You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@kafka.apache.org by "Mateusz Owczarek (JIRA)" <ji...@apache.org> on 2018/12/20 15:46:00 UTC

[jira] [Resolved] (KAFKA-7753) ValueTransformerWithKey should not require producing one value every message

     [ https://issues.apache.org/jira/browse/KAFKA-7753?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Mateusz Owczarek resolved KAFKA-7753.
-------------------------------------
    Resolution: Feedback Received

> ValueTransformerWithKey should not require producing one value every message
> ----------------------------------------------------------------------------
>
>                 Key: KAFKA-7753
>                 URL: https://issues.apache.org/jira/browse/KAFKA-7753
>             Project: Kafka
>          Issue Type: Wish
>          Components: streams
>            Reporter: Mateusz Owczarek
>            Priority: Minor
>
> Hi, speaking about: [https://github.com/apache/kafka/blob/trunk/streams/src/main/java/org/apache/kafka/streams/kstream/ValueTransformerWithKey.java]
> I have a quite simple case - I want to implement a Transformer which I will use later in my DSL-api-defined topology. Those are my requirements:
> - no repartition topic should be created since I do not change the keys
> - I can't forward messages on every message transformed. I have my internal state store which I'm using to ensure max. 1 message per window is sent.
> Basic transformer gives me possibility to not send messages down the stream (null internal handler), but it forces repartitioning which I think is unnecessary in my case. WDYT?



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