Terraform

Terraform

Terraform is the world's most popular Infrastructure as Code (IaC) tool. It uses HCL to provision resources on any cloud. 2025 introduces Terraform Stacks for easier component management.

When to Use

• Provisioning: Creating VPCs, Databases, K8s Clusters.
• Multi-Cloud: Learn one syntax (HCL), use it for AWS, Azure, GCP, Datadog, etc.
• State Management: It tracks resource state, allowing "Plan" (preview) and "Apply".

Quick Start

# main.tf
provider "aws" {
  region = "us-west-2"
}

resource "aws_s3_bucket" "b" {
  bucket = "my-tf-test-bucket"
  tags = {
    Name = "My bucket"
  }
}

Core Concepts

Providers

Plugins that talk to APIs (AWS, Azure, Kubernetes).

State

terraform.tfstate. The source of truth mapping your code to real-world resource IDs. Must be stored remotely (S3 + DynamoDB Locking) in teams.

Stacks (2025)

A new layer above Modules. Allows defined dependencies between deployments (e.g., Deploy VPC, then Deploy K8s using VPC ID output).

Best Practices (2025)

Do:

• Use Remote State: S3 backend or Terraform Cloud. Never local state.
• Use Modules: DRY. Write a "Company Standard Bucket" module and reuse it.
• Use tfsec / trivy: Scan HCL for misconfigurations (open security groups) before deploy.

Don't:

• Don't hardcode secrets: Use variable "db_password" {} and pass it via TF_VAR_ or a secret manager.

References

• Terraform Documentation