Re: SGI computer crash- resolved ...O2 problems

Michael Strain (strain@mango.uoregon.edu)
Thu, 15 Apr 1999 10:48:12 -0700 (PDT)

We have only one SGI-O2 (R5K) system in our lab and have had no problems
with it. However, I was just talking with the SGI administrator in the UO
Computer Science Dept. They have had numerous problems with their O2's
which were purchased soon after the O2 was introduced to the market.
Many of the problems have been similar to those mentioned in this AMMRL
thread. In particular, systems fail to power up, as evidenced by black
screens and dark-emitting diodes (DED's).
An example, from just this morning:

===================================
Subject: Re: osmium, still down...

> osmium is quite dead. I'm opening a case with SGI.

I opened a case and the first thing SGI told me to do was swap parts
with a working O2. Of course, both Lauradel and I thought it would be a
waste of time...

Well, here is what I did.

osmium = bad O2
iridium = good O2

1. put iridium Power Supply in osmium - osmium still wouldn't boot.

2. put iridium System Board in osmium - osmium still wouldn't boot.

3. put osmium monitor on iridium, iridium came up normal.

4. put osmium Power Supply in iridium, iridium came up normal.

5. put osmium System Board in iridium and iridium powered up as soon as
I plugged it in without pressing power button.

6. put osmium System Board back in osmium and osmium came up normal.

So, it appears that iridium jumpstarted (no, not roboinsted) osmium and
now both are good O2s.

Oh vacation is only a few days away...
=================================================

--Mike

-----------------------------------------------------------------------
Michael Strain strain@mango.uoregon.edu
Institute of Molecular Biology desk/voice-mail: 541-346-4605
& Department of Chemistry FAX: 541-346-5891
University of Oregon, Eugene, OR 97403
-----------------------------------------------------------------------