This article propose an new method based on weight of network delay by analysis the relationship of one nodes network delay and the storage performance of CEPH. The experiments show that the method could adjust the nodes weight depended on the network delay of that node
so that data files could be stored in good network performance with greater probability. This new method reduces the time cost caused by network delay and improves the performance of CEPH storage system in that case.
关键词
Keywords
references
Garcia H, Ludu A. The Google file system[J]. AcmSigops Operating Systems Review, 2003, 37(5): 29-43. DOI: 29
Mansouri N, Dastghaibyfard G H. Enhanced scheduling strategy in datagrid[J]. Journal of Parallel % Distributed Computing, 2013, 73(4): 534-543 DOI: 534
Weil S A, Brandt S A, Miller E L, et al. CRUSH: controlled, scalable, decentralized placement of replicated data[C]// ACM/IEEE Conference on Supercomputing. ACM, 2006: 31.
衣国垒. 基于中心节点架构的大规模数据对象存储系统[D]. 北京: 清华大学, 2013.
Ling Zhan, Fang Xieyun, Li Daping. The research and implementation of metadata cache backup technology based on CEPH file system[C]//IEEE International Conference on Cloud Computing and Big Data Analysis (ICCCBDA). IEEE, 2016: 72-77.
Jiang Y, Zhou Y, Wang W. Task allocation for undependable multiagent systems in social networks[J]. Parallel % Distributed Systems IEEE Transactions on, 2013, 24(8): 1671-1681. DOI: 1671
Decandia G, Hastorun D, Jampani M, et al. Dynamo: amazon's highly available key valuestore[J]. Acm Sigops Oper. syst. rev, 2007, 41(6): 205-220. DOI: 205