Cloud Programming Models (MapReduce)
Vladimir Dimitrov · 2016
MapReduce is a leading programming model for big data analytics. It uses pure functional concepts that benefit the highest level of parallelism granularity. Programming in this model is in nonfunctional programming languages such as Java, JavaScript and C++. The programmer has to write pure functional programming (without side effects) with nonfunctional programming language. MapReduce is a simple paradigm for programming large clusters of hundreds and thousands of servers that store many terabytes and petabytes of information. This chapter provides an overview of the MapReduce programming model, its variants, and its implementation.