Managing Amazon EBS Volumes and Snapshots Across Regions

 

🌩️ Lab Environment on Amazon Web Services using Amazon EC2

CPU   → 2–4 cores

RAM   → 4–6 GB

Disk  → 30–40 GB

Create:

• Ubuntu 22.04

• t2.medium

• SSH login

ssh ubuntu@<ip>

---

---

🔹 STEP 1 — Install Docker

✅ Why?

Minikube runs Kubernetes inside Docker containers

sudo apt update
sudo apt install docker.io -y
sudo systemctl enable docker
sudo systemctl start docker
sudo usermod -aG docker $USER
newgrp docker

Check:

docker ps

👉 works = good

---

---

🔹 STEP 2 — Install kubectl

✅ Why?

kubectl = control Kubernetes cluster

sudo snap install kubectl --classic
kubectl version --client

---

---

🔹 STEP 3 — Install Minikube

🤔 Why Minikube is necessary?

Before Minikube:
❌ need 3–5 servers
❌ costly
❌ complex

With Minikube:
✅ single node cluster
✅ local testing
✅ fast learning
✅ safe experiments

👉 Think:

Minikube = practice ground
EKS/Prod = real battlefield

Install:

curl -LO https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64
chmod +x minikube-linux-amd64
sudo mv minikube-linux-amd64 /usr/local/bin/minikube

Start:

minikube start --driver=docker

Check:

kubectl get nodes

Observe:

minikube   Ready

🎉 Cluster ready

---

---

🧠 Now Kubernetes Objects (one by one)

---

📦 POD

✅ What?

Smallest deployable unit

✅ Why?

Container cannot run alone → must be inside Pod

Think:

Docker container = engine
Pod = vehicle

---

Commands

kubectl run nginx --image=nginx --restart=Never
kubectl get pods
kubectl describe pod nginx
kubectl logs nginx
kubectl delete pod nginx

---

Observe

Create → Running
Delete → Gone permanently

❌ No self-healing

---

Learning

Pod = testing only
Not production

---

---

🔁 REPLICASET

✅ What?

Keeps fixed number of Pods

✅ Why?

If server crashes → auto recreate

---

Commands

kubectl create deployment nginx-rs --image=nginx --replicas=3
kubectl get pods

Delete one:

kubectl delete pod <name>

Observe:
👉 new pod auto-created

---

Learning

ReplicaSet = self-healing

---

---

🚀 DEPLOYMENT

✅ What?

Manages ReplicaSets

✅ Why?

Because we need:
✅ rolling update
✅ rollback
✅ scaling

ReplicaSet cannot update safely

---

Commands

kubectl create deployment nginx --image=nginx
kubectl scale deployment nginx --replicas=5
kubectl get pods -w

Observe:
pods 1 → 5

---

Update:

kubectl set image deployment/nginx nginx=nginx:latest
kubectl rollout status deployment nginx

Rollback:

kubectl rollout undo deployment nginx

---

Learning

Deployment = production controller

---

---

📊 METRICS SERVER

✅ What?

Collects:

• CPU

• Memory

✅ Why?

Required for:

• kubectl top

• HPA

Without metrics → no autoscaling

---

Install

minikube addons enable metrics-server

Check:

kubectl top nodes
kubectl top pods

Observe:
CPU/memory values

---

---

⚡ HPA (Horizontal Pod Autoscaler)

✅ What?

Auto increase/decrease pods

✅ Why?

Traffic changes daily

Without HPA:
❌ waste money
❌ app crashes

With HPA:
✅ scale automatically
✅ cost efficient

---

🔥 How to do HPA (step-by-step)

Step 1 — Create deployment first

kubectl create deployment nginx --image=nginx

---

Step 2 — Attach autoscaler

kubectl autoscale deployment nginx --cpu-percent=50 --min=2 --max=10

---

Step 3 — Monitor continuously

Open 3 terminals:

T1

kubectl get hpa -w

T2

kubectl get pods -w

T3 generate load

kubectl run -it load --image=busybox -- sh

Inside:

while true; do wget -q -O- http://nginx; done

---

Observe

Pods:

2 → 4 → 7 → 10

Stop load:

10 → 5 → 2

---

Commands to repeatedly check

kubectl get hpa
kubectl describe hpa nginx
kubectl top pods
kubectl get deploy
kubectl get pods -w

---

Learning

Metrics → HPA → Deployment → Pods scale

---

---

🖥️ DAEMONSET

✅ What?

Runs 1 pod on every node

✅ Why?

Some apps must run everywhere

Examples:

• logging

• monitoring

• security agents

---

Commands

kubectl create daemonset ds-nginx --image=nginx
kubectl get pods -o wide

Observe:
1 pod per node

---

Learning

DaemonSet = node-level service

---

---

🎨 kubecolor

Why?

Better readability

Install:

sudo apt install kubecolor -y
alias kubectl=kubecolor

Now outputs colorful

---

---

🧠 FINAL BIG PICTURE (memory trick)

Minikube → cluster
Pod → run container
ReplicaSet → maintain count
Deployment → update safely
Metrics → measure usage
HPA → scale automatically
DaemonSet → run everywhere
kubecolor → readable CLI

---

You’ve basically covered 80% of real Kubernetes used in companies already 😄🔥