[
https://issues.apache.org/jira/browse/KAFKA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17152984#comment-17152984
]
Randall Hauch commented on KAFKA-10242:
---------------------------------------
This is not a regression, so moving this to have a fix version of 2.7.0 since
2.6.0 is already in code freeze and approaching release candidates.
Please let me know if this is not correct.
> Adding metrics to track the total count of idempotent producers that Broker
> need to track
> -----------------------------------------------------------------------------------------
>
> Key: KAFKA-10242
> URL: https://issues.apache.org/jira/browse/KAFKA-10242
> Project: Kafka
> Issue Type: Improvement
> Components: producer
> Affects Versions: 2.5.0
> Reporter: Ming Liu
> Priority: Critical
> Fix For: 2.7.0
>
>
> We found it is very useful to track the total number of idempotent producers
> that broker is tracking.
> In our production environment, we have many idempotent producers for a
> cluster and sometimes that number increased to very high number which
> requires some attention to mitigate.
> This is especially true for client (< 2.4) where the client retry might
> generate too many different idempotent producers which can trigger broker GC.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
