2100
> and cabled into a BA-350 storage shelf. This went fine, the console
> sees the SCSI controller in the PCI bus and if you put a rz28 disk
> in slot 0 it can see it OK. I then booted the OS (OSF/1 V3.0) and
> configured the kernal to see the new disk and all was good, partioned
> the new disk, created a file system (using UFS) and mounted and it all
> seemed to work. Shut the OS down and the disk could no longer be seen
> at through the SHOW DEV or SHOW CONFIG commands, it was dead. OK bad
> disk we figured, replaced with a new one, yes it could be seen at the
> console and all worked OK UNTIL you halt the OS and then you can't
> see it again. 2 disks in as many days ! My field engineer is not game
> to put another one in yet until he does more research. the system
> functions perfectly otherwise ? All the disks on the first SCSI are fine.
>
> Anybody have any ideas or seen similar behaviour ?

Quite a few replies to this one...

   Check the Firmware Level (Yes it was OK, but a good idea)

   Check the terminations (Yes OK too, but an equally good idea)

   Check all the SCSI settings, partic fast SCSI (All OK)

   Upgrade to OSF/1 3.2 (Would be nice but our applications
                         aren't up there yet)

   Install the drive in some other SCSI machine and do diagnotics
                         ( FS Did this, the disk was really dead
                           would not spin up under any cirumstances)

   Similar problems have been encountered with third partt (Seagate)
   drives, it turned out it was a bad drive batch.

THE SOLUTION.

Drive number three, another RZ28 brick worked fine.

Possible causes....
         There has been a known bad batch of RZ28's and maybe I got
         two in a row !

         The power connector on the disks has been known to be flaky
         and may have introduced voltage spikes at startup, reseated
         all cables in the chain.

in any case we may never really know, but it works now.

Have a good weekend...

Chris Dale
--------------------------------------------

Many Thanks to...

Patrick O'Brien pobrien_at_draco.harvard.edu
                 alan_at_nabeth.cxo.dec.com
Del Merritt del_at_IntraNet.com
Jon Buchanan Jonathan.Buchanan_at_ska.com
John Stoffel john_at_WPI.EDU
Bob Wier wier_at_bobcat.etsu.edu
Tim McKee mckee_at_admin.sunbelt.net
                 golden_at_falcon.invincible.com
Knut Helleboe Knut.Hellebo_at_nho.hydro.com
Received on Fri Jul 14 1995 - 04:35:33 NZST

This archive was generated by hypermail 2.4.0 : Wed Nov 08 2023 - 11:53:45 NZDT