[
https://issues.apache.org/jira/browse/HADOOP-4347?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12636908#action_12636908
]
Vinod K V commented on HADOOP-4347:
-----------------------------------
This problem is immediately visible in schedulers like CapacityTaskScheduler
which lazily initialize jobs, and when jobs stay in PREP state for sufficient
enough time for clients to kill them.
> [mapred] Job client cannot kill jobs that are not initialized.
> --------------------------------------------------------------
>
> Key: HADOOP-4347
> URL: https://issues.apache.org/jira/browse/HADOOP-4347
> Project: Hadoop Core
> Issue Type: Bug
> Components: mapred
> Reporter: Vinod K V
>
> Job client cannot kill a job that is not yet initialized, JobTracker silently
> ignores the kill request (JobTracker.killJob(JobID)). Though the client
> reports that the job is killed, the job still resides with JT and in fact, it
> goes to running state and eventually finishes.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
