综合论文：ParallelDataProcessingwithMapReduce.Asurvey（英文）.pdfVIP

Parallel Data Processing with MapReduce: A Survey Kyong-Ha Lee Hyunsik Choi Bongki Moon Yoon-Joon Lee Yon Dohn Chung Department of Computer Department of Computer Department of Computer Science Science Science and Engineering University of Arizona KAIST Korea University bkmoon@ bart7449@ hyunsik.choi@korea.ac.kr yoonjoon.lee@kaist.ac.kr ydchung@korea.ac.kr ABSTRACT mains where the MapReduce framework is adopted and A prominent parallel data processing tool MapReduce is gain- discuss open issues and challenges. Finally, Section 7 ing significant momentum from both industry and academia concludes this survey. as the volume of data to analyze grows rapidly. While MapRe- duce is used in many areas where massive data analysis is re- 2. ARCHITECTURE quired, there are still debates on its performance, efficiency MapReduce is a programming model as well as a per node, and simple abstraction. This survey intends to framework that supports the model. The main idea assist the database and open source communities in under- of the MapReduce model is to hide details of parallel standing various technical aspects of the MapReduce frame- execution and allow users to focus only on data pro- work. In this survey, we characterize the MapReduce frame- cessing strategies. The MapReduce model consists of work and discuss its inherent pros and cons. We then intro- two primitive functions: Map and Reduce. The inp