Re: Unix machine re-boots at random

In article <xqx%f.1500$Lm5.1350@xxxxxxxxxxxxxxxxxxxxxxxxxx>,
FrankS <fshank@xxxxxxxxxxxxx> wrote:
Bill,

Thank you for responding. I forgot to mention that I also
re-seated all the boards already. I happen to have an extra
mother board that is identical and probably another processor
too. Looks like I'll be working on this thing over the weekend.
I didn't write the full PANIC error down, and know that I should
have.

I didn't say 'reseat all the boards' - at least that's not what I
inteded. I said to start with the minimum number of boards in the
machine, to see if there is any single board causing that problem.

I was just hoping that someone could pin this problem down based
on past experiences. I know that these types of problems are
often difficult to diagnose by any other means that replacing
parts or eliminating them. Unfortunately, I have people using the
server all day long.

I appreciate your response.

Is the machine available on weekends. I've found that on the
bigger SCO systems I used to maintain the only time we could
upgrade them was during a 3-day weekend, when everyone else was
enjoying the holiday, or watch the games [take your pick] on TV.

Sometimes your private life has to be sacrificed for your clients
if you wish to remain in business.

Bill
Frank


"Bill Vermillion" <bv@xxxxxxx> wrote in message news:Ixo9K8.K29@xxxxxxxxxx
In article <qau%f.69466$dW3.30457@xxxxxxxxxxxxxxxxxxxxxxxxxx>,
FrankS <fshank@xxxxxxxxxxxxx> wrote:
This machine is my current server that will soon be replaced, but, would
still like to fix for a backup. PII 400MHz. My problem with this machine
started last Thursday with a "PANIC" error message. From my past
experience
with this I replaced the 128 MB of memory and replaced it with two 256 MB
sticks. When I plugged the machine in there was nothing, no power. I got
a
new power supply, and it started up. Thought everything was fine.

This machine has been running 24X7 since 1999 and has never had so much
as a
hiccup. Now, it could run for 8 hours straight or as little as 30 min. It
just re-boots without any delay between boots (as if the reset button was
touched). Something may have gone wrong when I turned it off and let the
machine cool down for a day while getting the new memory.

Does anyone know where the most likely place for this problem is?

Mother Board?
Processor?
Raid Card?
Memory?
Power Supply? (I suppose I could have bought a brand new defective one)
Comtrol?

Thanks in advance for your help!

Well you have elminated most of the easy steps with memory and
power supply.

The first step in testing things like this is to pull all the cards
out and see hwo it runs, and then add them back one at a time.

I had at least two incidents in the past [most years ago] where the
serial board sub-system made the machine go bonkers.

One was giving errors that in the manual it said those could be
caused by hardware or software. No one had any more info. Pulling
the serial system and replacing them with a Specialix cured that
completely.

The other was a falied Stallion - that probably got hit by
lightning - and that was the only failed part.

If you can't make it run stable with all except the cards needed to
boot up out, then it's probably the mother board.

One thing you might try, and I've seen this only once, is to take
out the CPU and reseat it. That's pretty rare, but worht a shot.

If none of those do it, it's probably the motherboard.

What kind of message did you have besides PANIC. There is usually
more info at that point, often including register dumps. That
would be helpful for others to diagnose [but not me].

Bill

Bill


--
Bill Vermillion - bv @ wjv . com




--
Bill Vermillion - bv @ wjv . com
.

Quantcast