A data placement strategy based on clustering and consistent hashing algorithm in cloud computing
Qiang Li, Kun Wang, Suwei Wei, Xuefeng Han, Lili Xu, Min Gao · 2014
To reduce time delay of processing data and improve the efficiency of cloud computing, a clustering algorithm based on the principle of minimum distance is proposed to place user-based and item-based data, update cluster centre and the threshold dynamically. Besides, consistent hashing is combined to solve the fault tolerance and scalability issues. In addition, Case-Based Reasoning (CBR) algorithm and item-based Coordination Filtering (CF) algorithms are used for filling sparse matrix to achieve better effect on the user-based clustering. Simulation results show that compared with data placement strategy based on K-means algorithm, this data placement strategy significantly improves clustering accuracy, greatly reduces delay of processing data and increases database scalability and redundancy, thereby improving the efficiency of cloud computing.