☸️ Kubernetes & KOPS: Production-Grade Hands-on Learning

A comprehensive, hands-on repository documenting the journey of mastering Kubernetes using KOPS (Kubernetes Operations). This repository bridges the gap between basic container orchestration and production-ready cluster management on AWS.

---

🚀 Overview

This project demonstrates a real-world DevOps workflow for deploying and managing a Kubernetes cluster. It covers everything from setting up a management server to complex networking concepts like Pod IP isolation.

Key Highlights

• Infrastructure as Code: Managing AWS clusters via KOPS manifests.
• Cluster Lifecycle: Dry-runs, updates, and zero-downtime scaling.
• Deep Dive: In-depth exploration of Namespaces, Pod types, and Control Plane internals.
• Real-time Production Concepts: Leveraging Auto Scaling Groups for high availability.

---

🛠️ Environment Setup & Prerequisites

Before diving into Kubernetes, we need to prepare our Management Server (Ubuntu 24.04 recommended).

1. Management Server Preparation

We use a dedicated EC2 instance (t3.medium) to manage the cluster. No direct login to the nodes is required—everything is handled from here.

• IAM Role: Attach an IAM role with AdministratorAccess to the Management Server.
• SSH Keys: Generate keys specifically for the cluster.

  ssh-keygen

2. Tool Installation

Install kubectl and kops to the system path:

# Install KOPS
wget https://github.com/kubernetes/kops/releases/download/v1.34.0/kops-linux-amd64
chmod +x kops-linux-amd64
sudo mv kops-linux-amd64 /usr/local/bin/kops

# Install kubectl
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

3. Shell Configuration

Boost productivity by adding aliases and environment variables to your ~/.bashrc:

# Environment Variables
export NAME=example-cluster.com
export KOPS_STATE_STORE=s3://my-k8s-state-bucket
export AWS_REGION=us-east-1
export CLUSTER_NAME=example-cluster.com

# Shortcuts
alias ku='kubectl'

# Apply changes
source ~/.bashrc

---

🏗️ KOPS Cluster Lifecycle

KOPS is the best tool for deploying production-grade clusters on AWS because it automatically configures Auto Scaling Groups (ASG). If a node fails or is deleted, the ASG will instantly recreate it.

Cluster Creation Steps

1. Dry Run Manifest Generation:

   kops create cluster --name=${NAME} \
     --state=${KOPS_STATE_STORE} --zones=us-east-1a,us-east-1b \
     --node-count=2 --control-plane-count=1 --node-size=t3.medium \
     --control-plane-size=t3.medium --ssh-public-key ~/.ssh/id_ed25519.pub \
     --dns-zone=${NAME} --dry-run --output yaml > cluster.yml

2. Manifest Customization: Modify cluster.yml to adjust subnet CIDRs for better isolation:

   • Subnet A: 172.20.1.0/24
   • Subnet B: 172.20.2.0/24

3. Deploy & Validate:

   kops create -f cluster.yml
   kops update cluster --name ${NAME} --yes --admin
   kops validate cluster --wait 10m

Verification

A healthy cluster should report:

• 1 Master Node
• 2 Worker Nodes

---

📦 Kubernetes Hands-on Guide

1. Smoke Testing (Sanity Check)

ku get nodes         # List all nodes
ku cluster-info      # Check cluster status
ku get ns            # List namespaces

2. Namespace & Isolation

Namespaces allow multiple teams (e.g., Alpha, Bravo, Charlie) to work on the same cluster without interference.

• Tip: Communication between namespaces can be restricted or allowed using Network Policies.

3. Pod Concepts

A Pod is the smallest deployable unit. It can contain:

• Single Container: One app, one container.
• Multi-Container: Main app + Sidecar (logs) or Proxy (networking).

4. Imperative vs. Declarative

Feature	Imperative	Declarative
Command	ku run testpod1 --image=nginx	ku apply -f pod.yaml
Pros	Quick, good for testing	Reproducible, version controlled
Real-time Tip	Use --dry-run=client -o yaml to generate manifests quickly.

5️⃣ Realtime Inline Deployment Approach

In realtime, I sometimes use an inline YAML method directly from the terminal:

echo '
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: testpod3
  name: testpod3
spec:
  containers:
    - image: nginx:latest
      name: testpod3
' | ku apply -f -

---

🔍 Under the Hood: Control Plane

Inspect the brain of your cluster in the kube-system namespace:

ku get pods -n kube-system -o wide | grep -E 'api|etcd|scheduler|controller'

Observation: All control plane components run on the same private IP address of the Master Node, ensuring low latency and tight integration.

---

📁 Repository Structure

.
├── cluster.yml         # KOPS cluster configuration
├── manifests/          # Kubernetes YAML declarations
│   └── testpod1.yml    # Sample Nginx Pod
├── scripts/            # Setup and utility scripts
└── README.md           # You are here!

---

👋 Conclusion

This repository serves as a blueprint for anyone transitioning from local K8s experimentation to cloud-native cluster management. By combining the power of KOPS with structured kubectl operations, we achieve a scalable and resilient infrastructure.

Happy Architecting! 🚀