The CPU fan was replaced, correcting the angry buzzing sound. The old fan seems to spin more easily in one direction than the other when I blow air into it. I suspect the bearing is bad, but not letting the blades bump the casing. I don't know if the oops message will recur without a bad fan, but even if it doesn't recur I'm guessing there is some incompatibility. Hints on investigating the incompatibility might be helpful; a first lock at Google wasn't very helpful. Thanks very much.
Stewart Strait
If a fan's blade(s) hit(s) the fan casing, then the fan will stop[1], at least momentarily. Does the fan ever stop completely (pause) when it makes the loudest noises - those noises which lead you to surmise blades contacting case?
Failing fan -> increased heat -> higher resistance -> lower power efficiency + faster break-down of components. Yes, a failing fan could ultimately cause oops messages. Don't believe me? Try unplugging the fan from an Athlon or P4 for a while and wait for the fun. Trust me, it won't be a long wait.
I wouldn't expect a failing CPU fan to affect a hard drive. So, old drive? Faulty power supply? Or...
The first culprit to suspect is memory. Bad bit in memory -> bad bit to/from CPU -> bad bit to/from hard drive (as in wrong bit right place, right bit, wrong place, wrong bit, wrong place) -> bad bit to/from CPU -> ...
Case study:
My wife's computer was experiencing drive problems - reports at boot up complaining about bad spots on the drive, lost nodes, etc. Fdisk offer, and was allowed, to fix the problems.
But things only got worse. As in failure to boot. And this on a less than one year-old drive.
Smack on head with a "Duh!".
I should have followed my own rule before mucking with the drive: I ran memtest86 which indeed found problems in both of the two DIMMs. Since the odds of both going bad at the same time were reasonably low, I tested both in another computer where they both passed [2].
Conclusion: bad motherboard, likely involving the MMU. Bad data was being written to the hard drive, probably to bad places, spawning cascading HD failure.
Solution: I replaced the motherboard and all was fine, including the drive [3].
Moral: A flaky hard drive is not necessarily due to a failing hard drive.
----
Not bad memory? Could be a bad power supply (see above effects). I don't remember you saying which fan is making all the racket. A bad PS fan could have similar affects on other components, starting with the PS itself and cascading from there.
If you really suspect the HD as being bad, which can also start a chain reaction similar to that caused by bad memory, I suggest you download and run the drive manufacturer's diagnostic software on the drive. Most of these programs are run from a bootable floppy (one lets you make a bootable CD) and so are not affected by the drive being tested.
Those are all the simple things to look at. After that, it gets a lot tougher with the chances of an economical fix (i.e. repair vs. replacement) becoming vanishingly small. After all, there are only so many components which can be replaced. Once all those are gone, you're left with a new computer.
And a hole in your wallet. In the case of a laptop, more like a crater in your ass.
--
[1] Or if it's a really strong nasty fan blade, explode the casing and probably your computer.
[2] I always run all tests with memtest86 as I have had the advanced tests, especially test 8, fail parts which passed all other tests repeatedly.
[3] Some data on the drive was still lost, due mostly to Fdisk's attempted repairs while running on a failing motherboard, compounding the actual problem.
--
Best Regards,
~DJA.--
KPLUG-List mailing list [email protected] http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
