Diskless Topics on Aiven for Apache Kafka

Running Kafka in the cloud means paying cross-AZ replication fees on every byte you produce, whether it's read back or not. Diskless Topics remove that: they replicate through object storage instead of local disks. They're live in Aiven for Apache Kafka today.

What we'll cover:

  • How it works: the new storage layer abstraction, pluggable leaders, and the tradeoffs behind Diskless Topics.
  • A live walkthrough: creating a Diskless topic, then switching an existing classic topic over.
  • What it does to your bill: where the network savings come from. Available on Standard plans from 10 MB/s up, and classic and Diskless topics can share one cluster.

You'll leave knowing which of your topics are worth switching, and which are better left classic.

Get to know our speaker

Hugh Evans

Senior Product Advocate, Aiven

Hugh is a Senior Developer Advocate at Aiven with a background in software development, cloud infrastructure, and community building. He helps developers and data teams explore how open source technologies can power modern data and AI use cases. Outside of Aiven, Hugh organizes AI Signals, a community dedicated to real-world applications of AI, and advocates for vocational learning as an important pathway into technology careers.