Bug(C3): Midea Heat Pump Controller entities unavailable since last Home Assistant Core Update

[chasp152]

HA core version

2024.11.2

Current integration version

0.6.5

Last known working integration version

0.6.5

Device type and model

Heat Pump Wi-Fi Controller 171H120F (0) by Midea

Working mobile app

Smarthome

The detail description of bug/problem

1. description the device type and name（you may have multiple devices）
   Heat Pump Wi-Fi Controller 171H120F (0)
   by Midea

2. description the error action or error info (device have multiple data/features)
   Getting the following error when trying to switch On/Off one of the controls: "errno 32 broken pipe"
   All sensors haven't updated since the home Assistant core update
   

The logs

1. DEBUG log MUST be enabled (Please refer to README document)
2. Execute the action to trigger error log（you should description the error action）
3. Upload the debug log file as attachment。
4. Please don't paste the log content! upload as a attachment file.
5. Please don't paste the log content! upload as a attachment file.
6. Please don't paste the log content! upload as a attachment file.
   ...

[erikgieseler]

I have the same heat pump, HA 2024.11.2 and midea_ac_lan 0.6.5.
It works for me without any problems.
Can you please send a debug log, see: #387 (comment)

[chasp152]

It may not be the update that caused the issue.
I manage to connect it again by removing its configuration and setting it up again.
Yesterday night after a power outage, the heat pump doesn't reconnect to HA (it does appear in Midea's Smarthome app though). I checked the IP is still the same (I assigned a fixed one)
logs.zip

[wuwentao]

1. from the debug log, I can't found any useful info.

2. if your device is power off, HA will try to reconnect every 10 minutes after a long time offline, reboot HA can reset it or wait 10 minutes.

3. when you remove it and setup again, the same error exist?

@chasp152

if it's possible, enable debug log via README with configuration.yaml file and reboot your HA device, then wait for 1-3 minutes, upload full debug log to github issue.
as current debug log is a part and can't find the root cause.

[wuwentao]

@chasp152 please help to update the latest result.
if there is no any response for a long time, and we will consider to close current issue.
thanks.

[chasp152]

Sorry the issue resolved itself after a random restart and yesterday I lost all entitie updates once again for no apparent reason. So I don't think it's related to the latest Home Assistant Core update.

I was able to enable debug logs and capture some. I tried the restart again but it didn't fix the issue this time.
config_entry-midea_ac_lan-01JD1J1BE6MTQ3222A1RJC45YV-1.json
error_log-2.bin.zip

[wuwentao]

sorry, from your debug log, I found everything works well as expected, i don't found any error exist.
so what's your problem now ?
please consider to upgrade your HAOS or HA version, and make sure your HA is more stable.
from current midea_ac_lan or midealocal lib, I can't found error with your debug log.

if issue still exist, add your error screenshot or error log, and detail description.
if issue not exist, you can close current issue.
thanks
@chasp152

[chasp152]

I am using the latest HA OS and Core.
At some point I suspected a connectivity issue but I can see all the data in Midea's Smarthome app:

Midea Smarthome

Home Assistant (before reloading the integration)

Home Assistant (after reloading the integration)

After reload we can see the entities were "updated" but the data point isn't up to date and still displays an old value. Another weird behavior is that sometimes it will actually get a new value, creating huge leaps in the energy consumption (like yesterday it caught up on multiple days of consumption in juste one poll)

[chasp152]

logs Midea integration.zip

Logs of the reloading

[wuwentao]

@chasp152 thanks, I think I have found some error in your debug log now, but I still need to do more debug with why it works in this error status.

maybe you can contact me via discord group chat and we can quickly sync the issue , try to change some source code, add some debug output, and find out the root cause.

discord group chat link can be found in the top of README document.
if you can't join discord chat, we still need to do more debug checks via github issue step by step.

[chasp152]

Alright now it's in a strange state where

• I can control the thermostats and other control entities
• the data entities are not being updated though since the last restart of the HA server (it updated 1 time right after the restart and that's it, no refresh)

Here are some debug logs of me changing the control entities:

Let me know if there is anything else I can do to help

[wuwentao]

current error log can't got the root cause, so I have manual changes some file , send it via discord message, please check the detail and replace the file, once it done, send the full debug log to me.
we can confirm the next step with the new changes.

[wuwentao]

please update to latest version v0.6.6, confirm whethere it fix your issue.
as there is no any response for a long time,I will close current issue at first, if you still got the same error, you can reopen current issue with latest version and debug log file.
thanks!

[chasp152]

I had the issue again this morning with all entities becoming unavailable. Luckily I was capturing some logs when the system went back online.

[wuwentao]

@chasp152 thanks with your detail debug log, we can confirm the root cause is your device, HA can't connect to your device and can't got these data.

and it always loop in connection pass/failed/error/timeout and caused HA can't have a stable tcp connection to your device.but it only exist in your device (Midea PAC) [midealocal.device] [153931628531992]
as you still have another device (Comfee Déshumidificateur) [midealocal.devices.a1] and it works well without this issue.

so we can confirm there is no any error in HA or midea_ac_lan integration side, and it should be your device issue, maybe hardware or software error, maybe it rebooting in loop or any exception exist, it's hard for us to debug it in HA side.

below is the detail log, you can also check the time and connection pass/error loop from the debug log

connection_error.log

all the info just copy/paste for your origin debug log file error_log-7.txt
you can check it by yourself.

sorry, we can't fix it