Feature request: Add a status entity (e.g. binary_sensor) to allow automatic integration reload

[fx-host-assistant]

Hi,
Sometimes the tunnel stops working until Home Assistant is restarted. Since the Dataplicity integration currently doesn’t expose any entity, it’s not possible to detect this failure from within Home Assistant.

A simple binary_sensor reflecting the tunnel status (on = connected, off or unavailable = disconnected) would allow users to automate recovery. For example, if the sensor stays off or unavailable for 15 minutes, an automation could call homeassistant.reload_config_entry for that entity causing Dataplicity integration to reload and, hopefully, restoring the tunnel without restarting Home Assistant.

This would improve reliability and make failure recovery seamless.

Thank you for considering it

[ivano85]

What about a keepalive? Or heartbit?

[fx-host-assistant]

you can use an external system to monitor, e.g. uptimerobot, but it is something you have to add. Doing it internaly with an exposed entity, makes things more easy.

[riddik14]

same problem for me

[yarmak72]

Yep, it will be very helpful feature

[ivano85]

I've temporarily applied a workaround:

in configuration.yaml:

binary_sensor:
  - platform: rest
    name: Reachable through Dataplicity
    resource: https://<your-public-name>.dataplicity.io
    method: GET
    timeout: 10
    value_template: "{{ not 'Device not Connected' in value }}"
    scan_interval: 60

Then, an automation to reload the integration (easier if you build it through the UI):

alias: Reload Dataplicity
description: ""
triggers:
  - trigger: state
    entity_id:
      - binary_sensor.reachable_through_dataplicity
    from: null
    to: "off"
  - trigger: state
    entity_id:
      - binary_sensor.reachable_through_dataplicity
    from: null
    to: unavailable
    for:
      hours: 0
      minutes: 10
      seconds: 0
conditions: []
actions:
  - action: homeassistant.reload_config_entry
    metadata: {}
    data:
      entry_id: <entry id of Dataplicity integration>
mode: single

You can skip the unavailability check.

I think a similar periodic check could be included in the integration code at least.

[fx-host-assistant]

that is odd (the error says that it needs a target and you must select it)

and this is the reason why I asked for this feature request.

BUT, doing the same action into an automation as you did, it works.

No idea why

[Akszyn]

this should be a default part of this integration

Tymczasowo zastosowałem obejście:

w pliku configuration.yaml:

binary_sensor:

• platform: rest
  name: Reachable through Dataplicity
  resource: https://.dataplicity.io
  method: GET
  timeout: 10
  value_template: "{{ not 'Device not Connected' in value }}"
  scan_interval: 60
  Następnie automatyzacja służąca do ponownego załadowania integracji (łatwiejsze, jeśli zbudujesz ją za pomocą interfejsu użytkownika):

alias: Reload Dataplicity
description: ""
triggers:

• trigger: state
  entity_id:
  • binary_sensor.reachable_through_dataplicity
    from: null
    to: "off"
• trigger: state
  entity_id:
  • binary_sensor.reachable_through_dataplicity
    from: null
    to: unavailable
    for:
    hours: 0
    minutes: 10
    seconds: 0
    conditions: []
    actions:
• action: homeassistant.reload_config_entry
  metadata: {}
  data:
  entry_id:
  mode: single
  Możesz pominąć sprawdzanie niedostępności.

Myślę, że podobną okresową kontrolę można by uwzględnić przynajmniej w kodzie integracyjnym.

So could you add this sensor and mechanism so that it wouldn't be necessary to create the automation manually

[ebagnoli]

Dataplicity ID doesn't work anymore

[ebagnoli]

It will be very nice that the watchdog will be inside the integration.