In Kubernetes, Pods are ephemeral → if a Pod crashes or moves to another node, its local data is lost.

• PV + PVC solve this → by attaching persistent storage that survives Pod restarts, rescheduling, and even cluster upgrades.
• Think of it as:
  • Pod = your app
  • PVC = app’s request (“I need 20Gi storage”)
  • PV = actual storage resource (EBS, Ceph, NFS, etc.)

PV & PVC — the idea in 6 bullets

• PersistentVolume (PV): A cluster resource that represents actual storage (EBS, GCE PD, Ceph, NFS, local SSD, etc.).
• PersistentVolumeClaim (PVC): An app’s request for storage (size, access mode, class). The control plane binds PVC → PV.
• StorageClass: Describes how to provision storage (type, parameters). With dynamic provisioning, a PV is created on demand.
• Mounting: Pods mount the PVC to a path (e.g., /data). If the Pod restarts or is rescheduled, the data stays on the PV.
• Access Modes: ReadWriteOnce (most common), ReadWriteMany (shared across Pods), ReadOnlyMany.
• Backups & retention: Use snapshots/backup tools for PVs; deleting a PVC may or may not delete the underlying PV depending on the reclaimPolicy.

Real-time example

Order Service writes invoices to /data.

• The Deployment mounts a PVC of 20Gi from the gp3 StorageClass (AWS EBS).
• A rollout causes Pods to restart on new nodes. The data persists because the PVC remains bound to the PV, which points to the same EBS volume.
• If the Pod crashes, Kubernetes restarts it; when it comes back, the same PVC is mounted and files are intact.

Minimal YAML (StorageClass → PVC → Pod/Deployment)

# 1) StorageClass (example: AWS EBS gp3). Use your cloud's provisioner.
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: gp3
provisioner: ebs.csi.aws.com
parameters:
  type: gp3
reclaimPolicy: Delete          # or Retain (keeps the volume after PVC deletion)
allowVolumeExpansion: true
volumeBindingMode: WaitForFirstConsumer
---
# 2) PVC: app requests persistent storage
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: orders-pvc
spec:
  accessModes:
    - ReadWriteOnce
  storageClassName: gp3
  resources:
    requests:
      storage: 20Gi
---
# 3) Deployment mounting the PVC at /data
apiVersion: apps/v1
kind: Deployment
metadata:
  name: order-service
spec:
  replicas: 3
  selector:
    matchLabels:
      app: order-service
  template:
    metadata:
      labels:
        app: order-service
    spec:
      containers:
        - name: app
          image: ghcr.io/example/order-service:1.0.0
          ports:
            - containerPort: 8080
          volumeMounts:
            - name: data
              mountPath: /data
          # optional: probes omitted for brevity
      volumes:
        - name: data
          persistentVolumeClaim:
            claimName: orders-pvc

Business Use Cases

1. Databases (High Priority Use Case)

• Example: MySQL, PostgreSQL, MongoDB, Cassandra
• Reason: Databases must store data persistently, otherwise every restart wipes records.
• PVC ensures DB data stays on disk volumes even if Pods die.

---

2. E-commerce Applications

• Example: Order Service writes invoices to /data.
• Checkout microservice needs to save PDF invoices, receipts, or cart session data.
• With PVC → the data remains intact even when Pods autoscale up/down.

---

3. Content Management Systems (CMS)

• Example: WordPress, Drupal, Joomla running in Kubernetes.
• Media uploads (images, videos, docs) need persistent storage.
• Without PV/PVC, user-uploaded files would vanish after a Pod restart.

---

4. Data Science & AI/ML

• Example: Jupyter Notebooks, ML pipelines.
• Scientists save datasets, trained models, and checkpoints.
• PVC ensures large datasets survive across Pod lifecycles.

---

5. Logging & Monitoring Systems

• Example: Elasticsearch, Prometheus.
• Metrics & logs need to be retained for analysis.
• PVC ensures logs don’t disappear if Pods roll out.

#Kubernetes

#CloudComputing

#DevOps

#Containers

#CloudNative

#PersistentStorage

#Microservices

#CloudArchitecture