Description
We've seen a timing problem with the SSH action where the callback comes back too fast, before the action has transitioned to RUNNING and is still in PREP. This causes Oozie to ignore the callback, which means it won't find out that the action completed until it manually checks (default=10min). This happened in an HA setup, but I think it could happen even without HA. Adding a 30 second delay into the ssh scripts fixed the problem, but ideally we should come up with a better solution.
Here's the relevant logs:
2015-01-16 18:00:12,916 INFO org.apache.oozie.action.ssh.SshActionExecutor: SERVER[FOO] USER[foo] GROUP[-] TOKEN[] APP[${job_name}] JOB[0000027-150113223634420-oozie-oozi-W] ACTION[0000027-150113223634420-oozie-oozi-W@action-1] start() begins 2015-01-16 18:00:12,917 INFO org.apache.oozie.action.ssh.SshActionExecutor: SERVER[FOO] USER[foo] GROUP[-] TOKEN[] APP[${job_name}] JOB[0000027-150113223634420-oozie-oozi-W] ACTION[0000027-150113223634420-oozie-oozi-W@action-1] Attempting to copy ssh base scripts to remote host [foo@bar.com] 2015-01-16 18:00:15,769 INFO org.apache.oozie.servlet.CallbackServlet: SERVER[FOO] USER[-] GROUP[-] TOKEN[-] APP[-] JOB[0000027-150113223634420-oozie-oozi-W] ACTION[0000027-150113223634420-oozie-oozi-W@action-1] callback for action [0000027-150113223634420-oozie-oozi-W@action-1] 2015-01-16 18:00:15,774 ERROR org.apache.oozie.command.wf.CompletedActionXCommand: SERVER[FOO] USER[-] GROUP[-] TOKEN[] APP[-] JOB[0000027-150113223634420-oozie-oozi-W] ACTION[0000027-150113223634420-oozie-oozi-W@action-1] XException, org.apache.oozie.command.CommandException: E0800: Action it is not running its in [PREP] state, action [0000027-150113223634420-oozie-oozi-W@action-1] at org.apache.oozie.command.wf.CompletedActionXCommand.eagerVerifyPrecondition(CompletedActionXCommand.java:77) at org.apache.oozie.command.XCommand.call(XCommand.java:251) at org.apache.oozie.service.CallableQueueService$CallableWrapper.run(CallableQueueService.java:174) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615) at java.lang.Thread.run(Thread.java:745)
Attachments
Attachments
Issue Links
- breaks
-
OOZIE-2589 CompletedActionXCommand is hardcoded to wrong priority
- Closed