kafka-jira mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Guozhang Wang (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (KAFKA-6443) KTable involved in multiple joins could result in duplicate results
Date Fri, 12 Jan 2018 17:49:02 GMT

    [ https://issues.apache.org/jira/browse/KAFKA-6443?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16324283#comment-16324283
] 

Guozhang Wang commented on KAFKA-6443:
--------------------------------------

The two duplicate records will be the same, see my added tests in https://github.com/apache/kafka/pull/4331,
the expected result list for details.

> KTable involved in multiple joins could result in duplicate results
> -------------------------------------------------------------------
>
>                 Key: KAFKA-6443
>                 URL: https://issues.apache.org/jira/browse/KAFKA-6443
>             Project: Kafka
>          Issue Type: Bug
>          Components: streams
>            Reporter: Guozhang Wang
>
> Consider the following multi table-table joins:
> {code}
> table1.join(table2).join(table2);    // "join" could be replaced with "leftJoin" and
"outerJoin"
> {code}
> where {{table2}} is involved multiple times in this multi-way joins. In this case, when
a new record from the source topic of {{table2}} is being processing, it will send to two
children down in the topology and hence may resulting in duplicated join results depending
on the join types.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Mime
View raw message