[ https://issues.apache.org/jira/browse/FLINK-12122?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16812308#comment-16812308 ]
Tony Xintong Song commented on FLINK-12122: ------------------------------------------- Hi [~till.rohrmann], I agree with you that we should spread tasks evenly across TaskManagers. FYI, similar behavior change is already included in the to-be-merged [blink branch|[https://github.com/apache/flink/blob/blink/flink-runtime/src/main/java/org/apache/flink/runtime/resourcemanager/slotmanager/DynamicAssigningSlotManager.java]], where we support three policies (random, spreading based on available slots, and spreading based on available resource) for distribute slots to TaskManagers for session mode. > Spread out tasks evenly across all available registered TaskManagers > -------------------------------------------------------------------- > > Key: FLINK-12122 > URL: https://issues.apache.org/jira/browse/FLINK-12122 > Project: Flink > Issue Type: Sub-task > Components: Runtime / Coordination > Affects Versions: 1.6.4, 1.7.2, 1.8.0 > Reporter: Till Rohrmann > Assignee: Till Rohrmann > Priority: Major > Fix For: 1.7.3, 1.9.0, 1.8.1 > > > With Flip-6, we changed the default behaviour how slots are assigned to > {{TaskManages}}. Instead of evenly spreading it out over all registered > {{TaskManagers}}, we randomly pick slots from {{TaskManagers}} with a > tendency to first fill up a TM before using another one. This is a regression > wrt the pre Flip-6 code. > I suggest to change the behaviour so that we try to evenly distribute slots > across all available {{TaskManagers}} by considering how many of their slots > are already allocated. -- This message was sent by Atlassian JIRA (v7.6.3#76005)