Integration looses connection to Wallbox - Wallbox remains online in App

[nondrun]

Checklist

• I have installed the latest release (or BETA) version of the integration and home assistant.
• I have prepared DEBUG log output (for technical issues) | In most of the cases of a technical error/issue I would have the need to ask for DEBUG log output of the integration. There is a short tutorial/guide 'How to provide DEBUG log' here
• I understand the SHIT-IN - SHIT-OUT principle (in case of implausible data) | It will happen, that the data that is displayed by this integration does not make much sense (to you) - aka the data is not plausible. Of course it could be the case, that something in this integration has been messed up - but so far - in all reported issues the root cause of implausible data was/is, that the go-eCharger device itself already provided this data [you can check this by directly requesting the attribute from the wallbox]
• I confirm it's really an issue | In the case that you want to understand the functionality of a certain feature/sensor Please be so kind and make use if the discussion feature of this repo (and do not create an issue) - TIA

Add a description

Hi Marq,

Unfortunately I have been struggling with a new issue for some weeks now. The integration (I am now using the local Version of your integration) seems to loose connection to the go charger repeatedly, but only when charging with PV Surplus mode. Strangely, the go-e remains available over the app while offline in the Integration. Mobile network is off so connection to internet over wifi seems not affected.

Only way to regain connection in integration is a restart of the wallbox.

One additional info: I have also been playing with an go-e mqtt Integration (https://github.com/syssi/homeassistant-goecharger-mqtt). While this Integration loses connection, the MQTT Integration remains Access to the charger.

Any ideas?

Thanks for your Support!

home-assistant_2026-01-26T12-52-09.349Z.log

I have attached a logfille.

Add your DEBUG log output

[marq24]

You are using the new firmware 60.2 - right?

[nondrun]

Sorry forgot to mention, no still on 59.4

[marq24]

but you noticed in your log, that you have several other local network issues?

[nondrun]

Sorry I am not familar enough with the topic to really decipher all those logs entries. What are you referring to? I had to make some changes to my setup so several automations and stuff still have to be cleaned up. The go e issues have been there longer though…

[nondrun]

These are screenshots from right now, happening again. If I reload/reconfigure the integration, it says the charger cannot be found. MQTT all the while running with no issues…

[marq24]

IMHO the internal webserver of your charger is overloaded with too many request... IMHO you must make a decision, use MQTT or the API integration... there can be only one

[marq24]

also I just can recommend you review your surplus automation... IMHO makes no sense to send data to the wallbox all day long - letting the box know, that there is almost none pv available...

• Only send data when a vehicle is connected...
• Only send data when the PV production is above a certain threshold - and then continue sending for max of 1h or so...

[nondrun]

I had a similar suspicion. unfortunately, the problem persists even when I deactivate the MQTT integration. I only rectivated it just now to check the connectivity… and other suggestions?

[nondrun]

also I just can recommend you review your surplus automation... IMHO makes no sense to send data to the wallbox all day long - letting the box know, that there is almost none pv available...

• Only send data when a vehicle is connected...
• Only send data when the PV production is above a certain threshold - and then continue sending for max of 1h or so...

I have only setup an anutomation for the mqtt Integration for testing purposes, this mqtt integration normaly deactivated. Otherwise im only using the one provided by Your Integration as predefined…

[marq24]

well - your log is full of

  File "/config/custom_components/goecharger_api2/pygoecharger_ha/__init__.py", line 320, in _write_values_int
    async with self.web_session.get(f"{self.host_url}/api/set", headers=headers, params=args) as res:
               ~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.13/site-packages/aiohttp/client.py", line 1510, in __aenter__
    self._resp: _RetType = await self._coro
                           ^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.13/site-packages/aiohttp/client.py", line 779, in _request
    resp = await handler(req)
           ^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.13/site-packages/aiohttp/client.py", line 738, in _connect_and_send_request
    raise ConnectionTimeoutError(
        f"Connection timeout to host {req.url}"
    ) from exc
aiohttp.client_exceptions.ConnectionTimeoutError: Connection timeout to host http://192.168.0.166/api/set?ids=%7B%22pGrid%22:+4048.0,+%22pPv%22:+0.0,+%22pAkku%22:+0.0%7D

this is this (and not the MQTT) integration that is posting PVSurplus data to your wall box...

[nondrun]

this is this (and not the MQTT) integration that is posting PVSurplus data to your wall box...

Sorry but it’s well above my capabilities seriously deciphering this… :-(

But it seems to come from the api Integration no? Obviously its sending data as the PV Surplus loading is currently active and car ist connected, so its behaving as expected no? What do you suggest? I am bit lost what to do or where to Look for what….

[nondrun]

…please also note the log was created while the api integration had already lost connection…. I would be extremely helpful for a hands on support for someone with limited IT capabilities, I am happy to haven got this setup and the integration installed in the first place,.,.. ;-)

[marq24]

even IF the car is connected - and probably the sun have risen - the available PV is Zero - IMHO trying to charge your car via PV is not the smarted thing to do in Winter... (as long as you are located in Europe)

To give you some technical background info - with MQTT, the Wallbox send messages from the BOX to the MQTT Server - so information flows from your BOX to the MQTT-Server (and from there to your HA instance)...

This API Integration is requesting data from your BOX - so the communication goes into the opposite direction (from your network TO the BOX)... And based on your log output I have the impression, that the overall network is not stable... (might be as simple as the WLAN to which the BOX is connected)...

So while you don't see any ERROR (in case of MQTT), when the BOX can't reach the MQTT-Server, this integration reports an error, when it can't reach your BOX... So the absence of error message in the MQTT integration does not mean 'all is fine'...

The root problem for this integration remains - it frequently can't reach your BOX (or simply does not reply to the requests)... Based on your logs I was able to identify another type of exception that is worth catching & handling... But even with the adjusted code the problem will remain, when the network is not stable.