You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@kafka.apache.org by "Nicolas Guyomar (Jira)" <ji...@apache.org> on 2024/09/30 11:00:00 UTC

[jira] [Created] (KAFKA-17663) Add Metadata caching in admin.internals.PartitionLeaderStrategy

Nicolas Guyomar created KAFKA-17663:
---------------------------------------

             Summary: Add Metadata caching in admin.internals.PartitionLeaderStrategy
                 Key: KAFKA-17663
                 URL: https://issues.apache.org/jira/browse/KAFKA-17663
             Project: Kafka
          Issue Type: Improvement
          Components: admin
            Reporter: Nicolas Guyomar


Hi team,

To the best of my knowledge, there is no Metadata caching in AdminClient admin.internals.PartitionLeaderStrategy, and for use cases that keep long live AdminClient running with scheduled invocation of method such as the listOffsets, we could keep reusing the same Metadata and not request fresh one on each invocation

This is particularly true when the listOffsets is invoked for monitoring purpose, often querying a high number of partitions at the same time, leading to rather "large" Metadata requests to be processed on the cluster side

Thank you



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