Getting Started with Kind for Local Kubernetes Development

Kind was originally created to facilitate Kubernetes' own conformance testing, but it has evolved into a powerful tool for local development and continuous integration workflows. By using Docker containers to simulate Kubernetes nodes, Kind offers a remarkably lightweight yet fully functional Kubernetes environment.

The name "Kind" stands for "Kubernetes IN Docker" which aptly describes how it works. Instead of virtualizing entire machines, Kind uses Docker containers to represent Kubernetes nodes. This approach dramatically reduces resource consumption while preserving the essential behavior of a real Kubernetes cluster.

For developers working on Kubernetes-native applications, Kind provides an ideal environment to test deployments, debug issues, and validate changes before committing them to production environments.

In this article, we'll discuss setting up a local Kubernetes development environment with Kind.

Prerequisites

Before diving into Kind, you'll need to ensure your development environment meets the necessary requirements:

• Docker: Since Kind runs nodes as containers, Docker is essential. You'll need a recent version of Docker installed and running on your system.

• kubectl: The Kubernetes command-line tool is necessary for interacting with your Kind clusters after they're created.

• Go: While not strictly required for using Kind, it's needed if you want to build Kind from source or contribute to its development.

For the optimal experience, ensure that Docker has enough resources allocated. On Docker Desktop, you can adjust resource limits in the application settings.

Installation

Installing Kind is straightforward across all supported platforms. Let's explore the various installation methods.

The most universal way to install Kind is through its pre-built binaries, which are available for Linux, macOS, and Windows.

For Linux and macOS users, you can download and install Kind with a simple command:

curl -Lo ./kind https://kind.sigs.k8s.io/dl/v0.20.0/kind-$(uname)-amd64

chmod +x ./kind && sudo mv ./kind /usr/local/bin/kind

This command downloads the appropriate Kind binary, makes it executable, and moves it to a directory in your PATH for easy access.

For macOS users with Homebrew, the installation is even simpler:

brew install kind

Windows users can install Kind using Chocolatey:

choco install kind

Or with PowerShell:

curl.exe -Lo kind-windows-amd64.exe https://kind.sigs.k8s.io/dl/v0.20.0/kind-windows-amd64

Move-Item .\kind-windows-amd64.exe c:\some-dir-in-your-PATH\kind.exe

Installing kubectl

Since you'll need kubectl to interact with your Kind clusters, make sure it's installed on your system.

On Linux, you can install kubectl via the package manager:

sudo apt-get update && sudo apt-get install -y kubectl

On macOS with Homebrew:

brew install kubectl

On Windows with Chocolatey:

choco install kubernetes-cli

Alternatively, you can download kubectl directly from the Kubernetes release page for any platform.

Verifying installation

After installing Kind and kubectl, verify that they're working correctly by checking their versions:

kind version

This should output something like:

kind v0.20.0 go1.20.4 linux/amd64

Similarly, check kubectl:

kubectl version --client

With both tools successfully installed, you're ready to create your first Kind cluster.

Creating your first cluster

Creating a Kind cluster is remarkably simple, especially compared to setting up traditional Kubernetes environments. Let's explore the basic creation process and some of the available configuration options.

The simplest way to create a Kind cluster is with the default configuration:

kind create cluster

This command creates a single-node Kubernetes cluster running inside a Docker container. The output will look something like this:

Creating cluster "kind" ...
✓ Ensuring node image (kindest/node:v1.27.3) 🖼
✓ Preparing nodes 📦
✓ Writing configuration 📜
✓ Starting control-plane 🕹️
✓ Installing CNI 🔌
✓ Installing StorageClass 💾
Set kubectl context to "kind-kind"
You can now use your cluster with:

kubectl cluster-info --context kind-kind

Have a question, bug, or feature request? Let us know! https://kind.sigs.k8s.io/#community 🙂

This output tells you that Kind has:

1. Downloaded the node image (if not already present)
2. Prepared the node container
3. Created the necessary configuration
4. Started the control plane components
5. Installed the Container Network Interface (CNI)
6. Set up the default storage class
7. Configured kubectl to use this cluster

The cluster name defaults to "kind" if not specified. You can verify that your cluster is running with:

kubectl cluster-info --context kind-kind

This should show information about your Kubernetes control plane and CoreDNS service.

Multi-node clusters

One of Kind's strengths is its ability to create multi-node clusters, which more closely resemble production environments. To create a multi-node cluster, you'll need to define a configuration file.

