You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Alex Sorokoumov (Jira)" <ji...@apache.org> on 2023/03/23 22:30:00 UTC

[jira] [Commented] (FLINK-31599) Update Kafka dependency in flink-connector-kafka to 3.4.0

    [ https://issues.apache.org/jira/browse/FLINK-31599?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17704378#comment-17704378 ] 

Alex Sorokoumov commented on FLINK-31599:
-----------------------------------------

Patch - https://github.com/apache/flink-connector-kafka/pull/11.

> Update Kafka dependency in flink-connector-kafka to 3.4.0
> ---------------------------------------------------------
>
>                 Key: FLINK-31599
>                 URL: https://issues.apache.org/jira/browse/FLINK-31599
>             Project: Flink
>          Issue Type: Technical Debt
>          Components: Connectors / Kafka
>            Reporter: Alex Sorokoumov
>            Assignee: Alex Sorokoumov
>            Priority: Minor
>
> There is a number of reasons to upgrade to the latest version.
>  
> First, the Kafka connector uses reflection, so internal changes in Kafka clients' implementation might require changes in the connector. With more frequent upgrades, the amount of work per upgrade is smaller.
>  
> Second, there were a number of relevant bug fixes since 3.2.3:
> * [KAFKA-14303] - Producer.send without record key and batch.size=0 goes into infinite loop
> * [KAFKA-14379] - consumer should refresh preferred read replica on update metadata
> * [KAFKA-14422] - Consumer rebalance stuck after new static member joins a group with members not supporting static members
> * [KAFKA-14417] - Producer doesn't handle REQUEST_TIMED_OUT for InitProducerIdRequest, treats as fatal error
> * [KAFKA-14532] - Correctly handle failed fetch when partitions unassigned
> * 



--
This message was sent by Atlassian Jira
(v8.20.10#820010)