Hi, 

 

I've noticed that my system will suddenly jump to 100% CPU usage on all cores periodically, most recently after installing a new VM for HassOS. A quick restart tends to fix this. I also tried disabling Docker when this occurred earlier today, around 10am is when I noticed it, and there was no improvement. I'm wondering if there is an issue with my cache or other storage, or if the VM is hogging the CPU (I pinned two cores to the VM and have no other VMs running, but perhaps that's not adequate?). 

 

Diagnostics attached.

 

Thanks!

tower-diagnostics-20230420-1207.zip

Radarr/Sonarr appear to be the main culprits.

I see. They hadn't caused issues at all in the past... so I wonder what is different now.. I'll take a closer look. Perhaps there is an issue with the indexers or the like. Thank you

So I have an update. Even when stopping Radarr and Sonarr, the problem persists. This all began after installing HassOS on a VM. I have updated diagnostics here. Perhaps there is something wrong with Home Assistant as a VM? 
 

the process that eats up CPU is qemu-system-x86_64 when I checked the terminal. It looks like it’s being run by the “homeassistant” user

tower-diagnostics-20230423-1248.zip

Does it improve if you shutdown the VM?

@joshsalviI have this same issue. I do not use Radarr/Sonarr. I have rebooted and shutdown with power cycles a number of times. The system will occasionally run ok for a short time but return to maxing out the cpu, causing other applications to hang and crash. The system was fine and stable before the install in a vm per Home Assistant forum instructions: https://community.home-assistant.io/t/install-home-assistant-as-a-vm-on-unraid/270195 The issue of CPU 100% usage resolves and returned to normal the second the home assistant vm is quit. Two CPU cores are pinned per the instructions, i did alternate one, issue still persists.

 

Diagnostics removed after resolution.

 

Edited May 31, 20233 yr by Be-Art
Diagnostics removed after resolution.

So after some days of shutting down various vm's and testing only for the 100% usage to return, i found a RAM chip to be not showing in the system, after

1. re-seating the ram and

2. finding this comment on a thread speaking of a similar issue with CPU, i de-activated integrity plugin for 2 days. The issue did not return. I have no re-activated integrity plugin and will monitor further. but so far the issue has resolved. HA VM is running along with 4 other ubuntu vm's.

 @joshsalvimaybe this can assist with your issue?

following...  i'm getting similar issue as well. can't figure out what is causing cpu to run so high. 

@Be-Art  Just out of curiosity, do you have authentik containers running?  I stopped the containers one by one untill the cpu and read on the cache drive drop.

I then realise i was still running a 2022 version and then updated since.  been ok so far.

Hey @LeoRX , No i am not running authentik.

In my case I was using sonarr/radarr, and plex. When download is finished and sonarr was copying files to destination location - plex is starting indexing it and probably both process needs a lot of computation. Im disabling plex (during copy process) and monitoring it. Probably I will disable auto scanning folders in plex and add sonarr plex integration if it will fix the issue. Stay tuned or ping me for further updates.