Implementing AWS Auto Scaling Using EC2, ALB & Target Tracking

Introduction

In modern cloud environments, auto scaling is critical for high availability, performance, and cost optimization. Instead of running multiple servers all the time, AWS allows you to automatically scale EC2 instances based on load.

In this blog, we will configure:

1. Apache Web Server on EC2

2. Auto Scaling Group (ASG)

3. Launch Template

4. Target Tracking Policy

5. Scale out when CPU > 5%

6. Automatically launch a second Apache instance

This setup is ideal for:

• Web applications

• Traffic-based scaling

• Cost optimisation

• High availability

Architecture Overview

Prerequisites

Before starting, ensure:

1. AWS Account

2. IAM user with EC2, Auto Scaling, ELB access

3. Key Pair created

4. Basic knowledge of Linux & SSH

   
   

STEP 1: Launch First EC2 Instance (Apache Setup)

1. Go to EC2 Dashboard → Launch Instance

Setting	Value
Name	apache-asg-base
AMI	Rocky 9
Instance Type	t3.small
Key Pair	Select existing
Security Group	Allow HTTP (80) & SSH (22)
Auto-assign Public IP	Enabled

2. Add User Data (Auto Install Apache)

Paste this in User Data:

#!/bin/bash
dnf update -y
dnf install httpd -y
systemctl start httpd
systemctl enable httpd

echo "<h1>Builddevops Apache Server - $(hostname)</h1>" > /var/www/html/index.html

This ensures Apache installs automatically whenever a new instance is launched.

3. Launch the Instance

Wait until:

• Instance is running

• Status checks are passed

Browser test:

http://<PUBLIC-IP>

You should see:

STEP 2: Create an AMI from This Instance

We’ll use this AMI for auto scaling.

Steps:

1. EC2 → Instances

2. Select your instance

3. Actions → Image and templates → Create Image

4. Name: apache-asg-ami

5. Click Create Image

⏳ Wait until AMI status becomes Available

STEP 3: Create Launch Template

Why Launch Template?

Auto Scaling uses it to create new EC2 instances.

Steps:

1. EC2 → Launch Templates

2. Click Create Launch Template

Configuration:

Setting	Value
Name	apache-launch-template
AMI	Select your custom AMI
Instance Type	t3.micro
Key Pair	Same as earlier
Security Group	Allow HTTP & SSH

Click Create Launch Template

STEP 4: Create Target Group

1. EC2 → Target Groups

2. Click Create target group

3. Choose:

Type: Instance

Protocol: HTTP

Port: 80

4. Health check path: /

5. Create Target Group

STEP 5: Create Application Load Balancer (ALB)

Steps:

1. EC2 → Load Balancers → Create

2. Select Application Load Balancer

3. Settings:

   • Name: apache-alb

   • Scheme: Internet-facing

   • Listener: HTTP 80

4. Select:

VPC

At least 2 subnets

Attach Target Group

Create Load Balancer

STEP 6: Create Auto Scaling Group

Go to Auto Scaling → Create ASG

Step 1: Launch Template

• Select apache-launch-template

Step 2: Network

• Select VPC

• Choose minimum 2 subnets

Step 3: Attach Load Balancer

• Select existing ALB

• Choose Target Group

Step 4: Configure Group Size

Setting	Value
Desired	1
Minimum	1
Maximum	2

STEP 7: Configure Scaling Policy (CPU > 5%)

Choose:

✔ Target Tracking Scaling Policy

Configure:

Setting	Value
Metric	Average CPU Utilization
Target Value	5%
Instance Warm-up	60 seconds

👉 This means:

• If CPU > 5% → new instance launches

• If CPU < 5% → instance terminates automatically

  
  

STEP 8: Test Auto Scaling

Generate Load

SSH into the instance:

[root@siddhesh ~]# dnf install epel-release -y
[root@siddhesh ~]# dnf install stress-ng -y 
[root@siddhesh ~]# stress-ng --cpu 2 --timeout 300

What Happens?

• CPU crosses 5%

• Auto Scaling triggers

• New EC2 instance is created

• Added to Load Balancer automatically

  
  

Check:

EC2 → Auto Scaling Groups → Activity History

As shown above, a new instance was automatically launched when the instance load exceeded 5%.

After the load dropped below 5%, the newly created instance was automatically terminated by the Auto Scaling group, as shown in the screenshot above.

This confirms that the scale-in policy is working correctly and that AWS Auto Scaling is dynamically adjusting the number of running instances based on real-time CPU utilization, ensuring optimal resource usage and cost efficiency.

Best Practices

• Use ALB instead of direct EC2 access

• Enable CloudWatch alarms

• Use Auto Healing

• Use HTTPS (ACM SSL)

• Use Launch Templates (not Launch Configurations)

• Use IAM Role instead of access keys

Conclusion

With this setup, you have successfully:

• Created an Apache Auto Scaling environment

• Configured CPU-based scaling

• Automated infrastructure growth

• Improved availability and performance

  
  

This architecture is production-ready and scalable.