[ https://issues.apache.org/jira/browse/YARN-9714?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16917482#comment-16917482 ]
Tao Yang commented on YARN-9714: -------------------------------- {quote} Instead of comparing, how about checking for resourceManager.getZKManager() == null? This basically sync the code where zkManager initialization to closing it. {quote} Make sense to me. Attached v5 patch for this, thanks! > ZooKeeper connection in ZKRMStateStore leaks after RM transitioned to standby > ----------------------------------------------------------------------------- > > Key: YARN-9714 > URL: https://issues.apache.org/jira/browse/YARN-9714 > Project: Hadoop YARN > Issue Type: Bug > Components: resourcemanager > Reporter: Tao Yang > Assignee: Tao Yang > Priority: Major > Labels: memory-leak > Attachments: YARN-9714.001.patch, YARN-9714.002.patch, > YARN-9714.003.patch, YARN-9714.004.patch, YARN-9714.005.patch > > > Recently RM full GC happened in one of our clusters, after investigating the > dump memory and jstack, I found two places in RM may cause memory leaks after > RM transitioned to standby: > # Release cache cleanup timer in AbstractYarnScheduler never be canceled. > # ZooKeeper connection in ZKRMStateStore never be closed. > To solve those leaks, we should close the connection or cancel the timer when > services are stopping. -- This message was sent by Atlassian Jira (v8.3.2#803003) --------------------------------------------------------------------- To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org