Single request tend to fail after some inactivity

[Maczuga]

trafficstars

After not issuing any commands for some hours - next command fails (throwing error below). After that error - all following commands are working as intended.

Action I tried to perform was setting hvac mode from off to 'dry'. However - this also happens when switching from/to other modes, changing temp. etc.

Stack trace:

Traceback (most recent call last): File "/usr/src/homeassistant/homeassistant/components/websocket_api/commands.py", line 121, in handle_call_service connection.context(msg)) File "/usr/src/homeassistant/homeassistant/core.py", line 1150, in async_call self._execute_service(handler, service_call)) File "/usr/src/homeassistant/homeassistant/core.py", line 1172, in _execute_service await handler.func(service_call) File "/usr/src/homeassistant/homeassistant/helpers/entity_component.py", line 194, in handle_service required_features File "/usr/src/homeassistant/homeassistant/helpers/service.py", line 316, in entity_service_call future.result() # pop exception if have File "/usr/src/homeassistant/homeassistant/helpers/service.py", line 337, in _handle_service_platform_call await getattr(entity, func)(**data) File "/config/custom_components/midea/climate.py", line 276, in async_set_hvac_mode await self.apply_changes() File "/config/custom_components/midea/climate.py", line 95, in apply_changes await self.hass.async_add_executor_job(self._device.apply) File "/usr/local/lib/python3.7/concurrent/futures/thread.py", line 57, in run result = self.fn(*self.args, **self.kwargs) File "/usr/local/lib/python3.7/site-packages/midea/device.py", line 165, in apply data = self._cloud_service.appliance_transparent_send(self.id, data) File "/usr/local/lib/python3.7/site-packages/midea/cloud.py", line 174, in appliance_transparent_send 'applianceId': id File "/usr/local/lib/python3.7/site-packages/midea/cloud.py", line 92, in api_request return self.api_request(endpoint, args) File "/usr/local/lib/python3.7/site-packages/midea/cloud.py", line 94, in api_request raise RecursionError() RecursionError

[Maczuga]

It seems that 3 concurrent API request failed throwing following error. However from what I have noticed - next execution always work. Perhaps there should be a timeout between api call retries (1sec or configurable)? Configurable retry count would be also nice.

[andersonshatch]

I've noticed the same. Have you tried any of your ideas? Open to PRs!

[Maczuga]

Will try, but kinda busy at work lately. Gonna setup my python enviroment and test it hopefully before the weekend.

Is there a way to access the dependency lib when using hass.io? The path seems to be somehow hidden behind docker.

[andersonshatch]

Not sure I'm afraid, I am not using hass.io

On 30 Jul 2019, at 21:03, Maczuga [email protected] wrote:

 I had one, but kinda busy at work lately. Gonna setup my python enviroment and test it hopefully before the weekend.

Is there a way to access the dependency lib when using hass.io? The path seems to be somehow hidden behind docker.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

[andersonshatch]

I guess if we could fix the fan_only issue that happens for every update on some devices, you would not notice the session timeout as it would happen for background update requests.

Or, maybe we can find a state change that will not alter the device mode, so if it's off or not in fan_only, you won't notice. Maybe turn swing on and off again? This would update the state of the device in home assistant, and log in if the session has expired.

[Maczuga]

Sure, this would be great, but I'll take small steps. I got the AC the same day you created the fork so I still need to figure out some stuff from the code.

Testing the timeout between calls right now, gonna report hopefully tommorow if it did anything.

[andersonshatch]

Sounds good! Out of interest do you have the fan_only problem and need the workaround? In the component repo I am considering making the workaround optional and have a branch here with that, could use some testers for those without the issue.

[Maczuga]

I think I had this one. After I am done with testing the timeout thingy - I can help with this one.

[Maczuga]

I did some digging with the responses I get when trying to sync state with midea servers - ~~I guess some models needs different value readings~~. - nope, hex indexes and offsets seems fine, I was just looking at wrong bytearray.

Timeout solution didn't work sadly.

Gonna try the base Ruby package another day, see if that works, maybe something was skipped during the porting.

[alesnav]

Hello there.

I've just installed midea-ac-py for Home Assistant, from HACS. Everything worked at the first attempt but the Fan Only mode "bug". After using the published workaround use_fan_only_workaround: true, Home Assistant does not update its data if it was modified by any other method, for example, directly in the A/C using its remote.

I own two HTW Air Conditioners that works using a USB from Midea.

How can I help you with the troubleshooting?

Home Assistant version 0.97.2, inside a Docker Swarm cluster.

Thanks, Best regards

P.D.: I enabled debug mode for midea, and all HASS events are been sent to my Elasticsearch cluster.