Call Traces/Stops Responding

[Dreded]

Anyone have any pointers how to troubleshoot this sort of issue? Started after changing the motherboard/CPU... the ram was pulled from the previously working system but after having these issues I did a MemTest anyhow and ran the memtest for 6 passes(6 full passes not tests) without issue

 

The system might run 3 days without issue or might run 3 minutes(or less)

I never had these issue with my previous Motherboard/CPU but I also only had run unraid for about 3 days on it before changing the Motherboard/CPU before that I ran FreeNAS for 3yrs on it.

 

Is the most likely issue a bad motherboard? I checked the Motherboard website and there is no new BIOS's

 

Output from the log file: https://pastebin.com/fytzfLCm

[Dreded]

Just updated to 6.4.0-rc13 no fix unfortunately same issue

[Dreded]

Still trying to work this out... seems if I only read from HD's with minimal writing it runs fine.. as soon as I start a docker with either SabNZBD or Deluge the system reboots very shortly after.

 

Watching movies in plex(even in UHD with 30gb files) works just fine played about 4 movies this way and 3-4 tv shows even jumping around in the 4k works just fine

 

reading and writing files to and from the SMB shares works fine too.. I moved 300GB off and then back on without issue

 

so seems to be some sort of issue with writing to disk from docker.. Sabnzbd complains it cannot find some files when it starts(files it was in the middle of downloading) but deleting these from the cache and disk the re-adding them and rebooting SabNZBD without any errors showing still results in the Call Traces shortly after...

 

anyone have any sort of ideas?

Edited November 23, 2017 by Dreded

[Squid]

Powersupply?

[Dreded]

the whole system uses 40 Watts and has a GOOD 500 Watt PSU.. I also did try another PSU just in case so definitely not the issue I can use the hD's to my hearts content right up until a docker wants to write to the drive in little bits(newsnab/torrent) I even tried having plex move a ton of files around no issue

 

 in the meantime I have been downloading files manually and copying them over the network system has been up 4 days now without issue.. if I was to run the Sabnzbd or Deluge Docker it would crash in 5 seconds

 

also can do Pre-clear's and check parity without issue

Edited November 25, 2017 by Dreded

[Squid]

Quick google search shows 2 causes 

- overclocking

- BIOS update needed.

 

Could also be a corrupted docker.img / filesystem on the cache drive

 

A full diagnostics instead of the snippet might also help

Edited November 25, 2017 by Squid

[Dreded]

i have had quite the week with a few hospital visits here and there so this took the back burner... but it is still up and running without running deluge or sabnzbd so tomorrow I will be deleting the docker image and starting that from scratch... hopefully thats the issue(seems quite possible as I moved it back and forth a bit while trying different cache drives).. there is no BIOS updates for this motherboard and its not overclocked(unless it was overclocked from factory)

Edited November 29, 2017 by Dreded

[Dreded]

ok well its been 2 days since I "fixed" it and it was indeed a corrupt docker image... or something with the linuxserver builds doesn't play nice with my system as I switched to the binhex builds(much more unraid friendly/thought out I might add) so it was one or the other as everything works great now(I deleted all my app settings(except copied back the plex db) and deleted docker.img also)

 

appreciate the assistance

Edited December 5, 2017 by Dreded

[Zonediver]

Had a similar problem years ago - reason: My server draw less power (like in your case, ~10% of max. Power), but the PSU was not able to work stable under this low power conditions.

It was specified for a minimum load of 20%. This might be the reason of this "random" reboots. Try a PSU with at lesat 80+ Gold Spec, because they can work also with a powerload of less then 20% (~8-10%).

My Server draws ~50W idle and the PSU is a 700W Enermax Gold. It is working at 7,2% load - so this is far less then the ATX-Spec of 20%.

2 years ago, the PSU died and Enermax told me, that the reason was the low load.

ATX-PSUs are switching units and need a min. load of 10-20% but Gold-PSUs can also work stable with 5-10% load.

Edited December 5, 2017 by Zonediver

[Dreded]

On 12/4/2017 at 10:49 PM, Zonediver said:

Had a similar problem years ago - reason: My server draw less power (like in your case, ~10% of max. Power), but the PSU was not able to work stable under this low power conditions.

It was specified for a minimum load of 20%. This might be the reason of this "random" reboots. Try a PSU with at lesat 80+ Gold Spec, because they can work also with a powerload of less then 20% (~8-10%).

My Server draws ~50W idle and the PSU is a 700W Enermax Gold. It is working at 7,2% load - so this is far less then the ATX-Spec of 20%.

2 years ago, the PSU died and Enermax told me, that the reason was the low load.

ATX-PSUs are switching units and need a min. load of 10-20% but Gold-PSUs can also work stable with 5-10% load.

that is a really good point, I should have thought of this as a electrical engineer but I didn't  the issue seems fixed for now and was most likely a corrupt docker image but running the bronze 80 PSU at 8-12% load probably isnt a great idea.. I probably will replace it with a pico PSU in the near future anyhow just to lower fan noise(aka none) I made a temperature controller for the internal case fans so they only turn on when cpu exceeds 65c and/or Hds exceed 35c(effectively making the unit silent unless transcoding which it rarely does) but that PSU fan is still noisy

Edited December 7, 2017 by Dreded