Create a file named multi-node.yaml with the following content:

kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
- role: control-plane
- role: worker
- role: worker

This configuration defines a cluster with one control plane node and two worker nodes. To create this cluster:

kind create cluster --name multi-node-cluster --config multi-node.yaml

After the cluster is created, you can verify that all nodes are running:

kubectl get nodes --context kind-multi-node-cluster

You should see output similar to this:

NAME                                 STATUS   ROLES           AGE     VERSION
multi-node-cluster-control-plane     Ready    control-plane   2m30s   v1.27.3
multi-node-cluster-worker            Ready    <none>          2m      v1.27.3
multi-node-cluster-worker2           Ready    <none>          2m      v1.27.3

Configuration options

Kind offers numerous configuration options to customize your clusters. Let's explore some of the most useful ones.

You can specify which Kubernetes version to use by defining the node image in your configuration file:

kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
- role: control-plane
 image: kindest/node:v1.25.11
- role: worker
 image: kindest/node:v1.25.11

This configuration creates a cluster using Kubernetes v1.25.11. Kind maintains images for various Kubernetes versions, which you can find in the Kind documentation.

Port mapping

To expose services in your Kind cluster to the host system, you can configure port mappings:

kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
- role: control-plane
 extraPortMappings:
 - containerPort: 30080
   hostPort: 8080
   protocol: TCP

This configuration maps port 30080 in the control plane node to port 8080 on your host. This is particularly useful for testing NodePort and Ingress services.

Extra mounts

You can also mount files or directories from your host into the Kind nodes:

kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
- role: control-plane
 extraMounts:
 - hostPath: /path/to/my/files
   containerPath: /files

This configuration mounts /path/to/my/files from your host to /files in the control plane node, allowing you to easily share configuration files or other resources with your Kind cluster.

Working with the Kubernetes dashboard

The Kubernetes Dashboard provides a web-based UI for managing and monitoring your cluster. While Kind doesn't include the dashboard by default, you can easily deploy it to enhance your development experience.

Deploying the dashboard

To deploy the Kubernetes Dashboard to your Kind cluster, you can use the official manifest:

kubectl apply -f https://raw.githubusercontent.com/kubernetes/dashboard/v2.7.0/aio/deploy/recommended.yaml

This command creates the necessary resources for the dashboard in a dedicated namespace called kubernetes-dashboard.

Creating a dashboard user

For security reasons, you'll need to create a user to access the dashboard. Create a file named dashboard-adminuser.yaml with the following content:

apiVersion: v1
kind: ServiceAccount
metadata:
 name: admin-user
 namespace: kubernetes-dashboard
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
 name: admin-user
roleRef:
 apiGroup: rbac.authorization.k8s.io
 kind: ClusterRole
 name: cluster-admin
subjects:
- kind: ServiceAccount
 name: admin-user
 namespace: kubernetes-dashboard

Apply this configuration:

kubectl apply -f dashboard-adminuser.yaml

Accessing the dashboard

To access the dashboard, you need to create a secure channel to your cluster using kubectl's proxy feature:

kubectl proxy

This command starts a proxy server on localhost, allowing you to access the dashboard at:

http://localhost:8001/api/v1/namespaces/kubernetes-dashboard/services/https:kubernetes-dashboard:/proxy/

To log in, you'll need a token. Generate one with:

kubectl -n kubernetes-dashboard create token admin-user

Copy the generated token and paste it into the dashboard login page.

Once logged in, you'll see the main dashboard interface, which provides a comprehensive view of your cluster resources.

Deploying applications to Kind

With your Kind cluster up and running, you can now deploy applications to test and validate their behavior in a Kubernetes environment.

Let's start with a simple deployment of NGINX. Create a file named nginx-deployment.yaml:

apiVersion: apps/v1
kind: Deployment
metadata:
 name: nginx-deployment
 labels:
   app: nginx
spec:
 replicas: 3
 selector:
   matchLabels:
     app: nginx
 template:
   metadata:
     labels:
       app: nginx
   spec:
     containers:
     - name: nginx
       image: nginx:1.21
       ports:
       - containerPort: 80
---
apiVersion: v1
kind: Service
metadata:
 name: nginx-service
spec:
 type: NodePort
 selector:
   app: nginx
 ports:
 - port: 80
   targetPort: 80
   nodePort: 30080

