This is the summary of messages received about the exabyte tape drives and the warning messages "st: warning - tape may be wearing out or heads may need cleaning". First, it seems clearly a problem with Sun's st driver but not serious. In fact Sun addresses the problem on page 103 of the "SUNOS 4.1.1 Release and Install notes" and I quote them "these errors are fully recoverable and do not represent a failure of the system or threat to your data. The reason for this is descibed in the first two messages listed.
I apologize for the being late in the summary and I really thank everybody who responded.
SUMMARY
From: thompson@eds.com
My Artecon rep. told me this is normal. He says it is a
problem with the retry count. Apparently the retry count
at which you receive an error message is set artificially
low. He said a typical (allowable )error rate is 18 retries
per megabyte, so your retry count of 5171 is not bad if it
occurs after writing about 287 Mbytes (i.e., 5171/18).
What has happened is that the retry count at which it
flags an error has been set in the Sun st driver at 5000,
even though it is apparently perfectly valid to have that
many retries after 287 Mbytes. It is, in summary, a bogus
message.
You can apparently avoid this warning by inserting:
mt rew
mt eom
in between each of your dump files. I have not
tried this.
I would like to see a summary of any other responses
you receive.
Joel Thompson
thompson@eds.com
----------------------------------------------------------------------------
From: sjh@helicon.math.purdue.edu
We have the same problem. Fortunately George Goble ( the worlds
foremost expert on the exabyte drive(:-) works at Purdue. According
to George the problem is not in the tape or the drive, but in the
driver software. The sun driver begins reporting errors after a fixed
number of write retries (I think 5000). Normal is in the 3 to 4
percent range. It doesn't mean data is lost, just the the total
capacity of the tape is reduced somewhat. Since so much data can be
put on one of these tapes the fixed threshold might be reached every
time you use the drive. A driver which reports errors after they
reach some relative maximum (say 6%) would be more reasonable.
Steve.
Steve Holmes purdue!sjh
Systems Administrator sjh@math.purdue.edu
Dept. of Mathematics (317) 494-6055
Purdue University, W. Lafayette, Indiana 47907
---------------------------------------------------------------------------
From: "Ric Anderson" <ric@cs.arizona.edu>
ken:
The problem with 4/490 resident exabytes is an old one. We got
so digusted, that we moved ours to an external enclosure on the
back of a sparcstation.
it might be fixed in 4.1.1, but I'm not going to re-installl it
in the 4/490 just to find out :-)
Ric (ric@cs.arizona.edu <Ric Anderson>)
----------------------------------------------------------------------------
From: stern@sunne.East.Sun.COM (Hal Stern - Consultant)
try the on-line patch database; the patch you're looking for is in
there in the 4.1.1 SunOS section
--hal
------------------------------------------------------------------------------
From: Ray Ballisti <ray@ifh.ethz.ch>
Dear Ken, we have an Exabyte working with sunOS 4.1 correctly. However
we do not use st0. Instead we use the driver given by Exabyte (smt0).
I got it together with the unit. Sorry of non being able to help you more.
Ciao Ray
------------------------------------------------------------------
From: rick@wiau.medical-biophysics.manchester.ac.uk
I also got grossly fed up of this error.
I phoned sun.
Apparently it is only a bug in the 4.1_PSRA version of SunOS 4.1
(for 4/490's and the like)
After a not too long period of time my tape drive broke and I got a new
one, the problem went away.
I think what it is is that 4.0.3 does not report theses messages a well
so if you have a dogey tape drive and upgrade to 4.1 the messages apperar.
RICK DIPPER, Wolfson Image Analysis Unit, rick@uk.ac.man.mb.wiau
Department of Medical Bio-Physics, University of Manchester 061-275-5158
.............................Meemberrr the of Fumblee Ingers Culb..............
From: uunet!indetech!smw002!dev0601!steve (Steve Giuliano)
Ken...
Hope this mail gets to you.
I have a similiar Exabyte problem on a 490 with 4.1. Sun
sent me patch 100134-01. It didn't work for me, but I've
heard it's worked for other people.
------------------------------------------------------------
From: S_FITCH@vax.acs.open.ac.uk
Hi,
We had problems on our 490 boxes with 8mm decks.
After lots of hassle with Sun we got patch-id# 100134-01 that seems to have
done the trick. The problem description says, " Premature "st?:
warning tape may be wearing out...."", etc, etc.
Steve
----------------------------------------------------------------------
From: lbd@alux5.att.com
Subject: Exabyte and Sun 4.1
Hi, Ken --
The patch number you want for that Exabyte message
problem is 100134-01. I got it from the Sun BBS
patch database. It fixes bug id #1042822.
-------------------------------------------------------------------------
From: uunet!cadence!voder!nsc!berlioz.nsc.com!levine (David LeVine)
We used to get this problem too. It seems to be fixed in 4.1.1.
--------------------------------------------------------------------------
From: stan@nli.com (Stan Sawyer)
We had the problem, got the driver patch, reconfigured and still get
the message. No good news here. We have NOT upgraded our server to
4.1.1 though and I wonder if this may include a better patch.
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:11 CDT