Retaining resources after terraform destroy

Question

I'd like to retain CloudWatch logs after I spin down a bunch of resources created using terraform - which includes the CloudWatch log group. Is there a way to tell terraform destroy to spare some resources?

I suppose I could manually remove CloudWatch resources from tfstate before calling destroy, doesn't seem like the right approach.

Answer 1

There is a meta-argument called lifecycle for that.

resource "aws_s3_bucket" "MyPreciousBucket" {
  lifecycle {
    prevent_destroy = true
  }
}