DevOps Practice Lab

A hands-on Kubernetes learning repository covering core workloads, Helm packaging, service mesh, autoscaling, RBAC, CRDs, and a real FastAPI application — all tested on Kind and Minikube local clusters.

---

Table of Contents

• Overview
• Tech Stack
• Repository Structure
• Getting Started
• Cluster Setup
• Modules
  • Kubernetes Basics
  • Apache Workload + RBAC
  • MySQL StatefulSet
  • Notes App (FastAPI)
  • Helm Chart — Apache
  • Istio Service Mesh
  • Custom Resource Definitions
  • Kubernetes Dashboard

---

Overview

A structured reference repository covering the Kubernetes ecosystem end-to-end. Each directory is a self-contained module with manifests, configurations, and commands — spanning core workloads, stateful applications, Helm packaging, RBAC, autoscaling, CRDs, and service mesh.

All workloads run locally using Kind (Kubernetes in Docker) or Minikube — no cloud account required.

---

Tech Stack

Tool	Purpose
Kubernetes	Container orchestration
Kind / Minikube	Local cluster environments
Helm	Kubernetes package manager
Istio 1.29.2	Service mesh (traffic, security, observability)
FastAPI	Sample Python web application
MySQL	Stateful database workload
Apache HTTPD	Stateless web server workload
NGINX Ingress	Cluster ingress controller
SQLAlchemy	ORM for the notes app
Docker	Container runtime

---

Repository Structure

kubernetes/
├── kubernetes/          # Core K8s primitives (pod, deployment, job, namespace)
├── apache/              # Apache workload with HPA, VPA, and full RBAC
│   └── RBAC/            # ServiceAccount, Role, RoleBinding
├── mysql/               # MySQL StatefulSet with headless service and PVC
├── notes-app/           # FastAPI CRUD app + K8s manifests + ingress
│   └── k8s/
├── helm/                # Helm chart for Apache with full template set
│   └── apache-helm/
│       └── templates/
├── crd/                 # Custom Resource Definition + CR instance
├── dashboard/           # K8s Dashboard RBAC setup
├── install_kind.sh      # Automated Kind + kubectl installer
└── monitoring/          # (planned)

---

Getting Started

Prerequisites

• Docker
• Kind or Minikube
• kubectl
• Helm 3

Quick Cluster Setup

Kind (Kubernetes in Docker):

# Automated setup (installs Docker, Kind, kubectl)
chmod +x install_kind.sh
./install_kind.sh

# Create a cluster
kind create cluster --name devops-lab

Minikube:

minikube start --cpus=2 --memory=4g

Install Helm

chmod +x helm/get_helm.sh
./helm/get_helm.sh

---

Cluster Setup

All workloads were tested on:

• Kind v0.29.0 — multi-node local clusters
• Minikube — single-node local clusters

The install_kind.sh script handles Docker installation, Kind binary download (aarch64/x86_64 auto-detection), and kubectl setup automatically on Linux.

---

Modules

1. Kubernetes Basics

Directory: kubernetes/

Starting point — the fundamentals of Kubernetes objects.

File	What it covers
pod.yaml	Standalone pod definition
deployment.yml	NGINX deployment with 2 replicas
namespace.yml	Namespace isolation
job.yaml	One-shot Job resource

kubectl apply -f kubernetes/
kubectl get pods -n nginx-namespace

---

2. Apache Workload + RBAC

Directory: apache/

A complete Apache HTTPD workload demonstrating production-style resource management and access control.

Highlights:

• Deployment with CPU/memory requests and limits
• HPA — scales between 1–5 replicas, triggers at 5% CPU
• VPA — automatically right-sizes container resources (Auto mode)
• Full RBAC setup: ServiceAccount → Role → RoleBinding

kubectl apply -f apache/namespace.yml
kubectl apply -f apache/RBAC/
kubectl apply -f apache/deployment.yaml
kubectl apply -f apache/hpa.yml
kubectl apply -f apache/vpa.yml

RBAC layout:

ServiceAccount: apache-user
    └── RoleBinding
            └── Role (get/list/watch on pods, deployments, services)

---

3. MySQL StatefulSet

Directory: mysql/

Stateful workload using a StatefulSet with persistent storage — a common real-world pattern for databases.

