HABApp creates increasing load on OH4.0.1

I have installed Cadvisor for monitoring the containers hosting my home automation stack.
Currently, I am evaluating porting my rules from JSR223.

What I saw this morning right now is that the OH processor load shows a ramp-up behavior. I was able to put this in correlation to HABApp container running and restarting.

My first idea was that this might be correlated with the weird usage of HABApp in my test cases. On the other side if there is anything like a resource leak this should be reflected on the HABApp container monitors. On these, I do not see anything pointing to this.
Anyway, This morning I removed all Python files from the rules directory and restarted HABApp.
I see exactly the same behavior.

Here is the monitoring of OH4.0.1:

Here the HABApp monitoring:

The HABApp.log is not showing any further messages since the last restart, the event.log shows the ongoing OH events.

I am pretty sure it’s not an HABApp issue more a OH4.0.1 one. I need a verification from somebody else.

After some investigation, I found the root cause of the issue:

In order to have one VisualStudio Code project both the HABApp and JSR223 code, I have mapped HABApp docker volumes to openhab_conf/automation. That means rules and log directory were located in the automation folder. This seems to be a bad idea

I have now reconfigured the volume mounts away from the automation directory and the effect of increasing the processor load over time disappeared.

1 Like