HA constantly malfunctioning

Installation Home Assistant OS
1

My HA OS is frequently needing to be rebooted. It loses connection constantly. I’ve read countless threads herein and either can’t find relevance or missing some key point.

The only error I receive in logs is below.

My thought is to reinstall HA, though I want to keep my Automations, etc. Can I import a backup, but only my Automations/Scripts/Custom files/folders, etc. and not the OS files?

  File "/usr/src/homeassistant/homeassistant/helpers/frame.py", line 373, in _report_usage_no_integration
    raise RuntimeError(msg) from err
RuntimeError: Detected code that calls device_registry.async_update_device from a thread other than the event loop, which may cause Home Assistant to crash or data to corrupt. For more information, see https://developers.home-assistant.io/docs/asyncio_thread_safety/#device_registryasync_update_device.
2

Restoring a back up only restores your configuration not the os.

But reason your system is failing will probably have nothing to do with the os. So tell us about your hardware and installation.

3

Hello ScottS,

Also let us know if you are on WIFI or Ethernet wired.

4

I’m running HAOS in a Hyper VM on a mini PC; below are the basic specs.

I’m also directly running my Plex server on this PC, which includes my download scripts. That said, I notice the constant HA reconnections occurring when there’s no other significant processing.

Running on Ethernet.
600/600 FO connection.

HP mini PC
i5_6500T
16GB RAM

Edit: I will try w/o my Plex server nor downloads running for a couple days.

5

How much disk space and ram in your VM? Cant see plex affecting it much more likely to be a resource issue or malfunctioning integration or addon.

6

This error is about an integration that calls certain functions in a bad way. That should never happen. This was always a problem, but since the ill effects have gotten worse, the HA team have put in a warning about it. It is most likely an outdated custom integration misbehaving. The warning was introduced a couple of months ago, so one would expect all integration developers to have fixed the problems by now, but that is not always the case for old, abandoned integrations.

Your best bet is to disable integrations one by one to see if the problem goes away, starting with custom integrations without recent updates.

7

That is not a warning anymore.
It was a problem before, so the devs deprecated the function and gave 6 months before removing it, so integration owners could update their code.
During the 6 months there were a warning, but the 6 months is over and now it is an error instead.

8

Thank you, I will do so and report back.

9

Okay, though technically, it’s still a custom integration error?

10

Plenty of disk space in the VM and the host.
RAM is within parameters; I think the host has 8GB, giving HA VM 4.