Kubernetes for Developers: What You Actually Need to Know

Kubernetes (K8s) has become the standard platform for deploying containerised applications at scale. Most developers encounter it through their company’s infrastructure but understand it poorly. Here is what actually matters.

What Kubernetes Does

Kubernetes manages containers across multiple machines (nodes). Given a desired state (5 instances of my application, each with 2 CPU, 4GB RAM, update them with zero downtime), Kubernetes continuously works to achieve and maintain that state — starting containers, restarting failed ones, distributing traffic, and managing updates. The core value: instead of manually managing individual servers, you describe what you want and Kubernetes handles how.

The Objects Developers Care About

Pod: One or more containers that share network and storage. The smallest deployable unit. Deployment: Manages a set of identical Pods — handles replicas, rolling updates, and rollbacks. This is usually what you create when deploying an app. Service: A stable network endpoint that routes traffic to Pods (which have ephemeral IPs). ConfigMap/Secret: Configuration and credentials injected into containers as environment variables or mounted files. Ingress: Exposes Services to the external internet via HTTP/HTTPS with routing rules.

The kubectl Commands You’ll Use Daily

kubectl get pods -n namespace    # list pods in namespace
kubectl logs pod-name -f          # stream pod logs
kubectl describe pod pod-name     # debug pod events and status
kubectl exec -it pod-name -- bash # shell into a running pod
kubectl apply -f deployment.yaml  # apply a configuration file
kubectl rollout status deployment/name  # check update progress
kubectl rollout undo deployment/name    # rollback to previous

When Not to Use Kubernetes

Kubernetes is significant operational overhead. For small teams and small applications, managed alternatives (Render, Railway, Fly.io, Google Cloud Run) provide similar capabilities with far less complexity. Kubernetes makes sense when: you have a dedicated platform/infrastructure team, you run at a scale that requires custom resource scheduling, or you need features (custom operators, fine-grained network policies) only K8s provides.