From ClickOps to Terraform: what we measured

The allure of the AWS console is strong. It's visual, immediate, and perfectly designed to make you feel productive. But as any team scaling past their first few services quickly learns, "ClickOps" is the enemy of reliability.

The Hidden Cost of Manual Provisioning

Before our transition, spinning up a new environment was an oral tradition passed down from senior engineers to new hires. We documented it in Confluence, but the documentation was always a step behind reality. When incidents occurred, reproducing the exact state of a production service in a staging environment to debug it was nearly impossible. We were battling configuration drift on a weekly basis.

The Migration Strategy

We didn't boil the ocean. We started with the foundational layers: networking (VPCs, Subnets, Route Tables) and IAM. Once the base was solid, we moved our RDS instances and finally, the compute layer (EKS/ECS).

module "vpc" {
  source = "terraform-aws-modules/vpc/aws"
  version = "5.0.0"

  name = "${var.environment}-main-vpc"
  cidr = var.vpc_cidr

  azs             = ["us-east-1a", "us-east-1b", "us-east-1c"]
  private_subnets = ["10.0.1.0/24", "10.0.2.0/24", "10.0.3.0/24"]
  public_subnets  = ["10.0.101.0/24", "10.0.102.0/24", "10.0.103.0/24"]

  enable_nat_gateway = true
  single_nat_gateway = var.environment != "prod"
}

What We Measured (The ROI)

• MTTR Dropped by 60%

  Rollbacks became a simple git revert and terraform apply rather than a frantic click-hunt.

• Environment Parity

  Staging finally became an exact replica of production, drastically reducing "it works on my machine" bugs.