Apache Spark, the in-memory and real-time data processing framework for Hadoop, turned heads and opened eyes after version 1.0 debuted. The feature changes in 1.2 show Spark working not only to ...