[Info-vax] Another Alpha "TEMPERATURE_VECTOR" anomaly

[Rich Jordan]

On Aug 21, 9:40 pm, David Froble <da... at tsoft-inc.com> wrote:
> Rich Jordan wrote:
> > Just ran into another unusual behavior using the undocumented and
> > unsupported (yes I know) F$GETSYI("TEMPERATURE_VECTOR")  for
> > monitoring a couple of Alpha DS10 6/466 systems.  They are clustered,
> > and sit stacked in the same rack, getting the same environment,
> > airflow, etc.
>
> > Both systems are configured identically; one internal disk (same bay),
> > CDROM, blank plate, 1 each KZPEA, 2 each fiber channel cards, 1 each
> > GBe card, 768MB RAM.
>
> > One system consistently reports 43-45C.  The other reports 23-24C.
> > Both vary as room temp goes up/down, and I've asked for a test with a
> > fan blowing into the front panel (which on a DS10L quickly drops
> > reported temps 2-3 degrees) just to see if it moves down.
>
> > The cold unit reports the same temp range from the RMC console STATUS
> > command
>
> > For now we're going to fudge the thresholds by adding 20C to the cold
> > node's temps before checking against the warning, critical, and
> > shutdowns.
>
> > Anyone seen this before?
>
> I don't think I'm going to believe a powered on EV6 at 24C, 75F

I'd believe it if ambient was below freezing, but its not.  Room temp
is between 72 and 75F as reported.