Latent Behavior Based Replica Replacement Algorithm in Data Grid
The main problem in data grids is how to access the huge and widely distributed data timely given the limited number and capacity of storage element (SE), and high latency of the interconnection network. One approach to address this problem is to replicate the files locally, and then remote access overheads are avoided. Because the number and size of SE are limited, a good replica replacement algorithm (RRA) is needed, and it is also the focus of this paper. Specifically, we propose a new algorithm LBBRRA (Latent Behavior Based Replica Replacement Algorithm). We evaluate the LBBRRA using the data grid simulator OptorSim, and results show LBBRRA performs better than the LRU and LFU.
replica replacement algorithm optimization data grid data mining frequent item set latent behavior
Ji Huifang Xu Gaochao Jiang Jianhua Hanqing Zhou
College of Computer Science and Technology Jilin University Changchun, China Department of Information, Changchun Taxation College, Changchun, China School of Computer Science and Engineering, Beihang University, Beijing, China
国际会议
海口
英文
285-289
2011-02-22（万方平台首次上网日期，不代表论文的发表时间）