How do I manage persistent storage in Kubernetes?

Managing persistent storage in Kubernetes takes a few moving parts. Here’s how the pieces fit together.

1. Understand Kubernetes Storage Concepts

Kubernetes abstracts storage through three main objects:

PersistentVolume (PV): A piece of storage in the cluster — local disk, NFS, iSCSI, cloud block volume, etc.
PersistentVolumeClaim (PVC): A pod’s request for storage. PVCs bind to PVs either statically or dynamically.
StorageClass: Defines how volumes are provisioned on demand (AWS EBS, GCE PD, Ceph, etc.).

2. Choose a Storage Provider

The backend can be a cloud service, on-prem hardware, or a distributed system:

Cloud: AWS EBS, Google Persistent Disk, Azure Disk
On-prem: VMware vSphere, NFS, iSCSI, SAN
Distributed: Ceph, GlusterFS, OpenEBS, Longhorn, Rook

Pick based on what your workloads need for performance, redundancy, and cost.

3. Configure Storage Classes

StorageClasses enable dynamic provisioning. Example for gp2 EBS volumes:

yaml
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: fast-storage
provisioner: kubernetes.io/aws-ebs
parameters:
  type: gp2
  fsType: ext4

The provisioner points to the backend; parameters tune the volume (disk type, filesystem, etc.).

4. Create PersistentVolumeClaims

Pods request storage via PVCs. This one asks for 10Gi on the fast-storage class:

yaml
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: my-pvc
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 10Gi
  storageClassName: fast-storage

The PVC binds to an existing PV or triggers dynamic provisioning when a StorageClass is named.

5. Use Volumes in Pods

Mount the PVC in a pod’s volume spec:

yaml
apiVersion: v1
kind: Pod
metadata:
  name: my-pod
spec:
  containers:
  - name: my-container
    image: nginx
    volumeMounts:
    - mountPath: /data
      name: my-volume
  volumes:
  - name: my-volume
    persistentVolumeClaim:
      claimName: my-pvc

The volume appears at /data inside the container.

6. Manage Data Retention (Reclaim Policies)

A PV’s reclaimPolicy decides what happens after its PVC is deleted:

Retain: Data stays; you clean up manually.
Delete: Underlying storage is removed.
Recycle: Scrubs and re-provisions (deprecated).

Example PV with Delete policy (using hostPath for demo only):

yaml
apiVersion: v1
kind: PersistentVolume
metadata:
  name: my-pv
spec:
  capacity:
    storage: 10Gi
  accessModes:
    - ReadWriteOnce
  persistentVolumeReclaimPolicy: Delete
  storageClassName: fast-storage
  hostPath:
    path: /mnt/data

7. Monitor Storage Usage

Track capacity and performance with Prometheus, Grafana, or your cloud provider’s monitoring.

8. Ensure High Availability and Scalability

For production workloads:

Run distributed storage (Ceph, Rook, Longhorn) for replication and fault tolerance.
Spread across zones or clusters for disaster recovery.
Back up with Velero or an external solution.

9. Backup and Disaster Recovery

Set up a backup cadence:

Velero can snapshot and restore Kubernetes resources and persistent volumes.
Ship backups to object storage (S3, Azure Blob, etc.) on a schedule.

10. Security and Access Control

Lock down storage access:

Use RBAC to limit who can create or bind PVCs.
Encrypt volumes — cloud provider encryption or tools like Vault.
Enforce read-only mounts where appropriate via pod security standards.