[ https://issues.apache.org/jira/browse/HADOOP-9220?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13559135#comment-13559135 ]
Tom White commented on HADOOP-9220: ----------------------------------- It's true that the elector checks for a stale ZK client, but that doesn't prevent the problem here which is caused by i) having multiple watchers for the ZK client (due to the creation of a new watcher in monitorLockNodeAsync), and ii) a postponed call to recheckElectability unnecessarily forcing a new election (this call doesn't go through the watcher). > Unnecessary transition to standby in ActiveStandbyElector > --------------------------------------------------------- > > Key: HADOOP-9220 > URL: https://issues.apache.org/jira/browse/HADOOP-9220 > Project: Hadoop Common > Issue Type: Bug > Components: ha > Reporter: Tom White > Assignee: Tom White > Attachments: HADOOP-9220.patch, HADOOP-9220.patch > > > When performing a manual failover from one HA node to a second, under some > circumstances the second node will transition from standby -> active -> > standby -> active. This is with automatic failover enabled, so there is a ZK > cluster doing leader election. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira