Highly Available Web Application Infrastructure on AWS

Infrastructure-as-Code deployment of a resilient web app using VPC, NAT Gateways, an Application Load Balancer, and Auto Scaling.

Goal

Stand up a web application on AWS that survives both instance failure and traffic spikes, without any manual scaling or failover steps.

Architecture

VPC with public and private subnets across multiple AZs
NAT Gateways so private-subnet instances can reach the internet for updates without being directly exposed
Application Load Balancer distributing traffic across healthy instances only
Auto Scaling Group adding or removing capacity based on real load

resource "aws_autoscaling_group" "web" {
  min_size         = 2
  max_size         = 6
  desired_capacity = 2
  vpc_zone_identifier = module.vpc.private_subnets
  target_group_arns  = [aws_lb_target_group.web.arn]
}

Why it matters

This is the baseline pattern behind most production web infrastructure on AWS — the same shape scales from a side project to a multi-region service with the right module boundaries.