Project Status
Date: Fri, 1 Dec 1995 15:01:05 -0800 (PST)
From: Yonah Karp
Subject: 12/1 mead project update
Mead Project Report 12/01/95
Wells --> Horton
All services are off wells and onto horton. The exception
is li, which is running on three servers right now. We need
a home for a fourth li server.
Wells
From Doug:
Wells has been upgraded from a 370 -> 390. We were minus
a cable that connects the new single-ended-wide-fast
adapter connector to a
single-ended-narrow-fast-centronics style disk drive.
Ibm's bring one over in about an hour. It still needs
operating system upgrades.
Doug will upgrade the operating system to 3.2.5.1, and Tom R
will look at setting up batch on Wells so that it is only
accessible from Goodall.
RAID
Doug has been learning about DFS, which we will use for the
RAID box. About the RAID box, he says
I have found that when I run 3 or more reads on different
disks, it starts returning no-device-response. If backup
is running, the system/driver seems to recover fine,
though my test program detects the error and fails. The
A16 also encountered device timeout errors when it was
doing some initial testing (with more than one device).
I'm not surprised that we're also getting them on the
RS/6000's. I fax'ed a copy of the error log reports
yesterday to Macro, which is also fax'ing a copy to
Storage Computer Corp (the folks that made the raid
unit). We'll see what they say.
Yonah
Date: Fri, 3 Nov 1995 14:03:42 -0800 (PST)
From: Yonah Karp
Subject: Mead project update 11/3
Mead Project Report 11/03/95
RAID
Hugo[12] were added to the tulsa configuration files, and
bldfs was run to allow exports of the special filesystems
from the oz's (e.g. info and inst.maint, etc.), as well as
exports of jrr1, nineveh, and seuss1 filesystems.
The RAID unit is back up and running. Two power supplies, a
disk, and a SCSI adapter card (all of which had failed) were
replaced. Doug is working on the disk configuration.
Wells --> Horton
Early this week, staffdir was moved from wells to horton
(staffdir was also modified to write campus phone numbers
out to 7 digits and to include alternate phone number if it
exists). Moving staffdir is one of the steps required to
free wells for use as the student batch machine.
Horton replaced wells this week as the secondary load-based
cluster nameserver.
Misc.
The batch machines asimov, clarke, leguin, wells had DCE
clients installed. Also, tmr can now handle cd's from
mead1.
Yonah
Mead Project Report 10/27/95
This week, asimov and leguin (both IBM RS/6000-370's running
AIX 3.2.5) were replaced by IBM RS/6000-590's running
3.2.5.1. In addition, clarke, a RS/6000-370, was upgraded
to a 390, and the 3.2.5.1 software upgrade was applied. The
old asimov and leguin hardware became 390's, and their
destiny is to become fileservers for the RAID box. All of
the 370's also had dual attached fddi added as well as a
second 1.0 Gb internal disk. Address reassignment went
along with this -- front door addresses for all three boxes
are on the 140.142.12 subnet, and back door addresses are on
the 198.48.74 subnet.
Yonah
Date: Fri, 27 Oct 1995 14:54:03 -0700 (PDT)
From: Douglas Luft
Processor s/n 7012 26-53795 was asimov, and has now been upgraded from a
370 to a 390. A dual-attached fddi adapter and a second 1.0 Gb internal
disk have also been added. New UW Tag numbers are needed. This system
will be used as a mead file server, and called hugo1.
Asimov is now a 590 (s/n 7013 26-53914).
Processor s/n 7012 26-53792 was leguin, and has also been upgraded from
a 370 to a 390. A dual-attached fddi adapter and a second 1.0 Gb internal
disk have also been added. New UW Tag numbers are also needed. This
system will be used as a mead file server, and called hugo2.
Leguin is now a 590 (s/n 7013 26-31246).
Clarke has also been upgraded from a 370 to a 390 (s/n 7012 26-53793).
A dual-attached fddi adapter and a second 1.0 Gb internal disk have also
been added. New UW Tag numbers are needed.
A asimov disk (uw tag 1046431) and a leguin disk (uw tag 1046433) have
been transferred over to thurgood.
A spare asimov 2.9 Gb disk (uw tag 893439) has been put back into the
cage.
dbl
Back to AST Home Page