A data placement algorithm for data intensive applications in cloud

Abstract

Data layout is an important issue which aims at reducing data movements among data centers to improve the efficiency of the entire cloud system. This paper proposes a data-intensive application oriented data layout algorithm. It is based on hierarchical data correlation clustering and the PSO algorithm. The datasets with fixed location have been considered, and both the offline strategy and the online strategy for data layout have been given. As this proposed strategy is aimed at reducing the global amount of data transmissions, and the special permission of the datasets has been introduced, the cost of data transmission can be measured more reasonable. Simulation results show that compared with two classical strategies, our algorithm can reduce the amount of data transmission more effectively.