Scale a Kubernetes Deployment manually with kubectl scale (or automatically with kubectl autoscale)

Documented steps

1. Check current state: kubectl get deployment <name> -n <namespace>
2. Scale: kubectl scale deployment/<name> --replicas=<N> -n <namespace>. The --replicas flag is required
3. Optional safety precondition: kubectl scale deployment/<name> --current-replicas=<M> --replicas=<N> — only scales if the current count matches M; fails otherwise
4. kubectl scale returns as soon as the desired count is set — it does NOT wait for pods to be ready. Follow with: kubectl rollout status deployment/<name> or kubectl get pods -w
5. For metrics-driven scaling instead, create an HPA: kubectl autoscale deployment <name> --min=2 --max=10 --cpu-percent=80
6. Check HPA state: kubectl get hpa <name> — TARGETS shows <unknown> if the metrics pipeline isn't working
7. Official docs: https://kubernetes.io/docs/reference/kubectl/generated/kubectl_scale/ and https://kubernetes.io/docs/concepts/workloads/controllers/deployment/#scaling-a-deployment