Amazon EKS — StatefulSets & PVCs

1 minute read

StatefulSets & Persistent Volume Claims

StatefulSets are the Kubernetes workload for stateful applications — databases, message queues, and any app that needs stable identity and persistent storage.

StatefulSet vs Deployment

Feature Deployment StatefulSet
Pod names Random (web-7abc123-xyz) Ordered (db-0, db-1, db-2)
Storage Shared or none Each pod gets its own PVC
Scaling All at once One at a time (ordered)
Deletion All at once Reverse order
DNS Via Service only Individual pod DNS

StatefulSet with EBS

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: postgres
spec:
  serviceName: postgres
  replicas: 3
  selector:
    matchLabels:
      app: postgres
  template:
    metadata:
      labels:
        app: postgres
    spec:
      containers:
        - name: postgres
          image: postgres:15
          ports:
            - containerPort: 5432
          volumeMounts:
            - name: data
              mountPath: /var/lib/postgresql/data
          env:
            - name: POSTGRES_PASSWORD
              valueFrom:
                secretKeyRef:
                  name: postgres-secret
                  key: password
  volumeClaimTemplates:
    - metadata:
        name: data
      spec:
        accessModes: ["ReadWriteOnce"]
        storageClassName: ebs-gp3
        resources:
          requests:
            storage: 20Gi

This creates:

  • postgres-0, postgres-1, postgres-2
  • Each with its own PVC: data-postgres-0, data-postgres-1, data-postgres-2

Headless Service for StatefulSet

apiVersion: v1
kind: Service
metadata:
  name: postgres
spec:
  clusterIP: None      # Headless!
  selector:
    app: postgres
  ports:
    - port: 5432

With headless Service, each pod gets a DNS name:

postgres-0.postgres.default.svc.cluster.local
postgres-1.postgres.default.svc.cluster.local
postgres-2.postgres.default.svc.cluster.local

Common Mistakes

Mistake Impact
Using Deployment for databases Pods share storage, data corruption
Forgetting headless Service Can’t address individual pods
Deleting PVCs accidentally Data loss (use Retain reclaim policy)
Not setting resource requests Pod evicted under memory pressure

What’s Next?

High-performance storage: FSx for Lustre.

You may also enjoy