How to Install Kubernetes Manually?

Kubernetes has become the de facto standard for container orchestration, enabling seamless deployment, scaling, and management of containerized applications. While there are automated installation tools available, understanding how to install Kubernetes manually provides a deeper insight into its architecture and components. In this guide, we will walk you through the step-by-step process of manually installing Kubernetes, empowering you with a comprehensive understanding of the installation procedure.

Prerequisites:

Before diving into the installation process, ensure that you have the following prerequisites in place:

1. Linux Distribution:

   • Kubernetes is best supported on Linux distributions. Choose one that fits your preferences, such as Ubuntu, CentOS, or Fedora.

2. Docker:

   • Install Docker to create, deploy, and run containers. Kubernetes relies heavily on containerization, and Docker is a popular choice.

Step 1: Install kubectl

kubectl is the command-line tool for interacting with a Kubernetes cluster. To install it, use the following commands:

# For Linux
sudo apt-get update && sudo apt-get install -y kubectl

# For macOS
brew install kubectl

# For Windows (using Chocolatey)
choco install kubernetes-cli

Step 2: Install kubelet and kubeadm

kubelet is the primary node agent that runs on each node in the cluster, while kubeadm is a command-line tool for bootstrapping Kubernetes clusters. Install them using the package manager associated with your Linux distribution:

# For Ubuntu
sudo apt-get update && sudo apt-get install -y kubelet kubeadm

# For CentOS
sudo yum install -y kubelet kubeadm

# For Fedora
sudo dnf install -y kubelet kubeadm

Step 3: Initialize the Master Node

Run the following command on the designated master node:

sudo kubeadm init

Follow the on-screen instructions, and at the end of the process, you'll be provided with a command to join worker nodes.

Step 4: Set Up kubectl for Cluster Access

Configure kubectl to access the cluster:

mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config

Step 5: Join Worker Nodes

On each worker node, run the command obtained from the master node initialization process:

sudo kubeadm join <MASTER_NODE_IP>:<MASTER_NODE_PORT> --token <TOKEN> --discovery-token-ca-cert-hash sha256:<CERT_HASH>

Verification:

Check the status of the cluster components:

kubectl get nodes
kubectl get pods --all-namespaces

More Examples:

Explore more configurations and features by referring to the official Kubernetes documentation: Kubernetes Documentation.

Now you have successfully installed Kubernetes manually and have a working cluster ready for deploying containerized applications.