Merge to Master not kicking Open PRs to also build

[Jeffrey]

Good Afternoon Jenkins-Land

     Running into a weird issue in version Jenkins 2.53 and 2.57 (Latest plugins). 

Old Version: (I believe 2.45) (Git / SCM plugins previous to version 2.0)

Jenkins Multibranch Pipeline, Freestyle --> Jenkinsfile from ORG folder

1.   We have 5 open PRs. 

2.   We merge one of the PRs to master

3.   Master auto builds (webook)

4.   The remaining 4 PRs also build (merge with new master and build) 

Weird issue: (Versions 2.53, and 2.57) (Git and SCM plugins now post 2.0)

1.   We have 5 open PRs. 

2.   We merge one of the PRs to master

3.   Master auto builds (Webhook)

4.   Nothing happens to open PRs

5.   Re-scan of the ORG or the repository in Jenkins causes the PRs to build. 

Our configurations are stored on a central NFS mount and Jenkins Master run in Docker. 

Any help on where to look for the possible causes? 

[Jeffrey]

So -- Ive reverted Jenkins to 2.46.2 latest patches and the behavior is the same.

What I expect: 

1. Two open PRs
2. Merge one PR to master
3. Master auto builds
4. Repo is scanned and the other PR builds. This is still not happening. Any help much appreciated.

[Baptiste Mathus]

Must be missing something but failing to see why merging a PR should retrigger the builds of others.

Care to clarify?

Thanks

--
You received this message because you are subscribed to the Google Groups "Jenkins Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/d6db8731-d016-4821-b806-ceb637049cac%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Jeffrey]

Hey, Thanks for reading. 

All PRs build -- and post back a status message to GIT delivering build success or failure right? 

2 PRs like before -- both build successfully and show "Clear to ship"

PR1 Gets merged to Master (Which changes code that may or may not work on PR2 )

Unless someone goes back and forces PR 2 to build again -- there is no test of the newly introduced code in master (from PR1) -- you can ship untested / breaking code. 

Any time a merge occurs -- all PRs referencing that head should be forced to rebuild to ensure the merged version works right? 

To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-use...@googlegroups.com.

[Stephen Connolly]

I think this is https://github.com/jenkinsci/github-branch-source-plugin/pull/131 unless you have upgraded to ghbs 2.0.5

To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/25570dc9-af90-47e7-8f1b-d263b418d294%40googlegroups.com.

[Jeffrey]

@2.0.5 GHBS --- everything is patched to latest as of today. Still exhibits the incorrect behavior. Thanks for taking the time to read.