Get Kafka-Nated Episode 3

Aiven

Nov 27, 2025

Join us for Episode 3 of Get Kafka-Nated, where host Hugh Evans sits down with Greg Harris, Staff Software Engineer and the driving force behind KIP-1150, to explore the future of cloud-native Apache Kafka.

In this episode, we dive into diskless Kafka — a bold reimagining of Kafka’s architecture that moves data from broker disks to object storage. Greg explains the technical challenges, the operational benefits, and the broader implications for Kafka users.

Topics we’ll cover:
🔷 What “diskless” really means and the cloud storage pain points it solves.
🔷 The enormous technical challenge of rerouting Kafka’s replication from broker disks to object storage.
🔷 How diskless topics coexist with traditional disk-backed topics, enabling smoother migrations.
🔷 Eliminating Kafka’s leader-based model—what it means for hot partitions and cluster flexibility.
🔷 Greg’s advice on contributing to open source and why being an OSS engineer is a unique experience.

Timestamps:

0:00 – Introduction & Show Overview

1:01 – Guest Introduction: Greg Harris

1:35 – Life as an Open Source Engineer

3:01 – The Core Problem: KIP-1150

4:50 – Understanding ‘Discless’ CFKA

7:30 – Technical Challenges & Batch Coordinator

12:02 – Self-Hosting & CFKA Improvements

16:48 – Motivation vs Implementation

22:16 – Impact on Users & Migration

30:04 – Getting Involved in Open Source

32:06 – Conclusion & Next Steps