Contains files for deploying and managing the core infrastructure. We use terraform for the deployment of our infrastructure. These files describe how the infrastructure needs to be deployed on AWS, our cloud provider.
For each environment, there is a folder with the Terraform files. Per environment, we split the different resources into different folders. This is to keep the state separate and not let changes in one part effect others. Additionally, the change rate between resources might be different. For example, the VPC probably won't change as often as the kubernetes cluster. We followed, this guide: https://blog.gruntwork.io/how-to-manage-terraform-state-28f5697e68fa In the terraform-state-storage are the files with which the initial state storage resources were created.
Terraform should be run locally. To synchronise, follow these steps:
- Pull the latest version from the repository
- Move to the correct environment folder (example
cd acceptance) - Start with running the vpc folder as this creates all the networking for the other resources
- Now run apply all other resources You can run the terraform files with:
- Run
terraform initthis will download all required files and version - Run
terraform planto see what changes terraform proposes - Run
terraform applyto start apply the changesyesto approve the changes and run terraform
This should successfully update the infrastructure.
Be aware that the master password for the Document store (document-store) needs to be set manually.
It is secret, so we cannot push it to the GitHub repository.
For existing environments this secret is managed in the AWS Secret Store.
This needs to be fixed, see https://blog.gruntwork.io/a-comprehensive-guide-to-managing-secrets-in-your-terraform-code-1d586955ace1
Be aware that the route table needs an update after the peering. This is still a manual action and needs to be automated with Terraform.