My unRAID server has been unreliable for some time, causing docker containers to stall or become unresponsive. The unRAID GUI and even CLI/ssh response gets very sluggish when this occurs. Attempting to stop and restart containers rarely works. Usually I end up rebooting the server and things go back to normal for a few days or so.

 

Running the Fix Common Problems plugin indicates the following:

 

Out Of Memory errors detected on your server
Your server has run out of memory, and processes (potentially required) are being killed off. You should post your diagnostics and ask for assistance on the unRaid forums

Call Traces found on your 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

 

I've attached the diagnostic file and would really appreciate any recommendations. Thanks!

tower-diagnostics-20170914-1950.zip

You should probably read this post and the next ten of twelve posts that follow:

 

I appreciate the quick reply, @Frank1940. I had the impression that the vm settings here were for virtual machines and thus not applicable as I've turned off VMs (also the help tips seem to indicate that). Just in case, I modified the vm.dirty_background_ratio and vm.dirty_ratios from their defaults of 10% and 20% to the levels suggested in the post (2% and 4%).

 

Sep 14 22:01:09 Tower tips.and.tweaks: Tweaks Applied

 

The system is running very sluggishly so it's hard to tell whether this has made any effect. I'm concerned that rebooting will simply delay further troubleshooting.

If there were OOM errors you should reboot.

 

@Frank1940please recommend to set to those values to 1 and 2, 2 and 4 is not going to have any benefit over 1 and 2 and is still more likely to get OOM errors for users with large amounts of RAM.

4 hours ago, johnnie.black said:

@Frank1940please recommend to set to those values to 1 and 2, 2 and 4 is not going to have any benefit over 1 and 2 and is still more likely to get OOM errors for users with large amounts of RAM.

 

Updated my original post where I made this suggestion.

8 minutes ago, Frank1940 said:

 

Updated my original post where I made this suggestion.

 

?

Thanks @johnnie.black and @Frank1940! I've adjusted the vm.dirty_background_ratio and vm.dirty_ratios settings to 1 and 2. My server has 64GB of RAM. I'm rebooting and will monitor over the next days.

All looks good. Marking this issue as solved!