This is a long shot, but I thought I would post here just in case some one had run into this before.
I bought an Intel Core2 Extreme QX6700 based system when it was first being released through a reputable system builder. The system used the Adaptec 44300 Serial Attached SCSI Controller and had 2 Seagate Cheetah 15K.4 146GB 15,000 RPM SAS drives in a RAID 0. After minimal usage, the system started having problems, and then it went from bad to worse. On booting, the system would constantly spin up and down the drives at the SCSI bios. It quickly got worse to where it wouldn't boot past that...just constantly spinning up and down. So, I shipped it back to them. They replaced some components, tested it, everything seemed okay, and sent it back to me. Basically the same thing happened again. I sent it back. We ended up changing to Xeons instead, since they said the platform was built more for these types of setups. This time, it used a LSI Logic 3041X-R Serial Attached SCSI Controller. They said the drives were changed out and everything. So, it had a different motherboard, SCSI controller, and drives. Yet, the same thing basically happened. The only difference was that this time it started with BSoDs, but quickly developed into the same spin up/down problem. Whereas before the problem seemed to start after doing video encoding, this time it started with simply copying files onto the drives. I should mention that I have an external SCSI array connected to the system via a LSI Logic 21320 Ultra 320 SCSI Controller. It seemed to work fine throughout all of this when I could get into Windows, and that is what I was copying files from. The problem occurred with or without it connected once it developed. Does anyone have any thoughts on what might be behind this? It just seems too weird to me that the problem would repeat after so much being changed with the switch to the Xeon setup, especially. Thanks much.
Since you changed motherboard, SCSI controller, and drives and still have the problem, it points squarely to either:
1. Cooling solution (likely)
2. Power supply (less likely)
15K Cheetas put out a LOT of heat. If you don't have fans blowing directly over these drives, they'll overheat. These drives are meant to be used in servers where they have direct airflow, not desktops where the typical hard drive cooling solution is just ambient air in the case.
This sounds like what happened, given that the problem slowly got worse. Overheated drives typically degrade like that (quickly, if they're well overheated).
And, Enterprise servers sit in strongly air-conditioned rooms (62-65 degrees) equipped with two power supplies and four fans. Overheating is never an issue. There is rarely, if ever, a slowly degrading hard drive in that scenario. They just crash when their time is up. Ah, I harken back to my days at Price-Waterhouse looking for the yellow lights.
Joes' take is right on. You need to seriously improve your cooling. Throw a dedicated A/C in that room and add all the fans you can fit into that box. If you can, keep the cover open. Cheetahs are like Ford trucks. Be kind.
Powering up and down could be heat or power, after seeing how typical "professional builders" build their systems I'd favor it being a power issue, even if heat seems more likely in "most systems" (most being Dell, HP, etc).
Dell and HP are way over the top of typical "pro" builders when it comes to power supplies.
Very good to hear from Crash. Point well taken from a great technician.
We didn't ask mysti what power supply was in that box. He did say "reputable" builder though - who you wouldn't think is putting chancy power in there. Of all the things to go cheap on, with core2 & scsi.
PS - the cooling for the Xeon configuration was 2x Swiftech MCX-V Pro Cooler For Socket-771. The Extreme configuration had the Coolermaster Aquagate Mini R120 Water Cooling System. Bother configurations also had a Noise Reduction and Cooling Kit (Stage3 120mm).
I would go with the power supply being the issue, when I added more drives to my raid array it did the exact same thing and since I added a new power supply I haven't had any problems. I can't see you having problems right away on boot from it being a cooling issue as the drives wouldn't have time to even heat up but while you're at it make sure your cooling is up to par.
If you let the computer sit for a long time off, when it booted up next, it sometimes go further. For example, after the problem happened, if I tried to boot again right away, the same spin-up/down problem would repeat every time. However, at first if I let it sit for a long time, the next time it booted, it would sometimes go further (and usually BSoD later and then repeat the same spin-up/down on boot). The more attempts, the worse it seemed to get and harder to get any further it became. I knew it had to go back because I leave the system on almost all the time, but I did want to do some testing so that I could give them feedback and information on the issues.
I've never had hard drives act like this over heat...back in the day when I wasn't smart enough to know scsi drives needed to be actively cooled...I had 4 of them(3 10k's and a 7200) sandwiched together and I got a nasty burn from touching the top one. Even being that hot, they still ran flawlessly. However, usually when power is an issue the drives won't be able to spin up fully but from the sound of it they are for you, so it's hard to say. Maybe it's a mix of both lack of power and overheating? Throw a big fan over the drives and if that doesn't help I'd say invest in a bigger power supply.
I think with everything you just described 750w just isn't gonna cut it, but if you could give us a list of everything in the system I could give you a more definitive answer. Also if you could describe the cooling setup
750w should be plenty ample for two hd's, gpu, and a dvd. What's powering the drives in the external array?
With the angst you've gone through - and the money spent, my recommendation is to get a new Silverstone (just to put that puppy to bed). It's an easy swap, and, if that isn't the problem - you have a quality spare.
If increasing the air flow to cut down on heat, and changing the psu doesn't solve the problem - then you have a defective motherboard. One with a circuit somewhere that's opening & closing - impossible to troubleshoot. It's the least likely candidate - but it happens.