-
Bug
-
Resolution: Unresolved
-
Major
-
None
-
Windows 7 32bit
Using 1.433 of Jenkins, 1.37 of plugin.
Windows 7 32bit master jenkins install. Windows 7 32bit slave jenkins install, launched by manually running JNLP from the slave. Created job in Jenkins which is associated to only run on the slave. The job is configured with a Mercurial repository. If I configure the job to trigger based on SCM changes, I get this in the Mercurial polling log:
Started on Oct 5, 2011 9:26:37 AM
No workspace is available, so can't check for updates.
Done. Took 3 ms
No changes
The job has previously run fine and the workspace exists on the slave. The job is configured to download the repository with a subdirectory called 'source' in the workspace. The Remote FS Root on the slave is set to:
c:\jenkins\jobs
The workspace for the job exists as:
c:\jenkins\jobs\workspace\<job name>\source
- is duplicated by
-
JENKINS-11220 Mercurial Plugin doesn't support SCM polling on slaves
-
- Resolved
-
-
JENKINS-11221 Mercurial Plugin does not support SCM polling on slaves
-
- Resolved
-
It sounds to me like this is actually an issue with Jenkins core, not the mercurial plugin. When a job is tied to a particular slave and configured to use an SCM that requires a workspace for polling, Jenkins currently assumes that if the workspace doesn't exist, it should just wait for it to exist. While the description of this issue says that the workspace exists, the output ("No workspace is available, so can't check for updates.") seems to indicate that Jenkins thinks the workspace doesn't exist.
My suggestion would be to manually trigger the build by clicking "Build Now", and see if polling works after that.
https://github.com/jenkinsci/jenkins/blob/master/core/src/main/java/hudson/model/AbstractProject.java#L1282