Skip to content

1.01 Terraform: Installation & Introduction

Overview

Terraform is an open-source Infrastructure as Code (IaC) tool by HashiCorp that provisions and manages cloud and on-premise resources using a declarative configuration language called HCL.

Abstract

Terraform lets you define infrastructure as code using .tf files written in HCL (HashiCorp Configuration Language). A single binary handles everything — no agents, no daemons. It supports hundreds of providers across AWS, GCP, Azure, and on-premise platforms, making it the de facto standard for multi-cloud IaC.

Why It Matters in Production

Infrastructure managed manually is error-prone and difficult to reproduce. Terraform solves this by codifying your infrastructure, enabling version control, peer review, automated pipelines, and repeatable deployments across environments.

Key Concepts

Concept Description
HCL HashiCorp Configuration Language — declarative, human-readable syntax
Resource An object Terraform manages (VM, bucket, IAM user, DNS record, etc.)
.tf file Configuration file containing resource definitions
Provider Plugin that interfaces with a cloud or service API (AWS, GCP, Azure, etc.)
State Terraform's record of what it has deployed (terraform.tfstate)
Plan Preview of changes before applying (terraform plan)

Installation

Terraform ships as a single static binary. No dependencies required.

# Download the binary
wget https://releases.hashicorp.com/terraform/0.13.0/terraform_0.13.0_linux_amd64.zip

# Unzip the archive
unzip terraform_0.13.0_linux_amd64.zip

# Move to system PATH
mv terraform /usr/local/bin

# Verify installation
terraform version

brew tap hashicorp/tap
brew install hashicorp/tap/terraform
terraform version

choco install terraform
terraform version

Expected output after install:

Terraform v0.13.0

Note

All examples in this course use Terraform v0.13 on Linux. Commands and syntax may differ slightly on newer versions. Always pin the version in production pipelines.

Resource Basics

A resource is anything Terraform creates and manages. Resources are declared in .tf files using HCL syntax.

Resource Syntax

resource "<PROVIDER>_<TYPE>" "<LOCAL_NAME>" {
  argument = "value"
}

Example — AWS EC2 Instance (aws.tf)

resource "aws_instance" "webserver" {
  ami           = "ami-0c2f25c1f66a1ff4d"
  instance_type = "t2.micro"
}
  • aws_instance → provider (aws) + resource type (instance)
  • "webserver" → local name used to reference this resource within the config
  • ami and instance_type → arguments specific to this resource type

Resource Types Covered in This Course

Early sections use simple, easy-to-understand resources to build foundational knowledge:

Resource Purpose
local_file Creates a file on the local filesystem
random_pet Generates a random pet name (useful for testing)

Later sections introduce real-world resources: EC2 instances, S3 buckets, IAM users/groups/policies, DynamoDB tables, GCP Compute, Azure databases, and more.

HCL Configuration Files

  • Files use the .tf extension
  • Can be created in any text editor: Notepad++, VS Code, Vim, Emacs, etc.
  • Multiple .tf files in the same directory are merged by Terraform at runtime
  • HCL is declarative — you describe the desired end state, not the steps to get there

Supported Platforms

Platform Supported
Linux (amd64, arm) ✅
macOS ✅
Windows ✅

Note

Course labs run on Linux. If you're on Windows or macOS, behaviour is identical for HCL/Terraform logic, but file paths and shell commands may differ.

Best Practices

Best Practices

  • Pin the Terraform version in all projects using a required_version constraint in terraform block.
  • Store state remotely (S3 + DynamoDB, Terraform Cloud) — never commit terraform.tfstate to version control.
  • Split resources into separate .tf files by concern (e.g. network.tf, compute.tf, iam.tf).
  • Always run terraform plan before terraform apply in production.
  • Use a CI/CD pipeline to automate plan/apply with approvals.

Security Best Practices

Security

  • Never hardcode AWS credentials or secrets in .tf files — use environment variables, IAM roles, or HashiCorp Vault.
  • Apply least-privilege IAM policies to the identity running Terraform.
  • Enable state encryption at rest (e.g. S3 server-side encryption + KMS).
  • Use terraform plan output as a pull request artifact for peer review before any apply.
  • Scan .tf files with tools like tfsec or Checkov in CI pipelines.

Do and Don't

✅ Do ❌ Don't
Pin Terraform version per project Use whatever version happens to be installed
Use remote state with locking Commit terraform.tfstate to Git
Use terraform plan before every apply Run terraform apply blind in production
Store secrets in Vault or Secrets Manager Hardcode credentials in .tf files
Organise resources across multiple .tf files Dump everything into one massive main.tf

Common Mistakes

Common Mistakes

  • Binary not in PATH — terraform: command not found after install. Fix: confirm /usr/local/bin is in $PATH with echo $PATH.
  • Wrong AMI ID for region — AMIs are region-specific. An AMI from us-east-1 won't work in eu-west-1.
  • Forgetting terraform init — must be run before plan or apply to download providers and initialise the working directory.
  • Editing state manually — never edit terraform.tfstate by hand; use terraform state subcommands.

Troubleshooting

# Check installed version
terraform version

# Confirm binary is accessible
which terraform

# Re-initialise working directory (safe to re-run)
terraform init

# Validate HCL syntax without connecting to any provider
terraform validate

# Preview changes without applying
terraform plan

Quick Recap

  • Terraform is a single binary — download, unzip, move to PATH, done.
  • Infrastructure is defined in .tf files using HCL (declarative syntax).
  • A resource is anything Terraform manages: files, VMs, buckets, IAM users, DNS records, etc.
  • Early course examples use local_file and random_pet to teach core concepts before moving to cloud resources.
  • HCL files can be edited in any text editor and have a .tf extension.
  • This course targets Terraform v0.13 on Linux.

Interview / Revision Notes

Quick Revision

  • What language does Terraform use? HCL — HashiCorp Configuration Language (declarative).
  • What is a resource in Terraform? Any object Terraform creates and manages (VM, bucket, IAM policy, file, etc.).
  • What file extension do Terraform configs use? .tf
  • How do you install Terraform? Download the binary, unzip, move to system PATH.
  • What command checks the installed version? terraform version
  • What must you run before plan or apply? terraform init
  • What is Terraform state? A file (terraform.tfstate) tracking the real-world resources Terraform manages.