Since we are not resetting the cluster parameter during remove shard, it is possible that a customer has a sharded cluster that they restart as a replica set or standalone. In this case, we should stop checking for direct shard operations.

To do this, we should add a check for whether sharding is enabled to the service entry point checks.