[ https://issues.apache.org/jira/browse/FLINK-14674?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16971363#comment-16971363 ]
Zhu Zhu commented on FLINK-14674: --------------------------------- The job hangs due to FLINK-14701, which happen if {{SharedSlotOversubscribedException}} happens (A.T.M only if a shared slot cannot fulfill the managed memory requirements from all tasks in it). Insufficient managed memory for one task is also a problem for q28 and q77, however, jobs will fail but not hang in this case if FLINK-14701 is resolved. > some tpc-ds query hang in scheduled stage for long time > ------------------------------------------------------- > > Key: FLINK-14674 > URL: https://issues.apache.org/jira/browse/FLINK-14674 > Project: Flink > Issue Type: Sub-task > Components: Table SQL / Runtime > Affects Versions: 1.9.1 > Reporter: Leonard Xu > Priority: Major > Fix For: 1.10.0 > > > When run tpc-ds query in standalone mode, some query(4、11、28、31、77、88) hang > in scheduled stage for long time, > BTW,I use blink planner,bath mode shuffle. > Can reproduce this issue in > [https://github.com/leonardBang/flink/tree/FLINK-11491] > > cc [~zhuzh] -- This message was sent by Atlassian Jira (v8.3.4#803005)