Open Andre1Becker opened 2 days ago
I just noticed this myself: I'm using oh 4.2.2 with a 4.3.0.x binding. Maybe that's causing the problem?
I just noticed this myself: I'm using oh 4.2.2 with a 4.3.0.x binding. Maybe that's causing the problem?
I would not expect that to be the cause. You could upgrade to the latest snapshot for both openHAB and the binding.
That's not that easy because this is my "production" setup :-)
Let's do some other checks before we try something like that.
Had a quick look and did not find anything directly pointing towards a leak. I did notice that the refresh command is calling an async update of all data causing a remote api call.
If you have many channels linked, this might cause some DDOS-like actions towards the remote API. Not 100% iuf this is the cause, but this can be improved.
There are some karaf commands that you could use to minitor threads:
shell:threads | grep -i runnable | wc -l
In the OP you mention that you revert to the marketplace version. Do you also experience this issue with the stock version that comes with openHAB 4.2.2 ?
Hey, no - stock version runs fine. There are 40 Channels i think that i have linked to items.
I've just switched to the snapshot binding and try to look into it a bit further. i try to monitor the threads as you mentioned. Anything else i could do?
Current Behavior
Because of Fix the parsing of total consumption channel i've installed the snapshot version of this binding. Memory Usages creeps up from 2.3 GB to around 9 GB and CPU goes up to 98%ish of that container
I removed the Marketplace binding (Ver, 4.2.2) and copied the jar to the addons folder and restarted the docker container. After that it takes roughly 4-5 hours till openHAB becomes unresponsive. Removing the jar and installing the Marketplace addon, everything is working normal again.
I could reproduce this issue, but i cloud not troubleshoot it any further because lack of skills ;-) i cloud not see anything suspicious in the openhab logs regarding this problem.
Running openHAB Docker container 4.2.2 on unraid 6.12.13 with an Intel Core i5-6500 CPU @ 3.20GHz and 32 GB of RAM.