Terraform in GitLab CI - Mini Project: Build & Apply

Terraform in GitLab CI

📖 Scenario: You are working as a cloud engineer who wants to automate infrastructure deployment using Terraform. Your team uses GitLab for version control and CI/CD pipelines. You want to create a GitLab CI pipeline that runs Terraform commands to plan and apply infrastructure changes automatically.

🎯 Goal: Build a GitLab CI configuration file .gitlab-ci.yml that initializes Terraform, runs terraform plan, and applies the plan with terraform apply in a safe and automated way.

📋 What You'll Learn

Create a GitLab CI pipeline with three stages: init, plan, and apply

Use the official Terraform Docker image for all jobs

Initialize Terraform in the init stage

Run terraform plan in the plan stage and save the plan output

Run terraform apply in the apply stage using the saved plan

Use environment variables for Terraform backend configuration

Ensure the apply stage only runs after a successful plan

💡 Why This Matters

🌍 Real World

Automating Terraform deployments in GitLab CI saves time and reduces errors by running infrastructure commands automatically on code changes.

💼 Career

Cloud engineers and DevOps professionals use GitLab CI pipelines to manage infrastructure as code with Terraform in real projects.

1

Create the basic GitLab CI pipeline structure

Create a file named .gitlab-ci.yml with a stages section that lists the stages init, plan, and apply in that order.

Terraform

stages:
  # Your code here

Need a hint?

Define the stages key as a list with the three stages in order.

2

Add the init job to initialize Terraform

Add a job named terraform_init that runs in the init stage. Use the image hashicorp/terraform:latest. The job should run the command terraform init. Use environment variables TF_BACKEND_CONFIG for backend configuration by passing -backend-config=${TF_BACKEND_CONFIG} to terraform init.

Terraform

stages:
  - init
  - plan
  - apply

# Your code here

Need a hint?

Define the job with the correct name, stage, image, and script command using the environment variable.

3

Add the plan job to create a Terraform plan

Add a job named terraform_plan that runs in the plan stage. Use the same Terraform image. The job should run terraform init -backend-config=${TF_BACKEND_CONFIG} followed by terraform plan -out=tfplan to save the plan output to a file named tfplan. Make sure this job depends on the terraform_init job by using needs. Also, add artifacts to save the tfplan file for the next stage.

Terraform

stages:
  - init
  - plan
  - apply

terraform_init:
  stage: init
  image: hashicorp/terraform:latest
  script:
    - terraform init -backend-config=${TF_BACKEND_CONFIG}

# Your code here

Need a hint?

Use needs to depend on the init job and artifacts to pass the plan file.

4

Add the apply job to apply the Terraform plan

Add a job named terraform_apply that runs in the apply stage. Use the Terraform image. This job should depend on terraform_plan using needs. It should run terraform init -backend-config=${TF_BACKEND_CONFIG} followed by terraform apply -auto-approve tfplan to apply the saved plan without manual approval.

Terraform

stages:
  - init
  - plan
  - apply

terraform_init:
  stage: init
  image: hashicorp/terraform:latest
  script:
    - terraform init -backend-config=${TF_BACKEND_CONFIG}

terraform_plan:
  stage: plan
  image: hashicorp/terraform:latest
  needs:
    - terraform_init
  script:
    - terraform init -backend-config=${TF_BACKEND_CONFIG}
    - terraform plan -out=tfplan
  artifacts:
    paths:
      - tfplan

# Your code here

Need a hint?

Use needs to depend on the plan job and run terraform apply with the saved plan file.