beeeeeeeeep booooop beeeeeep booooooop

Wow, what a series of events.

Last night storms came through my area, and the power went out. My UPS normally keeps my two desktops powered for 5-10 minutes, and the outage was longer than that. The UPS daemon which powers down my systems wasn't configured properly, so I had to manually shutdown both desktops, which I did. Power comes back on, I begin booting the desktops again, and my AthlonXP-powered Linux fileserver is up for about 5 minutes when suddenly I get the dreaded high-low-high-low PC beep indicative of a dead fan, borked CPU, or other "uh oh" condition.

When I open the case, the HSF (Vantec Aeroflow) is super hot, and the fan is not freely spinnable. I take the fan apart and there's a piece of copper that's somehow dislodged itself from the fan's motor and caught itself so that the fan can't spin. After dislodging the copper, the fan spins freely by hand, but the motor is apparently dead, because when powered on, the fan doesn't spin. Not a good scenario for our hero.

Luckily, or so I thought, I've got a spare Athlon HSF lying around, though not the best in the world, it should get me by for a couple days. As I'm swapping it in, my power goes out again, this time for a good 5 hours. It just came back on an hour or so ago, during which time I've swapped in the new HSF, to no avail. The CPU POSTs, and I can just about get to a GRUB boot loader screen before the nasty H-L-H-L beep comes back.

Right now, my *guess* is that the fan I'm using (stolen from my idle Duron 800 box) isn't beefy enough to cool an AthlonXP 1800+. I've never had an AMD chip fail before, so I don't know if it's possible the CPU's already been damaged or not. Can an AMD CPU be damaged from overheating and still POST, or no? I'm trying to figure out whether I need to order just a replacement HSF or also a new CPU.

Also, anyone able to recommend a good, reasonably priced replacement HSF that I can order from NewEgg?

Oh, and just for added intrigue, my RAID-5 array is running in degraded mode, so if one of my drives dies during this exercise, I lose 500GB of data. Wooooooo!

Well, I'm not sure if you'll consider this reasonably priced, but I love my Zalman HSF. It seems to do the job, and looks quite impressive through my case window

Yeah, $40+ for a HSF is a bit more than I was looking to spend. I don't care about looks for something that's inside an opaque case. I just want good cooling, and something that won't break in a little over a year like the Vantec Aeroflow did.

Anyone with an answer to my other question? If the CPU passes the POST, is it highly unlikely I did any damage to it?

I suppose you won't really know about CPU damage until you get the thing booted and running. That RAID issue sounds even more pressing. Keeping your CPU happy sounds like something that's worth a few extra bucks. If you buy some fancy Zalman cooling system and it turns out your CPU is hosed, then you can always get another. At least you've got Moore's Law pushing down the prices for you...

Quote:
That RAID issue sounds even more pressing.

Sorta. It's weird. It's a 3Ware Escalade 7500-4LP, by all accounts, a good RAID card. It's been happily working for at least a year now, and a few weeks back, one of the drives (drive 2) showed up in the BIOS status screen as "not in use." Rebuilding the array brought it back. Then, a week later, a different drive (drive 4) did the same thing, another rebuild, and now it's drive 4 again. I'm left wondering if it's really the drives having problems or if the card's just flaking out. Each time, rebuilding the array brought it back in several hours with no data loss.

I have yet to find out how to get SMART status or any kind of readout from the drives themselves, though. I'm also wondering if it's a heat issue, I don't have any drive coolers or anything in there and the 4 drives are right on top of each other. A lot of variables to figure out.

I guess the "right thing to do" would be buy the HSF *and* some new drives to cover all my bases. Grrr.