[ https://issues.apache.org/jira/browse/FLINK-9458?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16501402#comment-16501402 ]
ASF GitHub Bot commented on FLINK-9458: --------------------------------------- GitHub user tillrohrmann opened a pull request: https://github.com/apache/flink/pull/6119 [FLINK-9458] Ignore SharedSlot in CoLocationConstraint when not using legacy mode ## What is the purpose of the change The SharedSlot in CoLocationConstraint is only set when using the legacy mode. Thus, CoLocationConstraint#isAssignedAlive should only check the SharedSlot if it was previously set. This fixes the NPE when restarting a job with a co-location constraint when using the new mode. ## Verifying this change - Added `ExecutionGraphCoLocationRestartTest#testConstraintsAfterRestart` which verifies that an `ExecutionGraph` with co-location constraints can be restarted with the legacy as well as the new mode. ## Does this pull request potentially affect one of the following parts: - Dependencies (does it add or upgrade a dependency): (no) - The public API, i.e., is any changed class annotated with `@Public(Evolving)`: (no) - The serializers: (no) - The runtime per-record code paths (performance sensitive): (no) - Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Yarn/Mesos, ZooKeeper: (yes) - The S3 file system connector: (no) ## Documentation - Does this pull request introduce a new feature? (no) - If yes, how is the feature documented? (not applicable) You can merge this pull request into a Git repository by running: $ git pull https://github.com/tillrohrmann/flink fixCoLocationConstraintRecovery Alternatively you can review and apply these changes as the patch at: https://github.com/apache/flink/pull/6119.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #6119 ---- commit acf70e8cca10caa2fe0cd8d5e95885b39ed0b187 Author: Till Rohrmann <trohrmann@...> Date: 2018-06-05T07:12:49Z [FLINK-9458] Ignore SharedSlot in CoLocationConstraint when not using legacy mode The SharedSlot in CoLocationConstraint is only set when using the legacy mode. Thus, CoLocationConstraint#isAssignedAlive should only check the SharedSlot if it was previously set. This fixes the NPE when restarting a job with a co-location constraint when using the new mode. ---- > Unable to recover from job failure on YARN with NPE > --------------------------------------------------- > > Key: FLINK-9458 > URL: https://issues.apache.org/jira/browse/FLINK-9458 > Project: Flink > Issue Type: Bug > Affects Versions: 1.5.0 > Environment: Ambari HDP 2.6.3 > Hadoop 2.7.3 > > Job configuration: > 120 Task Managers x 1 slots > > > Reporter: Truong Duc Kien > Assignee: Till Rohrmann > Priority: Blocker > Fix For: 1.6.0, 1.5.1 > > > After upgrading our job to Flink 1.5, they are unable to recover from failure > with the following exception appears repeatedly > {noformat} > 2018-05-29 04:56:06,086 [ jobmanager-future-thread-36] INFO > o.a.f.r.e.ExecutionGraph Try to restart or fail the job xxx > (23d9e87bf43ce163ff7db8afb062fb1d) if no longer possible. > 2018-05-29 04:56:06,086 [ jobmanager-future-thread-36] INFO > o.a.f.r.e.ExecutionGraph Job xxx (23d9e87bf43ce163ff7db8afb062fb1d) switched > from state RESTARTING to RESTARTING. > 2018-05-29 04:56:06,086 [ jobmanager-future-thread-36] INFO > o.a.f.r.e.ExecutionGraph Restarting the job xxx > (23d9e87bf43ce163ff7db8afb062fb1d). > 2018-05-29 04:57:06,086 [ jobmanager-future-thread-36] WARN > o.a.f.r.e.ExecutionGraph Failed to restart the job. > java.lang.NullPointerException at > org.apache.flink.runtime.jobmanager.scheduler.CoLocationConstraint.isAssignedAndAlive(CoLocationConstraint.java:104) > at > org.apache.flink.runtime.jobmanager.scheduler.CoLocationGroup.resetConstraints(CoLocationGroup.java:119) > at > org.apache.flink.runtime.executiongraph.ExecutionGraph.restart(ExecutionGraph.java:1247) > at > org.apache.flink.runtime.executiongraph.restart.ExecutionGraphRestartCallback.triggerFullRecovery(ExecutionGraphRestartCallback.java:59) > at > org.apache.flink.runtime.executiongraph.restart.FixedDelayRestartStrategy$1.run(FixedDelayRestartStrategy.java:68) > at > java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511) at > java.util.concurrent.FutureTask.run(FutureTask.java:266) at > java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180) > at > java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293) > at > java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) > at > java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) > at java.lang.Thread.run(Thread.java:748) > {noformat} -- This message was sent by Atlassian JIRA (v7.6.3#76005)