[ https://issues.apache.org/jira/browse/IGNITE-13126?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17153544#comment-17153544 ]
Maxim Muzafarov commented on IGNITE-13126: ------------------------------------------ {code:title=CacheContinuousQueryEventBuffer.java} if (batch == null || cntr < batch.startCntr) {code} We can skip backupQ.add() if we already have acked counter more then current counter. > Refactor of Continuous Query buffers to reduce heap usage > --------------------------------------------------------- > > Key: IGNITE-13126 > URL: https://issues.apache.org/jira/browse/IGNITE-13126 > Project: Ignite > Issue Type: Improvement > Reporter: Maxim Muzafarov > Assignee: Maxim Muzafarov > Priority: Major > Fix For: 2.9 > > > There are several unnecessary changes needs to be done to reduce continuous > query buffer usage: > - CacheContinuousQueryEventBuffer.part is unnecessary; > - CacheContinuousQueryEventBuffer.backupQ must be changed to > ConcurrentLinkedDeque; > - Use lambda instead for partition counter resolving instead on anonymous > class creation for CacheContinuousQueryEventBuffer; > - Batch local variables may be {{short}} type; > etc. -- This message was sent by Atlassian Jira (v8.3.4#803005)