Astronomical Data Application Research Based on MapReduce

Abstract

MapReduce as an abstract distributed computing programming model could solve the issues of parallel computing, such as load balancing, network storage, data distribution, resource allocation, fault tolerance. This makes it easy for people to manipulate large scale cluster systems without considering hardware details. The paper discusses completely how to apply MapReduce. In the construction of the experimental platform, this paper successfully designs and implements the cone search service based on MapReduce, The final result proves that the astronomical data application method based on MapReduce greatly improves the processing capacity.