Thanks to changes in Apache Kafka 2.4.0, consumers are no longer required to connect to a leader replica to consume messages. In this article, I introduce you to Apache Kafka’s new
ReplicaSelector interface and its customizable
RackAwareReplicaSelector. I’ll briefly explain the benefits of the new rack-aware selector, then show you how to use it to more efficiently balance load across Amazon Web Services (AWS) availability zones.
Continue reading “Consuming messages from closest replicas in Apache Kafka 2.4.0 and AMQ Streams”