[kwlug-disc] Troubleshooting and Fixing a PC locally?

[unsolicited]

John Johnson wrote, On 12/03/2010 2:24 AM:
> At 17:28 2010-12-02, Khalid wrote:
>> One of the test servers I use has died ....
> 
> Suggestions have included:
> Motherboard. Capacitors. Power Supply. etc.
> 
> I am reminded of the following:
> 
> " ... as we know, there are known knowns; there are things we know we 
> know. We also know there are known unknowns; that is to say we know 
> there are some things we do not know. But there are also unknown 
> unknowns -- the ones we don't know we don't know."

You know?

What you have not included in this risk analysis regurgitation is that
the likelihood and scope of the risk is relevant. (And determines the
extent of the contingency budget.) It may rain (for an inside
activity), vs. a meteor strike, vs. the cost of a keyboard.

Risks expression, without assessment, qualification, quantification, 
contingency plan, and fall back plan, are just gas.

>> So I'd rather hire a pro. And I got some pointers from people on this 
>> list.

> Given the "unknown unknowns" hiring a pro can be prudent.

Probably not. But also probably depends upon inclination, and most
especially opportunity cost. The latter being most important to
Khalid, not just from a $ perspective, but from a customer
satisfaction perspective.

In these cases:
- there is some number of hours and activity required before one
reaches the decision to take it in. During this time, things like
memory tests can be run. (Fire and forget - get on with other things
in mean time.) [Assuming he can. On another machine, if not this one - 
problematic, as he has pointed out.]
- e.g. Disconnecting USB keyboard to find the problem fixed saves
self-recrimination later.
- the hardware is never going to be trusted again, anyways.
- for the cost of 2 hours of service one can put in a new motherboard.
- any such system, beyond pick up and put down of disk and
reconfiguring video, is going to consume WAY more hours than anything
else. (Reconfiguring, etc.) In *addition* to any repair diagnostic, 
resolution, and remediation, time.

Note - we have been distinguishing, thus far, cost of time to
diagnose, not - slap another motherboard in that puppy and we'll be
done with it. (You'd want to run the memory tests, regardless. The
success of which governs further steps, but most of all, whether you
use and trust it in that replacement motherboard or not.)

Hardware is a PITA, and in these situations there's just no getting
away from the requirement for personal time and attention - if you
want to have confidence in the hardware after the fact.

Um ... wouldn't computer recycling be perfect for this effort?

IIRC, the volunteers there are looking for challenges and experience,
under the guidance of certain other well known and experienced ... people.

Not suggesting you get them to do it for free (think Timmy's cards and
CR donation), nor that you wouldn't have to supply any necessary
hardware. (I don't expect, for example, they have too many quad core
MBs kicking around to test against.)