Random freezes

[akrobet]

I have bought my current configuration in February and since then it produces random freezes, sometimes once a week or once in two weeks.
Originally it was manually built around 2008 August from the following components:

Asus Maximus Extreme (X38)
Intel Core 2 Quad Q9450
NorthQ Siberian Tiger CPU Water Cooler
Zotac GTX280 AMP!
4GB Corsair XMS3 DHX DDR3 1066MHz
Corsair TX 750Watt Power Supply
1xVeloci Raptor 300GB, 1xSamsung 1TB, 1xWD Green Power 1TB, 2xSamsung 500GB in RAID
Creative X-Fi Xtreme Gamer
Lian-Li PC-V2100B PlusII
Pioneer DVD burner

I have used it with Vista x64 Ultimate up until three weeks ago when I switched to Windows 7 RC x64 and it already froze on win7 too.

I happens while doing random things, surfing, watching movies, I don't remember maybe while playing too.

On Vista after it froze the HDD started seeking but it was unresponsive and only hard reset works. And if there was sound that keeps turning on every ca.3 seconds for 0.25 seconds or so.
It haven't been overclocked. (since I have it at least...)

I don't really know where to look for the error since I can't reproduce it on purpose. Any suggestions how to tackle the problem?

 

[CeruLucifus]

FYI, on my previous system builds where I've had random freezing, it was usually a bad hard drive. I had one motherboard where the RAM could get loose, that also caused freezing. Anyway, how to test ...

Since you can't reproduce the freezes on purpose, your test method becomes surfing and watching movies for X hours where X is how long it takes to be sure the freezing would have happened already.

Since it happens with multiple OSes, the problem is most likely hardware. You need to isolate and remove components one by one. When you can run without freezing, the disconnected part is the one with the problem. You may be able to focus your efforts based on messages in Event Viewer, or maybe not.

You have a water cooling system. Are you overclocking the CPU or GPU? Start by going back to stock speeds.

Next check RAM by removing half your RAM and running with the other half. To test the first half, swap. (If you only have one RAM stick, go buy more, time to go to 8 GB!)

For your optical drive, just disconnect it (that's probably not it).

For your sound card, just remove it (I doubt that's it). Use motherboard sound if you have it, otherwise go silent long enough to test.

For your video ... if you have another video card (or onboard motherboard video), you can swap it out long enough to test. Otherwise, leave it for last.

For your hard drives, disconnect all but the OS drive. If the freezes are gone, it's one of the other drives (breathe a sigh of relief). Reconnect them one at a time (the pair for the RAID one of course) and see if the freezes come online with it.

If you do have freezes from the OS drive, you need to clone your OS partition either onto a new drive or onto one of the other drives (and set it up to boot). If you've disconnected the original OS drive but the freeze follows to the new OS partition, it's your OS partition (unlikely though).

If you have isolated it to a drive, back up its data, then use the manufacturer's diagnostics. These can fix some conditions by rewriting blocks of data, but not all.

If it's none of the above, run temperature monitors on your motherboard components and set them to log. If you get a freeze, check the log for that time.

I doubt it's a PSU problem.

 

[akrobet]

Thanks for your response.

Well, the situation is not that easy since I can't really take out the HDD's for a week or so until the problem reappeares because I use them on a daily basis, so I'd need to force the problem somehow or run a test if there is any?!

It's on stock speed nothing is overclocked. A screenshot about the voltages:



The ram and the video card is hard to test since I don't have any spare ones, and I don't know anyone to borrow from :S

I understand that your process would be the way to go, if I could reproduce the error, since I can't maybe the easiest solution is to wait until it becomes more frequent.

But the thread is still open for discussion, and any ideas are welcome

 

[CeruLucifus]

Yeah, I've been there, trying to troubleshoot with no spares.

For the RAM -- I'm looking at specs on the Asus web site right now -- that motherboard supports dual channel, so you probably have a pair of modules. You can take one out, it should revert to single channel, which is slightly slower but will run okay with 2 GB and let you troubleshoot (if it doesn't run, put the module back).

For the drives, you could still try running the manufacturer's diagnostics on each hard drive. Usually these are available as a download that can make a bootable CD. They'll say which ones can destroy data -- don't use those. Run them overnight on each drive, check out the results.

For the rest of the system, same idea -- get some spec testing software. I'm not really up on these but I think a lot of people here use Sandra from SiSoft. Set up some comprehensive tests and tell them to log and run continuously. Start that at night when you go to bed. In the morning, check the logs.

Keep banging on it until you find something that looks wrong.