Hi,
I hope this isn';t an out of scope posting. I have an ABIT BP6 dual
Celeron system running RH6.1. I started off with one bank of 64MB memory
(no ECC) from Crucial and had no problem booting up.
A couple of months later I added a 2nd 64MB Crucial no-ECC DIMM. I then
began to SOMETIMES hear a short zunng sound as (I think) the system
unsuccessfully tries to access the hard drive (a 10MB WD) when I push in
the power button.
I tried reseating, rotating, the DIMMS, etc. Ater a while I thought I
noticed a correlation between power-up problem and temperature, having seen
this possibility mentioned on another list. So I created a log and noticed
that, whenever the room temperature fell below a certain level (my
thermometer is crappy so the actual temp probably doesn't matter), power-up
is a problem. If I were to do a non-parametric test, I'm fairly sure I
would conclude that this is a systematic effect.
Well, to put an at least temporary end to this saga, I ran memtest86 over
the last several days. In particular, I did one pass of the extended suite
and there was no problem. So, I think I've pretty much ruled out memory
itself as the culprit (BTW, I've had almost no system hangs over the past
4/5 months).
Here are the tentative conclusions I've reached. I would appreciate your
comments on which one(s) are most plausible.
(1) This is not a Linux problem (i.e., smp-power-off in apm.c).
(2) This is not an SMP problem.
(3) This is not a memory problem.
(4) This might be a problem due to assembly. When I assembled the system,
I missed putting washers on all the risers for the mobo. When it gets cold,
I hypothesize, there is contraction, leading to the power-up problem. I saw
this invoked as a possible explanation for BP6 random hangs.
(5) This might be a power-supply problem: too much fluctuation.
If, I hate to say it, other more likely possibilities come to mind, please
advise. Or, maybe this belongs on a different list (I seem to have argued
that it does).
Roy Wilson
[EMAIL PROTECTED]
-
Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/dmentre/smp-howto/
To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
