Abstract: Google proved that two basic concepts from functional programming—the “map” and “reduce” functions from Lisp—are enough to express many important algorithms in a large-scale, distributed setting, by creating MapReduce. Hadoop open-sourced and popularized this approach. More recently, challenges in implementing classes of algorithms important in machine learning, as well as the desire for a unified programming model for batch processing, query processing à la Hive, and stream processing, have motivated the development of Spark, which is written in the object-functional language Scala. In this Meetup, we’ll take a quick spin through the Spark stack, and touch lightly on why Scala is the perfect language in which to develop Spark applications.
Bio: Paul Snively is an architect at Verizon OnCue, where he uses Scala and related technologies daily to build TV-related services. He’s a four-year veteran professional Scala developer, with several more years as a hobbyist. He was also the technical editor of the first edition of “Beginning Scala.”