An application is hosted on a three-node cluster in which each server has identical compute and network performance specifications. A fourth node is scheduled to be added to the cluster with three times the performance as any one of the preexisting nodes. The network architect wants to ensure that the new node gets the same approximate number of requests as all of the others combined. Which of the following load- balancing methodologies should the network architect recommend?