Deployments

Relax. We'll talk through this in plain words — no textbook voice.

What you'll walk away with

Understand Deployments without any of the intimidation
Get comfortable running kubectl commands and YAML yourself
Apply this concept right away in a real project

Take a moment to think about this

A Deployment manages Pods indirectly, through a 'ReplicaSet' — you tell it 'I always want 3 pods running from this image' as the desired state, and the Deployment keeps that state maintained at all times. If one Pod crashes, the Deployment immediately creates a new one to fill in (so you're back to 3 out of 3). Updating a Deployment (say, to a new image version) also triggers an automatic rolling update (covered in a later chapter).

Let's connect this to a real scenario

Writing `replicas: 3` in a Deployment YAML keeps 3 pods running at all times. If you manually delete one pod with `kubectl delete pod`, you'll see the Deployment immediately create a replacement — a hands-on demo that proves the Deployment's self-healing power for yourself.

Deployment = a resource that gives Pods self-healing + scaling
replicas: N = the desired state of always having N pods running
Even if you delete a pod, the Deployment creates a new one to replace it

Let's look at it together

yaml

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

You should see

$ kubectl apply -f deployment.yaml
deployment.apps/web-deployment created

$ kubectl get pods
NAME                              READY   STATUS    RESTARTS   AGE
web-deployment-7d9f8c6b5-abc12    1/1     Running   0          5s
web-deployment-7d9f8c6b5-def34    1/1     Running   0          5s
web-deployment-7d9f8c6b5-ghi56    1/1     Running   0          5s

5-minute try-it

Apply the Deployment above and watch 3 pods running — then delete one with `kubectl delete pod <name>` and confirm it reappears on its own (self-healing).

A quick word of caution

Deleting a Deployment (`kubectl delete deployment`) takes down all the pods it controls too — if you only want to delete a single pod, target the pod directly (not the Deployment).

Easy traps

Writing a Deployment's metadata.labels and selector.matchLabels so they don't match — they need to match for the Deployment to know which pods are its own
Setting the replicas count without matching it to actual production traffic (too few or too many)

Now try it yourself

Apply the Deployment above and watch 3 pods running — then delete one with `kubectl delete pod <name>` and confirm it reappears on its own (self-healing).

You'll know it worked when: $ kubectl apply -f deployment.yaml deployment.apps/web-deployment created $ kubectl get pods NAME READY STATUS RESTARTS AGE web-deployment-7d9f8c6b5-abc12 1/1 Running 0 5s web-deployment-7d9f8c6b5-def34 1/1 Running 0 5s web-deployment-7d9f8c6b5-ghi56 1/1 Running 0 5s