Added flag to invoke doStop on WorkflowRun instead of throwing FlowInterruptedException

[cconnert]

Currently the scmSkip does not work as expected in our multi branch pipeline. The build continues to run and only is properly aborted sometimes. Still aborting the build in the JenkinsUI does work. Thus I added a backwards compatible flag to invoke the doStop on a WorkflowRun instead of throwing an execption.

Testing done

Added new unit test to cover new flag

Submitter checklist

Beta Give feedback

1. Make sure you are opening from a topic/feature/bugfix branch (right side) and not your main branch!
2. Ensure that the pull request title represents the desired changelog entry
3. Please describe what you did
4. Link to relevant issues in GitHub or Jira
5. Link to relevant pull requests, esp. upstream and downstream changes
6. Ensure you have provided tests - that demonstrates feature works or fixes the issue

[zbynek]

Interesting, my experience was the opposite: with doStop some multibranch pipelines kept running, with the exception they stop. Would be nice to have one solution that "just works" instead of a flag. I somehow suspect this may be related to agent type. What kind of agents do you use -- ssh / websocket / ... ?

[cconnert]

Indeed interesting.
Well I use containerized build agents based on the standard jenkins agent container. The agent itself is then managed by the jenkins swarm plugin (https://plugins.jenkins.io/swarm/).
Inside those containers I make use of DockerInDocker (DinD) in order to write run the different stages of the declarative multibranch pipeline.
In regards to the pipeline, I do have a custom checkout step. The scmSkip step is executed directly after doing the checkout. I also tried to create a followup stage but that did not change anything.
In regards to the doStop, I did not test it yet but I strongly guess it will work, as it does within the UI.

I agree, that the flag is not a pretty solution. Just wanted to try if it works... If possible (server is not entierly under my control), I'll install a custom build in order to verify my assumption

[cconnert]

@zbynek did you manage to figure something out? Otherwise it would be great to if consider merging this issue as a workaround. I know that a flag is not to pretty but it is better then nothing

[zbynek]

@cconnert sorry, I forgot about this.

[cconnert]

@zbynek: Currently it seems that the aborts to work. I'll monitor our release pipelines and get back to this issue