razsultana
commented
1 year ago I just noticed issue #11 , which seems very closely related to this one.
You asked the reporter of issue #11 to provide the output of GoSungrow ha lovelace, so here is what I got:
bash-5.1# GoSungrow ha lovelace
Can't find DeviceType 22 attached to ps_id 1151314.which is similar to what he got but the DeviceType reported in his case by the error was 43 (he had a DeviceType 22 though listed by GoSungrow show ps list).
Here is what I get listing my devices:
bash-5.1# GoSungrow show ps list
┏━━━━━━━━━━━━━━━━┳━━━━━━━━━┳━━━━━━━━━━━━━┳━━━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┓
┃ Ps Key ┃ Ps Id ┃ Device Type ┃ Device Code ┃ Channel Id ┃ Serial # ┃ Factory Name ┃ Device Model ┃
┣━━━━━━━━━━━━━━━━╇━━━━━━━━━╇━━━━━━━━━━━━━╇━━━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━╇━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━┫
┃ 1151314_14_1_1 │ 1151314 │ 14 │ 1 │ 1 │ A1912190548 │ SUNGROW │ SH5K ┃
┗━━━━━━━━━━━━━━━━┷━━━━━━━━━┷━━━━━━━━━━━━━┷━━━━━━━━━━━━━┷━━━━━━━━━━━━┷━━━━━━━━━━━━━┷━━━━━━━━━━━━━━┷━━━━━━━━━━━━━━┛I can see from the example output in your README.md that Device Type 22 represents the WiNet-S WiFi stick. My system is probably older as it has a WiFi stick that is simply referred as "WiFi" in the iSolarCloud interface, when you try a local connection and click the "manual connection" button - there are many other sticks there e.g. WiNet-S, EyeS4-EU, WiFi-P2, EyeM4, Logger1000 and similar Bluetooth devices. So if the code is looking for DeviceType 22, it won't find it :(
flare04
shelldor
Hi Mick, GoSungrow looks awesome. I have the same problem that you had - I am trying to get my Sungrow SH5K-20 inverter integrated in Home Assistant. I managed to get GoSungrow to connect to the iSolarCloud API and report data back using standalone GoSungrow. When I tried though to get it working in HomeAssistant, it fails at the
GoSungrow mqtt runstage with the error stack below, which I got by running manually the docker entrypoint script inside the docker container. The MQTT integration works fine with the same parameters (host, user, pass) that are used by the GoSungrow add-on.I have no experience with golang, so it's a bit tricky to even attempt to debug the panic error below - it sounds like it's a variable that's supposed to have a non-nil value and it's nil, but I'm not quite sure which one is that. The panic seems to be triggered by a call to the GetValue function defined here: (https://github.com/MickMake/GoSungrow/blob/master/iSolarCloud/api/GoStruct/struct_reflect.go#L984) called from somewhere around here: (https://github.com/MickMake/GoSungrow/blob/master/iSolarCloud/AppService/queryDeviceList/data.go#L427) but the line in data.go is not the one that's in the error stack, since there is no GetValue call on line 434 in the current commit of the repo. This raises the question of which exactly is the commit of GoSungrow that is used to generate the docker image that is deployed by the add-on and I don't know the answer to that question either. This also means that even if I can add some code to avoid processing nil pointers, I don't know how to generate a new docker image with that code.
Since you mention in the README that the API changes occasionally, it could be a result of such a change. It could also be a result of a badly configured or defective inverter, which doesn't report properly certain values. I noticed that even in iSolarCloud, the "load" values are sometimes absent, sometimes they show up for a while, so the report is very confusing, which is one of the reasons I want to get away from it. Unfortunately, GoSungrow still relies on the data that is reported by the iSolarCloud API, so the GIGO principle will make this data useless.
Can you have a look and see if there's any obvious culprit, so I can at least get some data in HomeAssistant, even if it's missing some values (e.g. the load)? Thanks, Razvan