Hi Olga Washburn

is your job run on a slave node ? Does the path /mnt/mesos/sandbox/workspace/olga-test exists on it ?

Hi,

I have exactly the same issue and I can confirm that the job is running on a slave node.  And the path given in the stack trace does exist with all rights enabled for the user that connected over ssh to the slave.

[WeblogicDeploymentPlugin] - The JDK /ec/sw/sygma/java/64/jdk1.7.0_91/jre will be used.
[WeblogicDeploymentPlugin] - Retrieving artifacts recorded [filtered resources on ecs-ear.ear]...
[WeblogicDeploymentPlugin] - the base directory specified [/ec/local/data/jenkins-data/workspace/expert-common-services-current/expert-common-data/ecs-ear/target/] is invalid (doesn't exists or is not a directory or has insufficient privilege). Please check the job configuration
java.lang.RuntimeException: The base directory specified [/ec/local/data/jenkins-data/workspace/expert-common-services-current/expert-common-data/ecs-ear/target/] is invalid (doesn't exists or is not a directory or has insufficient privilege)
{{ at org.jenkinsci.plugins.deploy.weblogic.FreeStyleJobArtifactSelectorImpl.selectArtifactRecorded(FreeStyleJobArtifactSelectorImpl.java:58)}}
{{ at org.jenkinsci.plugins.deploy.weblogic.task.DeploymentTaskServiceImpl.perform(DeploymentTaskServiceImpl.java:129)}}
{{ at org.jenkinsci.plugins.deploy.weblogic.WeblogicDeploymentPlugin.perform(WeblogicDeploymentPlugin.java:305)}}
{{ at hudson.tasks.BuildStepMonitor$1.perform(BuildStepMonitor.java:20)}}

...

ls from the build machine:build@xxxxxxxxx /ec/local/data/jenkins-data/workspace/expert-common-services-current/expert-common-data/ecs-ear/target 15$ ls -al
total 62640
drwxrwxrwx 6 build weblrtd 4096 Aug 8 11:18 .
drwxr-xr-x 5 build weblrtd 4096 Aug 8 11:17 ..
rw-rr- 1 build weblrtd 64118528 Aug 8 11:18 ecs-ear.ear
drwxr-xr-x 4 build weblrtd 4096 Aug 8 11:17 expert-common-data
drwxr-xr-x 2 build weblrtd 4096 Aug 8 11:18 failsafe-reports
drwxr-xr-x 2 build weblrtd 4096 Aug 8 11:18 surefire
drwxr-xr-x 6 build weblrtd 4096 Aug 8 11:17 temp

Master and slave are using a different directory structure, it makes me think that Jenkins' weblogic deployment plugin is trying to read the deployable from the master node instead of the slave.

Hi Patrick Leyman and Olga Washburn

You're right, the plugin use Java File to read which is not the right class when you use node to execute task (because only works with master directory structure). In the future release 4.0 which support deployment from node, this limitation will be fixed.

Regards,

Hi Olga Washburn and Patrick Leyman

I've just released the version 4.0. Let me know if it's ok for you.

Hi Raphael CHAUMIER,

Many thanks! I just tested the new version of the plugin and it worked flawless.

Good news. I close the issue.

Regards