Hello, I recently moved and had to re-set up my server. For the most part there didn't seem to be any issues after booting everything up, but I also was upgrading my network in the new home so I initially had some network issues. I feel like I solved most of them and now I am not seeing any abnormal logs, but my server just seems to lock up every once in a while. The CPU/RAM usage looks okay, and I don't see any crazy issues. I do see some odd stuff like this flash backup error pop up more than usual

flash_backup: adding task: /usr/local/emhttp/plugins/dynamix.my.servers/scripts/UpdateFlashBackup update

Fix common problems says that there was a machine check error, but I couldn't see it in the logs. Any help?

brooksnook-diagnostics-20230815-2229.zip

Try enabling the syslog server to see if it catches something, MCE log is not reporting what the error is, kind of strange.

@JorgeB I appreciate the response. While that is collecting data, anything else out of the ordinary? I did notice before while dealing with networking changes that the server would get less responsive while UpdateFlashBackup was going, although I was making a lot of system changes so it was understandable. Any idea why it ran every couple of hours while I had the server untouched?

 

Also, as mentioned I had network changes. I updated my pfSense settings to VLANs, but I should have an allow all traffic rule for it so not sure what could be the issue there. I now have my server on a managed switch so I was able to get 802.3ad set up. Initially I had a 8 port SFP+ switch, but turns out you can't do auto navigation with SFP ports so Unraid didn't like that, but I have since swapped that out and haven't seen any major logs about networking since. So not sure if there still is something odd there that is causing it.

 

Finally, I enabled local syslog. Should I just let it run for the day and post that file only or just do another diagnostic dump?

5 minutes ago, MooseCadets said:

anything else out of the ordinary?

Nothing jumped out to me in the syslog, looks normal.

@JorgeB I just got another diagnostic dump since I saw my server was (and still is) almost completely unresponsive for the past hour or so. Could there be any sign of any reason why between the two diagnostics?

brooksnook-diagnostics-20230816-1921.zip

Syslog looks normal, but Connect plugin is using a lot of CPU:

 

  PID USER      PR  NI    VIRT    RES    SHR S  %CPU  %MEM     TIME+ COMMAND
10387 root      20   0   11.1g 365824  48020 S  61.1   0.4 151:15.65 unraid-api

 

Try rebooting, also make sure it updated.

Thanks. I had to force restart as nothing was responding.

 

I think I found the issue however. I did notice increased temps in my servers new location, so I moved my cache ssd from the back of the server to the front to prevent it from overheating. The back drives do get a bit hot, and it looks yesterday everything was not responding due to things getting too hot. I have additional scripts to increase fan speed but it might not been enough. The temps of my CPU and drives got a bit warm but nothing I would think that would cause these issues, however I just thought that maybe the flash drive got too hot as it is located within the machine... Does temperature issues/correlations show up anywhere in the logs to confirm? 

Also at a high level I know that Unraid is ran on RAM, but if the USB was overheating that could cause issues like this correct?

10 hours ago, MooseCadets said:

but if the USB was overheating that could cause issues like this correct?

It could, but it seems unlikely to me.