Thanks for any help. First off, I’m aware of the problem with AMD processors and XP SP3. This computer only has SP2.
AMD 2GHZ, 512 RAM, XP Home SP2.
It will fire up and run. At the beginning I hear the hard drive spin up madly, and when I look at the Task Manager, “winauclt.exe” is taking 99% of the CPU. This is Windows Update, and I think it’s screwed anyway. I’ll shut the process down and the hard drive will return to normal.
In 5-10 minutes or so, the machine will proceed to spontaneously reboot. At this point it will get to the XP Welcome screen, then go to black and reboot. Rinse and repeat over and over again.
If I turn it off using the power switch, wait a minute, then start it, it will run fine for the 5-10 minutes then do the reboot boogie again.
It even does it in Safe Mode.
I can’t keep it running long enough to do any meaningful scan, virus, or otherwise. CHKDSK does not work, it will ask me if I want to run it on reboot, I’ll say yes, and it will not run.
This is pointing to a hardware problem, I’m thinking. Power supply or hard drive.
Could be a virus, but assuming it’s not it sounds like possible binary damage on a failing drive. If this is the case save all critical data to a thumb-drive and and get a new hard drive.
One thing I do when diagnosing mysterious hardware failures is to remove and re-seat the RAM. You’d be surprised how often this cures weird hardware glitches.
During the 5 minutes of up time go to the system configuration utility and under "services "disable (uncheck) “background intelligent transfer Service” - This might/should prevent winauclt.exe from executing.
Stupid question, but are you sure it’s the hard drive you’re hearing, and not a fan? When the bearings go in a fan, it often manifests as a loud whirring noise at startup, that settles down after a short period. The fan may still be turning when you open the case, but perhaps at a much lower speed, causing overheating - which would cause exactly the kind of symptoms you’re describing.
It’s not necessarily unusual or sinister for a process to take 99% of the CPU - so that part may be coincidental.
I second the suggestion of reseating your RAM. While you have the case opened, blow out any dust that may have gathered over the past year. Do these simple things first before further diagnosis.
Good idea. If the problem persists, and if your RAM comprises more than one module, also try removing one module at a time (with the power off), then running the system - it could be failure in one memory module only.
Another thing you could try is boot up with a Linux live CD if you have access to another computer where you could download one. If it reboots while running a live CD, it’s definitely hardware.
It could indeed be a update that doesn’t play along with the rest of the system. As the PC is booting into Windows, press F8. You’ll only have a second or two to do this so be quick!
There are two ways of progressing from here:
(1) From the screen which then appears, just turn off the auto-reboot. Let Windows crash again and note the results. Debug from there as detailed in many previous threads. Or proceed as below.
(2) From the screen which then appears do a roll-back. Then let the PC boot normally. Once the PC has booted, go into the Properties of My Computer and turn off the auto-reboot. Then launch Windows Update manually, look at the updates available, and turn them off. Then look carefully at the rest of your computer to see if there’s anything amiss - drivers not up to date, disk full, disk corrupt, etc. When you’re happy, create a savepoint and enable and apply each update in turn, rebooting after each.
The computer never runs long enough to run any kind of diagnostic.
I turned auto-reboot off. When it shut down, it went to a black screen. No codes or indication of any kind why it shut off.
It still reboots on Safe Mode. And I’ve rolled back to several points already.
I’m sure it’s the hard drive spinning. Like I said, when I kill the winauclt.exe process, the heavy whirring spins down, then stops. I’ve heard enough hard drives in my time…
I’ll try re-seating the RAM and running SpeedFan if I can get it loaded that fast, thanks. I’m also planning on having my multimeter reading a power plug to see if there is any drastic drop in power from the supply.
Off question: why wouldn’t it run CHKDSK? Is there possibly a way I could set the flag manually, because the normal way of doing it is not letting it run on reboot.
I think I would suggest, at this point (after testing the RAM) booting to a recovery console and running the system file checker (SFC) - which should check and restore the integrity of all the key system files, but I’m floating that as a suggestion - what do the others advising in this thread think?
SFC sounds good, but I worry that there’s not enough time for it to run. Two more things to try:
Go into the BIOS and turn on the temperature warning.
Then let it sit in the BIOS and watch the temperature sensor.
Your earlier comment about opening the case will make little difference if the CPU fan isn’t working or the fins are clogged with dust. If you have a source of compressed air, give the heatsinks and fans a good squirt. Do the PSU as well, but be very careful. I used to use a breathing aparatus at the local fire brigade for cleaning their PCs.
