Feature Request: Reload from cloud on failure or on demand

[davethetallguy]

I normally operate my system with prefer_cache = true.

Twice in the last few months, I've made a change in a Kumo level system parameter (using my MHK controller) and the hass-kumo integration has subsequently stopped working.

By going into the core configuration, I edited (via VIM) "prefer_cache": from true to false, saved the file, rebooted home assistant, and the integration reloaded and immediately connected successfully. I couldn't find any other way to change this preference than editing the core file.

While this works, it might be nice to either have a default timeout on "unavailable" status (1 hour?) or even a selection to "use kumo cloud on next boot" in order to force this without diving into a config file that isn't normally intended for meat-sacks to be editing directly.

Thanks as always for a solid integration.

[dlarrick]

The v3 code does attempt to contact the Kumo servers at startup and fetch new values, even if prefer_cache is True. It uses values from the cache if they can't be fetched from the servers.

There could still be improvement here to trigger a reload on error or on demand, which might remove the need to re-check the servers on startup (and the resulting delay). This is not an itch I currently feel the need to scratch myself, but I'd accept patches.