Better Handle a Power Cut when Pro (possibly Mini) is connected

[gedger]

We recently had a power cut but as the house is battery backed the Octopus Pro remained powered but couldn't connect to the meter as it had no power. The integration started to log warnings about this. Fortunately the grid outage was fairly short but if it was for multiple hours there would be a lot of logging. I'm also not sure what gets logged if the Pro losses power?

This is a bit of an edge case but I was going to suggest that the logging could be a bit cleverer. Perhaps the first warning could say can't connect to 'xxx' will try every minute and report when the connection is restored. That way each event would only generate 2 warnings rather than the potential 100's for an extended outage.

Any thoughts?

2024-08-15 11:59:11.589 WARNING (MainThread) [custom_components.octopus_energy.api_client_home_pro] Failed to connect. Timeout of ClientTimeout(total=None, connect=None, sock_read=20, sock_connect=20, ceil_threshold=5) exceeded.
2024-08-15 11:59:11.589 WARNING (MainThread) [custom_components.octopus_energy.coordinators.current_consumption_home_pro] Failed to retrieve current consumption data from home pro data - using cached version. Next attempt at 2024-08-15 11:59:41.682649+01:00
2024-08-15 12:00:11.558 WARNING (MainThread) [custom_components.octopus_energy.api_client_home_pro] Failed to connect. Timeout of ClientTimeout(total=None, connect=None, sock_read=20, sock_connect=20, ceil_threshold=5) exceeded.
2024-08-15 12:00:11.558 WARNING (MainThread) [custom_components.octopus_energy.coordinators.current_consumption_home_pro] Failed to retrieve current consumption data from home pro data - using cached version. Next attempt at 2024-08-15 12:01:41.682649+01:00
2024-08-15 12:06:33.509 WARNING (MainThread) [custom_components.octopus_energy.api_client_home_pro] Failed to connect. Timeout of ClientTimeout(total=None, connect=None, sock_read=20, sock_connect=20, ceil_threshold=5) exceeded.
2024-08-15 12:06:33.509 WARNING (MainThread) [custom_components.octopus_energy.coordinators.current_consumption_home_pro] Failed to retrieve current consumption data from home pro data - using cached version. Next attempt at 2024-08-15 12:07:12.682408+01:00
2024-08-15 12:07:33.506 WARNING (MainThread) [custom_components.octopus_energy.api_client_home_pro] Failed to connect. Timeout of ClientTimeout(total=None, connect=None, sock_read=20, sock_connect=20, ceil_threshold=5) exceeded.
2024-08-15 12:07:33.507 WARNING (MainThread) [custom_components.octopus_energy.coordinators.current_consumption_home_pro] Failed to retrieve current consumption data from home pro data - using cached version. Next attempt at 2024-08-15 12:09:12.682408+01:00

[BottlecapDave]

This is by design to provide information to users on why their sensors might not be updating. The warnings get less frequent as the updates slowly back off if they can't connect. If this was reduced then this information would be lost when connections can not be continually met. The warnings are also raised for more than just no connection (e.g. OE servers are down).

If you're not wanting to see these you can reduce your log level for the integration to error.