You are viewing a plain text version of this content. The canonical link for it is here.
Posted to yarn-issues@hadoop.apache.org by "Carlo Curino (JIRA)" <ji...@apache.org> on 2015/07/24 23:50:05 UTC
[jira] [Commented] (YARN-3977) NPE in
ProportionalCapacityPreemptionPolicy
[ https://issues.apache.org/jira/browse/YARN-3977?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14641125#comment-14641125 ]
Carlo Curino commented on YARN-3977:
------------------------------------
Running a large gridmix run to test the reservation system, I stumbled on this:
{code}
15/07/24 14:35:23 ERROR yarn.YarnUncaughtExceptionHandler: Thread Thread[SchedulingMonitor (ProportionalCapacityPreemptionPolicy),5,main] threw an Exception.
java.lang.NullPointerException
at org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy.getContainersToPreempt(ProportionalCapacityPreemptionPolicy.java:643)
at org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy.containerBasedPreemptOrKill(ProportionalCapacityPreemptionPolicy.java:233)
at org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy.editSchedule(ProportionalCapacityPreemptionPolicy.java:185)
at org.apache.hadoop.yarn.server.resourcemanager.monitor.SchedulingMonitor.invokePolicy(SchedulingMonitor.java:84)
at org.apache.hadoop.yarn.server.resourcemanager.monitor.SchedulingMonitor$PreemptionChecker.run(SchedulingMonitor.java:94)
at java.lang.Thread.run(Thread.java:722)
{code}
It seems to be related to the getQueuePartitions() returning a null, although the code itself is not very clear to me,
why are we looping first, and then take the last value (possibly null) of leafQueue and do something with it?
> NPE in ProportionalCapacityPreemptionPolicy
> -------------------------------------------
>
> Key: YARN-3977
> URL: https://issues.apache.org/jira/browse/YARN-3977
> Project: Hadoop YARN
> Issue Type: Bug
> Reporter: Carlo Curino
>
> There seems to be an NPE in the ProportionalCapacityPreemptionPolicy near 643, likely due to getQueuePartitions() returning a null.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)