Your search cluster is made up of one or more partitions, each of which is made up of one or more nodes. As your search collection becomes larger, the collection can be partitioned into smaller pieces to facilitate more efficient access to the data. As the Search Service becomes more heavily utilized, replicas of the existing partitions, in the form of additional nodes, can be used to distribute the load. Additional nodes also provide fault-tolerance; if a node becomes unavailable, queries are automatically issued against the remaining nodes.