[bug]: rerunning of master pipeline after dev release via release-manager

[SimonGolms]

Describe the bug
After a release with the release manager to the dev environment, the generated release branch is merged back into the master branch and pushed. This push activates the Bitbucket webhook and triggers the Jenkins master pipeline again for a rerun.

To Reproduce
Steps to reproduce the behavior:

1. Go To your release-manager pipeline and start a new build with parameter environment: dev
2. After finishing the release, look to your master pipeline and see it running again.

Expected behavior
No unnecessary re-running of the master pipeline

Screenshots
N/A

Affected version (please complete the following information):

• OpenShift: 3.11
• OpenDevStack >=3

Log Output (ensure to remove any confidential information like tokens, project names, etc.
N/A

Additional context
The keyword [ci skip] is in one of the commits, but the check of whether the pipeline should be started or skipped only considers the title of the merges commit (Merge branch release/20211207.003 vs Pull Request 672: docs update readme [ci skip]) in which the keyword is missing.

[SimonGolms]

@metmajer & @clemensutschig

[clemensutschig]

@michaelsauter @serverhorror - can you take a look this must be in the CI skip routine :(

https://github.com/opendevstack/ods-jenkins-shared-library/blob/master/src/org/ods/services/GitService.groovy#L140

Or we add a merge message that contains the [ci skip] in

https://github.com/opendevstack/ods-jenkins-shared-library/blob/master/src/org/ods/services/GitService.groovy#L326

or @SimonGolms - create a PR :)

[SimonGolms]

isCiSkipInCommitMessage should not be edited, because [ci skip] should only be respected in the subject of the commit message! Otherwise there will be problems if Bitbucket automatically summarizes the previous commit messages during a PR and a [ci skip] appears there!

or @SimonGolms - create a PR :)

Unfortunately, it's not obvious to me what the preferred merge strategy is and if it causes any other side effects, so I'd rather leave it to the experts 😉

[clemensutschig]

Ok we'll fix it on master and 4.x with a merge commit with a message that contains [ci skip].

You are on 4.x right?

[SimonGolms]

Not yet, this also still affects 3.x. I don't know if this was the case in the past as well, since we hadn't done any releases in the last 3 months.

[clemensutschig]

@SimonGolms - assuming you tal about the re-integration to master .. here is the PR at work :)

redish: earlier
yellow: now

[clemensutschig]

@SimonGolms is this a must fix for 3.x?

[SimonGolms]

In regard to the triggering of unwanted side effects (murphy's law) and the unnecessary resource consumption and blocking of the system I would see it as such.