Load peaks - what is HassOS doing?

[gilbert-grape]

Hello

I am running an old Rasperry 3b with the newest HassOS. I have not really much plugins running (DuckDNS, MariaDB, deCONZ, Samba, ADB Debug Bridge) - but in the end with the months and years the number of sensors increased. I have sometimes load peaks and exactly at this peaks my zigbee sensors (I am using a conbee 2 stick) are unavailable for 1-2 minutes (until the peak is away).

So my next step is to find out, what HassOS is doing at this peak. Is there a log with timestamp? A monitoring or a top for the processes what I can run?
Thanks for some tips.
Gilbert

[ciotlosm (Marius)]

Maybe try https://developers.home-assistant.io/docs/operating-system/debugging/

[gilbert-grape]

Thats clear: Debugging with ssh. But how? How do I see, on what time what HassOS is doing?

[ciotlosm (Marius)]

I thought you had some rough idea, and could have process monitors up or a docker logs on some containers that you think might be causing problems.

[Recte]

It could be IO_Wait, see my topic on this.

Troubleshooting deCONZ unavailability and high CPU loads

Hi All, It feels like since v0.114.x I have an issue which first showed as unavailability of all bulbs connected to my deCONZ. Digging a little showed relatively high loads on CPU. Since I noticed a high memory consumption (80%+) caused by the UniFi addon, I decided to first replace the RPi3+ for a RPi4b-8GB (I missed out on the fact that 4GB at most is currently supported, so am running a development build as mentioned here. While doing that, I thought it would be good to use a brand new Samsu…

[Krojack]

I had this on my system as well. I noticed in Proxmox CPU usage was hovering at a solid 70%. When I SSH in and looked around all I could see was the Docker process using all the CPU.

As you can see in the image, I did a Host Reboot which seems to have cleared up the problem. I’ll keep an eye on it to see if it starts up again and if so try to get more information.

[Recte]

It’s important to understand that CPU usage percentage is something different than CPU load.
Since when do you have this @Krojack ?
If it is since v 0.115, then perhaps it is caused by the way you “query” states, see the post below.

0.115: B-Day release! Media browser, tags, automations & WTH

This could be due to the changes to templates mentioned in Breaking Changes. Template entities will now correctly track all entities that you reference. So look out for templates that use states (i.e. everything), those templates are now re-calculated whenever anything changes. It is better to use e.g. states("light.office") because this will only track changes to that single entity.

[Krojack]

I haven’t had the problem since I did that restart. CPU has been hovering between 5 and 10. I can’t say when it started because the Proxmox only goes back 2 hours for the graphs.