There are two main operations that can be performed on a RDD in Spark:
1. Transformation: This is a function that is used to create a new RDD out of an existing RDD.
2. Action: This is a function that returns a value to Driver program after running a computation on RDD.