Maximilian Michels created FLINK-4540:
-----------------------------------------
Summary: Detached job execution may prevent cluster shutdown
Key: FLINK-4540
URL: https://issues.apache.org/jira/browse/FLINK-4540
Project: Flink
Issue Type: Bug
Components: YARN Client
Affects Versions: 1.2.0, 1.1.2
Reporter: Maximilian Michels
Priority: Minor
Fix For: 1.1.3, 1.2.0
There is a problem with the detached execution of jobs. This can prevent
cluster shutdown 1) when eager jobs are executed, i.e. the job calls
`collect()/count()`, and 2) when the user jar doesn't contain a job.
1) For example, {{./flink -d -m yarn-cluster -yn 1
../examples/batch/WordCount.jar}} will throw an exception and only disconnect
the YarnClusterClient afterwards. In detached mode, the code assumes the
cluster is shutdown through the {{shutdownAfterJob}} method which ensures that
the YarnJobManager shuts down after the job completes. Due to the exception
thrown when executing eager jobs, the jobmanager never receives a job and thus
never shuts down the cluster.
2) The same problem also occurs in detached execution when the user jar doesn't
contain a job.
A good solution would be to defer cluster startup until the job has been fully
assembled.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
