Deploying a Secure & Scalable Web App on AWS EC2 (No ECS, No Beanstalk!)

Abdiel Jonathan — Wed, 15 Oct 2025 20:43:01 +0000

Starting the Tutorial

In this tutorial, we’ll go step-by-step through deploying a Dockerized web application on AWS EC2 using only EC2 services — no Elastic Beanstalk, no ECS, no shortcuts.

By the end, you’ll have:

A launch template that auto-installs Docker and deploys your app.
An Auto Scaling Group for high availability.
A custom domain pointing to your EC2 instance.

Prerequisites
Before we start, make sure you have:

An AWS account
Access to ChatGPT (optional, but great for debugging along the way!)
IAM roles with permissions for EC2, ECR, SSM, and Auto Scaling
Basic knowledge of GitHub (to clone your project repo)
Docker installed on your local machine

Part 1 – Configuring EC2 Using a Launch Template

We’ll start by setting up an EC2 Launch Template — this defines how each instance should be created, configured, and secured.
a - On your AWS Management Console, search for “Launch Templates”.

b - We will then be redirected to this page and proceed to click on Create Launch Template

You’ll now see a form to define the template’s details. Let’s fill it out together 👇

Basic Configuration

Template name: cloud-tutorial

Application and OS Image (AMI):

Ubuntu Server 24.04 LTS (64-bit x86)

Instance type: t3.small

Network settings:

Create a new security group

Allow port 80 (HTTP) and port 443 (HTTPS)
Disable SSH access — we’ll use SSM Session Manager for secure logins.

Storage: 20 GB (change from the default 8 GB)
Monitoring: Enable CloudWatch monitoring

This helps with debugging and tracking CPU, disk I/O, and network usage.

IAM instance profile:

Choose your IAM role (e.g., LabRole if you’re using AWS Academy).

Finally, click Create launch template ✅

You should now see your new template listed.

Part 2 - Setting Up Auto Scaling for High Availability

Next, we’ll configure an Auto Scaling Group (ASG) to automatically manage EC2 instances.

Go to EC2 → Auto Scaling Groups → Create Auto Scaling Group

Name it ASG-cloud-tutorial
Select your launch template (cloud-tutorial)
Choose two Availability Zones and their subnets for high availability.(In the network section)

Leave the distribution as “Balanced best effort”.
Integrating Load Balancing In the “Load Balancing” section:
- Enable Elastic Load Balancing (ALB).

* Turn on ELB health checks, so EC2 Auto Scaling can replace unhealthy instances.(Then configure scaling:Minimum capacity: 1,Desired capacity: 1,Maximum capacity: 3,Target tracking policy: Average CPU utilization at 50%.)
  This ensures your app scales up under load and back down when idle — saving costs while staying responsive.

Once done, click Create Auto Scaling Group ✅
Your configuration should look like this:
Part a -

Part b -

Part c -

Part 3 – Deploying the Application

Now that your infrastructure is ready, let’s deploy a simple app.
Project Structure
Your app repository will look like this:

Deployment steps

Clone your repo to your local machine:

git clone https://github.com/AbdielJonathan007/cloud-tutorial
cd cloud-tutorial

Use Docker Buildx for multi-platform builds (for M1/M2 Mac compatibility and when we upload to EC2 instances later):

Create an ECR repository for each component:
- backend
- frontend
- nginx At Amazon search bar look for : ECR repository Choose: Visibility: Private Repository name: hello-world-app Leave encryption and scanning default Click Create repository After creating you get a url like this:339712882354.dkr.ecr.us-east-1.amazonaws.com/hello-world-app
- After creating the repository we will need the following information

Which in our case we are using the canvas lab to access the information would be under AWS details then clicking on AWS CLI , then we fill out that informationB Then we run this command to Retrieve an authentication token and authenticate your Docker client to your registry. Use the AWS CLI:
- Then we run this command

aws ecr get-login-password --region us-east-1 | \
docker login --username AWS --password-stdin 339712882354.dkr.ecr.us-east-1.amazonaws.com

We will need to push 3 Docker images

frontend

backend

nginx

Part 4 - Automating Deployment via Launch Template

Now we’ll make sure new EC2 instances automatically install Docker and run the app.
Steps:

Go to EC2 → Launch Templates → Actions → Create new version
Scroll to Advanced details → User data
Paste this script:

#!/bin/bash
apt update -y
apt install -y docker.io unzip curl
systemctl enable docker --now
aws ecr get-login-password --region us-east-1 | \
docker login --username AWS --password-stdin 339712882354.dkr.ecr.us-east-1.amazonaws.com

docker pull 339712882354.dkr.ecr.us-east-1.amazonaws.com/backend:latest
docker pull 339712882354.dkr.ecr.us-east-1.amazonaws.com/frontend:latest
docker pull 339712882354.dkr.ecr.us-east-1.amazonaws.com/nginx:latest

docker network create myapp-network
docker run -d --name backend --network myapp-network -p 5000:5000 backend:latest
docker run -d --name frontend --network myapp-network -p 3000:3000 frontend:latest
docker run -d --name nginx --network myapp-network -p 80:80 nginx:latest

4 - Create the new version and attach it to your Auto Scaling Group (ASG).
Now, whenever ASG launches a new EC2 instance, it’ll run this script automatically. 💪
We should see something like this :

Part 5 - Connecting Your Domain (Using Name.com Only)

Now that our app is up and running on EC2, let’s make it accessible through a custom domain instead of the public IP.

We’ll be using a domain from Name.com, and we’ll connect it directly using DNS “A” records — no Route 53 needed.

Step 1: Get Your EC2 Public IP

1️⃣ Open the AWS Management Console → EC2 → Instances.
2️⃣ Select the instance that’s running your app.
3️⃣ Copy the Public IPv4 address (for example, 3.90.248.104).

Step 2: Add DNS Records on Name.com

1️⃣ Log in to your Name.com
account.
2️⃣ Go to My Domains → abdielcloud.live → Manage DNS.
3️⃣ Under DNS Records, add two A records:

Click Save Changes when you’re done.

Once propagation finishes, navigating to
http://abdielcloud.live
(or http://www.abdielcloud.live)
should open your frontend running on the EC2 instance 🎉

Thank you for reading thus far!