[
https://issues.apache.org/jira/browse/FLINK-16705?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17063913#comment-17063913
]
Zili Chen commented on FLINK-16705:
-----------------------------------
It seems the problem is that we want to support {{executeAsync}} while keep
{{JobClient}} "stateless" so that we don't have to manage resource. Thus, we
add a terminate future in {{LocalExecutor}} to terminate the cluster when the
job finished, which, as described here, possibly race with a separated result
request.
My opinion would be still we(and the user) manage {{JobClient}} lifecycle so
that we ensure the result delivered in {{execute}} (blocking version).
Alternative is we make something like {{MiniDispatcher}} to special handle it.
What do you think? cc [~kkl0u] [~aljoscha].
> LocalExecutor tears down MiniCluster before client can retrieve JobResult
> -------------------------------------------------------------------------
>
> Key: FLINK-16705
> URL: https://issues.apache.org/jira/browse/FLINK-16705
> Project: Flink
> Issue Type: Bug
> Components: Client / Job Submission
> Reporter: Maximilian Michels
> Assignee: Maximilian Michels
> Priority: Major
> Labels: pull-request-available
> Time Spent: 10m
> Remaining Estimate: 0h
>
> There is a race condition in {{LocalExecutor}} between (a) shutting down the
> cluster when the job has finished and (b) the client which retrieves the
> result of the job execution.
> This was observed in Beam, running a large test suite with the Flink Runner.
> We should make sure the job result retrieval and the cluster shutdown do not
> interfere.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
