change language around sharded clusters and mongos configs

Author: i80and

source/core/sharded-cluster-components.txt

@@ -46,32 +46,33 @@ for a production sharded cluster deployment:
 Where possible, consider deploying one member of each replica set 
 in a site suitable for being a disaster recovery location.
 
-Sharding requires at least two shards to distribute sharded data. Single 
-shard sharded clusters may be useful if you plan on enabling sharding in the 
-near future, but do not need to at the time of deployment.
-
-Deploying multiple :binary:`~bin.mongos` routers supports high availability
-and scalability. A common pattern is to place a :binary:`~bin.mongos` on
-each application server. Deploying one :binary:`~bin.mongos` router on each
-application server reduces network latency between the application and
-the router.
-
-Alternatively, you can place a :binary:`~bin.mongos` router on each shard
-primary. This approach also reduces network latency between the
-application and the router: applications use a :doc:`connection
-string </reference/connection-string>` listing all the hostnames of each shard primary. The MongoDB
-driver then determines the network latency for each :binary:`~bin.mongos`
-and load balances randomly across the routers that fall within a set
-:ref:`latency window <selection-discovery-options>`. Ensure that the
-server hosting the shard primary and :binary:`~bin.mongos` router has
-sufficient capacity to accommodate the extra CPU and memory
-requirements.
+Sharding requires at least two shards to distribute sharded data. Single
+shard sharded clusters may be useful if you plan on enabling sharding in
+the  near future, but do not need to at the time of deployment.
+
+Deploying multiple :binary:`~bin.mongos` routers supports high
+availability and scalability. A common pattern is to place a
+:binary:`~bin.mongos` on each application server. Deploying one
+:binary:`~bin.mongos` router on each application server reduces network
+latency between the application and the router.
+
+Alternatively, you can place a :binary:`~bin.mongos` router on dedicated
+hosts. Large deployments benefit from this approach because it decouples
+the number of client application servers from the number of
+:binary:`~bin.mongos` instances. This gives greater control over the number
+of connections the :binary:`~bin.mongod` instances serve.
+
+Installing :binary:`~bin.mongos` instances on their own hosts allows these
+instances to use greater amounts of memory. Memory would not be shared
+with a :binary:`~bin.mongod` instance. It is possible to use primary shards
+to host :binary:`~bin.mongos` routers but be aware that memory contention may
+become an issue on large deployments.
 
 There is no limit to the number of :binary:`~bin.mongos` routers you can
-have in a deployment. However, as :binary:`~bin.mongos` routers communicate
-frequently with your config servers, monitor config server performance
-closely as you increase the number of routers. If you see performance
-degradation, it may be beneficial to cap the number of
+have in a deployment. However, as :binary:`~bin.mongos` routers
+communicate frequently with your config servers, monitor config server
+performance closely as you increase the number of routers. If you see
+performance degradation, it may be beneficial to cap the number of
 :binary:`~bin.mongos` routers in your deployment.
 
 .. include:: /images/sharded-cluster-production-architecture.rst