[ 
https://issues.apache.org/jira/browse/NIFI-8021?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17259967#comment-17259967
 ] 

ASF subversion and git services commented on NIFI-8021:
-------------------------------------------------------

Commit 783633cac5e02e96086ab2d36c92d8057df93c01 in nifi's branch 
refs/heads/main from Mark Payne
[ https://gitbox.apache.org/repos/asf?p=nifi.git;h=783633c ]

NIFI-8021: Fixed bug in ConsumeKafka_2_6 and ConsumeKafkaRecord_2_6 where 
explicit partition assignment causes issues with more than 1 concurrent task. 
Also fixed bug that prevented more nifi nodes than partitions because it didn't 
properly handle empty string for the list of partitions

Signed-off-by: Pierre Villard <[email protected]>

This closes #4744.


> Allow ConsumeKafka processors to use static partition mapping
> -------------------------------------------------------------
>
>                 Key: NIFI-8021
>                 URL: https://issues.apache.org/jira/browse/NIFI-8021
>             Project: Apache NiFi
>          Issue Type: Improvement
>          Components: Extensions
>            Reporter: Mark Payne
>            Assignee: Mark Payne
>            Priority: Major
>             Fix For: 1.13.0
>
>          Time Spent: 1h
>  Remaining Estimate: 0h
>
> With the ConsumeKafka processors, we use the KafkaConsumer.subscribe() 
> method, which is generally what is desirable. However, for a case where 
> strict FIFO order per partition is critical, as may be the case with CDC use 
> cases among others, we should allow users to pin a particular partition to a 
> particular node. This way, even if the node is restarted or stops pulling 
> data for a while, its partitions are not re-assigned, so there's no risk of 
> messages being delivered out of order.



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

Reply via email to