containerization-and-devops

Class 11 - Introduction to Kubernetes & Local Setup (Hands-on)

In this class, we covered the fundamentals of Kubernetes, why it is needed for container orchestration, its core architecture, and how to set up a local Kubernetes cluster using k3d and kubectl on WSL.

1. Why Kubernetes?

Modern applications use multiple services (backend, frontend, database, cache, message queues, monitoring). Without orchestration, managing these becomes difficult:

Problem Description
Startup Order Services must start in correct sequence
Scaling Handle traffic spikes dynamically
Crash Recovery Restart failed containers automatically
Networking Containers need to discover and talk to each other
Rolling Updates Deploy new versions without downtime

Kubernetes automates all of this — deployment, scaling, self-healing, service discovery, load balancing, and rolling updates.

2. Core Architecture

Control Plane (Master)

| Component | Role | |—|—| | API Server | Entry point for all commands | | Scheduler | Decides where containers run | | Controller Manager | Maintains desired state (e.g., keeps 3 replicas running) | | etcd | Key-value store for cluster data |

Worker Nodes

| Component | Role | |—|—| | Kubelet | Communicates with control plane | | Container Runtime | Runs containers (Docker, containerd) | | Kube Proxy | Handles networking |

3. Local Kubernetes Tools

Tool Description Best For
Minikube Single-node cluster Beginners
k3s Lightweight Kubernetes Edge/IoT
k3d k3s inside Docker Fast local dev
kind Kubernetes in Docker CI/CD testing

Our stack: WSL + kubectl + k3d — lightweight, fast, and runs entirely inside Docker.

4. Installing kubectl

Download, make executable, and move to path:

curl -LO https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl
chmod +x kubectl
sudo mv kubectl /usr/local/bin/
kubectl version --client

Terminal Output

kubectl Installation

5. Installing k3d

curl -s https://raw.githubusercontent.com/k3d-io/k3d/main/install.sh | bash
k3d version

Terminal Output

k3d Installation

6. Creating a Kubernetes Cluster

k3d cluster create mycluster
kubectl get nodes

k3d creates a single-node cluster running k3s inside Docker. The kubectl get nodes command shows the node with STATUS: Ready.

Terminal Output

Cluster Creation and Nodes

7. Running a Pod

A Pod is the smallest deployable unit in Kubernetes. Here we run a standalone nginx pod:

kubectl run nginx --image=nginx
kubectl get pods

Terminal Output

Running First Pod

8. Inspecting a Pod

Use describe to see full details and logs to see container output:

kubectl describe pod nginx
kubectl logs nginx

Terminal Output

Pod Describe and Logs

9. Creating and Scaling a Deployment

A Deployment manages replicas and rolling updates. Here we create one and scale to 3 replicas:

kubectl create deployment web --image=nginx
kubectl get deployments
kubectl scale deployment web --replicas=3
kubectl get pods

After scaling, we can see 4 pods total — 1 standalone nginx pod + 3 web-xxxxx deployment pods.

Terminal Output

Deployment and Scaling

10. Exposing the Application

Expose the deployment as a NodePort service to make it accessible:

kubectl expose deployment web --port=80 --type=NodePort
kubectl get services

Terminal Output

Expose Service

11. Port Forwarding & Browser Access

Use port-forward to access the service from localhost:

kubectl port-forward service/web 8080:80

Open browser at http://localhost:8080 — the Nginx welcome page confirms the app is running!

Terminal Output & Browser

Port Forward and Nginx Welcome Page

12. Exploring kubeconfig

The kubeconfig file (~/.kube/config) contains cluster connection details. It has 3 parts:

# 1. Clusters — server address & CA
clusters:
- name: k3d-mycluster
  cluster:
    server: https://0.0.0.0:43925

# 2. Users — credentials
users:
- name: admin@k3d-mycluster
  user:
    client-certificate-data: DATA+OMITTED

# 3. Contexts — links cluster + user
contexts:
- name: k3d-mycluster
  context:
    cluster: k3d-mycluster
    user: admin@k3d-mycluster

current-context: k3d-mycluster

kubectl config view
kubectl config get-contexts
kubectl config current-context

How kubectl Works Internally

kubectl command
  → reads kubeconfig
    → finds current context
      → gets cluster address + user credentials
        → connects to API server
          → executes command

Terminal Output

Kubeconfig Exploration

13. Cleanup

Delete all resources and destroy the cluster:

kubectl delete pod nginx
kubectl delete deployment web
kubectl delete service web
k3d cluster delete mycluster

Terminal Output

Cleanup

Summary

Concept What We Did
kubectl Installed CLI tool to interact with Kubernetes
k3d Created lightweight local cluster inside Docker
Pod Ran a standalone nginx container
Deployment Created managed replicas with scaling
Service Exposed the app with NodePort
Port Forward Accessed nginx from browser at localhost:8080
kubeconfig Explored cluster, user, and context configuration
← Previous Class Theory Index