TisonKun commented on issue #7889: [FLINK-11665] Wait for job termination before recovery in Dispatcher URL: https://github.com/apache/flink/pull/7889#issuecomment-481278785 @tillrohrmann > I would like to make the Dispatcher not being used across leader sessions and instead creating a new Dispatcher for every new leader session. By not sharing the SubmittedJobGraphStore between different Dispatchers we should be able to solve this issue here as well. sounds good to me. We might not maintain a failing contender(the leader) but just crash it and restart a new one(contender). Following this pattern we can exit on dispatcher lost leadership but it looks like we need a cluster manager to take care of creating a new `Dispatcher`
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
