You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Haibo Sun (JIRA)" <ji...@apache.org> on 2019/07/30 09:38:00 UTC

[jira] [Comment Edited] (FLINK-13491) AsyncWaitOperator doesn't handle endInput call properly

    [ https://issues.apache.org/jira/browse/FLINK-13491?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16895958#comment-16895958 ] 

Haibo Sun edited comment on FLINK-13491 at 7/30/19 9:37 AM:
------------------------------------------------------------

So for one-input operator, `endInput()` repeats `close()` semantically. Maybe we don't need `BoundedOneInput`, but we need to call `close()` immediately when the input is finished?


was (Author: sunhaibotb):
So for one-input operator, `endInput ()` repeats `close ()` semantically. Maybe we don't need `BoundedOneInput`, but we need to call `close()` immediately when the input is finished?

> AsyncWaitOperator doesn't handle endInput call properly
> -------------------------------------------------------
>
>                 Key: FLINK-13491
>                 URL: https://issues.apache.org/jira/browse/FLINK-13491
>             Project: Flink
>          Issue Type: Bug
>          Components: API / DataStream
>    Affects Versions: 1.9.0
>            Reporter: Piotr Nowojski
>            Priority: Blocker
>             Fix For: 1.9.0
>
>
> This is the same issue as for {{ContinousFileReaderOperator}} in https://issues.apache.org/jira/browse/FLINK-13376. {{AsyncWaitOperator}} will propagate {{endInput}} notification immediately, even if it has some records buffered.
> Side note, this also shows that the current {{BoundedOneInput#endInput}} API is easy to mishandle if an operator buffers some records internally. Maybe we could redesign this API somehow [~aljoscha] [~sunhaibotb]?



--
This message was sent by Atlassian JIRA
(v7.6.14#76016)