The HTTP RA does not do any kind of load balancing by itself. You will need a separate HTTP load balancer product. There are several well-known open source ones around, such as HAProxy or Nginx. Or there are commercial ones like F5 BIG-IP.
The load balancer can be deployed in front of your Rhino cluster, so that incoming HTTP requests are distributed across all Rhino nodes. Alternatively if you want to load balance outgoing requests from the HTTP RA to some other cluster, then you would deploy the load balancer between Rhino and the other servers so that it can receive the requests and distribute them across the server nodes.
Does this answer your question? If you can provide more information about what you want to achieve we might be able to give you some more suggestions.