As these list often have questions dealing with system/hardware problems causing
system crashes and/or the like I thought that I would relate a recent experience that
I had.
I have built several systems and consider myself to be quite knowledgeable in both
system arch and components.
My latest system I put together in late Dec, my Xmas present to myself.
The specs of said system are not relevant to my tale so I shall omit those.
Recently this system began to suffer from random and total lockups.
Because these lockups would occured in both linux and windows it was apparent that I
was afflicted with a hardware problem.
When booting cold into linux the system would run for roughly 5hrs without problem but
then the lockups would begin occurring, had it even happen at the GUI login screen.
Once the point was reached that the lockups began the system was useless unless I
shutdown and left it idle for several hours.
Even worse than the lockups was that data corruption was occurring.
MY first thought of a cause was either bad memory or hd.
I checked all the drive cables and plugs.
Ensured that components were securely seated.
I ran memtest86 and an IBM drive fitness test with both showing no errors.
I changed memory modules and slots, changed hd and cables, even ide controllers; and
after each change keep the system up and after each change after about 5hrs the
lockups would still start.
Having eliminated memory and hds from the list of possible causes the only ones I saw
remaining were either a faulty MOBO or CPU or possible graphics card.
Switching from a GeForce3 to a Radeon produced no better results.
Well I did not have another MOBO that I could use in this system but I
did have a CPU that I could take from a working system that could be used in this MOBO.
I removed the current CPU and heatsink and installed those from the other system.
Because on this MOBO the CPU fits just below the power supply connector my hand rested
against the connector while affixing the heatsink.
It was then that I noted the the cable did not feel secure in the socket.
Has anyone guessed the punchline yet?
That's right, but you don't get a prize, after ensuring that the power cable was
seated as securely as possible and restarting the system I have not suffered a lockup.
Added together, between changing out components, reinstalling a half dozen time and
then waiting for the system to either pass or fail the lockup tests I probably spent
50+ hrs over the course of a week fixing a
system to which I had access, which is 49hrs longer than it would have required to
completely dismantle the system and rebuild it, and then stumbled upon the solution
when I did only by chance.
I suppose that the morale to this tale is to not forget to check the innocuous little
details.
As I have shown it is not alway hardware failures that cause system failure and in
many instances finding and eliminating the cause of the problem is often trial and
error and yes, even Luck.
Charles
Want to buy your Pack or Services from MandrakeSoft?
Go to http://www.mandrakestore.com
