-
Bug
-
Resolution: Not A Defect
-
Major
-
None
If you have a slave node that goes offline when idle that holds the workspace of a build, any GitHub push automatically triggers a build with the message:
Started on May 24, 2013 8:15:43 PM
Workspace is offline.
Scheduling a new build to get a workspace. (nonexisting_workspace)
Done. Took 1 ms
Changes found
Two issues here:
- If the job is triggered by a "Test Hook" call from GitHub and the slave is offline, it was always make a build job even if the there isn't any changes.
- (unconfirmed but very likely) The job would build even if the branch it is supposed to build isn't the one that was updated because of the "nonexisting_workspace" error.
This could be resolved if all the tracking took place on the master node.
Thank you for your consideration.
- depends on
-
-
- Open
-
[JENKINS-18079] GitHub Push always builds if workspace is on offline slave node
Joseph Lawson
created issue -
Joseph Lawson
made changes -
| Description |
Original:
If you have a slave node that goes offline when idle that holds the workspace of a build, any GitHub push automatically triggers a build with the message: {quote} Started on May 24, 2013 8:15:43 PM Workspace is offline. Scheduling a new build to get a workspace. (nonexisting_workspace) Done. Took 1 ms Changes found {quote} Two issues here: # If the job is triggered by a "Test Hook" call from GitHub and the slave is offline, it was always make a build job even if the there isn't any changes. # (unconfirmed but very likely) The job would build even if the branch it is supposed to build isn't the one that was updated because of the "nonexisting_workspace" error. Thank you for your consideration. I would like to propose that the status of the |
New:
If you have a slave node that goes offline when idle that holds the workspace of a build, any GitHub push automatically triggers a build with the message: {quote} Started on May 24, 2013 8:15:43 PM Workspace is offline. Scheduling a new build to get a workspace. (nonexisting_workspace) Done. Took 1 ms Changes found {quote} Two issues here: # If the job is triggered by a "Test Hook" call from GitHub and the slave is offline, it was always make a build job even if the there isn't any changes. # (unconfirmed but very likely) The job would build even if the branch it is supposed to build isn't the one that was updated because of the "nonexisting_workspace" error. Thank you for your consideration. |
Joseph Lawson
made changes -
| Description |
Original:
If you have a slave node that goes offline when idle that holds the workspace of a build, any GitHub push automatically triggers a build with the message: {quote} Started on May 24, 2013 8:15:43 PM Workspace is offline. Scheduling a new build to get a workspace. (nonexisting_workspace) Done. Took 1 ms Changes found {quote} Two issues here: # If the job is triggered by a "Test Hook" call from GitHub and the slave is offline, it was always make a build job even if the there isn't any changes. # (unconfirmed but very likely) The job would build even if the branch it is supposed to build isn't the one that was updated because of the "nonexisting_workspace" error. Thank you for your consideration. |
New:
If you have a slave node that goes offline when idle that holds the workspace of a build, any GitHub push automatically triggers a build with the message: {quote} Started on May 24, 2013 8:15:43 PM Workspace is offline. Scheduling a new build to get a workspace. (nonexisting_workspace) Done. Took 1 ms Changes found {quote} Two issues here: # If the job is triggered by a "Test Hook" call from GitHub and the slave is offline, it was always make a build job even if the there isn't any changes. # (unconfirmed but very likely) The job would build even if the branch it is supposed to build isn't the one that was updated because of the "nonexisting_workspace" error. This could be resolved if all the tracking took place on the master node. Thank you for your consideration. |
Probably should be fixed with migration to notifyCommit (TODO)
| Assignee | New: Kanstantsin Shautsou [ integer ] |
| Link | New: This issue is related to JENKINS-30345 [ JENKINS-30345 ] |
| Link | New: This issue depends on JENKINS-30345 [ JENKINS-30345 ] |
| Link | Original: This issue is related to JENKINS-30345 [ JENKINS-30345 ] |
Two years after is still reproducing. Is there any possibility to look at this bug? I'm working with docker-plugin starting slaves and is an ugly thing to launch a build of the same revision anytime a push is done to other branches.