supervisorctl: stop process before removing it

[DeLoWaN]

SUMMARY

Stop the processes before removing them.

This is needed to create a idempotent playbook. You should not have to explicitly stop a supervisor process before removing it. If you would, then the stop action on all subsequent calls will fail since the process does not exists on supervisor since you removed it at the first call.

It's like any process manager. If you apt remove <some-daemon>, you don't have to stop explicitly stop the deamon before.

This way, even with a running process, you just have to mark it absent to stop/remove it.

ISSUE TYPE

• Feature Pull Request

COMPONENT NAME

supervisorctl

ADDITIONAL INFORMATION

Before:

TASK [Delete supervisor process] *************************************************************************************************************************************************************
fatal: [localhost]: FAILED! => changed=false 
  msg: |-
    ERROR: process/group still running: foo
  name: foo
  state: absent

Now:

TASK [Delete supervisor process] *************************************************************************************************************************************************************
changed: [localhost]

[ansibullbot]

cc @inetfuture @mattupstate
click here for bot help

[felixfontein]

Thanks for your contribution! Could you please add a changelog fragment? Thanks.

[DeLoWaN]

Thanks for your contribution! Could you please add a changelog fragment? Thanks.

Sure thing.

[felixfontein]

Generally, this is a breaking change. The module explicitly documented what it does when state=absent is provided, and you are changing the behavior in that case.

I think it's OK to make this configurable (add a boolean option stop_before_removing or something like that, with default value false, and only stop the processes if the user supplies true.) Does that sound good to you?

[DeLoWaN]

Generally, this is a breaking change. The module explicitly documented what it does when state=absent is provided, and you are changing the behavior in that case.

I think it's OK to make this configurable (add a boolean option stop_before_removing or something like that, with default value false, and only stop the processes if the user supplies true.) Does that sound good to you?

Yes I agree. I introduced stop_before_removing as you suggested, but I left the fragment as minor_change since it keeps now a backward compatibility.

[felixfontein]

Looks good to me. If nobody objects, I'll merge this in a week.

[patchback]

Backport to stable-7: 💚 backport PR created

✅ Backport PR branch: patchback/backports/stable-7/b88b04593fac7a41c58640b8e4d7ad360b2aeaf8/pr-7284

Backported as #7325

🤖 @patchback
I'm built with octomachinery and
my source is open — https://github.com/sanitizers/patchback-github-app.

[felixfontein]

@DeLoWaN thanks for your contribution!