Hi,

 

My unraid is locking up after a bit of investigating I found that one of the cpu cores is always at 100%.

 

Then when I look again it changes to another, it does not seem to matter if a vm is allocated to that cpu or not. I tried a reboot but once I started the vm's or docker back up it just max's out again. 

 

I have included my diagnostic files, I am hoping someone here can help me figure this out as it is causing the computer to not work.

 

apollo-diagnostics-20200617-0324.zip

Hi,

 

My unraid is locking up after a bit of investigating I found that one of the cpu cores is always at 100%.

 

Then when I look again it changes to another, it does not seem to matter if a vm is allocated to that cpu or not. I tried a reboot but once I started the vm's or docker back up it just max's out again. 

 

I have included my diagnostic files, I am hoping someone here can help me figure this out as it is causing the computer to not work.

 

apollo-diagnostics-20200617-0324.zip

It’s most likely the wsdd process. Check in “top’ from the command line.

 

It’s Windows discovery and the only way to permanently get rid of that problem is to disable WSD in SMB settings.

 

Personally I prefer WSD to enabling SMB 1.0 so I just use the “i- br0” parameter in WSD options in SMB settings. That seems to minimize the 100% CPU problem for me.

 

It will take a reboot (or a restart of SMB from the command line) to fix the pegged CPU issue.

 

That problem has never locked up my server so it may not be the cause of your server locking up, but, it is still a problem.

 

 

Sent from my iPhone using Tapatalk

Hi Hoopster

 

Thank you for getting back to me.

 

I think I have found the culprit, which turns out to be the plex docker. I'm not sure what happend but it looks like because there was two tv shows and two movies listed in the config (one of each was empty) it throw a wobbly. A update must have updated the config and added additional fields, not really sure as I have not touched it, except for updates in a long time.

 

That's interesting about the wsd option as I have never noticed that before, I will take a look at it.

 

I will keep an eye on unraid and see what happens over the next week or so, but i'm hopeful (maybe foolish) that this has resolved the issue.

 

Thanks again.

 

6 hours ago, Houdini_2nd said:

That's interesting about the wsd option as I have never noticed that before, I will take a look at it.

Here is the bug report about wsdd causing one CPU/thread to hit 100% and bouncing around to different CPUs.

 

I don't think it is anything limetech can fix as they don't write the WSD daemon.

 

As I said, adding the i- br0 parameters seems to have greatly reduced the issue for me and I rarely see it anymore although it does still occur from time to time.