⚡ AWS Auto-Scale API

Production-grade, self-healing infrastructure that handles 1000 concurrent users at 1,268 req/s with zero errors - fully automated from git push to live deployment.

---

🏗️ Architecture

                        ┌─────────────────────────────────────────┐
                        │              GitHub Actions              │
                        │   push → build → dockerize → ECR push   │
                        │         → ASG Instance Refresh          │
                        └──────────────────┬──────────────────────┘
                                           │
                                    ┌──────▼──────┐
                                    │  AWS ECR    │
                                    │Docker Images│
                                    └──────┬──────┘
                                           │
          ┌──────────────────────────────── ▼ ────────────────────────────────┐
          │                     AWS VPC (ap-south-1)                          │
          │                                                                    │
          │          ┌─────────────────────────────────────┐                  │
          │          │    Application Load Balancer (ALB)   │ ← HTTP :80      │
          │          └──────────┬──────────────┬───────────┘                  │
          │                     │              │                               │
          │          ┌──────────▼──┐      ┌────▼────────┐                     │
          │          │  EC2 :8080  │      │  EC2 :8080  │   ← Auto Scaling   │
          │          │ ap-south-1a │      │ ap-south-1b │      Group (2-6)    │
          │          └─────────────┘      └─────────────┘                     │
          │                                                                    │
          │   CloudWatch → CPU > 50% → Scale Up (+2)                          │
          │   CloudWatch → CPU < 20% → Scale Down (-1)                        │
          └────────────────────────────────────────────────────────────────────┘

---

🚀 Tech Stack

Layer	Technology
Application	Spring Boot 3.5, Java 21
Containerization	Docker, AWS ECR
Infrastructure	Terraform, AWS VPC, Subnets, IGW
Compute	AWS EC2 (m7i-flex.large), Auto Scaling Group
Load Balancing	AWS Application Load Balancer
CI/CD	GitHub Actions
Monitoring	AWS CloudWatch, Prometheus, Actuator
Load Testing	k6
OS	Ubuntu 22.04 LTS

---

📊 Load Test Results - 1000 Concurrent Users

  █ THRESHOLDS
    errors              ✓ rate=0.00%
    http_req_duration   ✓ p(95)=56.67ms

  █ TOTAL RESULTS
    Total Requests......: 229,074
    Requests/sec........: 1,268 req/s 🚀
    Error Rate..........: 0.00% ✅
    Avg Response Time...: 38.24ms ⚡
    P95 Response Time...: 56.67ms ⚡
    Data Received.......: 77 MB
    Duration............: 3 minutes
    Max VUs.............: 1,000

229K requests. Zero errors. 38ms average response. 1000 users. Infrastructure auto-scaled and self-healed throughout.

---

🔥 Key Features

⚡ Auto Scaling

• Minimum 2 EC2 instances always running across 2 Availability Zones
• Scales up by +2 instances when CPU > 50%
• Scales down by -1 instance when CPU < 20%
• Maximum 6 instances at peak load

🔄 Self Healing

• ALB health checks hit /actuator/health every 30 seconds
• Unhealthy instance? → Automatically deregistered from ALB
• ASG detects instance count drop → Launches replacement automatically
• Zero manual intervention required

🚀 Zero-Downtime Deployments

• Git push → GitHub Actions triggers
• New Docker image built and pushed to ECR
• ASG Instance Refresh with MinHealthyPercentage=50
• Rolling update - old instances drain, new ones warm up

---

📁 Project Structure

aws-autoscale-api/
├── src/
│   └── main/java/com/devops/aws_autoscale_api/
│       ├── AwsAutoscaleApiApplication.java
│       └── AppController.java
├── terraform/
│   └── main.tf                 # Complete AWS infrastructure as code
├── k6/
│   └── loadtest.js             # 1000 user load test script
├── .github/
│   └── workflows/
│       └── deploy.yml          # CI/CD pipeline
└── Dockerfile

---

🛠️ Infrastructure (Terraform)

Everything provisioned as code one command to create, one command to destroy:

cd terraform
terraform init
terraform apply    # provisions everything in ~5 minutes
terraform destroy  # tears down everything

Resources provisioned (22 total):

• VPC + 2 Public Subnets (ap-south-1a, ap-south-1b)
• Internet Gateway + Route Tables
• Security Groups (ALB + EC2)
• Application Load Balancer + Target Group + Listener
• Launch Template (Ubuntu 22.04, m7i-flex.large, 20GB gp3)
• Auto Scaling Group (min: 2, max: 6)
• CloudWatch Alarms (high CPU + low CPU)
• Auto Scaling Policies (scale up + scale down)
• ECR Repository
• IAM Role + Instance Profile

---

🔌 API Endpoints

Endpoint	Description
GET /api/hello	Returns server hostname + timestamp
GET /api/info	Returns CPU, memory, server stats
GET /actuator/health	Health check (used by ALB)
GET /actuator/prometheus	Prometheus metrics

Sample Response - /api/hello

{
  "message": "Hello from AWS Auto-Scaling Infrastructure!",
  "server": "ip-10-0-1-245",
  "timestamp": "2026-06-03T14:32:11.123456",
  "status": "healthy"
}

The server field changes on each refresh proof the ALB is routing across multiple EC2 instances.

---

🧪 Run Load Test

# Install k6
winget install k6

# Run 1000 user load test
k6 run -e ALB_URL=YOUR_ALB_DNS k6/loadtest.js

Test stages:

0:00 → 0:30  →  ramp up to 100 users
0:30 → 1:30  →  ramp up to 500 users
1:30 → 2:30  →  ramp up to 1000 users  ← ASG scales here
2:30 → 3:00  →  ramp down to 0

---

🔧 CI/CD Pipeline

git push origin main
       │
       ▼
GitHub Actions
       │
       ├── Build JAR (Maven)
       ├── Build Docker Image
       ├── Push to AWS ECR (:latest + :commit-sha)
       └── Trigger ASG Instance Refresh
                  │
                  ▼
         Rolling deployment
         Zero downtime ✅

---

🛠️ Chaos Engineering - Self Healing Demo

1. Go to AWS Console → EC2 → Instances
2. Terminate any autoscale-ec2 instance manually
3. Watch ASG detect the drop
4. New instance launches automatically within 60 seconds
5. Health checks pass → Traffic resumes

The system heals itself. No alerts. No manual work.

---

⚙️ Setup & Deployment

Prerequisites

• AWS CLI configured
• Terraform installed
• Docker installed
• Java 21

Deploy

# 1. Clone
git clone https://github.com/Sumeet-Y1/aws-autoscale-api

# 2. Provision infrastructure
cd terraform && terraform apply

# 3. Push Docker image to ECR
aws ecr get-login-password --region ap-south-1 | docker login --username AWS --password-stdin ACCOUNT_ID.dkr.ecr.ap-south-1.amazonaws.com
docker build -t aws-autoscale-api .
docker tag aws-autoscale-api:latest ECR_URL:latest
docker push ECR_URL:latest

# 4. Push to GitHub → CI/CD auto-deploys
git push origin main

Destroy

cd terraform && terraform destroy

---

👤 Author

Sumeet - GitHub