Failure to POST to airco causes HA to become unavailable ?

[alcohol]

This is merely a deduction based on practical experience and I have no good data to back it up, but I've noticed that sometimes this addon fails to update my AC (POST request fails), and then for reasons unknown to me, HA in its entirety becomes unavailable / unreachable. I can see the container is still running, but it no longer accepts any incoming requests. The logs show very little useful information:

Aug 13 13:35:55 universe docker[1207]: 2025-08-13 13:35:55.615 WARNING (MainThread) [custom_components.mitsubishi_wf_rac.wfrac.repository.http] Error POSTing to http://192.168.0.129:51443/beaver/command/getAirconStat: HTTPConnectionPool(host='192.168.0.129', port=51443): Max retries exceeded with url: /beaver/command/getAirconStat (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7fe2701f16d0>: Failed to establish a new connection: [Errno 111] Connection refused'))
Aug 13 13:35:55 universe docker[1207]: 2025-08-13 13:35:55.615 WARNING (MainThread) [custom_components.mitsubishi_wf_rac.wfrac.device] Error: something went wrong updating the airco [AC 3.5] values
Aug 13 14:35:59 universe docker[1207]: 2025-08-13 14:35:59.749 WARNING (MainThread) [custom_components.mitsubishi_wf_rac.wfrac.repository.http] Error POSTing to http://192.168.0.129:51443/beaver/command/getAirconStat: HTTPConnectionPool(host='192.168.0.129', port=51443): Max retries exceeded with url: /beaver/command/getAirconStat (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7fe2701f1f40>: Failed to establish a new connection: [Errno 111] Connection refused'))
Aug 13 14:35:59 universe docker[1207]: 2025-08-13 14:35:59.750 WARNING (MainThread) [custom_components.mitsubishi_wf_rac.wfrac.device] Error: something went wrong updating the airco [AC 3.5] values
Aug 13 15:36:05 universe docker[1207]: 2025-08-13 15:36:05.708 WARNING (MainThread) [homeassistant.helpers.entity] Update of climate.ac_3_5 is taking over 10 seconds

I have to restart the container/process at this point, after which HA becomes reachable again.

[vimu86]

This is most likely related to the issue #137

I have the same issue that HA becomes unavailable. Docker container log contains the following messages

2025-08-17 10:56:11.910 WARNING (MainThread) [custom_components.mitsubishi_wf_rac.wfrac.repository.http] Error POSTing to http://192.168.178.xxx:51443/beaver/command/getAirconStat: HTTPConnectionPool(host='192.168.178.xxx', port=51443): Max retries exceeded with url: /beaver/command/getAirconStat (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7f4bf65f40>: Failed to establish a new connection: [Errno 111] Connection refused'))
2025-08-17 10:56:11.910 WARNING (MainThread) [custom_components.mitsubishi_wf_rac.wfrac.device] Error: something went wrong updating the airco [Airco Wohnz] values

However, this error is not the last message in the docker logs. HA became unavailable 10 minutes later and there was no such log entry.
The last log entry before HA becoming unavailable is in my case one indicating that the temperature setting of the aircon was changed:

2025-08-17 11:05:08.549 INFO (MainThread) [homeassistant.components.automation.airco_wiga_temp_update] Airco Wiga Temp Update: Running automation actions
2025-08-17 11:05:08.549 INFO (MainThread) [homeassistant.components.automation.airco_wiga_temp_update] Airco Wiga Temp Update: Executing step call service

Which HA version are you using?

[vimu86]

I have upgraded today my HA docker from version 2024.12 to 2025.08... the problem with HA freezing when using this integration persists.

[vimu86]

I have looked into this issue more deeply.
I am able to reproduce the issue of HA becoming unavailable.

In my HA, I have the following aircon:

If I click on the ring every 3 seconds, i.e. set the temperature of the aircon every 3 seconds, this is fine. If I click set/change the temperature with short duration (less than second), this will freeze the HA webserver (F5 will not work anymore). Interestingly, a few message are still being logged after the freeze into home-assistant.log, so HA is not completely dead. Docker is reported as running nominally, but maybe the websocket component or so is frozen.

I have set the Logger to log in debug mode and also used the homeassistant: debug=true setting.
There are no errors visible in the logfiles of HA or Docker.

[jeatheak]

Could you check if it is resolved after updating to the latest release? 2024.12