Server crashing...comskip?

[nettech_gt]

I have been having an issue with my server shutting off every night (between 11:10-11:50) for the past two weeks. My server is running Win 7 x64 on an AMD Athlon II quad core CPU (3.1 Ghz) and 4 GB of RAM. Three HDPVRs (my only tuners) Sage 7.19.256. and commercial detector 1.00.
The event viewer (system) shows the same four events every-time it happens. All telling me the system "uncleanly shut down". I have actually witness it happening. Comskip is processing three files and then with no warning the server shuts off. There is nothing that is scheduled to run at that time other then compskip processes recordings. Commercial detector is configured to process 3 recordings at the same time. "Play nice" is enabled and live TV processing is turned off. Also the are no BSOD dump files.
I have tried comskip (donators) 81_023 and 81_052 both with the same results. However if I disable Commercial detector and all relating plugins, the server stays over night. I have scanned my server with Windows security essentials and Defender and nothing was found. I have scanned all my HDs with a short self test and they have all passed. I have double checked all my power settings and its configured to never sleep or hibernate unless running on my UPS battery. I have attached my log (Trace) can some take a look? Right now I'm at a loss...this is very frustrating!! THANX!

[craigap]

Could the processor be overheating and causing a shutdown?

[nettech_gt]

Possible but not likely. The CPU is not overclocked. Both sides of the case are removed, so there is great air flow through the entire case. Also I would expect to see BSOD dump files for something like that, and I don't.

[Denali]

I just fixed a problem with my system that would randomly shutdown without warning one or more times a week. It was caused by bad electrolytic capacitors on the motherboard and video card. Capacitors were bulged out. Might be worth looking into.

[Fuzzy]

Is this the only time you are processing 3 comskip recordings at a time? I've never seen a bit of software be able to take down a win7 system like that - it is almost certainly a hardware issue - either a power supply, MB regulator, or a cpu/north bridge heat issue.

[nettech_gt]

Originally Posted by Denali

I just fixed a problem with my system that would randomly shutdown without warning one or more times a week. It was caused by bad electrolytic capacitors on the motherboard and video card. Capacitors were bulged out. Might be worth looking into.

Originally Posted by Fuzzy

Is this the only time you are processing 3 comskip recordings at a time? I've never seen a bit of software be able to take down a win7 system like that - it is almost certainly a hardware issue - either a power supply, MB regulator, or a cpu/north bridge heat issue.

@Denali

I don't have any popped or bulging caps on my MB. In my experience popped or bulging caps cause more random unpredictable issues. My issue occurs at a specific range of time every night consistently.

@Fuzzy

This was my initial thought as well. But short of running a full diagnostic of every piece of hardware I have not been able to pin point the issue. Also usually there are BSOD when hardware crashes and there are no dump files.

A couple days ago I decided to change my list of restricted times (from everything restricted except 11pm-5am) to everything restricted except 7am-4pm) At 11:15 pm comskip is not running. I check the commercial detector queque and there are recordings in the queue. The the next morning @6am my server was still running and all recordings in the queque had already been processed. This has continued for at least three days. This is good because my recordings are getting processed and my server isn't crashing but I don't know when the recordings are actually being processed. I want to make sure its not during my daily backup.

Can someone check my logs? THANX!

[OneOfMany]

Sure sounds like CPU overheating. Maybe repaste the CPU/heatsinnk. Also having the sides of can often slow the movement of air in many cases causing them to run hotter.

Heavy disk activity could tip a marginal power supply. Other symptoms include not always firing up the disk during a cold boot.

Grant

[Fuzzy]

Originally Posted by nettech_gt

Also usually there are BSOD when hardware crashes and there are no dump files.

This is actually not very true.
Things like a bad memory chip, or bad peripheral, will most often result in a BSOD. It actually takes a still functioning processor to generate a BSOD, and sump memory to disk. However, a problem in the power supply, or mainboard voltage regulator, will usually result in a reeset of the CPU itself midstream, which sometimes will cause the monitoring system on the motherboard to shut down the power supply, or at best, force a reboot. Neither action will generate a BSOD, or any log entries.

[nettech_gt]

Originally Posted by OneOfMany

Sure sounds like CPU overheating. Maybe repaste the CPU/heatsinnk. Also having the sides of can often slow the movement of air in many cases causing them to run hotter.

Heavy disk activity could tip a marginal power supply. Other symptoms include not always firing up the disk during a cold boot.

Grant

Originally Posted by Fuzzy

This is actually not very true.
Things like a bad memory chip, or bad peripheral, will most often result in a BSOD. It actually takes a still functioning processor to generate a BSOD, and sump memory to disk. However, a problem in the power supply, or mainboard voltage regulator, will usually result in a reeset of the CPU itself midstream, which sometimes will cause the monitoring system on the motherboard to shut down the power supply, or at best, force a reboot. Neither action will generate a BSOD, or any log entries.

I definitely think if could be a hardware issue.This certainly fits the best. In fact my first thought was it was the power supply. However when I moved the restricted times for commercial detector so my recordings would not process during the time I was consistently having the sever crash.(11pm-11:50pm) The issue "seems" to have gone away. The server is not crashing and my recordings are being processed. However as far I can tell comskip is not obeying the new restriction times. So if the issue is the power supply, motherboard or CPU is getting overloaded/overheated doesn't it make sense that the same amount of load/heat but at a different time would produce the same results? Yet this is not what is happening...

I will run a diagnostic on all the hardware. Just so I verify or rule it out. In the mean can some please check my logs (a couple posts up) and see if they can tell when my recordings got processed? THANX!