Install Confluent Platform Using Docker

Using Docker images, you can install the entire Confluent Platform or individual components.

The Confluent Platform Docker images are available on DockerHub, and the source files for the images are available on GitHub. From GitHub you can extend and rebuild the images and upload them to your own DockerHub repository. See Docker Image Reference for Confluent Platform for more information about the Confluent Platform images.

Looking for a fully managed cloud-native service for Apache Kafka®?

Sign up for Confluent Cloud and get started for free using the Cloud quick start.

Considerations

You should consider the following before using the Docker images.

Multi-node Environment
For more information, see Configure a Multi-Node Environment with Docker.
Persistent Data (Mounted Volumes)
When deploying Kafka and ZooKeeper images, you should always use Mounting Docker External Volumes for the file systems those images use for their persistent data. This ensures that the containers will retain their proper state when stopped and restarted. The other images maintain their state directly in Kafka topics, so mounted volumes are not usually required for those containers.
Bridge Networking vs. Host Networking

Bridge networking is currently only supported on a single host. To run a bridge network over multiple hosts, you must use overlay networks which are not currently supported. To expose Kafka to clients outside of the bridge network, you must find the container IP and put it in advertised.listeners. This can be difficult to achieve depending on how you’re using the images. Furthermore, it can add a network hop and may not be as performant as the host network, which shares the network stack.

Host networking is the recommended option in the following cases:

  • Multi-host clusters without using Swarm/Kubernetes.
  • You need clients to be able to access Kafka outside the bridge/overlay network.
Adding Connectors to the Kafka Connect Image

Here are the methods to add new connectors to the Kafka Connect image.

  • Build a new Docker image that has the connector installed. You can follow the examples found in Extend Confluent Platform images. You must make sure that the connector JARs are on the CLASSPATH for the Connect service (the default location of /usr/share/java/kafka-connect-* is the recommended location).
  • Add the connector JARs via volumes. If you don’t want to create a new Docker image, see the documentation on Extend Confluent Platform images to configure the cp-kafka-connect container with external JARs.
Supported Java
The Confluent Docker images are tested and shipped with Azul Zulu OpenJDK. Other JDK’s (including Oracle Java) are supported, but you must extend the images yourself to implement that change.
Untested Features
The images are not currently tested on Docker Swarm.
ZooKeeper
As of Confluent Platform 7.5, ZooKeeper is deprecated for new deployments. Confluent recommends KRaft mode for new deployments. For more information, see KRaft Overview.

Supported platforms

Confluent Platform Docker images are provided for the following architectures:

  • Linux AMD64 & x86_64

  • Linux ARM64 & aarch64

    Important

    In preview support for evaluation and non-production testing purposes. These images are not for production workloads.

    Linux ARM64 Docker images can be used on Apple macOS with M1.

To pull a Confluent Platform image for a specific architecture, use of the following command. The container runtime auto selects the platform:

docker pull confluentinc/<image-name>:<tag>

For example:

docker pull confluentinc/cp-server:latest

Optionally, you can use the --platform option to pull your platform-specific image, but most container runtimes automatically select this for you.

For example:

docker pull confluentinc/cp-server:latest --platform linux/arm64

Related content