WordPress on AWS with Terraform

This project is a Terraform-based proof of concept for migrating a simple WordPress blog from on-premises infrastructure to AWS.

The goal is to validate whether AWS can run WordPress with higher availability and lower operational burden, then remove all resources after evaluation.

Project Objectives

• Deploy a highly available WordPress infrastructure on AWS.
• Support roughly 1,000 daily visitors.
• Estimate monthly infrastructure costs.

Target AWS Architecture

The infrastructure maps WordPress requirements (PHP app + MySQL DB + shared file storage) to these AWS services:

• Application Load Balancer (ALB)

  • Distributes HTTP/HTTPS traffic across EC2 instances.
  • Performs health checks and stops routing to unhealthy targets.
  • Highly available by default.

• Amazon EC2

  • Runs Linux virtual machines hosting Apache, PHP, and WordPress.
  • Minimum of 2 instances for high availability.

• Amazon RDS for MySQL

  • Managed MySQL database for WordPress data.
  • Supports backups, patching, and optional Multi-AZ for higher availability.

• Amazon EFS

  • Shared NFSv4.1 file system for WordPress files and user uploads.
  • Accessible by multiple EC2 instances.

• Security Groups

  • Restrict traffic between ALB, EC2, RDS, and EFS.
  • Principle of least privilege (only required ports/protocols).

Availability Goal

Target uptime: 99.99%.

High-level strategy:

• Place at least two EC2 instances behind an ALB.
• Use health checks for automatic failover at the load balancer layer.
• Use managed database and shared filesystem services.

Current Terraform Baseline

Current provider configuration:

• Terraform CLI: >= 1.6.0
• AWS provider: hashicorp/aws ~> 6.34.0
• Default region variable: eu-west-2

Cost Estimation

Estimate costs before finalizing:

1. Run a Terraform plan and list all resources.
2. Use AWS Pricing Calculator for selected region and sizes.
3. Record assumptions (instance class, RDS size, storage, data transfer).