Provider does not destroy IAM roles and subsequent runs result in broken Federated OIDC provider configurations

[InfoSec812]

I recently created/destroyed a cluster several times to test the terraform manifests we are using. The first cluster seemed to work OK, but we needed to change some things and test again... Subsequent runs of the provider resulted in clusters which were not very functional in multiple ways and this morning I discovered that all of the Federated roles in IAM for that cluster were referring to an OIDC identity provider which no longer exists!

I re-ran my destroy and discovered that the IAM roles are not being destroyed/deleted, and so when I run a new apply command it skips creating/updating those resources and builds a cluster which is somewhat non-functional.

IMHO, the rhcs provider (or the underlying AWS providers) should destroy those roles or at the very least update them when creating a new cluster with the same name/tags.