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

Posted to jira@kafka.apache.org by "Matthias J. Sax (Jira)" <ji...@apache.org> on 2020/10/07 00:57:00 UTC

[jira] [Commented] (KAFKA-10578) Convert KTable to a KStream using the previous value

    [ https://issues.apache.org/jira/browse/KAFKA-10578?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17209253#comment-17209253 ] 

Matthias J. Sax commented on KAFKA-10578:
-----------------------------------------

[~javier.freire] – I added you to the list of contributors and assigned the ticket to you. You can know also self-assign tickets.

> Convert KTable to a KStream using the previous value
> ----------------------------------------------------
>
>                 Key: KAFKA-10578
>                 URL: https://issues.apache.org/jira/browse/KAFKA-10578
>             Project: Kafka
>          Issue Type: Improvement
>          Components: streams
>            Reporter: Javier Freire Riobó
>            Priority: Minor
>              Labels: kip
>
> Imagine that we have an entity for which we want to emit the difference between the current and the previous state. The simplest case would be that the entity was an integer number and you want to emit the subtraction between the current and previous values.
> For example, for the input stream 4, 6, 3 the output 4 (4 - 0), 2 (6 - 4), -3 (3 - 6) is expected.
> The way to achieve this with kafka streams would be through an aggregate.
> The main problem, apart from needing more code, is that if the same event is received twice at the same time and the commit time is not 0, the difference is deleted and nothing is emitted.
>  KIP-675: [https://cwiki.apache.org/confluence/display/KAFKA/KIP-675%3A+Convert+KTable+to+a+KStream+using+the+previous+value]



--
This message was sent by Atlassian Jira
(v8.3.4#803005)