Kubernetes From Zero to Hero – Part 5: Deployments, ReplicaSets & Scaling Your Applications

---

A Quick Recap

In our previous blog, we explored:

• The Pod Lifecycle and common states like Pending, Running, CrashLoopBackOff

• Real-time debugging using kubectl describe, logs, and exec

• How to simulate pod failures and investigate them

• Tools and commands to effectively troubleshoot pods

That gave us the skills to manage and debug a single instance of our app. But modern apps rarely run with a single pod. What if it crashes? What if you need to scale?

That’s where today’s concepts come in — Deployments and ReplicaSets.

---

What is a Deployment?

A Deployment is a higher-level Kubernetes object that manages Pods for you.

It provides:

• Declarative updates (change state with YAML)

• Self-healing (if pod crashes, it recreates)

• Scaling (manually or auto)

• Rolling updates (zero downtime app updates)

You don’t create Pods directly in production—you use Deployments.

---

Behind the Scenes: Deployments & ReplicaSets

A Deployment creates a ReplicaSet, and that ReplicaSet manages the actual Pods.

Hierarchy: (Other Words we say Ranking)

Deployment → ReplicaSet → Pod

Each ReplicaSet ensures the right number of identical Pods are running.

---

Sample Deployment YAML

Let’s write a simple Nginx Deployment with 3 replicas.

nginx-deployment.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-deployment
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
        - name: nginx
          image: nginx:latest
          ports:
            - containerPort: 80

Apply it:

kubectl apply -f nginx-deployment.yaml

---

Explore What Happened

Check Deployments

kubectl get deployments

Check ReplicaSets

kubectl get rs

Check Pods

kubectl get pods -l app=nginx

You should see 3 pods running.

---

Scaling the Deployment

Need to increase capacity?

kubectl scale deployment nginx-deployment --replicas=5

Need to decrease?

kubectl scale deployment nginx-deployment --replicas=2

---

Rolling Updates in Action

Let’s update the Nginx version.

kubectl set image deployment/nginx-deployment nginx=nginx:1.21.0

Watch the rollout:

kubectl rollout status deployment/nginx-deployment

Check history:

kubectl rollout history deployment/nginx-deployment

Rollback if needed:

kubectl rollout undo deployment/nginx-deployment

---

Self-Healing in Action

Delete one pod:

kubectl delete pod <pod-name>

Run:

kubectl get pods

Kubernetes will automatically recreate the deleted pod to maintain the replica count.

---

Bonus: Autoscaling with HPA

You can even autoscale deployments based on CPU usage:

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

We’ll dive deeper into HPA (Horizontal Pod Autoscaler) in a future post.

---

Cleanup

To delete the deployment:

kubectl delete deployment nginx-deployment

This also deletes the ReplicaSet and all its pods.

---

What’s Next?

In the next blog, we’ll:

• Expose your app externally using a Service

• Understand ClusterIP, NodePort, and LoadBalancer

• Setup Ingress controllers for clean URL routing