Suggested workarounds for discovering Terragrunt dependency changes during plans?

[iantybo]

If I have the following configuration, where my Aurora cluster is dependent on my VPC, how can I let the Aurora cluster know during the plan phase that it should also be recreated?

# vpc/terragrunt.hcl 

terraform {
  source = "[email protected]:iangrunt/terraform-fake-modules.git//modules/aws/vpc?ref=v1.0.0"
}

inputs = {
  # Old value at the time when we applied:
  #  cidr_block = "10.111.0.0/16"

  # New value that will recreate my VPC:
  cidr_block = "10.222.0.0/16"
}

# aurora-postgres/terragrunt.hcl 

terraform {
  source = "[email protected]:iangrunt/terraform-fake-modules.git//modules/aws/aurora?ref=v1.0.0"
}

dependency "vpc" {
  config_path = "../vpc"
}

inputs = {
  vpc_id = dependency.vpc.outputs.id
}

When I run a terragrunt run-all plan I see my VPC being recreated correctly as the cidr_block has changed to force that, but I need to let my Aurora cluster know it will also be recreated:

---

Tracked in ticket #109200

[yorinasub17]

This is unfortunately a known limitation of run-all plan where there is no clear workaround we have implemented yet. This specific issue and suggested alternative workflows are documented in the associated warning for run-all plan in the documentation.