What is a paired RDD in Spark?

In Spark, a paired RDD is an RDD that stores data as a key-value pair. Paired RDDs are implemented as a collection of Tuple2 objects. Tuple2 is a Scala class that is an ordered set of two elements. The first element of a Tuple2 object is considered to be the key and the second the value. The groupByKey, sortBykey, reduceBykey, countByKey and join are some of the useful transformations used for manipulating paired RDDs in Spark.