Highlights:

• 3-replica StatefulSet with 1Gi PersistentVolumeClaim per pod
• Headless Service for stable pod DNS (mysql-0.mysql, mysql-1.mysql, ...)
• ConfigMap for database configuration
• Secret for root password (base64 encoded)

kubectl apply -f mysql/namespace.yml
kubectl apply -f mysql/configmap.yml
kubectl apply -f mysql/secret.yml
kubectl apply -f mysql/service.yml
kubectl apply -f mysql/statefulset.yml

---

4. Notes App (FastAPI)

Directory: notes-app/

A real CRUD application to practice deploying custom workloads on Kubernetes.

App features:

• FastAPI backend with SQLite (SQLAlchemy ORM)
• Jinja2 templates for server-side rendered frontend
• REST API: GET / POST / PUT / DELETE /api/notes/
• Dockerized and pushed to krishna5534/notes-app:latest

K8s manifests:

File	Resource
k8s/namespace.yml	Dedicated namespace
k8s/deployment.yml	1-replica deployment on port 8000
k8s/service.yml	ClusterIP service
k8s/ingress.yml	NGINX ingress with path routing

kubectl apply -f notes-app/k8s/
kubectl get ingress -n notes-app

---

5. Helm Chart — Apache

Directory: helm/apache-helm/

A fully templated Helm chart for the Apache HTTPD workload — demonstrating real chart authoring practices.

Chart: apache-helm v0.1.0 | App: httpd 2.4

Templates included:

Template	Purpose
deployment.yaml	Parameterized Apache deployment
service.yaml	Service with configurable type/port
ingress.yaml	Ingress with TLS support
httproute.yaml	Gateway API HTTPRoute
hpa.yaml	HPA with CPU/memory targets
serviceaccount.yaml	Optional service account
_helpers.tpl	Shared label/name helpers
tests/	helm test connection test

Key defaults (values.yaml):

replicaCount: 3
image:
  repository: httpd
  tag: "2.4"
autoscaling:
  enabled: false
  minReplicas: 1
  maxReplicas: 100
  targetCPUUtilizationPercentage: 80

Usage:

# Install from local chart
helm install my-apache ./helm/apache-helm

# Install from packaged chart
helm install my-apache ./helm/apache-helm-0.1.0.tgz

# Override replicas
helm install my-apache ./helm/apache-helm --set replicaCount=5

# Run tests
helm test my-apache

---

6. Istio Service Mesh

Istio 1.29.2 installed for hands-on service mesh practice.

Covers:

• Sidecar proxy injection (Envoy)
• Traffic management (VirtualService, DestinationRule)
• mTLS and security policies
• Observability (Kiali, Jaeger, Prometheus integration)
• Ztunnel (ambient mesh mode)

# Install Istio
cd istio/istio-1.29.2
./bin/istioctl install --set profile=demo

# Enable sidecar injection on a namespace
kubectl label namespace default istio-injection=enabled

---

7. Custom Resource Definitions

Directory: crd/

Demonstrates extending the Kubernetes API with a custom resource.

CRD: MlBatch (group: mlmodels.io)

# Shortnames: ai, ml, models
kind: MlBatch
spec:
  name: string
  model: string        # e.g. resnet50
  environment: string  # e.g. cloud
  platform: string     # e.g. Colab

kubectl apply -f crd/devops-crd.yaml
kubectl apply -f crd/devops-cr.yaml
kubectl get mlbatch        # or: kubectl get ai / kubectl get ml
kubectl describe mlbatch mlbatch-3

---

8. Kubernetes Dashboard

Directory: dashboard/

Admin-level access setup for the Kubernetes Dashboard UI.

kubectl apply -f dashboard/dashboard-admin-user.yaml

# Get the login token
kubectl -n kubernetes-dashboard create token admin-user

# Start the proxy
kubectl proxy
# Open: http://localhost:8001/api/v1/namespaces/kubernetes-dashboard/services/https:kubernetes-dashboard:/proxy/

---

Built for learning.