This configuration creates a deployment with three NGINX pods and a NodePort service that exposes them on port 30080. Apply this configuration:

kubectl apply -f nginx-deployment.yaml

Verify that the deployment is running:

kubectl get deployments

You should see output like:

NAME               READY   UP-TO-DATE   AVAILABLE   AGE
nginx-deployment   3/3     3            3           30s

And check that the service is created:

kubectl get services

Output:

NAME            TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
kubernetes      ClusterIP   10.96.0.1       <none>        443/TCP        10m
nginx-service   NodePort    10.96.131.245   <none>        80:30080/TCP   45s

If you created your cluster with port mapping as shown earlier, you should now be able to access the NGINX service at http://localhost:8080.

Using Ingress resources

For more sophisticated routing, you can use Kubernetes Ingress resources. First, you'll need to enable the Ingress controller in your Kind cluster.

Create a new configuration file named ingress-cluster.yaml:

kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
- role: control-plane
 kubeadmConfigPatches:
 - |
   kind: InitConfiguration
   nodeRegistration:
     kubeletExtraArgs:
       node-labels: "ingress-ready=true"
 extraPortMappings:
 - containerPort: 80
   hostPort: 80
   protocol: TCP
 - containerPort: 443
   hostPort: 443
   protocol: TCP

Create a new cluster with this configuration:

kind create cluster --name ingress-cluster --config ingress-cluster.yaml

Install the NGINX Ingress controller with specific patches for Kind:

kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/main/deploy/static/provider/kind/deploy.yaml

Wait for the Ingress controller to be ready:

kubectl wait --namespace ingress-nginx \
 --for=condition=ready pod \
 --selector=app.kubernetes.io/component=controller \
 --timeout=90s

Now, deploy a sample application:

apiVersion: apps/v1
kind: Deployment
metadata:
 name: hello-app
spec:
 replicas: 3
 selector:
   matchLabels:
     app: hello
 template:
   metadata:
     labels:
       app: hello
   spec:
     containers:
     - name: hello
       image: gcr.io/google-samples/hello-app:1.0
       ports:
       - containerPort: 8080
---
apiVersion: v1
kind: Service
metadata:
 name: hello-service
spec:
 selector:
   app: hello
 ports:
 - port: 80
   targetPort: 8080

Apply this configuration:

kubectl apply -f hello-app.yaml

Finally, create an Ingress resource to route traffic to your application:

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
 name: hello-ingress
spec:
 rules:
 - http:
     paths:
     - path: /hello
       pathType: Prefix
       backend:
         service:
           name: hello-service
           port:
             number: 80

Apply the Ingress configuration:

kubectl apply -f hello-ingress.yaml

Now you should be able to access your application at http://localhost/hello.

Volume mounting and persistent storage

For applications that require persistent storage, Kind supports Persistent Volumes and Persistent Volume Claims.

First, create a PVC:

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
 name: data-pvc
spec:
 accessModes:
   - ReadWriteOnce
 resources:
   requests:
     storage: 1Gi

Apply this configuration:

kubectl apply -f pvc.yaml

Now create a deployment that uses this PVC:

apiVersion: apps/v1
kind: Deployment
metadata:
 name: stateful-app
spec:
 replicas: 1
 selector:
   matchLabels:
     app: stateful
 template:
   metadata:
     labels:
       app: stateful
   spec:
     containers:
     - name: app
       image: nginx
       volumeMounts:
       - mountPath: /data
         name: data-volume
     volumes:
     - name: data-volume
       persistentVolumeClaim:
         claimName: data-pvc

Apply this configuration:

kubectl apply -f stateful-app.yaml

Kind's built-in storage provisioner will automatically create a Persistent Volume to satisfy the PVC, and your application will have access to persistent storage that survives pod restarts.

Managing Kind resources

One of Kind's strengths is its ability to run multiple clusters simultaneously, which is useful for testing multi-cluster scenarios or isolating different projects.

To list all your Kind clusters:

kind get clusters

You can switch between clusters by specifying the context in your kubectl commands:

kubectl config use-context kind-cluster-name

Or by providing the context flag:

kubectl --context kind-cluster-name get pods

When you're done with a cluster, you can delete it:

kind delete cluster --name cluster-name

Loading Docker images into Kind

One common development workflow involves building Docker images locally and using them in your Kind cluster. Since Kind runs in its own Docker network, you need to explicitly load images into the Kind cluster.

