Possible timeout loading node info from plugin?

20 views
Skip to first unread message

Bradley Atkins

unread,
Nov 7, 2014, 10:51:45 AM11/7/14
to rundeck...@googlegroups.com
Hi

I have a job that calls five other jobs:

I have created an option in the first job called TargetServer.

This is passed down to the child jobs as the node to run the job on.

This works fine when I have a file based Resource Model Source to supply the node names etc.



Now I have created a plugin to read the node names from foreman and I am using that to list my nodes.

If I go into my nodes page in the UI I see the nodes listed, (after a few seconds while the plug-in interrogates foreman).

But when I run my job it fails:

[workflow] Begin execution: rundeck-workflow-node-first context: null
15:35:56 com.dtolabs.rundeck.core.NodesetEmptyException: No matched nodes: MultiNodeSelector{nodenames=[t30-test-node1.odd.xxx.com]}
15:35:56 at com.dtolabs.rundeck.core.execution.workflow.NodeFirstWorkflowStrategy.validateNodeSet(NodeFirstWorkflowStrategy.java:332)
15:35:56 at com.dtolabs.rundeck.core.execution.workflow.NodeFirstWorkflowStrategy.executeWorkflowImpl(NodeFirstWorkflowStrategy.java:84)
15:35:56 at com.dtolabs.rundeck.core.execution.workflow.BaseWorkflowStrategy.executeWorkflow(BaseWorkflowStrategy.java:120)
15:35:56 at com.dtolabs.rundeck.core.execution.WorkflowExecutionServiceThread.run(WorkflowExecutionServiceThread.java:55)
15:35:56 Exception: class com.dtolabs.rundeck.core.NodesetEmptyException: No matched nodes: MultiNodeSelector{nodenames=[t30-test-node1.odd.xxx.com]}
15:35:56 [workflow] Finish execution:  rundeck-workflow-node-first: [Workflow : exception: com.dtolabs.rundeck.core.NodesetEmptyException: No matched nodes: MultiNodeSelector{nodenames=[t30-test-node1.odd.xxx.com]}]
15:35:56 No matched nodes: MultiNodeSelector{nodenames=[t30-test-node1.odd.xxx.com]}




If I go from the error report to my nodes they are blank and take a second or two to reload.

Do you think this is failing because the plugin takes a few seconds to retrieve the YAML from Foreman and parse it when executing each child job?

If so is there some timeout or option I can set?

Thanks

Brad 

Reply all
Reply to author
Forward
0 new messages