You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@flink.apache.org by "Sihua Zhou (JIRA)" <ji...@apache.org> on 2018/03/12 05:04:00 UTC

[jira] [Commented] (FLINK-8918) Introduce Runtime Filter Join

    [ https://issues.apache.org/jira/browse/FLINK-8918?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16394817#comment-16394817 ] 

Sihua Zhou commented on FLINK-8918:
-----------------------------------

Hi [~fhueske] could you please have a look at this after flink 1.5 is released ?

> Introduce Runtime Filter Join
> -----------------------------
>
>                 Key: FLINK-8918
>                 URL: https://issues.apache.org/jira/browse/FLINK-8918
>             Project: Flink
>          Issue Type: Bug
>          Components: Table API &amp; SQL
>            Reporter: Sihua Zhou
>            Assignee: Sihua Zhou
>            Priority: Major
>             Fix For: 1.6.0
>
>
> Right now, for every record that need to be joined, we need to query both `left stream's state` and `right stream's state`. I proposal to introduce RF join to reduce the `query count` of state, which could improve the performance of `stream join`, especially when the joined rate is low. A simple description for RF join can be found [here|https://www.cloudera.com/documentation/enterprise/5-9-x/topics/impala_runtime_filtering.html] (even though it not for stream join original, but we can easily refer it to `stream join`).



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)