For questions about join optimizing in Apache Hive when one the join tables is small enough to fit into the Mapper's memory, then there is no need to launch the Reducer and the Join can be done on Mapper.
