About

A CI/CD pipeline where where .tfstate files are saved on an S3 bucket and updated at every GitHub push/PR using GitHub Actions.

Set up the CI/CD pipeline

1. Design your AWS infrastructure.
2. Code the infrastructure using Terraform. I named my file main.tf.
3. Create an S3 bucket. This will be later used to store the terraform state.
4. Launch AWS IAM and create 2 policies:

• to read and write the S3 bucket that we created in step 1.
• to read and write other AWS tools that will be used by our main.tf file.

5. Create a new IAM role and attach the policies we created.
6. Create a GitHub Actions file. This will connect to your IAM role using OpenID Connect protocol and run your terraform code. You can copy the file in ./github/workflows/main.yml
7. Launch the GitHub repo to create our GitHub Action secrets. Go to Settings > Secrets and variables > Actions. Create 4 new repo secrets:

• AWS_ROLE: ARN value of the role that we created in step 5.
• AWS_BUCKET_NAME: Name of the S3 bucket that we created in step 3.
• AWS_REGION: Region of the S3 bucket that we created in step 3.
• AWS_BUCKET_KEY_NAME: Name of the terraform state file that will be created in the S3 bucket. I named mine infra.tfstate.

8. Stage files and commit. Either:

• push to main branch
• make a new PR request and merge it

9. GitHub Actions will start running. Wait until it finishes. Hopefully no errors will be produced. If you get any errors, create a new issue in this repo.

Destroy terraform resources

1. Delete everything in main.tf.
2. Push committed changed to remote repo.

Notes

1. Video explanation of the infrastructure