Manually refreshing template update entity after finished update

Configuration
, , ,
1

So I’m currently in the process of setting up templated update entities to remotely update my Docker containers since doing that manually is annoying enough with an increasing number of containers.

This is what I currently have:

- triggers:
    - trigger: "time_pattern"
      minutes: "/5"
  actions:
    - service: "shell_command.get_docker_version"
      data:
        compose_path: "~/docker/portainer"
        container_name: "portainer"
      response_variable: result
  update:
    - name: "Docker: Portainer"
      unique_id: "85fcc27f-7e93-4b22-83e7-fe4143fc511e"
      installed_version: "{{ result.stdout | default('unknown') }}"
      latest_version: "{{ states('sensor.docker_portainer_latest') }}"
      release_url: "{{ state_attr('sensor.docker_portainer_latest', 'html_url') }}"
      title: "{{ state_attr('sensor.docker_portainer_latest', 'name') }}"
      release_summary: >
        {% set body = state_attr('sensor.docker_portainer_latest', 'body') %}
        {{ body | default("No description available.") }}
      availability: |-
        {{
            result is defined and result.stdout is defined
            and has_value("sensor.docker_portainer_latest")
        }}
      install:
        action: "script.update_docker_container"
        data:
          update_entity: "{{ this.entity_id }}"
          compose_path: "~/docker/portainer"
          to_version: "{{ states('sensor.docker_portainer_latest') }}"
          from_version: "{{ result.stdout }}"

It basically works with getting and matching the versions and actually doing the update. What I’m currently struggling with is getting the update entity to switch to OFF after doing the update. I have no way of knowing when it’s complete except for retrieving the installed version again.

When trying to use homeassistant.update_entity, I only get the error NotImplementedError: Update method not implemented.

What’s a good way to fix that? I refrained from setting the time pattern to low since the scripts I call rely on SSH calls. Not sure if that would even be a problem setting the pattern to every 30 seconds or so?

Bonus question:
I’ve noticed that the release summary I get for the update is cut off. It cuts off just after “It’s not possible to” with the full being

## Known issues
- On Async Edge environments, an invalid update schedule date can be displayed when browsing a snapshot

### Known issues with Podman support
- Podman environments aren't supported by auto-onboarding script
- It's not possible to add Podman environments via socket, when running a Portainer server on Docker (and vice versa)
- Support for only CentOS 9, Podman 5 rootful

## Changes:
- Fixed an issue where the environment status was not updating in a timely fashion for Standard Edge Agent
- Fixed an issue where the `--tlscert` and `--tlskey` CLI options did not work unless `--sslcert` and `--sslkey` were also provided
- Fixed an issue where Edge Stacks with GitOps enabled were not being updated correctly
- Fixed an issue where the container engine defaulted to Docker when associating the first Kubernetes environment
- Resolved the following CVEs:
    - CVE-2025-22871
    - CVE-2025-22868
    - CVE-2025-22869
    - CVE-2025-4673
    - CVE-2024-45341
    - CVE-2024-45336
    - CVE-2025-0913
    - CVE-2024-45338
    - CVE-2025-22872
    - CVE-2024-40635
    - CVE-2025-22870
    - CVE-2025-22866
    - CVE-2025-54410
    - GHSA-2464-8j7c-4cjm

## Deprecated and removed features
**Deprecated features**
- None

**Removed features**
- None
2

put another action that’s a custom event that triggers the template sensor to update.

i.e.

- triggers:
    - trigger: "time_pattern"
      minutes: "/5"
    - trigger: event
      event_type: portainer_updated
  actions:
    - service: "shell_command.get_docker_version"
      data:
        compose_path: "~/docker/portainer"
        container_name: "portainer"
      response_variable: result
  update:
    - name: "Docker: Portainer"
      unique_id: "85fcc27f-7e93-4b22-83e7-fe4143fc511e"
      installed_version: "{{ result.stdout | default('unknown') }}"
      latest_version: "{{ states('sensor.docker_portainer_latest') }}"
      release_url: "{{ state_attr('sensor.docker_portainer_latest', 'html_url') }}"
      title: "{{ state_attr('sensor.docker_portainer_latest', 'name') }}"
      release_summary: >
        {% set body = state_attr('sensor.docker_portainer_latest', 'body') %}
        {{ body | default("No description available.") }}
      availability: |-
        {{
            result is defined and result.stdout is defined
            and has_value("sensor.docker_portainer_latest")
        }}
      install:
      - action: "script.update_docker_container"
        data:
          update_entity: "{{ this.entity_id }}"
          compose_path: "~/docker/portainer"
          to_version: "{{ states('sensor.docker_portainer_latest') }}"
          from_version: "{{ result.stdout }}"
      - event: portainer_updated

add a delay if needed.

Or if you really want to get creative, add a while loop where you check until it’s actually updated and then fire the custom event.

You have many options.

1 Like
3

Awesome that worked, thanks!

Do you perhaps know anything about the “bonus question” I edited in earlier with the release summary being cut off?

Fwiw the raw text I get from the REST call is: ## Known issues\r\n\r\n- On Async Edge environments, an invalid update schedule date can be displayed when browsing a snapshot\r\n\r\n### Known issues with Podman support\r\n\r\n- Podman environments aren't supported by auto-onboarding script\r\n- It's not possible to add Podman environments via socket, when running a Portainer server on Docker (and vice versa)\r\n- Support for only CentOS 9, Podman 5 rootful\r\n\r\n## Changes:\r\n\r\n- Fixed an issue where the environment status was not updating in a timely fashion for Standard Edge Agent\r\n- Fixed an issue where the –tlscertand–tlskeyCLI options did not work unless–sslcertand–sslkey were also provided\r\n- Fixed an issue where Edge Stacks with GitOps enabled were not being updated correctly\r\n- Fixed an issue where the container engine defaulted to Docker when associating the first Kubernetes environment\r\n- Resolved the following CVEs:\r\n - CVE-2025-22871\r\n - CVE-2025-22868\r\n - CVE-2025-22869\r\n - CVE-2025-4673\r\n - CVE-2024-45341\r\n - CVE-2024-45336\r\n - CVE-2025-0913\r\n - CVE-2024-45338\r\n - CVE-2025-22872\r\n - CVE-2024-40635\r\n - CVE-2025-22870\r\n - CVE-2025-22866\r\n - CVE-2025-54410\r\n - GHSA-2464-8j7c-4cjm\r\n\r\n## Deprecated and removed features\r\n\r\n**Deprecated features**\r\n\r\n- None\r\n\r\n**Removed features**\r\n\r\n- None

But the REST sensor where I fetch it from has the full text when I look in the developers tools