Deployments – Keep Your Pizzas Running Automatically 🍕¶

In the last section, we created a single Pod using kubectl run.
It worked : but if it crashes or is deleted, it’s gone forever.

In real life, we want apps to:

• Stay running even if something goes wrong
• Automatically restart if a Pod dies
• Scale up or down when demand changes
• Update to a new version without downtime

A Deployment does exactly that.

🎯 Goal for this Section¶

Create a Deployment that:

• Runs 3 copies of the nginx web server
• Automatically restarts Pods if one fails
• Allows easy scaling up or down
• Uses a small YAML file to define desired state

Think back to the pizza chain analogy:
Deployment = “Always keep 3 margherita pizzas ready. If one burns or demand increases, make more. When the recipe improves, roll out the new version gradually.”

Step 1: Create a Deployment YAML File ✏️¶

In your terminal, create a file called nginx-deployment.yaml using any editor (Notepad, VS Code, etc.):

id="nginx-deployment-yaml"
apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-deployment
spec:
  replicas: 3                # Always keep 3 pizzas ready
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.25     # Use a specific version for stability
        ports:
        - containerPort: 80

Plain English explanation:

• Create a Deployment named nginx-deployment
• Always run 3 replicas (3 Pods)
• Use the label app: nginx so Kubernetes knows which Pods belong to this Deployment
• Each Pod runs one container using nginx:1.25
• Listen on port 80 inside the container
• Save this file in the same folder where you’re running commands.

Step 2: Apply the Deployment 🚀¶

Apply the YAML to create the Deployment:

kubectl apply -f nginx-deployment.yaml

Expected output:

deployment.apps/nginx-deployment created

💡 kubectl apply creates or updates resources defined in YAML files.

Step 3: Check That It’s Working 👀¶

Check the Deployment status:

kubectl get deployments

Sample output:

NAME                 READY   UP-TO-DATE   AVAILABLE   AGE
nginx-deployment     3/3     3            3           45s

• READY 3/3 → All 3 desired Pods are running and healthy

Check the actual Pods:

kubectl get pods

Expected output (names will vary):

NAME                                  READY   STATUS    RESTARTS   AGE
nginx-deployment-5d8f7c6b9f-abcde     1/1    Running   0          50s
nginx-deployment-5d8f7c6b9f-defgh     1/1    Running   0          50s
nginx-deployment-5d8f7c6b9f-ijklm     1/1    Running   0          50s

✅ All Pods share the same prefix : Kubernetes manages them automatically.

Step 4: Scale the Deployment 📈¶

Change the number of pizzas (Pods) with kubectl scale:

kubectl scale deployment nginx-deployment --replicas=5
kubectl get pods

You should now see 5 Pods running.

Scale back down if desired:

kubectl scale deployment nginx-deployment --replicas=2
kubectl get pods

Step 5: Clean Up 🧹¶

Delete the Deployment (also deletes all its Pods):

kubectl delete -f nginx-deployment.yaml

Or delete by name:

kubectl delete deployment nginx-deployment

Verify it’s gone:

kubectl get pods

🧠 What We Learned

• A Deployment manages multiple Pods for you
• You define the desired state in YAML (e.g., 3 replicas)
• Kubernetes ensures reality matches that state:
• Creates Pods
• Restarts them if they crash
• Scales automatically on command
• kubectl apply → create or update resources
• kubectl get deployments and kubectl get pods → monitor status
• Scaling is instant with kubectl scale
• This is how real applications run in Kubernetes : not single Pods, but Deployments.

Next Step 🚀¶

Right now, your nginx Pods are running inside the cluster, but they are not reachable from your browser. We need a Service to provide a stable entry point : the “single phone number” for ordering pizza.

→ Continue to Part 5: Services – Make Your App Accessible