Scaling to a Million partitions with KRaft and Confluent

Stop managing legacy complexity. Learn how to unify your architecture, scale to millions of partitions, and migrate to a serverless, KRaft-native experience in minutes.

The era of the external controller is over. KRaft (Kafka + Raft Metadata Consensus) simplifies Kafka’s architecture by consolidating metadata management into Kafka itself. The result? 10x faster failover performance and the ability to scale your data streaming to millions of partitions.

Join our experts as we dive deep into the KRaft transition. We’ll show you exactly how to navigate this change—from a technical deep dive into KRaft modes to a live demo of an automated migration using Confluent’s migration tools.

What You’ll Learn:

• The KRaft Revolution: An introduction to KIP-500 and why the Kafka community is moving away from ZooKeeper.
• Architectural Deep Dive: Comparing ZooKeeper vs. KRaft Isolated and Combined modes—which is right for your production environment?
• Live Demo: A step-by-step tutorial on setting up a KRaft cluster from scratch.
• Migration Made Simple: How to use the Confluent Migration Tool (KCP) and Confluent for Kubernetes (CFK) to migrate from hosted Kafka to a KRaft-native architecture in just 5 minutes.
• The Confluent Advantage: How the Kora Engine goes beyond open-source KRaft to deliver a fully managed, serverless, and enterprise-grade data streaming platform

Speakers:

• David Marsh, Technical Product Marketing, Confluent