What to do in case of api.zonneplan unreachable?

[Paul1865]

RecentIy had two http response code 500 (service unavailable) when the zonneplan integration was trying to get data from zonneplan.
Result: lots of errors, automations broken etc because the zonneplan current electricity tariff was unavailable.

Is it possible to not touch that sensor (so leave it as it was) when getting a response code >= 400?

I hope that way my derived sensors and automations stay intact.

Trying in the meantime to create a persistant sensor and restore the zonneplan sensor with the data of the persistant one if it's unavailable. No luck so far :D .

[Paul1865]

Best thing I can think of now: a new sensor that only updates when the zonneplan current electricity tariff is available and replace all occurences of the original with this new one...

  - name: "Zonneplan current electricity tariff (last valid)"
    unique_id: zonneplan_current_electricity_tariff_last_valid
    availability: >
      {{ has_value('sensor.zonneplan_current_electricity_tariff') }}
    state: "{{ states('sensor.zonneplan_current_electricity_tariff') }}"
    state_class: measurement
    attributes:
      forecast: >
        {{ state_attr('sensor.zonneplan_current_electricity_tariff', 'forecast') }}
      source: >
        zonneplan_my.yaml
    unit_of_measurement: €
    icon: mdi:currency-eur

[fsaris]

Hi,

just ignoring all that's higher than status 400 isn't possible.

A response with HTTPStatus.UNAUTHORIZED (401) is handled different as this triggers the re-authentication flow.

If I read the HA code correctly the current logic of HA is that on a first error the data isn't processed but if the next update attempt also fails the sensors go to unavailable.

Do you see this behavior often? That the api returns 500 errors?

[Paul1865]

No, happened twice and didn't see it any more since then. On Apr 13, 2024 14:27, Frans Saris ***@***.***> wrote: Hi, just ignoring all that's higher than status 400 isn't possible. A response with HTTPStatus.UNAUTHORIZED (401) is handled different as this triggers the re-authentication flow. If I read the HA code correctly the current logic of HA is that on a first error the data isn't processed but if the next update attempt also fails the sensors go to unavailable. Do you see this behavior often? That the api returns 500 errors? —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you authored the thread.Message ID: ***@***.***>