Build your image normally:

docker build -t my-app:latest .

Then load it into your Kind cluster:

kind load docker-image my-app:latest

You can specify which cluster to load the image into:

kind load docker-image my-app:latest --name cluster-name

Now you can create deployments that reference this image:

apiVersion: apps/v1
kind: Deployment
metadata:
 name: my-app
spec:
 replicas: 1
 selector:
   matchLabels:
     app: my-app
 template:
   metadata:
     labels:
       app: my-app
   spec:
     containers:
     - name: my-app
       image: my-app:latest
       imagePullPolicy: Never

Note the imagePullPolicy: Never, which tells Kubernetes to use the local image rather than trying to pull it from a registry.

Export logs from cluster

When troubleshooting, it's often useful to collect logs from your Kind cluster. Kind provides a command to export logs to a directory:

kind export logs ./kind-logs

This command collects logs from all nodes in the cluster and saves them to the specified directory. You can specify which cluster to collect logs from:

kind export logs ./kind-logs --name cluster-name

The collected logs include:

• Kubernetes component logs (API server, scheduler, etc.)
• Container runtime logs
• Node logs
• System logs

These logs can be invaluable for diagnosing issues in your cluster or the applications running on it.

Advanced Kind features

Kind offers several advanced features that make it a powerful tool for Kubernetes development and testing.

Custom node configurations

For advanced scenarios, you might need to customize the configuration of individual nodes. Kind allows you to apply kubeadm configuration patches to control various aspects of node initialization.

Here's an example of a custom node configuration:

kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
- role: control-plane
 kubeadmConfigPatches:
 - |
   kind: InitConfiguration
   nodeRegistration:
     kubeletExtraArgs:
       system-reserved: memory=2Gi
- role: worker
 kubeadmConfigPatches:
 - |
   kind: JoinConfiguration
   nodeRegistration:
     kubeletExtraArgs:
       eviction-hard: 'memory.available<10%'

This configuration customizes resource reservation on the control plane and eviction thresholds on the worker node. These sorts of customizations allow you to test how your applications behave under specific node configurations.

Integration with container registries

For more complex development workflows, you might want to use a container registry within your Kind cluster. You can deploy a local registry and configure Kind to use it:

First, create a Docker network for the registry and Kind to communicate:

docker network create kind-registry

Run a local registry:

docker run -d --name registry --network kind-registry -p 5000:5000 registry:2

Then create a Kind cluster configured to use this registry:

kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
containerdConfigPatches:
- |-
 [plugins."io.containerd.grpc.v1.cri".registry.mirrors."localhost:5000"]
   endpoint = ["http://registry:5000"]
nodes:
- role: control-plane
 extraPortMappings:
 - containerPort: 5000
   hostPort: 5000

Create the cluster:

kind create cluster --name registry-cluster --config registry-cluster.yaml

Connect the Kind cluster to the registry network:

docker network connect kind-registry registry-cluster-control-plane

Now you can push images to your local registry:

docker build -t localhost:5000/my-app:latest .
docker push localhost:5000/my-app:latest

And use them in your deployments:

apiVersion: apps/v1
kind: Deployment
metadata:
 name: my-app
spec:
 replicas: 1
 selector:
   matchLabels:
     app: my-app
 template:
   metadata:
     labels:
       app: my-app
   spec:
     containers:
     - name: my-app
       image: localhost:5000/my-app:latest

This setup allows you to test the full CI/CD workflow, including image pushing and pulling, within your local development environment.

Testing Kubernetes operators

Kind provides an excellent environment for testing Kubernetes operators—software extensions that use custom resources to manage applications and their components.

To test an operator in Kind, first create a cluster:

kind create cluster --name operator-test

Install the operator SDK's OLM (Operator Lifecycle Manager):

operator-sdk olm install

Deploy your operator:

kubectl apply -f operator-deployment.yaml

Create custom resources for your operator to manage:

kubectl apply -f custom-resource.yaml

You can then monitor the operator's behavior and verify that it correctly manages the custom resources in response to changes.

This approach allows operator developers to rapidly iterate on their code without needing to deploy to a remote cluster for each test cycle.

Final thoughts

As Kubernetes continues to evolve as the industry standard for container orchestration, Kind bridges the crucial gap between local development and production deployment, cementing its place as an essential tool in the modern developer's toolkit.

Thanks for reading!

Got an article suggestion? Let us know

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.