[
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)
