How to configure retry of the service call if failed?

[eugenet8k]

I have an essential flow where I want to turn on my heater in the morning

The problem is that the NEST thermostat could be unstable sometimes (offline for few secs), so occasionally the set_temperature service call results in an error in the HomeAssistant log:

[4449896000] Error from API: 400: FAILED_PRECONDITION: Device is offline.: Bad Request
Traceback (most recent call last):
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/google_nest_sdm/auth.py", line 80, in _raise_for_status
    resp.raise_for_status()
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/aiohttp/client_reqrep.py", line 1000, in raise_for_status
    raise ClientResponseError(
aiohttp.client_exceptions.ClientResponseError: 400, message='Bad Request', url=URL('https://smartdevicemanagement.googleapis.com/v1/enterprises/1234556789:executeCommand')

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/homeassistant/components/websocket_api/commands.py", line 141, in handle_call_service
    await hass.services.async_call(
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/homeassistant/core.py", line 1488, in async_call
    task.result()
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/homeassistant/core.py", line 1523, in _execute_service
    await handler.job.target(service_call)
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/homeassistant/helpers/entity_component.py", line 204, in handle_service
    await self.hass.helpers.service.entity_service_call(
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/homeassistant/helpers/service.py", line 642, in entity_service_call
    future.result()  # pop exception if have
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/homeassistant/helpers/entity.py", line 681, in async_request_call
    await coro
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/homeassistant/helpers/service.py", line 679, in _handle_entity_call
    await result
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/homeassistant/components/climate/__init__.py", line 544, in async_service_temperature_set
    await entity.async_set_temperature(**kwargs)
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/homeassistant/components/nest/climate_sdm.py", line 313, in async_set_temperature
    await trait.set_heat(temp)
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/google_nest_sdm/thermostat_traits.py", line 124, in set_heat
    return await self._cmd.execute(data)
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/google_nest_sdm/traits.py", line 23, in execute
    return await self._auth.post(f"{self._device_id}:executeCommand", json=data)
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/google_nest_sdm/auth.py", line 73, in post
    return await AbstractAuth._raise_for_status(resp)
  File "/Users/pi/.homeassistant-venv/lib/python3.8/site-packages/google_nest_sdm/auth.py", line 85, in _raise_for_status
    raise ApiException(": ".join(detail)) from err
google_nest_sdm.exceptions.ApiException: Error from API: 400: FAILED_PRECONDITION: Device is offline.: Bad Request

so I thought the retry strategy would be a possible solution here. But I don't want to just make my flow over-complex, like having a timer for few secs and with attempts to restart the service node. I wonder if there is something like this builtin? Or perhaps an elegant way of solving this fairly generic problem?

[zachowj]

I am guessing HA isn't responding to the service call with an error, it happens after the fact?

Two possible options that come to mind are:

1. to use a tail node and watch the log for the error and if it occurs in the time frame when the set_temperature was made resend it.

2. use a wait-until node after the call-service node to check the temperature of the climate entity with a timeout x seconds and on the timeout circle back to the call-service node

[eugenet8k]

Thanks for the suggestion @zachowj. I think I came up with something that works, but it is not perfect. I created a subflow that I can reuse in different parts of my Node-Red automation.

[{"id":"c9ab060a.5989e","type":"subflow","name":"update state","info":"","category":"","in":[{"x":60,"y":100,"wires":[{"id":"747bb322.8bdc5c"}]}],"out":[],"env":[{"name":"DOMAIN","type":"str","value":""},{"name":"SERVICE","type":"str","value":""},{"name":"ENTITY_ID","type":"str","value":""},{"name":"DATA","type":"str","value":""},{"name":"WAIT_UNTIL","type":"str","value":""},{"name":"WAIT_VALUE","type":"str","value":""}],"color":"#3FADB5","icon":"node-red-contrib-home-assistant-websocket/ha-call-service.svg","status":{"x":940,"y":120,"wires":[{"id":"2c903cee.6ca354","port":0}]}},{"id":"747bb322.8bdc5c","type":"api-call-service","z":"c9ab060a.5989e","name":"try update state","server":"16e7a87a.4bb278","version":1,"debugenabled":false,"service_domain":"${DOMAIN}","service":"${SERVICE}","entityId":"${ENTITY_ID}","data":"${DATA}","dataType":"json","mergecontext":"","output_location":"","output_location_type":"none","mustacheAltTags":false,"x":260,"y":60,"wires":[["b84b27b8.ee4a18"]]},{"id":"2c903cee.6ca354","type":"status","z":"c9ab060a.5989e","name":"","scope":null,"x":760,"y":160,"wires":[[]]},{"id":"b84b27b8.ee4a18","type":"ha-wait-until","z":"c9ab060a.5989e","name":"wait 10 sec until state changed","server":"16e7a87a.4bb278","outputs":2,"entityId":"${ENTITY_ID}","entityIdFilterType":"exact","property":"${WAIT_UNTIL}","comparator":"is","value":"${WAIT_VALUE}","valueType":"re","timeout":"10","timeoutType":"num","timeoutUnits":"seconds","entityLocation":"","entityLocationType":"none","checkCurrentState":true,"blockInputOverrides":true,"x":250,"y":180,"wires":[["1f332575.77a31b"],["feea7690.ce2e88"]]},{"id":"feea7690.ce2e88","type":"counter-loop","z":"c9ab060a.5989e","name":"try 5 times","counter":"il30eda314b0f3e4","counterType":"msg","reset":true,"resetValue":"value-null","initial":0,"initialType":"num","operator":"lt","termination":"5","terminationType":"num","increment":1,"incrementType":"num","x":550,"y":180,"wires":[["1f332575.77a31b"],["747bb322.8bdc5c"]]},{"id":"1f332575.77a31b","type":"api-call-service","z":"c9ab060a.5989e","name":"final attempt to update state","server":"16e7a87a.4bb278","version":1,"debugenabled":false,"service_domain":"${DOMAIN}","service":"${SERVICE}","entityId":"${ENTITY_ID}","data":"${DATA}","dataType":"json","mergecontext":"","output_location":"","output_location_type":"none","mustacheAltTags":false,"x":760,"y":60,"wires":[[]]},{"id":"16e7a87a.4bb278","type":"server","name":"Home Assistant","legacy":false,"addon":false,"rejectUnauthorizedCerts":true,"ha_boolean":"y|yes|true|on|home|open","connectionDelay":true,"cacheJson":true}]

Wich uses an additional plugin (https://github.com/s1r-J/node-red-contrib-loop-processing#counter-loop)

With these ENV props to set

It will run HA "call service node" then HA "wait until node" for 10 secs if the state didn't change then retry it 5 times.

The solution works for a single entity id. In my original question the ENV params to subflow ended up being: