[ https://issues.apache.org/jira/browse/FLINK-4094?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15340447#comment-15340447 ]
Anıl Chalil commented on FLINK-4094: ------------------------------------ https://github.com/capacman/flinkoffheaptest Above github repository contains batch job and example data i used to test the issue. I also put my flink conf yaml as well. In my tests i used jvm 1.8.0_66 on ubuntu 16.04 with flink 1.0.3. Also if you use taskmanager.sh file as is(leaving MaxDirectMemorySize=8388607T) you will spot the problem more easily. In my tests i also checked whether limiting -XX:MaxDirectMemorySize to intended memory usage helps or not. Usually it defer the problem a little bit since ,as i understand, hitting limit of MaxDirectMemorySize can cause full GC and as [~ram_krish] said it clear some memory. Usually problem become more visible when you repeat test with a machine which have more than 24 GB memory. > Off heap memory deallocation might not properly work > ---------------------------------------------------- > > Key: FLINK-4094 > URL: https://issues.apache.org/jira/browse/FLINK-4094 > Project: Flink > Issue Type: Bug > Components: Local Runtime > Affects Versions: 1.1.0 > Reporter: Till Rohrmann > Priority: Critical > Fix For: 1.1.0 > > > A user reported that off-heap memory is not properly deallocated when setting > {{taskmanager.memory.preallocate:false}} (per default) [1]. This can cause > the TaskManager process being killed by the OS. > It should be possible to execute multiple batch jobs with preallocation > turned off. No longer used direct memory buffers should be properly garbage > collected so that the JVM process does not exceed it's maximum memory bounds. > [1] > http://apache-flink-mailing-list-archive.1008284.n3.nabble.com/offheap-memory-allocation-and-memory-leak-bug-td12154.html -- This message was sent by Atlassian JIRA (v6.3.4#6332)