Terragrunt
As you develop larger deployments, your Terraform code becomes more complex. You will also start to have teams working on the same deployment simultaneously. As discussed in previous chapters, it is advisable to break larger Terraform deployments into more manageable parts, each residing in its own subdirectory and each having its own state file. Unfortunately, that creates overhead, and you find yourself repeating yourself. For example, each subdirectory has its own provider.tf and backend.tf files. As we have already experienced, we cannot use interpolation in the backend.tf file and we need to edit each backend.tf file explicitly. Furthermore, we need to run Terraform in each subdirectory independently.
This is where Terragrunt comes into the picture. Terragrunt is defined as a thin wrapper to keep your Terraform code DRY – as in, in line with the don’t repeat yourself (DRY) principle. Terragrunt achieves this by preparing each subdirectory, for example...
