Skip to main content
In this lesson you’ll learn how to set up a Kubernetes cluster locally or access one remotely. We cover common local tools, managed cloud services, and browser-based playgrounds so you can choose the fastest path to start learning and experimenting with Kubernetes. There are several ways to get started with Kubernetes:
  • Install locally on your laptop or VMs using tools like Docker Desktop, Minikube, or kubeadm.
  • Use a managed Kubernetes service from a cloud provider such as GKE, EKS, or AKS.
  • Use an online playground that provides instant access to hosted Kubernetes clusters for quick practice and labs.
Each approach has trade-offs in ease, realism, and resource requirements. Choose based on your goals: quick experiments, local development, or learning production-style multi-node clusters.
A slide titled "Setting Up Kubernetes" showing three columns of options: local tools (Docker Desktop, Minikube, Kubeadm), cloud managed services (GKE, EKS, AKS), and KodeKloud Playgrounds with the URL kodekloud.com/k8s.
Summary comparison Docker Desktop Docker Desktop provides the quickest local Kubernetes experience for many users because it bundles a single-node Kubernetes control plane with Docker itself. After installation, enable Kubernetes in Docker Desktop’s Settings to create a local cluster and a kubeconfig entry so kubectl can access it.
  • Platforms: macOS (Intel & Apple Silicon), Windows (x86_64; ARM support varies), Linux (Docker Desktop available on some distros).
  • Use case: quick development and manifest testing without managing a VM or separate hypervisor.
A screenshot of the Docker Desktop Settings screen focused on the Kubernetes tab. The page shows the "Enable Kubernetes" toggle (currently off) and related sidebar options like Resources, Docker Engine, and Extensions.
Once enabled, Docker Desktop updates or writes a kubeconfig entry (usually merged into ~/.kube/config) so kubectl commands target the local cluster. This is an excellent starting point for developers who want low-friction access to Kubernetes. KodeKloud Public Playgrounds If you prefer browser-based access to a hosted Kubernetes cluster, public playgrounds like KodeKloud let you launch one-click clusters with a terminal in the browser. These are ideal for learning, completing coursework, or comparing behavior across Kubernetes versions without installing anything locally.
  • Benefits: instant access, preconfigured tasks, multiple Kubernetes versions.
  • Limitations: ephemeral sessions and potential concurrency limits.
A screenshot of the KodeKloud "Public Playgrounds" interface showing a Kubernetes multi-node clusters card on the left, task instructions in the center, and a terminal on the right with ASCII KodeKloud branding.
Playgrounds are particularly useful when you want to follow tutorials, run lab tasks, or avoid local configuration and resource usage. Minikube Minikube creates a single-node Kubernetes cluster on your workstation. It packages the control plane (API server, etcd, controller-manager, scheduler) and node components (kubelet, container runtime) into an image and boots it either inside a VM or as containers, depending on the driver you choose. Key points:
  • Drivers: VM-based drivers (VirtualBox, Hyper-V, KVM) or container-based drivers (Docker, Podman).
  • Minikube handles downloading the VM/ISO and starting the node for you when using VM drivers.
  • For container drivers, Minikube runs the components as containers on the host runtime.
A schematic diagram of Minikube showing a cloud icon connected to a VirtualBox-hosted single-node Kubernetes cluster. Client tools like minikube.exe and kubectl are shown linking into the cluster.
Typical Minikube setup requirements
Prerequisites checklist:
Example: start Minikube and verify cluster access
What these commands do:
  • minikube start downloads and boots the Minikube VM or starts containers depending on the driver.
  • kubectl get nodes confirms the single node is Ready and accessible.
  • kubectl version --client shows the kubectl client version installed locally.
From here you can create Deployments, Services, ConfigMaps, and other Kubernetes resources locally to learn how Kubernetes scheduling, networking, and resource management work. Further reading and resources Use the approach that matches your learning goals: lightweight and instant (playground), local development (Docker Desktop / Minikube), or production-like multi-node clusters (kubeadm or managed cloud services).

Watch Video