# Exploring ZooKeeper-less Kafka

DevFeed: [Exploring ZooKeeper-less Kafka](<https://devfeed.tech/articles/exploring-zookeeper-less-kafka-18814.md>)

Original publisher: [Read original article](<https://www.morling.dev/blog/exploring-zookeeper-less-kafka/>)

Published: 2021-05-17T17:45:00Z

Content type: tutorial

Language: en

Sources: [Gunnar Morling](<https://devfeed.tech/sources/gunnar-morling.md>)

Topics: [Kafka](<https://devfeed.tech/topics/kafka.md>), [Raft](<https://devfeed.tech/topics/raft.md>), [Deployment](<https://devfeed.tech/topics/deployment.md>), [Scalability](<https://devfeed.tech/topics/scalability.md>), [configuration](<https://devfeed.tech/topics/configuration.md>), [Testing](<https://devfeed.tech/topics/testing.md>)

Tags: [apache-kafka](<https://devfeed.tech/tags/apache-kafka.md>), [configuration](<https://devfeed.tech/tags/configuration.md>), [consensus](<https://devfeed.tech/tags/consensus.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [getting-started](<https://devfeed.tech/tags/getting-started.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [kafka](<https://devfeed.tech/tags/kafka.md>), [raft](<https://devfeed.tech/tags/raft.md>), [scalability](<https://devfeed.tech/tags/scalability.md>), [testing](<https://devfeed.tech/tags/testing.md>)

## AI overview

This article explores ZK-less Kafka, also known as KRaft, an early access Kafka 2.8 feature that removes ZooKeeper by storing metadata in Kafka and using a Raft-based controller election process. It provides setup guidance and initial observations rather than a systematic evaluation.

## Source excerpt

Table of Contents Trying ZK-less Kafka Yourself Taking Brokers Down Wrap-Up Sometimes, less is more. One case where that's certainly true is dependencies. And so it shouldn't come at a surprise that the Apache Kafka community is eagerly awaiting the removal of the dependency to the ZooKeeper service, which currently is used for storing Kafka metadata (e.g. about topics and partitions) as well as for the purposes of leader election in the cluster. The Kafka improvement proposal KIP-500 ("Replace ZooKeeper with a Self-Managed Metadata Quorum") promises to make life better for users in many regards: Better getting started and operational experience by requiring to run only one system, Kafka, instead of two Removing potential for discrepancies of metadata state between ZooKeeper and the Kafka controller Simplifying configuration, for instance when it comes to security Better scalability, e.g. in terms of number of partitions; faster execution of operations like topic creation