Skip to main content


In this workshop, you will deploy Kratix on a local Kubernetes cluster, and deploy Kratix workloads on another Kubernetes cluster. You will need the following tools:

  1. kind CLI / Kubernetes-in-Docker (KinD):
    Used to create and manage local Kubernetes clusters in Docker.
    See the quick start guide to install.

  2. docker CLI / Docker:
    Used to orchestrate containers. kind requires that you have Docker installed and configured.
    See Get Docker to install.

  3. kubectl / Kubernetes command-line tool:
    The CLI for Kubernetesallows you to run commands against Kubernetes clusters.
    See the install guide.

  4. yq / YAML parsing command-line tool:
    The CLI for Kubernetesallows you to run parse and transform YAML files. This is only used in Part II.
    See the install guide.


You can run the workshop without KinD. Ideally, you will have access to two Kubernetes clusters.

One cluster will be the Platform cluster, where Kratix will be installed. Whenever you see --context kind-platform, replace it with the context of your platform cluster.

The other worker will be the Worker cluster. Whenever you see --context kind-worker, replace it with the context of your worker cluster.

If you want to try it on a single cluster, you can omit the --context flag and argument entirely.

Docker Resources

In order to complete all tutorials in this series, you must allocate enough resources to Docker. Ensure you allocate at least:

  • 5 CPU
  • 12GB Memory
  • 4GB swap

This can be managed through your tool of choice (e.g. Docker Desktop, Rancher, etc).

Delete pre-existing kind clusters

Ensure no clusters are currently running:

$ kind get clusters
No kind clusters found.

If you have clusters named platform or worker please delete them with:

kind delete clusters platform worker

You are now ready to start. Navigate to the next section and get started!