-
Bug
-
Resolution: Duplicate
-
Major
-
None
-
Platform: All, OS: All
-version 1.221
-windows 2003 r2 server configured as Hudson Master
-suse 10.1 enterprise server configured as slave
-subversion vcs
I have a job configured on my Windows master Hudson to run on my Suse slave box
at 8:30pm. The job always start successfully. However, shortly after 10pm each
night, while my slave build is still going, the Windows master does a workspace
cleanup and deletes the "builds" directory on my slave which of course fails the
build. I can see this by seeing the timestamp on the workspace-cleanup.log file
on the master system in HUDSON_HOME. The text of the file says:
"Scanning <my-suse-slave-machine>
Deleting /hudson/builds/workspace/builds"
... where "/hudson/builds" is the defined FS Root of my slave configuration.
I have searched for a setting to prevent this cleanup but have not found one
yet. It seems that hudson is not recognizing that a slave build is in process.
Note that I have not updated slave.jar since earlier 1.2 releases as I assumed
it has not changed. Thank you.
- duplicates
-
JENKINS-3653 Slave workspaces deleted too often
- Resolved