Replies to post #35065 on The Question and Answer Board (IHUB)

[Bob Zumbrunnen]

An overheated component *should* cause the whole OS to crash, not just one app. "Should" is an iffy word, though.

The problem that I read about applies to the 1650 servers. This one's a 2650.

I'm looking at the temperature probes right now and everything's nice and cool. The hottest items are the CPU's, which are running at 39 and 40 Celsius right now, and configured to log a warning if they hit 70.

[gotmilk]

Message In Reply To: eWeek reported [Dell] server
[shut down] when a component overheated...

Posted by: Bob Zumbrunnen
In reply to: Koikaze (who wrote above msg to BobZ)
... problem [to] 1650 servers. This one's a 2650.
... looking at the temperature probes right now
and everything's nice and cool...

Fred,
Bob's reply "everything's nice and cool" is inclusive
of everything monitored by the BIOS
or whetever does the weather reports on the server,
but I'am sure hundreds of other electrical components
are not monitored, like the "have a faulty inductor" mentioned
in the Dell article as the guilty part.
#msg-2210164
Some Dell PowerEdge servers die after smoking...
Defective motherboards... [PowerEdge 1650 servers]
... the problem occurs because certain motherboards
have a faulty inductor--a component designed to regulate
the voltage--that overheats, causing the component
to emit smoke and then shut down....