Uploaded image for project: 'Jenkins'
  1. Jenkins
  2. JENKINS-27399

CLONE - Cancel Job and Build Number buttons appear under main feedback panel

    • Icon: Bug Bug
    • Resolution: Unresolved
    • Icon: Major Major
    • heavy-job-plugin
    • None
    • Windows system, running 1.596.1. Happens on IE, Chrome and Firefox browsers

      For some jobs the drop-down for the job number and the cancel job button appear under the main feedback panel. When they appear under the panel it is not possible to use them. In addition we just found out that if you are editing settings for a job the SAVE and APPLY buttons appear under the "Build Executor Status" information and are not selectable either.

          [JENKINS-27399] CLONE - Cancel Job and Build Number buttons appear under main feedback panel

          Brian Smith added a comment -

          If you collapse the build executor status before you start editing the job you can get past the SAVE and APPLY buttons not working problem (doing it after causes you to lose your changes when the page refreshes).

          Brian Smith added a comment - If you collapse the build executor status before you start editing the job you can get past the SAVE and APPLY buttons not working problem (doing it after causes you to lose your changes when the page refreshes).

          Daniel Beck added a comment -

          Which "Other Jira"?

          Daniel Beck added a comment - Which "Other Jira"?

          Brian Smith added a comment -

          Sorry, I guess clone didn't work. I have linked to the other JIRA.

          Brian Smith added a comment - Sorry, I guess clone didn't work. I have linked to the other JIRA.

          Daniel Beck added a comment -

          Are all your job names all caps, and using underscores between words?

          Could you provide a sample job name that shows the problem when a build is running?

          Daniel Beck added a comment - Are all your job names all caps, and using underscores between words? Could you provide a sample job name that shows the problem when a build is running?

          Brian Smith added a comment -

          Yes, our naming convention calls for all caps with underscores between the main sections.

          Sample job names:
          M6_TEST_MAIN_COMMIT_POSCHECK_WAASL1
          M6_TEST_MAIN_COMMIT_POSCHECK_PSRDIFF

          Brian Smith added a comment - Yes, our naming convention calls for all caps with underscores between the main sections. Sample job names: M6_TEST_MAIN_COMMIT_POSCHECK_WAASL1 M6_TEST_MAIN_COMMIT_POSCHECK_PSRDIFF

          Daniel Beck added a comment -

          This looks like the names of the jobs didn't get broken into separate terms on your system.

          Could you use your browser's dev tools to check whether there are <wbr></wbr> tags between parts of the job name in the queue?

          If the AJAX reloading of the queue interferes with that, disconnect the network form your machine or (on a test instance) shut Jenkins down. This should stop replacing the queue DOM elements.


          If you're able to reproduce the problem on a test instance, that would be best. I failed to see this problem on 1.596.1 with the job names you provided.

          Please also make sure there is no custom Javascript running on the instance (e.g. Simple Theme Plugin).

          Daniel Beck added a comment - This looks like the names of the jobs didn't get broken into separate terms on your system. Could you use your browser's dev tools to check whether there are <wbr></wbr> tags between parts of the job name in the queue? If the AJAX reloading of the queue interferes with that, disconnect the network form your machine or (on a test instance) shut Jenkins down. This should stop replacing the queue DOM elements. If you're able to reproduce the problem on a test instance, that would be best. I failed to see this problem on 1.596.1 with the job names you provided. Please also make sure there is no custom Javascript running on the instance (e.g. Simple Theme Plugin).

          Brian Smith added a comment - - edited

          It looks like this problem has gone away after we upgraded some of our plugins (and there is a <wbr> tag pair before each underscore now). Not sure which one it was as none of them would seem to have an effect on the main display (maybe the Claim plugin?). If needed we can go back and see which one it was but with the latest plugins and with the latest LTS release everything seems to be working.

          Please let me know if you would like us to try and track it down.

          Upgraded plugins (from what we can remember upgrading):
          artifact deployer plugin
          build pipeline
          claim
          copy artifact
          copy to slave
          credentials
          disk usage
          email extension
          LDAP
          monitoring
          parameterized trigger
          perforce
          SSH slaves
          subversion
          workspace cleanup

          Brian Smith added a comment - - edited It looks like this problem has gone away after we upgraded some of our plugins (and there is a <wbr> tag pair before each underscore now). Not sure which one it was as none of them would seem to have an effect on the main display (maybe the Claim plugin?). If needed we can go back and see which one it was but with the latest plugins and with the latest LTS release everything seems to be working. Please let me know if you would like us to try and track it down. Upgraded plugins (from what we can remember upgrading): artifact deployer plugin build pipeline claim copy artifact copy to slave credentials disk usage email extension LDAP monitoring parameterized trigger perforce SSH slaves subversion workspace cleanup

          Daniel Beck added a comment -

          Alright, resolving.

          If you have the time, try to figure out what combination of plugins and configuration is needed.

          Daniel Beck added a comment - Alright, resolving. If you have the time, try to figure out what combination of plugins and configuration is needed.

          Brian Smith added a comment - - edited

          OK... turns out I spoke too soon again as it started happening again this morning. I think we managed to track it down as there were only two differences between what ran yesterday when I checked and this morning. We had several matrix jobs running this morning and several jobs using the "Heavy Job" plugin.

          Based on the debug output from the browser (attached) it looks like the Heavy Job plugin might be the culprit as it doesn't have any <wbr> tags.

          Brian Smith added a comment - - edited OK... turns out I spoke too soon again as it started happening again this morning. I think we managed to track it down as there were only two differences between what ran yesterday when I checked and this morning. We had several matrix jobs running this morning and several jobs using the "Heavy Job" plugin. Based on the debug output from the browser (attached) it looks like the Heavy Job plugin might be the culprit as it doesn't have any <wbr> tags.

          Daniel Beck added a comment - Seems to be correct, see: https://github.com/jenkinsci/heavy-job-plugin/blob/master/src/main/resources/hudson/plugins/heavy_job/HeavyJobProperty/ExecutableImpl/executorCell.jelly

            Unassigned Unassigned
            frozen_man Brian Smith
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

              Created:
              Updated: