[
https://issues.apache.org/jira/browse/FLINK-5231?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16381428#comment-16381428
]
vinoyang commented on FLINK-5231:
---------------------------------
Hi [~StephanEwen] the JDK8, (more precisely since JDK 8u92, pelase see :
http://www.oracle.com/technetwork/java/javase/8u92-relnotes-2949471.html),
there are two new JVM options:
* *ExitOnOutOfMemoryError* - When you enable this option, the JVM exits on the
first occurrence of an out-of-memory error. It can be used if you prefer
restarting an instance of the JVM rather than handling out of memory errors.
* *CrashOnOutOfMemoryError* - If this option is enabled, when an out-of-memory
error occurs, the JVM crashes and produces text and binary crash files (if core
files are enabled).
I think for old JDK we can use 'OnOutOfMemoryError' option, however, higher
JDK these options would be better to fix the OOMError.
What's your opinion?
> JobManager should exit on "OutOfMemoryError"
> --------------------------------------------
>
> Key: FLINK-5231
> URL: https://issues.apache.org/jira/browse/FLINK-5231
> Project: Flink
> Issue Type: Sub-task
> Components: JobManager
> Affects Versions: 1.1.3
> Reporter: Stephan Ewen
> Priority: Major
>
> A JobManager that ever encountered an {{OutOfMemoryError}} is with a high
> probability not functional any more. However, the {{OutOfMemoryError}} kills
> only the thread where it is thrown by default, not the entire JVM.
> We should add {{-XX:OnOutOfMemoryError="kill -9 %p"}} both to the standalone
> bash options and to the YARN and Mesos launch options.
> On the TaskManager side, the decision is more tricky and less critical, so I
> suggest to add this only to the JobManager by default.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
