Phase 1 : Local Cluster with KIND + PostgreSQL Setup

---

Welcome, DevOps Learners!

In this hands-on series, we’re learning by building. No boring theory—just real projects, exactly how they are built in production.

We’ll start simple, deploy PostgreSQL on a local Kubernetes cluster using KIND, and scale our stack with secrets, monitoring, CI/CD, backups, and more as we go.

---

What You’ll Learn in Phase 1

By the end of this phase, you'll:

---

Tasks Completed in Phase 1 (with snippets)

---

Created a Local Kubernetes Cluster with KIND

Instead of using heavy tools like Minikube, we chose KIND (Kubernetes IN Docker)—a production-grade local cluster tool perfect for DevOps prototyping.

Cluster Config Preview:

# kind-config.yaml
nodes:
  - role: control-plane
    extraPortMappings:
      - containerPort: 30080
        hostPort: 8080

Command:

kind create cluster --name pg-devops --config kind-config.yaml

---

Created a Dedicated Namespace for PostgreSQL

Namespaces simulate real-world multi-env setups (like dev, staging, prod):

kubectl create namespace postgres-dev

---

Secured PostgreSQL Credentials using Secrets

We base64 encoded our DB credentials to store safely in Kubernetes Secrets.

Code Snippet:

apiVersion: v1
kind: Secret
metadata:
  name: postgres-secret
type: Opaque
data:
  POSTGRES_USER: cG9zdGdyZXM=
  POSTGRES_PASSWORD: c2VjdXJlcGFzcw==

---

Used ConfigMaps for Environment Configuration

We added a database name using a ConfigMap—great for non-sensitive environment variables.

Snippet:

apiVersion: v1
kind: ConfigMap
metadata:
  name: postgres-config
data:
  POSTGRES_DB: devdb

---

Provisioned Persistent Volumes

We created a PersistentVolumeClaim to retain database data even if the container crashes or restarts.

resources:
  requests:
    storage: 1Gi

---

Deployed PostgreSQL using StatefulSet

PostgreSQL isn’t just another app—it's stateful. So we used StatefulSet instead of Deployment, which helps with persistent identity and volume consistency.

Interesting bit:

volumeClaimTemplates:
  - metadata:
      name: postgres-storage

---

Created a Headless Service

This enables network identity for StatefulSet pods. We used clusterIP: None to make sure it works as a stable DNS for the DB instance.

clusterIP: None

---

Where’s the Full Code?

I’ve structured the full codebase day-by-day for learners. You can explore it in my GitHub repo:

👉 Kubernetes Diaries PostgreSQL – GitHub Repository

🌟 Star the repo if it helps you. Each ⭐ motivates me to build more content for you!

Navigate to branches like day1-setup, day2-secret-config for complete code walkthroughs!

---

What’s Coming in Phase 2?

In the next phase, we will:

• Deploy pgAdmin to visualize and manage our DB

• Connect to PostgreSQL from inside the cluster

• Prepare for cloud migration and backups

• Start documenting each phase visually with Grafana

---

💬 Final Words

This isn’t just a project—it’s a journey to become production-ready. Whether you're a student, DevOps beginner, or aspiring cloud engineer, this project will grow with you.

So don’t just copy—understand, break, fix, and deploy. That's how real DevOps is done.