PC Randomly Restarts

[awergh]

So I have a PC that lives in my cupboard which serves as a domain controller, file server, hyper-v server

except it seems to randomly restart during the course of the day for no apparent reason.

The specs are:

Gigabyte GA-G31M-S2L (Revision 1.0)

Intel Pentium Dual Core E6300 @2.8Ghz

2x2GB Mismatched Kingston DDR2 800

2x500GB Seagate Barracuda 7200.11

Nvidia Geforce 7300LE (when I used to use it, it ran really hot)

Corsair CX400 PSU

Its true the box doesn't have particularly good cooling but I'm not convinced that is the problem I'm sure it was working fine for a little while without problems, I know the PSU isn't the problem because I recently replaced the generic one in there. I also thought it could have been the dodgy CDROM drive because that wasn't connected for ages and I only started having a problem recently but that wasn't the problem either.

Perhaps I need something that monitors temperatures to a log file so I can see what the temperatures are when it restarts, because I can't accurately see the temperatures because Server 2008R2 takes forever to startup.

Or perhaps I should take it out of my cupboard and attach a monitor to watch what happens when it restarts.

Edited February 22, 2010 by awergh

[Zenskas]

Perhaps I need something that monitors temperatures to a log file so I can see what the temperatures are when it restarts, because I can't accurately see the temperatures because Server 2008R2 takes forever to startup.

Or perhaps I should take it out of my cupboard and attach a monitor to watch what happens when it restarts.

Good ideas there. Could be overheating, seems likely if you had trouble with your GPU before.

[awergh]

Well the GPU was hot but when I used it in my PC but in there it shouldn't really be in use much should it since I just remote in, the drivers for the video card aren't even installed. I can't actually see the GPU temperature. I think this would be easier if it didn't take what feels like 20 minutes to startup because it sits at Please wait for the Group Policy Client for quite a while.

Temperatures

CPU 28 °C

CPU #1 / Core #1 36 °C

CPU #1 / Core #2 30 °C

Seagate ST3500320AS 36 °C

Seagate ST3500320AS 36 °C

Cooling Fans

CPU 940 RPM

System 1513 RPM

[awergh]

So I ran RealTemp with logging although I don't really like the piece of software very much and I had a restart and the temperature going down of the cpu cores weren't significant at all so I'm still not sure what the problem is, also realtemp was doing something weird it reported CPU Load significantly higher then Processor Explorer.

I attached the log for you to see. I can't see anything useful from it, perhaps its something else thats getting too hot but the motherboard doesn't seem to have a temperature. Any suggestions would be nice since I'm going to goto the computer fair Saturday so would be good to work out wheter I just need a new 12cm fan or if I need something else.

RealTempLog.zip

[ripken204]

what do you mean by restart exactly?

does it BSOD and restart?

if so don't have it restart on BSOD.

also, do take it out of the cupboard, it has no right to be in there.

open up the side panel and even blow a house fan at it for a day to see if problems repeat.

[puntoMX]

Why not use the onboard Video only? You need DVI or so?

[bonestonne]

dried up/crappeh TIM!

take it apart outside (wear a coat) with a can (or 3 or 4) of compressed air.

get arctic silver (common and cheap, even radio shack has it), and reapply the TIM on the graphics card.

should work wonders for a plague of high temps.

[awergh]

TIM is that Thermal Paste?

Maybe its a BSOD I'd forgotten that causes automatic restart.

I suppose I could use onboard I don't use DVI I just put it in there once for something and didn't take it out.

[puntoMX]

should work wonders for a plague of high temps.

Just normal thermal paste would do fine, AS isn't that special these days. By the way, most 7300LE cards come without a fan so if it runs hot it's due to not enough airflow, but removing the pink gum between the cooler and GPU and applying new paste would be for sure an upgrade on the card.

I suppose I could use onboard I don't use DVI I just put it in there once for something and didn't take it out.

Take it out and I'm sure you have one problem less (besides having a card laying around).

[awergh]

Yeah I think taking it out would be the best thing just I want to see if it BSODs when it restarts or if it just restarts first except it didn't seem to want to happen yesterday even after I decided to run Prime95 for 4 hours, must of been too cool by then.

Having video cards lieing around doesn't seem a problem to me I have piles of expansion cards, pretty sure there is some EISA disk controllers there.

[awergh]

It appears that the problem is caused by a BSOD.

I have changed to using onboard video and briefly replaced the akasha 12cm fan with a coolermaster one but that actually made it worse.

It was running fine until today its been BSODing

Technical information:

*** STOP: 0x000000F4 (0x0000000000000003,OxFFFFFA80045B1360,0xFFFFFA80045B1640,0xFFFFF8000198B240)

I'm not sure if its overheating, I put a fan on one of the heatsinks on the motherboard to see if that would help.

The only thing that is a bit hot is the second heatsink near the bottom of the board.

It wasn't in the cupboard when this last BSOD happened but there case side was on.

[awergh]

Heres another one I got which is different to the previous one.

STOP: c0000135 The program can't start because %hs is missing from your computer

. Try reinstalling the program to fix this problem.

[puntoMX]

Heres another one I got which is different to the previous one.

STOP: c0000135 The program can't start because %hs is missing from your computer

. Try reinstalling the program to fix this problem.

Things that are fast to check are the data cables to the harddisk and memtestx86 to check that RAM. Did you check any of it?

[awergh]

I'd checked the SATA cables and they were fine I think well they were in properly I didn't try replacing them. The problem is it isn't a set pattern or because until power management recently turned the monitor off so I couldn't actually see the BSOD. I probably should memtestx86 I hadn't cause I didn't want to have to go buy more RAM. DDR2 isn't super cheap anymore. Ill run memtestx86 if it decideds to BSOD on me again today.

[cluberti]

Stop 0xF4 is critical object termination (lsass.exe, csrss.exe, or winlogon.exe - if you had the .dmp from that crash we could say for sure which it was), and a Stop 0x135 isn't a Microsoft Windows bugcheck, so that would mean some non-Microsoft driver called KeBugCheck to crash the box with a custom code.