Add publish container job depending on preparing artifacts

[ppatierno]

Looking at this operator release pipeline build https://dev.azure.com/cncf/strimzi/_build/results?buildId=186729&view=results I noticed that the first stage (prepare_release_artifacts) failed but it anyway ran the containers_publish one.
I guess it's because the containers_publish depends only on containers_publish_with_suffix. But the last one can be skipped during an RC so the containers_publish runs anyway.
This PR adds the dependency on the prepare_release_artifacts as well.

[scholzj]

Does that work with the conditions?

[ppatierno]

Does that work with the conditions?

Good catch. I add the condition that prepare artifacts has to be successful.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 74.63%. Comparing base (d311680) to head (1e5b894).
⚠️ Report is 21 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##               main   #11967   +/-   ##
=========================================
  Coverage     74.63%   74.63%           
- Complexity     6397     6398    +1     
=========================================
  Files           342      342           
  Lines         24238    24238           
  Branches       3194     3194           
=========================================
+ Hits          18089    18090    +1     
  Misses         4877     4877           
+ Partials       1272     1271    -1     

see 4 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[scholzj]

Does that work with the conditions?

Good catch. I add the condition that prepare artifacts has to be successful.

And does it work? Did you test it? I was doing the condition like 5 times before I got it into a working state.

[im-konge]

So as we want to migrate to GHA, should we close this PR? We lived without that condition even before, so maybe we should switch our focus to GHA instead of fixing something that will not be needed for 0.49.0 and on?

[ppatierno]

I agree. Going to close this one, it could be not worth fixing it.