terraplate VS terraformsh

I won't be able to share the Terraform files (customer work), but we use the AWS EKS module for the clusters, and our own quite specific modules, and then use Terraplate (a project I authored) to keep the Terraform configs dry.

People have been solving the DRY problem for many years. Terragrunt probably being the most well-known tool for it. I am working on a project Terraplate to solve this whilst keeping vanilla terraform and allowing me to use Terraform Cloud.

This is a problem me and my colleagues keep coming across and we were using Terragrunt fairly successfully. We have recently been working on a simple templating tool to keep everything DRY and as vanilla Terraform, called Terraplate: https://github.com/verifa/terraplate

I created Terraformsh to automatically pass multiple config files in different directories so symlinks aren't needed. The idea is to use a hierarchy, so the full file path to the tfvars files in parent directories get passed on with `-var-file` options. The end result is that there aren't really "global variables", but instead just a series of variables that are split up into different hierarchies, so that any given module can use them if they need them, or ignore them if they don't need them. To differentiate potentially-conflicting variables I just give the variables unique names.

It looks like we've both created the same solution! :) ltf looks very similar to terraformsh. Good luck with your project!

Since people seemed to like it last time I mentioned it here, I just released another version of my Terraform wrapper, Terraformsh: https://github.com/pwillis-els/terraformsh

I noticed Terragrunt was kind of complicated, so I made this to make it easier to keep Terraform code DRY without getting complicated: https://github.com/pwillis-els/terraformsh

This is similar to how I lay out my projects for use with my terraform wrapper, Terraformsh: https://github.com/pwillis-els/terraformsh/blob/main/DRY_CODE.md