On Fri, Apr 20, 2007 at 11:24:18PM -0400, Patrick Cummings wrote: > > So I finally had some time and tested the memory with memtest. I got 5 passes without errors, so I guess it was not that. It's also ECC memory. memtest won't catch a lot of problems. I've had memory test 100 passes just fine, but still end up being the problem. > However now the problem is getting worst. Sometimes it crashes ans I lose access to /var. It crashes every 30 minutes and I must restart the computer. maybe you've got a temperature problem? Have you been monitoring the temps? do you have a stuck fan? > Is there any way to get out of this without a clean install? I think that's what I will do. That will be very complex since there is so much on this computer. I guess I should have continued with sarge. > yes: find out what your hardware problem is. Start by opening up the box and cleaning everything, especially the fans. Then setup up lmsensors and monitor your voltages and temps. If that doesn't give you results then start swapping memory stick. Go through it piece by piece until you figure it out but I'm betting, based on the frequency, that its either temperature or power supply. A
Attachment:
signature.asc
Description: Digital signature