That's short-sighted and poor engineering. You need your service to be able to keep going no matter what any single vendor does. Never intentionally create a single point of failure like that.