Skip to main content
Version: 8.5

Local Kubernetes cluster

You can deploy Camunda 8 Self-Managed on your Kubernetes local cluster for development purposes using kind.

In this guide, we will use kind. However, the concept is the same for other tools like K3s, minikube, or MicroK8s. The goal in this guide is to reduce the resources required by Camunda components so they can work on a personal machine.


Based on your system, install the CLI tools used in this guide if you don't already have them:

Create a local Kubernetes cluster

At this stage, you should consider how to connect to the future Camunda 8 cluster. If you are setting up Camunda 8 for the first time, we recommend using port-forwarding which will be described later in this article, but for now, continue to create the kind cluster.

If you are familiar with Camunda 8 deployment and are looking to start process automation development, consider using Ingress. The first step will be to complete the Ingress configuration section prerequisites in connect to Camunda 8 components before continuing.

If you have not done so already, create a local Kubernetes cluster with the following command:

kind create cluster --name camunda-platform-local

Next, switch to the new cluster context using the following command:

kubectl config use-context kind-camunda-platform-local


Now it's time to deploy Camunda 8 on the local Kubernetes cluster:

  1. Add the Camunda 8 Helm repository using the following command:
helm repo add camunda
helm repo update
  1. Download the Camunda 8 Helm chart values file designed for the kind cluster: camunda-platform-core-kind-values.yaml.

If you are deploying Camunda 8 with Ingress configuration, make sure to add additional values to the file you just downloaded camunda-platform-core-kind-values.yaml as described in connecting to Camunda 8 components.

  1. Install Camunda 8 using the camunda-platform-core-kind-values.yaml file you downloaded previously. This file might contain additional values if you are adding Ingress, TLS, or using a variety of other configuration properties. See Camunda Helm chart parameters.

  2. Execute the following command:

helm install camunda-platform camunda/camunda-platform \
-f camunda-platform-core-kind-values.yaml

This will deploy Camunda 8 components (Optimize, Connectors, and Zeebe), but with a set of parameters tailored to a local environment setup.

Depending on your machine hardware and internet connection speed, the services might take some time to get started as it will download the Docker images of all Camunda 8 components to your local kind cluster.

  1. Check that each pod is running and ready with kubectl get pods. If one or more of your pods are pending for long time, it means it cannot be scheduled onto a node. Usually, this happens because there are insufficient resources that prevent it. Use the kubectl describe <POD NAME> command to check its status.

Connecting to Camunda 8 components

Camunda services deployed in a Kubernetes cluster are not accessible from outside the cluster. To connect to your Camunda 8 cluster, use either port-forwarding or Kubernetes Ingress.


The setup described here skips Identity setup and uses a default basic authentication with username and password as 'demo/demo'.

To interact with the Camunda services inside a Kubernetes cluster without Ingress setup, you can use the kubectl port-forward command to route traffic from your local machine to the services running in the kind cluster. This is useful for quick tests or for development purposes.

First, port-forward each of the components. Use a separate terminal for each command. Then, each component can be accessed at http://localhost:PORT (for example, Optimize will have the address http://localhost:8083). To get a full list of port mappings, run helm status camunda-platform.

Connecting to the workflow engine

To interact with the Camunda workflow engine via Zeebe Gateway using zbctl or a local client/worker from outside the Kubernetes cluster, run kubectl port-forward to the Zeebe gateway as follows:

kubectl port-forward svc/camunda-zeebe-gateway 26500:26500

The command helm status camunda-platform will print port-forward command examples for each deployed Camunda 8 component as reference.


To get a full list of the deployed Camunda components and their network properties, run kubectl get services. In Kubernetes, a service is a method for exposing a network application that is running as one or more pods in your cluster.


If you don't need the cluster anymore, you can just delete the local KIND cluster:


This is a destructive action and will destroy all data of Camunda 8 in the local development cluster.

kind delete cluster --name camunda-platform-local

For more details about deployment options, visit the full Helm deployment guide.