Uploaded image for project: 'Jenkins'
  1. Jenkins
  2. JENKINS-32984

Reload Configuration from disk disrupts build elements and pipeline

    XMLWordPrintable

Details

    • Bug
    • Status: Open (View Workflow)
    • Minor
    • Resolution: Unresolved
    • core
    • None
    • Jenkins ver. 1.642.1
      build-pipeline-plugin 1.4.9
      swarm 2.0
      nodelabelparameter 1.7.1
      hipchat 1.0.0

    Description

      Creating 2 new build jobs one dependent on the other by "trigger parameterized build on other projects" or "Build after other projects are built". This results in the relationship being observed here:

      Trigger the upstream build (in our case takes hours to complete) and observe the downstream relationship within the pipeline view:

      Now reload configuration from disk, and reload the build pipeline view:

      The resulting effects are as follows:
      Hipchat notification for upstream job (start) is sent
      CLI Exec on swarm slave runs and completes
      Hipchat notification for upstream job (complete) is never sent
      Downstream build job is never executed.

      Attachments

        Issue Links

          Activity

            james_powis James Powis added a comment - - edited

            Additional information: I am using puppet (and the jenkinssci/puppet-jenkins) for administration of Jenkins build jobs and service users, due to a heavy handed approach of reloading configuration from disk after the execution of jenkins_cli commands, and the non-idempotent user function, reload configuration from disk occurs at every puppet run. I have filed a bug report with them: https://github.com/jenkinsci/puppet-jenkins/issues/501 so that the method being used can be refined.

            james_powis James Powis added a comment - - edited Additional information: I am using puppet (and the jenkinssci/puppet-jenkins) for administration of Jenkins build jobs and service users, due to a heavy handed approach of reloading configuration from disk after the execution of jenkins_cli commands, and the non-idempotent user function, reload configuration from disk occurs at every puppet run. I have filed a bug report with them: https://github.com/jenkinsci/puppet-jenkins/issues/501 so that the method being used can be refined.
            jglick Jesse Glick added a comment -

            Reload Configuration from Disk has always been pretty buggy; it is best to avoid it when at all possible. You can always reload individual jobs, which is somewhat safer.

            jglick Jesse Glick added a comment - Reload Configuration from Disk has always been pretty buggy; it is best to avoid it when at all possible. You can always reload individual jobs, which is somewhat safer.

            People

              Unassigned Unassigned
              james_powis James Powis
              Votes:
              0 Vote for this issue
              Watchers:
              4 Start watching this issue

              Dates

                Created:
                Updated: