Autoscaling and Resource Management

Requests and Limits

Every container should declare how much CPU and memory it needs (requests) and its hard ceiling (limits). The scheduler uses requests; limits protect neighbours:

resources:
  requests:
    cpu: 250m
    memory: 256Mi
  limits:
    cpu: 500m
    memory: 512Mi

Horizontal Pod Autoscaler (HPA)

The HPA changes the number of replicas based on observed CPU or custom metrics:

apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: web-hpa
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: web
  minReplicas: 2
  maxReplicas: 10
  metrics:
  - type: Resource
    resource:
      name: cpu
      target:
        type: Utilization
        averageUtilization: 60

kubectl apply -f hpa.yaml
kubectl get hpa

Vertical and Cluster Autoscaling

• VPA - adjusts CPU/memory requests of pods based on usage.
• Cluster Autoscaler - adds/removes whole nodes as pods can't fit.

Key Points

• Always set requests and limits on every container.
• HPA scales pods; Cluster Autoscaler scales nodes.
• Test autoscaling with load before relying on it in production.