SUMMARY: HP 97548S on a Sparc 10 ?

From: whb@hoh-2.att.com
Date: Thu Dec 17 1992 - 16:54:34 CST


My original query was:

>I've got an HP 97548S SCSI drive which I'm trying to move from a
>SparcStation-1 to a SparcStation-10 which keeps saying "H*ll no,
>I won't go!"

and the symptoms were that when the HP was plugged into the Sparc-10's
SCSI bus, it caused problems even when a different device was being
accessed.

Well, the votes are in, and there are several, mostly non-conflicting
conclusions:

        1) It's probably a cable problem. The high-speed, synchronous
           SCSI on the Sparc-10 is more demanding of cables than that of
           a Sparc-1's SCSI. This is probably it; the cable on the HP
           is 2 meters long and of uncertain shielding.

        2) It might be a termination problem. Several people said that
           the Sparc-10's SCSI bus needed to be *actively* terminated
           (but nobody suggested specific vendors!). I have my doubts,
           since the CD-ROM is passively terminated and works just fine.

        3) It might be a firmware problem. Steve Holmes gave the best
           summary of this (sjh@math.purdue.edu):

>We just had similar problems. Things to check: make sure the PROM
>revision is 103 or 206. Other numbers won't work reliably on the
>HP97548. There is a B series and a C series at least. Our PROMS say
>C2SUB206 (or will when the new ones arrive tomorrow!). Look at left
>side of the drive (from the front view) inside near the bottom there
>should be a PROM standing up facing you hiding behind a part of the
>chassis. It should be labeled.

So how did I cure the problem? I gave up! The right cable was not
available (grr - we now have *three* different SCSI connectors!),
and still don't know where to get an active terminator, and getting
the right PROMs could take weeks, so the disk stays with the Sparc-1
(for now). Fortunately, this shouldn't be a problem for the owners
of either the new or old workstations.

I extend the usual abundant thanks to the speedy, although varying in
quality, replies from:

Richard Kurtz snowhit!rmk
Rob Scott zeke@mpl.ucsd.edu
Guntram Wolski gwolsk@sei.com
Andy Toy andy@mks.com
sater@cs.vu.nl
Glenn Satchell ups!glenn@fourx.Aus.Sun.COM
Steinar Haug Steinar.Haug@delab.sintef.no,
ilkka@ee.oulu.fi
Steve Holmes sjh@math.purdue.edu
Bruce Kall Kall@Mayo.edu
Ari shamash@boxhill.com
Michael Fischbein msf@fsg.COM
dworkin@shiara.rootgroup.com
John DiMarco jdd@cdf.toronto.edu
Chuck chuck@fsg.com

Wilson H. Bent, Jr. whb@hoh-2.ATT.COM
AT&T - Bell Laboratories HOH L-274A
Holmdel, NJ 07733 (908) 888-7129



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:54 CDT