This document evaluates the performance of Spark SQL using the BigBench benchmarking framework, comparing it to MapReduce and Hive. It discusses the scalability of various queries, provides insights into the configuration of the testing cluster, and examines the findings from running BigBench on Spark and MapReduce. Ultimately, it highlights that Spark SQL performs faster than Hive for stable queries, though some Spark SQL queries show scalability issues due to potential join optimization problems.