[ https://issues.apache.org/jira/browse/HBASE-6260?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-6260: ---------------------------------- Attachment: HBASE-6260.patch * Attached HBASE-6260.patch * I haven't run this through the test suite yet, let's see what HadoopQA says. - Master reads balanceSwitch state from ZK - Adds two test cases: 1. If balancer is on and master dies, and new master takes over, balancer is still running 2. 1. If balancer is off and master dies, and new master takes over, balancer is still not running > balancer state should be stored in ZK > ------------------------------------- > > Key: HBASE-6260 > URL: https://issues.apache.org/jira/browse/HBASE-6260 > Project: HBase > Issue Type: Task > Components: master, zookeeper > Affects Versions: 0.96.0 > Reporter: Gregory Chanan > Assignee: Gregory Chanan > Priority: Blocker > Attachments: HBASE-6260.patch > > > See: > https://issues.apache.org/jira/browse/HBASE-5953?focusedCommentId=13270200&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13270200 > And: > https://issues.apache.org/jira/browse/HBASE-5630?focusedCommentId=13399225&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13399225 > In short, we need to move the balancer state to ZK so that it won't have to > be restarted if the master dies. -- 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