How to install Apache Kafka on Linux without ZooKeeper (KRaft mode)

Kafka KRaft installation on Linux without ZooKeeper. Set up Java, format storage with kafka-storage, and verify a single-node KRaft cluster step by step.

By Stéphane Derosiaux · Updated October 16, 2026

Install and run Kafka in KRaft mode on Linux

KRaft mode runs Kafka without ZooKeeper, simplifying deployment and reducing resource requirements. This guide walks you through setting up a single-node KRaft cluster.

What you'll learn:

  • How to install Java (required dependency)
  • How to generate a cluster ID and format storage
  • How to start Kafka in KRaft mode
  • How to configure your PATH for CLI access

KRaft vs ZooKeeper

KRaft is the only mode in Kafka 4.0 and later: ZooKeeper support was removed. If you have to run an older ZooKeeper-based setup, see the Linux ZooKeeper installation guide (legacy).

Installation overview

Linux Kafka KRaft install flow: install Java, download Kafka, generate cluster ID, format storage, start Kafka

Step 1: Install Java JDK

Kafka 4 brokers require Java 17 or later. This guide uses Java 25, the version Kafka recommends today.

For Ubuntu/Debian:

wget -O - https://apt.corretto.aws/corretto.key | sudo gpg --dearmor -o /usr/share/keyrings/corretto-keyring.gpg
echo "deb [signed-by=/usr/share/keyrings/corretto-keyring.gpg] https://apt.corretto.aws stable main" | sudo tee /etc/apt/sources.list.d/corretto.list
sudo apt-get update
sudo apt-get install -y java-25-amazon-corretto-jdk

For other distributions:

Visit the Amazon Corretto 25 Linux install page for RPM, Alpine, and other packages.

Verify the installation:

java -version

Expected output (your patch version may differ):

openjdk version "25.0.4.1" 2026-08-18 LTS
OpenJDK Runtime Environment Corretto-25.0.4.10.1 (build 25.0.4.1+10-LTS)
OpenJDK 64-Bit Server VM Corretto-25.0.4.10.1 (build 25.0.4.1+10-LTS, mixed mode, sharing)

Step 2: Download and extract Kafka

Download Kafka 4.3.1 (Scala 2.13 binary) into your home directory:

cd ~
wget https://archive.apache.org/dist/kafka/4.3.1/kafka_2.13-4.3.1.tgz
tar -xzf kafka_2.13-4.3.1.tgz

A newer 4.x release from kafka.apache.org/downloads works the same way: replace 4.3.1 with its version in every command below.

Step 3: Generate cluster ID

KRaft clusters require a unique identifier:

~/kafka_2.13-4.3.1/bin/kafka-storage.sh random-uuid

This returns a UUID like 76BLQI7sT_ql1mBfKsOk9Q. Save this value.

Step 4: Format storage

Format the log directory using your cluster ID (replace <uuid> with your generated UUID):

~/kafka_2.13-4.3.1/bin/kafka-storage.sh format --standalone \
  -t <uuid> \
  -c ~/kafka_2.13-4.3.1/config/server.properties

This formats the directory specified in log.dirs (default: /tmp/kraft-combined-logs). The --standalone flag sets up this single node as its own controller.

Step 5: Start Kafka

~/kafka_2.13-4.3.1/bin/kafka-server-start.sh ~/kafka_2.13-4.3.1/config/server.properties

Keep this terminal window open. Kafka is now running at localhost:9092 in KRaft mode.

Step 6: Configure PATH

Add Kafka binaries to your PATH for convenient access.

Edit your shell configuration file (~/.bashrc or ~/.zshrc):

export PATH="$PATH:$HOME/kafka_2.13-4.3.1/bin"

Reload your shell:

source ~/.bashrc

Verify the setup:

kafka-topics.sh --version

See it in practice with Conduktor

Conduktor Console can connect to your KRaft cluster at localhost:9092 for visual topic management.

Next steps