[ https://issues.apache.org/jira/browse/FLINK-24489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17428173#comment-17428173 ]
Yuepeng Pan edited comment on FLINK-24489 at 10/14/21, 2:00 PM: ---------------------------------------------------------------- Thanks [~dwysakowicz] for the reply. Sorry for the inaccurate description of this jira. I am not sure the max capacity of the cache in some extreme cases. But it encountered the HEAP OOM in my CEP job, which contains great many NodeId & EventId & sharedBufferEdge objects in the hproof file. And please let me know what's your opinion about the significance of introducing the threshold. Thank you. was (Author: rocmarshal): Thanks [~dwysakowicz] for the reply. Sorry for the inaccurate description of this jira. I am not sure the max capacity of the cache in some extreme cases. But it encountered the HEAP OOM in my CEP job, which contains great many NodeId & EventId & sharedBufferEdge objects in the hproof file. > The size of entryCache & eventsBufferCache in the SharedBuffer should be > defined with a threshold to limit the number of the elements in the cache > -------------------------------------------------------------------------------------------------------------------------------------------------- > > Key: FLINK-24489 > URL: https://issues.apache.org/jira/browse/FLINK-24489 > Project: Flink > Issue Type: Improvement > Components: Library / CEP > Affects Versions: 1.10.0, 1.11.0, 1.12.0, 1.13.0, 1.14.0 > Reporter: Yuepeng Pan > Priority: Major > Attachments: 截屏2021-10-13 20.28.17.png > > > source code : > [here|https://github.com/apache/flink/blob/c3cb886ee73b5fee23b2bccff0f5e4d45a30b3a1/flink-libraries/flink-cep/src/main/java/org/apache/flink/cep/nfa/sharedbuffer/SharedBuffer.java#L79] > -- This message was sent by Atlassian Jira (v8.3.4#803005)