Repository navigation
Replies: 4 comments 2 replies
|
Makes sense as a request. Portainer-managed stacks don't respond well to the plain Docker recreate action since Portainer owns the stack definition, so a first-class integration (or update-capable command actions with the right context passed in) is the right shape. It's on the roadmap now. In the meantime the command action can drive Portainer's API with a webhook or a script if you want something working today. |
2 replies
|
Saw it, thanks. It's going in through your PR on dev/v1.8 so the attribution stays with you. Notes are on #941. |
0 replies
|
Amazing! Thank you for your work on this, and for picking this up so quickly. Its marvelous that we can enjoy live in this age of AI, agentic workflows are a blessing. |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
I am experimenting with using Drydock to manage updates for Portainer-managed Compose stacks.
The use case:
image: pihole/pihole:${PIHOLE_TAG:-2026.05.0}image: gitlab/gitlab-runner:${GITLAB_RUNNER_TAG:-v19.0.0}Technically, this can already be prototyped with a
commandaction. For example, I have a command action calledcommand.portainerthat:container_jsoncom.docker.compose.project.working_dirdd.portainer.version-var=PIHOLE_TAGThe problem is UX.
Right now, to update a Portainer-managed stack through this userland command action, the workflow is:
command.portainer.Compared to the built-in Docker Compose update action, this is too hidden and too cumbersome for day-to-day use.
I see two possible ways to solve this:
Implement Portainer as a first-class update action provider in Drydock.
This would be the cleanest model. Drydock could understand that the action is a real update action, show the normal update button in the container list, and apply the update through Portainer's API.
I would be willing to work on a PR for this if you think this belongs in Drydock.
Allow selected command actions to declare themselves as update actions.
This could be a fallback if you prefer to keep Portainer-specific logic in userland. For example, a command action could opt into update-action semantics and then appear wherever normal update buttons appear, with the same eligibility/policy behavior as docker/dockercompose update actions.
Some open questions:
mqtt.hassappear with a generic “Run” button next to action triggers in the same Associated Triggers list, or should notification and action triggers be visually separated?For my setup, the ideal outcome would be: Portainer-managed stacks get the same one-click update flow as Docker Compose-managed stacks, but the actual redeploy is performed through Portainer so Portainer remains the owner of the stack.
All reactions