dvcrn Create a Local Kubernetes Test Cluster
Overview
Create a local Kubernetes cluster for development and testing using Kind, Minikube, or k3d. All three tools run K8s inside Docker containers on the local machine.
Official docs: https://kubeblocks.io/docs/preview/user_docs/references/prepare-a-local-k8s-cluster
Tool Comparison
| Feature | Kind | Minikube | k3d |
|---|---|---|---|
| Engine | K8s in Docker | VM or Docker | k3s in Docker |
| Multi-node | Yes (via config) | Limited | Yes (via flags) |
| Resource usage | Low | Medium | Low |
| Startup speed | Fast | Slower | Fast |
| Best for | CI/CD, testing | Learning, local dev | Lightweight testing |
Default recommendation: Kind (widely used, lightweight, supports multi-node easily).
Workflow
- [ ] Step 1: Check Docker is running
- [ ] Step 2: Choose and install a tool
- [ ] Step 3: Create a cluster
- [ ] Step 4: Verify the cluster
Step 1: Check Docker
All three tools require Docker. Verify it is installed and running:
docker info > /dev/null 2>&1
If Docker is not available, tell the user to install it first:
- macOS/Windows: https://www.docker.com/products/docker-desktop/
- Linux: https://docs.docker.com/engine/install/
Step 2: Choose and Install a Tool
Ask the user which tool they prefer. If no preference, default to Kind.
Kind
# macOS
brew install kind
# Linux (amd64)
[ $(uname -m) = x86_64 ] && curl -Lo ./kind https://kind.sigs.k8s.io/dl/v0.24.0/kind-linux-amd64
chmod +x ./kind && sudo mv ./kind /usr/local/bin/kind
# Linux (arm64)
[ $(uname -m) = aarch64 ] && curl -Lo ./kind https://kind.sigs.k8s.io/dl/v0.24.0/kind-linux-arm64
chmod +x ./kind && sudo mv ./kind /usr/local/bin/kind
# Windows
choco install kind
Minikube
# macOS
brew install minikube
# Linux (rpm-based)
curl -LO https://storage.googleapis.com/minikube/releases/latest/minikube-latest.x86_64.rpm
sudo rpm -Uvh minikube-latest.x86_64.rpm
# Linux (deb-based)
curl -LO https://storage.googleapis.com/minikube/releases/latest/minikube_latest_amd64.deb
sudo dpkg -i minikube_latest_amd64.deb
# Windows
choco install minikube
k3d
# macOS
brew install k3d
# Linux
curl -s https://raw.githubusercontent.com/k3d-io/k3d/main/install.sh | bash
# Windows
choco install k3d
Step 3: Create a Cluster
Kind (default)
Single-node cluster (simplest):
kind create cluster --name kb-test
Multi-node cluster (recommended for KubeBlocks):
Create a config file kind-config.yaml:
kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
- role: control-plane
- role: worker
- role: worker
Then create the cluster:
kind create cluster --name kb-test --config kind-config.yaml
Minikube
# Single-node (default)
minikube start
# With Docker driver (recommended)
minikube start --driver=docker
# With more resources
minikube start --driver=docker --cpus=4 --memory=8192
k3d
# Single-node
k3d cluster create kb-test
# Multi-node (1 server + 2 agents)
k3d cluster create kb-test --servers 1 --agents 2
Step 4: Verify the Cluster
kubectl get nodes
Expected output (Kind multi-node example):
NAME STATUS ROLES AGE VERSION
kb-test-control-plane Ready control-plane 30s v1.31.0
kb-test-worker Ready <none> 20s v1.31.0
kb-test-worker2 Ready <none> 20s v1.31.0
Also verify kubectl context is set correctly:
kubectl config current-context
Delete the Cluster
When the cluster is no longer needed:
# Kind
kind delete cluster --name kb-test
# Minikube
minikube delete
# k3d
k3d cluster delete kb-test
Next Steps
Once the cluster is running, you can install KubeBlocks using the install-kubeblocks skill.