Data Pipeline VS. Glue: Data pipeline has more control over the orchestration and code (not just Spark) Glue automatically scales so does its underlying Spark cluster