Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 28 Mar 2001 00:02:04 +0200 (CEST)
From:      "Hartmann, O." <ohartman@klima.physik.uni-mainz.de>
To:        <freebsd-stable@freebsd.org>
Subject:   Overheated PIII in SMP system
Message-ID:  <Pine.BSF.4.33.0103280001420.1075-100000@klima.physik.uni-mainz.de>

next in thread | raw e-mail | index | archive | help

Dear Sirs.

I need a little bit help and some technical hints.

One of our SMP servers, a FreeBSD 4.-RC box, is not willing
to compile a world. It faults with SIG 11.

The problem is caused by a overheated Intel Pentium III/600MHz
with KATMAI core.

The main PCB is a ASUS P2B-D. Rear CPU slot is CPU 0. the inner
slot is for CPU 1. Today I swapped both CPUs and equipted both with
expensive fans and coolers. The case has a very, very good air
circulation, the rear is ventilated by two additional 80mm collers.

Secanrio: CPU 1, in the last configuration the inner one, has a
temperature of about 32 -38 degrees Celsius. The outer one, CPU 0
has 50 degress and up!! Yesterday I took measuremnets with swapped
CPUs, and figured out, that CPU 1, the inner one has 50 degrees and
the outer one, CPU 0 has only 30 to 40 degress.

I switched kernel from SMP to UP and ran the system while compiling
a world. Both CPUs were not over 35 degress Celsius!

Then I compiled a SMP kernel again and tryed to start make world.
CPU 0, the outer one, has now constantly about 50 degress Celsius.

Well, due the fact of changing fan and coller elements by better ones
one fan is very close to the next SECC-2 case of CPU 1 (the inner one).
This handicaped fan is for CPU 0, the overheated one. But yesterday
exactly this CPU has the place of the "cooler" CPU, the inner one,
so I think not that this could be a real heatsink problem.

I think the problem has to be targeted either by the mainboard (maybe
some kind of weakness in voltage regulation? But why only in SMP
mode of the kernel and not in UP mode?). Or the CPU has some faults.

I switched again to UP kernel to see, whether temeprature is decreasing
or not.

this is the actual output of "healthd -d -I" and "heat" as it reflects the
configuration at this moment:


HEAT:
System Temperature       69F (21.0C)
CPU1 Temperature        125F (52.0C)
CPU2 Temperature         95F (35.5C)
FAN2                    314 RPMs (functional)
FAN3                    312 RPMs (functional)


HEALTHD:

************************
* Hardware Information *
************************
Asus: AS97127F
************************

Temp.= 21.0, 52.0, 35.5; Rot.=    0, 5113, 4821
Vcore = 2.08, 2.05; Volt. = 3.22, 4.92, 12.04, -11.77, -5.11
Temp.= 21.0, 52.0, 35.5; Rot.=    0, 5357, 4963
Vcore = 2.06, 2.05; Volt. = 3.22, 4.89, 12.04, -11.77, -5.11
Temp.= 21.0, 52.0, 35.5; Rot.=    0, 5113, 4963
Vcore = 2.06, 2.05; Volt. = 3.22, 4.89, 11.98, -11.77, -5.11


yesterday these values were vise-versa with exchanged CPUs ...


Please tell me your opinion: should I exchange mainboard first or the suspected CPU?

Thanks,

oliver

--
MfG
O. Hartmann

ohartman@klima.physik.uni-mainz.de
----------------------------------------------------------------
IT-Administration des Institut fuer Physik der Atmosphaere (IPA)
----------------------------------------------------------------
Johannes Gutenberg Universitaet Mainz
Becherweg 21
55099 Mainz

Tel: +496131/3924662 (Maschinensaal)
Tel: +496131/3924144
FAX: +496131/3923532


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.33.0103280001420.1075-100000>