[rescue] Re: Old sparcstations & overheating

Janet L. Campbell janet at foonly.com
Sat Feb 19 03:45:24 CST 2005

On Thu, 17 Feb 2005, John Redmond wrote:

> Given the recent conversations about old sparcs, particularly
> overheating...  as per a previous post (couple of months back)
> to this list I am (still) planning to put a couple of old sparcs in a
> co-lo. One would be ss10 with dual sm81s the other ss20 with dual HS
> 133s. Both would have 2 smallish seagate  4.2 gig drives ST34371W in the
> SS20, not sure yet about the SS10. Is it just the bigger, newer drives
> that pose the overheating problems? I haven't been running either box
> too much currently (in fact the SS10 cpus are an upgrade I expect to
> receive in a couple of days) but they haven't seemed too bad when I have
> run them.

Since the drives and the mainboard are located in different air paths in
the SS20, hot drives are unlikely to kill the processors.  However, with
hot drives, the failure rate of the upper drive seems to be noticably
higher.  Make sure you have the hard drive fan FCO installed (looks like a
mini fan between the CDROM/floppy and hard drive bays). You're running
SM81s, which weren't technically qualified in the SS20, but I've done it
too and they work fine in reasonable weather.

Removing the side panels is easy and gives you a little extra airflow, 
late-model SS20s (and SS4/SS5s) had side panels with a more open flow 
design.  I seem to remember them being mandated with certain high-heat 
cards, like the TZX.

SS10 drive cooling isn't great.  I've burned up several early 7200RPM 
drives.  I'd stick to one drive and watch it, unless you're using really 
cool-running drives.

The biggest head problems don't come from the largest capacity drives, but
rather from those with the most cutting technology at the time.  The
early-generation Barracudas (pioneering 7200RPM drive technology) put out
a lot of heat and would burn up without proper airflow.


More information about the rescue mailing list