With manual polling, how do I update the optimiser entities?

[mtedds]

I have two inverters (cascaded) battery on each and 8 optimisers.
Having seen lots of issues with sensors going unavailable very frequently, I have turned off the automatic polling and created a one minute manual polling automation which includes an optimiser refresh every 5 minutes.
The optimiser refresh doesn't work. I have tried both of these:

- target:
  entity_id: sensor.jv22a9092609_output_power
  action: homeassistant.update_entity
- delay: "00:00:10"
- target:
  entity_id: sensor.master_inverter_optimizers_online
  action: homeassistant.update_entity

but neither seem to work.

The first sensor is the output power from one of the optimisers.
The second sensor is the number of actual optimisers online from the master inverter.

Anyone know which is the correct sensor to use?

Thanks!

[mohammadrezwankhan]

sensor.jv22a9092609_output_power is the correct type of entity to update. sensor.master_inverter_optimizers_online is not a trigger for the individual optimizer data.

The reason is that the integration uses two different coordinators:

• master_inverter_optimizers_online is read by the normal inverter-register coordinator.
• Each optimizer's output_power, voltage, current, temperature, etc. is attached to a separate optimizer-history coordinator.

Home Assistant's homeassistant.update_entity service requests a refresh of the coordinator behind the entity. Therefore, one enabled optimizer-detail entity is enough to refresh all optimizer-detail entities that share that coordinator; you do not need to call all eight sensors.

For example:

action: homeassistant.update_entity
target:
  entity_id: sensor.jv22a9092609_output_power

Because you have two cascaded inverters, there is one extra detail to check: the integration creates an optimizer coordinator per inverter. If both inverters have optimizer entities, target one optimizer-detail entity belonging to each inverter. If all eight optimizers are behind the same inverter, the single entity above is sufficient.

Calling master_inverter_optimizers_online ten seconds later only refreshes the normal inverter data; it does not cause another optimizer-history refresh, so you can remove that call unless you separately want the online-count register refreshed.

Also, a manual refresh cannot force the inverter to generate a new optimizer-history file. The integration intentionally uses a five-minute optimizer interval, and polling faster will not give finer data. An unchanged power value may simply mean that the latest file/value is unchanged.

To distinguish that from a failed refresh, run the action once from Developer Tools and check the Home Assistant log for the <serial>_optimizer_data_update_coordinator. The integration reports optimizer timeouts, interrupted connections, decode failures, and other Huawei Solar errors there. If you see timeouts or ServerDeviceBusy on the cascaded connection, offset the optimizer action from the one-minute inverter/battery actions so they do not start together, then include the exact log message in this Discussion.

Relevant source: optimizer coordinator, optimizer detail entity binding, and five-minute data-resolution note.