[
https://issues.apache.org/jira/browse/FLINK-2508?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14699672#comment-14699672
]
Aljoscha Krettek commented on FLINK-2508:
-----------------------------------------
I think this is a manifestation of the problem:
https://travis-ci.org/aljoscha/flink/jobs/75944062
This only occurs in some Travis runs. What happens is that if TranslationTest
is executed after TypeFillTest the Env will still contain operators for which
the TypeInformation cannot be determined.
This maybe only happens in my PR that reworks the StreamGraph generation
because I explicitly verify that each operator has a valid type upon generation
of the StreamGraph.
> Confusing sharing of StreamExecutionEnvironment
> -----------------------------------------------
>
> Key: FLINK-2508
> URL: https://issues.apache.org/jira/browse/FLINK-2508
> Project: Flink
> Issue Type: Bug
> Components: Streaming
> Affects Versions: 0.10
> Reporter: Stephan Ewen
> Assignee: Márton Balassi
> Fix For: 0.10
>
>
> In the {{StreamExecutionEnvironment}}, the environment is once created and
> then shared with a static variable to all successive calls to
> {{getExecutionEnvironment()}}. But it can be overridden by calls to
> {{createLocalEnvironment()}} and {{createRemoteEnvironment()}}.
> This seems a bit un-intuitive, and probably creates confusion when
> dispatching multiple streaming jobs from within the same JVM.
> Why is it even necessary to cache the "current" execution environment?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
