[ https://issues.apache.org/jira/browse/FLINK-2287?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15103477#comment-15103477 ]
Ufuk Celebi commented on FLINK-2287: ------------------------------------ The active JobManager is discovered via ZooKeeper. There is no need to provide the JobManager end points diretly at the moment. The ZooKeeper endpoint is configured in the flink-conf file. I agree that it would be nice to be able to overwrite this directly from the CLI (also when starting the cluster). +1 The client is always going to connect to the master job manager (eventually). If it connects to a standby job manager it will (eventually) be notified about the leader change and re-connect to the new master job manager. Hope this helps. > Implement JobManager high availability > -------------------------------------- > > Key: FLINK-2287 > URL: https://issues.apache.org/jira/browse/FLINK-2287 > Project: Flink > Issue Type: Improvement > Components: JobManager, TaskManager > Reporter: Ufuk Celebi > Fix For: 0.10.0 > > > The problem: The JobManager (JM) is a single point of failure. When it > crashes, TaskManagers (TM) fail all running jobs and try to reconnect to the > same JM. A failed JM looses all state and can not resume the running jobs; > even if it recovers and the TMs reconnect. > Solution: implement JM fault tolerance/high availability by having multiple > JM instances running with one as leader and the other(s) in standby. The > exact coordination and state update protocol between JM, TM, and clients is > covered in sub-tasks/issues. > Related Wiki: > https://cwiki.apache.org/confluence/display/FLINK/JobManager+High+Availability -- This message was sent by Atlassian JIRA (v6.3.4#6332)