Config entry from integration weatherflow_forecast has an invalid unique_id of type int

[dhover]

Describe the bug

Error in logfile:

2024-12-01 14:11:10.081 ERROR (MainThread) [homeassistant.config_entries] Config entry 'xxx' from integration weatherflow_forecast has an invalid unique_id '12345' of type int when a string is expected, please create a bug report at https://github.com/briis/weatherflow_forecast/issues

To Reproduce
Restart integration

Expected behavior
No errors

[StarGateGit]

It seems to have lost some entities

I did edit all the not available ones out...sorry, I didn't make a note of what changed!

[mstovenour]

I also had this message during initial startup when installing WeatherFlow on a new home assistant supervised installation:

2024-12-23 10:40:15.042 INFO (SyncWorker_3) [homeassistant.util.package] Attempting install of pyweatherflow-forecast==1.1.0
2024-12-23 10:42:21.092 ERROR (MainThread) [homeassistant.config_entries] Config entry 'Rottino Dr' from integration weatherflow_forecast has an invalid unique_id '146439' of type int when a string is expected, please create a bug report at https://github.com/briis/weatherflow_forecast/issues
2024-12-23 10:42:21.092 ERROR (MainThread) [homeassistant.config_entries] Config entry 'Rottino Dr' from integration weatherflow_forecast has an invalid unique_id '146439' of type int when a string is expected, please create a bug report at https://github.com/briis/weatherflow_forecast/issues

[briis]

I have seen this on a few occasions, also with other integrations. I do suspect it is some kind of caching issue with Home Assistant, or some kind of network issue, and there is not anything I can do.
I can guarantee that the module it is trying to pull does exist and has done so all the time.

[Erik-MG]

After updating 1.0.12 I'm still getting the same error message:

2025-01-09 13:27:17.513 ERROR (MainThread) [homeassistant.config_entries] Config entry 'xxx' from integration weatherflow_forecast has an invalid unique_id '135295' of type int when a string is expected, please create a bug report at https://github.com/briis/weatherflow_forecast/issues

I'm running Home Assistant 2024.12.5 in a Docker container, HACS 2.0.1 and Weatherflow integration 1.0.12.

[briis]

When setting up your station from scratch, I don't get this error.
My best assumption is that, even though you upgraded to the latest version, the Station ID is still stored as an Integer, and not a string.
Unless you have modified a lot of sensor names, please try and delete your station, and make a new installation, and let me know if you still see the error.

[daviddesmet]

Is this something we can update at the DB level or something without deleting and creating a new station?

[briis]

Have you updated to 1.0.13?
It should finally be fixed in that release

[daviddesmet]

Yes, I'm on 1.0.13 and I've just rebooted the HA server and still get the log message when it boots:

2025-01-16 15:21:08.053 ERROR (MainThread) [homeassistant.config_entries] Config entry 'Home' from integration weatherflow_forecast has an invalid unique_id '123456' of type int when a string is expected, please create a bug report at https://github.com/briis/weatherflow_forecast/issues

[Erik-MG]

I just updated Weatherflow integration to 1.0.13 and I can confirm I also still get the 'invalid unique_id' error message.

I'm hesitant to delete the station and re-create it again. The station name is also linked to my Weatherflow company account. There is also a risk of ending up with lots of duplicate entities which I have to correct manually. Will I loose the link to my history on influxdb?

[briis]

I use the Station ID as unique key, and that has worked a few years now, but it seems HA will not accept this anymore. So unfortunately there is only one thing to do, and this to make a new unique id that contains string characters also. But that also means that you most likely will have to remove and re-add the integration after upgrading.
Should be no big deal if you have not modified anything on the entities, but else you will have to redo that.

[mwareman]

Seeing this on my instance as well while troubleshooting Home Assistant crashing/restarting. I don't think the root cause is this integration yet - but I did come across it.
Everything 100% up to date non-beta.