View on GitHub

This repository demonstrates the evolution of Infrastructure as Code (IaC) and Configuration Management. It is structured as a 3-part journey, taking a raw Linux environment and progressively upgrading the deployment methodology to solve the limitations of the previous stage.

Tech Stack

  • Provisioning & Orchestration: Terraform, Vagrant
  • Configuration Management: Ansible, Bash
  • Cloud Provider: Amazon Web Services (AWS)
  • Application Stack: Node.js (Express), PostgreSQL, Redis, Nginx, PM2
  • Local Hypervisor: KVM / libvirt

Goal

To provision the infrastructure for a standard 3-tier web application using three entirely different DevOps philosophies.

A single, dummy-node-app is located at the root of this repository. This exact same application is deployed across all three levels to demonstrate how different infrastructure tools handle and scale the exact same workload.

3 Stages of Evolution

Level 1: A Single Bash Script

  • The Architecture: All services (Frontend, Cache, Database, Proxy) running on a single Ubuntu VM.
  • The Tooling: A procedural Bash bootstrap.sh script.
  • The Engineering Focus: Imperative execution.
  • The Lesson: Demonstrates the foundational Linux commands required to configure OS-level services (systemd, UFW, PM2) and integrate official GPG-secured package repositories. It ultimately exposes the fragility, lack of native idempotency, and scaling limitations inherent in imperative shell scripting.

Level 2: Configuration Management With Ansible

  • The Architecture: A distributed 3-node environment consisting of a Web Server (192.168.56.12), Redis Server (192.168.56.11), and Postgres Server (192.168.56.10).
  • The Tooling: Ansible Playbooks, modular Roles, and an optimized ansible.cfg with SSH Pipelining.
  • The Engineering Focus: Declarative, idempotent configuration.
  • The Lesson: Upgrades the infrastructure to a declarative state. Demonstrates native idempotency by abstracting the configuration away from the hardware. Showcases modular server configuration via the roles/ directory structure and enforces secure internal communication using a “Default Deny” UFW policy across separate network interfaces.

Level 3: Cloud-Native IaC on AWS (Terraform + Ansible)

  • The Architecture: Production-grade AWS Cloud Infrastructure featuring a Custom VPC, Public Subnets (Web/Proxy), Private Subnets (Database/Cache), and a NAT Gateway for secure outbound traffic.
  • The Tooling: Terraform for AWS hardware and network provisioning, handing off to Ansible for software configuration via Bastion Host SSH tunneling.
  • The Engineering Focus: Separation of Concerns (Immutable Provisioning + Mutable Configuration) and State Management.
  • The Lesson: Demonstrates how Terraform interacts directly with the AWS API to architect the cloud environment (tracking resources via terraform.tfstate), while Ansible securely connects to the live EC2 instances to configure the application stack. Successfully migrated the infrastructure by reusing 100% of the Level 2 Ansible roles and utilizing Ansible magic variables ({{ groups['db'][0] }}) to dynamically inject live AWS IPs into the Node.js environment. Overcame strict cloud network boundaries via ProxyCommand tunneling and private subnet isolation.

Testing Environments

To ensure a clean, reproducible setup while managing costs:

  • Levels 1 & 2 (Local Testing): Utilizes Vagrant with the KVM/libvirt hypervisor to spin up isolated, cost-free VMs locally. This allows for rapid iteration of configuration scripts before touching cloud resources.
  • Level 3 (Cloud Deployment): Targets Amazon Web Services (AWS) using standard API credentials to provision real cloud infrastructure, mapped dynamically via Terraform state outputs. Leverages Free Tier eligible EC2 instances and temporary NAT Gateways.