[ https://issues.apache.org/jira/browse/HDFS-1623?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13070584#comment-13070584 ]
Yang Yang commented on HDFS-1623: --------------------------------- bq. Yang: the delivery of the "disconnected" event from ZK is asynchronous. So, there is a small window of time when the old NN still "thinks" it owns the znode while the new node may have taken over. I agree, I asked this question on ZK recently: http://zookeeper-user.578899.n2.nabble.com/help-on-Zookeeper-code-walk-through-tp6589163p6595469.html I guess it helps to quantify how "small" this window needs to be and whether current ZK is able to provide the notification fast enough. if not, maybe implementing the ZAB protocol as part of the namenode/backup-nodenode communication ? (in that case it would be nice if ZK exports its protocol as a library) > High Availability Framework for HDFS NN > --------------------------------------- > > Key: HDFS-1623 > URL: https://issues.apache.org/jira/browse/HDFS-1623 > Project: Hadoop HDFS > Issue Type: New Feature > Reporter: Sanjay Radia > Assignee: Sanjay Radia > Attachments: HDFS-High-Availability.pdf, NameNode HA_v2.pdf, NameNode > HA_v2_1.pdf, Namenode HA Framework.pdf > > -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira