[TriLUG] Hardware problem

Max TenEyck Woodbury max at mtew.isa-geek.net
Wed Apr 18 19:22:13 EDT 2012


On 04/18/2012 12:25 PM, Peter Neilson wrote:
> On Wed, 18 Apr 2012 11:43:06 -0400, Seva Adari <oddissyus at gmail.com> wrote:
>
>> I have a legacy Redhat linux server that abruptly halts with
>> absolutely no hint of any the problem in any of the logs. ...
>
> I'd reseat any boards that I could pry loose. If that failed, I might
> wave a hammer about, but not actually hit anything.

Waving a hammer about is only helpful when you have enough magical
talent.  If you have the right stuff, it works, otherwise it's a waste
of time.

Reseating the boards is usually helpful when you've been messing around
inside the box recently. Prognosis: might help, might not.

Since you replaced the power supply that is not likely to be a problem
unless you put in a lower capacity one or got a bad one.

On the bad capacitor front, a careful inspection around the base of
each of them might show corrosion.  It sounds like the board might have
come from the era when that was a significant problem.  If you find
that kind of problem, you might as well trash the mo-bo.  Some cards
also have electrolytic capacitors, so check them also.

Another simple thing to check is the power cord and wall socket.

A bit more difficult to diagnose is a 'brown power' sensitivity.  If it
does _not_ happen in the early morning (4-6 AM), then that is probably
not the problem.  If you think it is a likely cause, rent a power line
monitor.

Also check for CPU over-heating.  Blow the dust out of the CPU cooling
fins.

Max



More information about the TriLUG mailing list