[ https://issues.apache.org/jira/browse/YARN-3286?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14490515#comment-14490515 ]
Jian He commented on YARN-3286: ------------------------------- I think this has changed the behavior that without any RM/NM restart features enabled, earlier restarting a node will trigger RM to kill all the containers on this node, but now it won't ? > Cleanup RMNode#ReconnectNodeTransition > -------------------------------------- > > Key: YARN-3286 > URL: https://issues.apache.org/jira/browse/YARN-3286 > Project: Hadoop YARN > Issue Type: Improvement > Components: resourcemanager > Affects Versions: 2.6.0, 2.7.0 > Reporter: Rohith > Assignee: Rohith > Attachments: 0001-YARN-3286.patch, YARN-3286-test-only.patch > > > RMNode#ReconnectNodeTransition has messed up for every ReconnectedEvent. This > part of the code can be clean up where we do not require to remove node and > add new node every time. > Supporting to above point, in the issue discussion YARN-3222 mentioned in the > comment > [link1|https://issues.apache.org/jira/browse/YARN-3222?focusedCommentId=14339799&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14339799] > and > [link2|https://issues.apache.org/jira/browse/YARN-3222?focusedCommentId=14344739&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14344739] > Clean up can do the following things > # It always remove an old node and add a new node. This is not really > required, instead old node can be updated with new values. > # RMNode#totalCapability has stale capability after NM is reconnected. -- This message was sent by Atlassian JIRA (v6.3.4#6332)