We are now using the llink database within our gcp datbase instances, so likely won't be needing this workaround