Call trace

[ptirmal]

 

Had this message pop up on my server: 

Your server has issued one or more call traces. This could be caused by a Kernel Issue, Bad Memory, etc. You should post your diagnostics and ask for assistance on the unRaid forums

Can someone take a look at this for me?

 

unraid6-diagnostics-20180218-1922.zip

[ptirmal]

Anybody?

[Frank1940]

I did take a look at the syslog in your diagnostics file.  I am not expert on analyzing syslogs but I can recognize some problems.  You started this system on Jan 29th and the log was taken on on Feb 18th.  The only call trace occurred on Feb 12th.   There did not seem that there was an recovery action required.   Does your system appear to be running normally?  

 

Now for a bit of history...  The tagging of these events was first done about a year ago for some reason.  Until that time, I can't recall seeing any mention of them.  (Probably because no one was seeing them in the syslogs.)  The problem is that many times these events are inconsequential and the system recovers without a problem.  Sometimes they are an indicator of a hardware issue that might be a one-time event or, possibly, an early indicator of eventual failure of that hardware or, in the worst state, that the hardware has failed.  Sometimes, it is the results of a error condition that the software is not prepared to handle.    Diagnosing them is difficult and can require an expert who can figure out exactly what was the most likely cause.  

 

In you case, I would not be panicking.  One instance almost eight days ago after the system had been up for over two weeks is not an indicator that you have some major problem.  It might have just been a power gitch that gave the CPU a 'brain phrat'.   Now, if you get them every half hour, there is something wrong.   

 

I have the suspicion that other folks have look at your post and diagnostic file, and have not found anything to indicate a possible cause.  Hence, they did not post back.

[ptirmal]

After this call trace I attempted updating Dockers, it wouldn't let me stop them and I was forced to hard reset my server. I started it up and parity check came out fine. I did have an issue about 2-3 months ago, the system was unresponsive, couldn't access anything via web gui, no IP address, nothing, also had to hard reset then. 

 

My initial thought were this is a docker/software issue not a hardware one, the hardware has been running 24/7 for 4+ years, I did add more RAM about a year ago and I added a sata controller maybe 6-9 months ago. 

 

Is there anything I should watch out for to diagnose it further? Is a memtest a good idea? 

 

Edit: Also, my apartment has been having power issues and my server has been getting on battery more often. The last time was probably 30 days ago when it shut itself off. The issue should have been fixed then though (electrician was there and did some work).

[Frank1940]

You could run a memtst but to really test the memory, you have to let it run for a minimum of 24 hours.  

 

About the Dockers.  Did you try to see which one  wouldn't shutdown?  That information could be a clue as to the issue.  You have to realize that this is (apparently) a very intermittent problem.  Fixing those is always very difficult.  

 

You said the UPS shut the server down about a month ago.  Was that a clean shutdown at that time?   

[ptirmal]

Not sure when I can schedule a downtime of 24 hours! It was a clean shutdown, the UPS initiated it. I was trying to update the Dockers, I started with duplicati and then did Ombi and Plex next, before duplicati finished. They all just hung, I think the log showed they couldn't be stopped, then they were inaccessible. They also locked my ability to stop the array.