Recreate container if container definition has changed and restart is true, simplify logic

[va1entin]

Hi team,

this PR is my proposal to fix #816. It changes make_started in podman_container_lib.py so that when a container definition has changed and restart/force_restart are True the container will be recreated and run to reflect the changes rather than restarted without reflecting any changes.

I also removed a call to container.restart later in the same function because the case it covers is sufficiently handled above imho.

Curious as to what you think! 😺 🙏

[sshnaidm]

Sorry, I'm not sure how it solved this case. Can you please provide a simple playbook to illustrate this?
Thanks

[va1entin]

Hi @sshnaidm sure here is a test process:

Playbook ubuntu-test.yml

---
- name: Ubuntu test
  hosts: "cloud"
  tasks:
    - name: Run Ubuntu container
      containers.podman.podman_container:
        name: "ubuntu-test"
        image: docker.io/library/ubuntu:22.04
        detach: true
        hostname: "ubuntu"
        pull: always
        force_restart: true 
        restart_policy: always
        volumes:
          - "/etc/localtime:/etc/localtime:ro"
        command: "sleep infinity"
        state: started

Test change to container with force_restart: true

1. Run ansible-playbook ubuntu-test.yml to create the container with Ubuntu 22.04
2. Change image version from 22.04 to 24.04 and run ansible-playbook ubuntu-test.yml again
3. Container created in step 1 will be restarted but image will still be 22.04

Test change to container with force_restart: false (or no force_restart in playbook)

1. Run ansible-playbook ubuntu-test.yml to create the container with Ubuntu 22.04
2. set force_restart to false, change image version from 22.04 to 24.04 and run ansible-playbook ubuntu-test.yml again
3. Container created in step 1 will be recreated and image will be 24.04

To summarize:
New image version (or other changes to the container) will not be applied as long as the container already exists and force_restart is true

After applying the patch in this PR you should instead see:

1. Run ansible-playbook ubuntu-test.yml to create the container with Ubuntu 22.04
2. Change image version from 22.04 to 24.04 and run ansible-playbook ubuntu-test.yml again
3. Container created in step 1 will be recreated and image will be 24.04

This PR allows users to keep force_restart: true in their container definition to have the container be restarted when there are no changes while also recreating the container if there are changes. Currently, only the first works, effectively forcing users to either not use force_restart, change it to false every time they're making any change to their container (and then change it back to true) or add a separate step just for restarting to the playbook.

This is also true if recreate: true by the way. So with recreate: true the user is actually not getting a recreate as long as force_restart: true even when they change container attributes like image version.

[Crapshit]

Thank you for that pull request @va1entin
I'm looking forward for a final solution.
Thanks

[va1entin]

Hi @sshnaidm do you have any feedback on the changes?