Update program to binary directive

Author: kay-kim

source/faq.txt

@@ -22,11 +22,11 @@ To promote data locality,
   system or one per shard for sharded systems.
 
 - Use a :readmode:`nearest` read preference to read from the local
-  :program:`mongod`.
+  :binary:`~bin.mongod`.
 
-- For a sharded cluster, you should have a :program:`mongos` on the
+- For a sharded cluster, you should have a :binary:`~bin.mongos` on the
   same nodes and use :ref:`localThreshold <spark-input-conf>`
-  configuration to connect to the nearest :program:`mongos`. 
+  configuration to connect to the nearest :binary:`~bin.mongos`. 
   To partition the data by shard use the 
   :ref:`conf-mongoshardedpartitioner`.
 
@@ -43,7 +43,7 @@ For an example, see :mongo-spark:`SparkStreams.scala
 How do I resolve ``Unrecognized pipeline stage name`` Error?
 ------------------------------------------------------------
 
-In MongoDB deployments with mixed versions of :program:`mongod`, it is
+In MongoDB deployments with mixed versions of :binary:`~bin.mongod`, it is
 possible to get an ``Unrecognized pipeline stage name: '$sample'``
 error. To mitigate this situation, explicitly configure the partitioner
 to use and define the Schema when using DataFrames.