[2306] in SIPB-AFS-requests
Craig Koller: RE: Seagate disk problem
daemon@ATHENA.MIT.EDU (Greg Hudson)
Tue Feb 20 16:40:01 1996
To: sipb-afsreq@MIT.EDU
Date: Tue, 20 Feb 96 16:38:49 EST
From: Greg Hudson <ghudson@MIT.EDU>
I guess I'll go ahead and install the 4GB Seagate drive we have on
reynelda. Swapping the drive doesn't seem likely to help, since this
is the third disk we've have problems with. Does anyone object?
------- Forwarded Message
Received: from PACIFIC-CARRIER-ANNEX.MIT.EDU by po6.MIT.EDU (5.61/4.7) id AA24035; Tue, 20 Feb 96 16:17:37 EST
Received: from uucp-1.csn.net by MIT.EDU with SMTP
id AA15325; Tue, 20 Feb 96 16:15:40 EST
Received: from r2.UUCP (uucp@localhost) by uucp-1.csn.net (8.6.12/8.6.12) with UUCP id OAA00917 for ghudson@MIT.EDU; Tue, 20 Feb 1996 14:15:57 -0700
Received: from r2mail (r2mail.r2.com) by r2.com (4.1/SMI-4.1)
id AA13683; Tue, 20 Feb 96 10:28:48 MST
Received: by r2mail with Microsoft Mail
id <312A0E99@r2mail>; Tue, 20 Feb 96 10:10:33 PST
From: Craig Koller <craig@r2mail.r2.com>
To: ghudson <ghudson@MIT.EDU>
Subject: RE: Seagate disk problem
Date: Tue, 20 Feb 96 10:17:00 PST
Message-Id: <312A0E99@r2mail>
Encoding: 55 TEXT
X-Mailer: Microsoft Mail V3.0
Sorry for the delay, I sent this to our engineering dept. and they said they
have not seen this before. We don't support the DEC platform, but I would go
ahead and use the drive on your SUN system. If it gives you any problems
just give us a call and we will swap out the drive. If you have any
questions please call or E-Mail me. Thanks.
Craig Koller
303-784-7020
Ext.3705
E-Mail:
craig@r2.com
----------
From: ghudson
To: craig
Subject: Seagate disk problem
Date: Friday, February 09, 1996 2:37PM
This is in reference to the problem we're having at MIT with a Seagate
4GB disk slowing down after reaching a certain capacity.
The history of this problem is as follows: about six months ago, we
bought a Seagate 4GB disk and installed it on a Decstation AFS server.
After the disk filled beyond a certain point (about half of the first
2GB partition on the disk), we found that the disk became so slow that
some AFS operations on the disk (e.g. moving a volume onto the disk)
would time out, and file accesses on that disk were very slow.
We had the disk replaced, twice, and each time we've had the same
problem. The third time, after verifying that the problem still
occurred after filling the disk partway through the first 2GB
partition, we moved the disk to a different Decstation AFS server.
This seemed to help, but recently we found that we still had the same
problem after filling the second 2GB partition about halfway. That
is, the threshold seems to have moved, but the problem remains.
Replacing or reordering the SCSI cables in the chain (using shorter
cables), using active terminators, and other changes did not cause the
problem to go away, although reordering the SCSI chain did seem to
cause the threshold to go up slightly.
We haven't had this problem on any other of our disks, including
several other kinds of 4GB disks (a 4GB Micropolis disk from R^2, a
5400rpm Seagate 4GB disk recently bought from ADS) as well as 2GB and
1GB disks. Our AFS servers have been running Ultrix 4.2A throughout
this process.
We've taken the disk out of production. If we put this disk or a
replacement back into production, it will probably be on a Sun
Sparcstation 5 running Solaris 2.3. I would rather not experiment by
simply putting the current disk into production on the new server.
Is this problem familiar? It doesn't seem like it could be an
enclosure problem.
------- End of Forwarded Message
