Render Farm Crashing

[Garret Diduck]

I just recently licensed my Farm. Ever since adding a license to my main system, the Farm will crash every so often on that system. About half the time, the Farm will give an error message stating that it crashed but it will still show as running in the Processes even though I don't believe it is actually running.

The other half of the time it will BSOD and reboot. All of the other systems on the Farm run as expected with the same render. The system will render just fine if the Farm is not used (obviously not ideal).

All versions of the Farm are 290a and are on Win7 x64 Pro or WinXP x64 Pro. The system it crashes on is Win7 X64 Pro and has 8gb RAM. The last time it crashed, only 3.8gb of memory were being used.

The farm ran perfectly on this system prior to adding the license.

Any ideas on what might be happening? 

[Roy Hirshkowitz]

Nope.  It's very unlikely to be related to the license-- coincidence, I would think.  If you've got a model that crashes consistently I can take a look at it.  Otherwise this can be a very difficult problem to track down.  On 64 bit systems memory is not an issue.  It's possible the system is overheating or has some hardware problem.  I hate to send you down that route, but drivers can sometimes crash (you could check the display driver to see if it's up to date.)

It is, of course, possible that there's a real bug in the farm.  I guess I would expect to see other machines crashing if that were the case, though.  Like I said, this can be very difficult to debug.

[Garret Diduck]

It is not one particular model that will crash, so I guess I'm left to check out the drivers/hardware. I'll keep you posted on what, if anything, I find.

[Roy Hirshkowitz]

I should add that when I say memory is not an issue, what I mean is that you can't run out of address space on 64 bit systems like you can on 32 bit systems (running out of address space will cause a crash on 32 bit systems.)  However, an intermittent memory hardware problem could certainly blue screen the thing.

[Garret Diduck]

I changed video cards and drivers this morning as well as a bigger PSU. So far I haven't experienced any crashes and I'm well past the time when it would crash.

[Roy Hirshkowitz]

Could be heat-- the new PSU may just have a better fan.  Make sure the air coming out of the thing doesn't seem ridiculous.

Could be the video, of course.

[Garret Diduck]

That's possible for sure. I do have a case full of fans though, so the ambient should be good. The cpu temp is 60°C @ 100% after 4+ hrs.

I'm hoping it's fixed after the mini-upgrade.