terraforming VS rain

Terraforming

I would also strongly recommend an iterative triage process: don't feel that you need to solve everything all once. It's a huge amount of progress if you can get to a stage where you understand the relationships between services, have enough monitoring to identify failures, and can iteratively move services to more granular IAM policies even if it's all still running in one big account. Simply using a tool like terraformer or terraforming to move the manual configuration into an IaC workflow is an accomplishment, especially since it helps you both make changes more confidently and identify where new changes are still being made. Depending on the politics, that last part might be important: you'll be unpopular if you take away everyone's rights but you'll probably have better luck first nudging people to make them using a sensible flow rather than by hand.

Haven't tried this but if I ever find myself in your situation I'll give it a spin https://github.com/dtan4/terraforming

There are a couple of tools to automate creating terraform configuration and prepopulate state from existing resources, like terraforming, terraformer or cf-terraforming. But it is still cumbersome and laborious and in my experience, it's usually way easier to just recreate everything within terraform from scratch wherever that's possible.

🧪 Terraforming export existing AWS resources to Terraform style (tf, tfstate).

Similar and also written in Go: Rain https://github.com/aws-cloudformation/rain

Some similar IaC dynamic analysis tools written in golang:

https://github.com/terraform-linters/tflint-ruleset-aws/blob...

https://github.com/aws-cloudformation/rain/pull/93

🌧 Rain A development workflow tool for working with CloudFormation. ❗️Rain is currently in preview and shouldn't yet be considered stable enough for production use.