[
https://issues.apache.org/jira/browse/NIFI-14470?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17945179#comment-17945179
]
Paul Grey commented on NIFI-14470:
----------------------------------
Hi. I just tested this on apache/main (as of a0ca48b). PublishKafka seems to
behave as expected. I added two attributes to the GenerateFlowFile:
* kafka.key = "a-kafka-key"
* another.key = "another-key"
When I publish and consume the FlowFile (via PublishKafka, ConsumeKafka):
* if I leave the attribute PublishKafka/"kafka.key" blank, then ConsumeKafka
reads the kafka key as "a-kafka-key"
* if I set the attribute PublishKafka/"kafka.key" as "other.key", then
ConsumeKafka reads the kafka key as "another-key"
Maybe you can share a little more information about your scenario?
> PublishKafka not able to add kafka key from attribute
> -----------------------------------------------------
>
> Key: NIFI-14470
> URL: https://issues.apache.org/jira/browse/NIFI-14470
> Project: Apache NiFi
> Issue Type: Bug
> Affects Versions: 2.0.0-M4
> Reporter: Rabii
> Priority: Major
>
> Using 2.0.0-M4 to migrate from PublishKafka2_6/ConsumeKafka2_6 to the new
> PublishKafka/ConsumeKafka Processors since M4 supports both type of
> processors.
>
> Generate flowFile as json, attach an attribute to the messages, publish
> message using PublishKafka and kafka key from the created attribute.
> Result message without key. It should be easy to reproduce.
> The same logic works with PublishKafka2_6
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
