Abstract :
The parallel processing technologies develop vigorously in the recent decade, along with the increasing challenges of Big Data. In particular, many institutions prefer to manage their massive data with the MapReduce paradigm, which is proposed by Google in 2003, because of its simplicity and remarkable scalability. However, from Day One MapReduce is proposed, the argument between it and parallel DBMSs never stops since it over-focuses on the scalability but overlooks the efficiency. Consequently, the MapReduce extensions and variants are studied continuously in order to overcome the shortcomings without disrupting the scalability. This paper reviews such systems, from Google and the other communities, trying to indicate the directions for the future research.
