Hi, newb here is need of some help.  Built my server from an old AMD eMachines desktop.  4 data drives + parity. Using all 4 onboard SATA ports and one HD on a flat IDE cable. The parity drive is on one of the SATAs.  Whole shebang's been working great for about 6 months.  Suddenly the system is rebooting at random, and often won't start-up during these random reboots (reboots again before beeps or login prompt).  Totally unpredictable when it will shut down by itself, and if it will reboot completely if it does.

 

So far, I changed CMOS battery, re-seated both sticks of RAM, and installed a new totally overkill PSU.  Parity check = 0 errors.  Still random reboots.  My thoughts at this point is either the RAM went bad or my motherboard is dying.  Thanks in advance for all advice.  My syslog is attached.

 

syslog-2011-11-21.txt

First guess would have been the power supply, but you've changed that. 

 

It could be motherboard.  How are temperatures?  Is the CPU fan/heatsink clogged with dust (common in older machines)?  Any other electrical equipment nearby been added recently?  (thinking about mains spikes, etc).

 

What happens if you leave Memtest running for a couple of days?

Everything's the same at near the server as before.  I cleaned the entire unit a when I swapped out the PSU so all is clean and dust free.  Still, I don't know for sure what the temps are since I don't know how to check it running unraid.  Memtest con't complete as it won't run for more than a few minutes before random shut-down. 

You have some bad hardware. Remove all PCI cards and try the memtest. If you have 2 RAM sticks remove one at a time and try memtest. If that does not work try a new MB.

Insstead of a new MB, try setting the BIOS options for memory voltage, timing and clock speed. Many MB set the wrong if left on auto settings.

 

Incorrectly set values result in exactly your symptoms.

Thanks for the input, will try to isolate components further when I get a chance.  Cheap, fast, or good?  In this case I choose good and fast. My thoughts are to upgrade the whole thing if I have to replace anything at all.  So in the meantime, I ordered new mobo, cpu, ram, and another HD to use for cache.  Lets see if I can figure this out. 

BTW, one of my drives (WD Black 1TB) is typically at 40 degrees C.  Is this too high?  My other WD and Samsung drives are in the mid 30's.  Not sure what I can do about it though.