摘要
研究了数据网格中数据副本分布与存取的关键问题.基于对动态副本访问模式预测机制的分析,同时考虑QoS,提出“读-复制”和“测试-删除”机制,实现了一种基于隐含QoS-簇映射的分布式自适应副本分布算法(QCMR-DG).模拟试验表明,QCMR-DG算法能够以相当低的代价为数据网格应用提供至少满足各自QoS要求的较高数据存取速率.
In this paper,we study the key problems of the replica distribution associated with the data access in data grids.Having analyzed the forecasting mechanism of the dynamic replica access pattern while taking QoS into account,we implement QCMR-DG,a distributed self-adaptive replica distribution algorithm based on implicit 'QoS-Cluster' mapping,with the help of replicate-read/test-delete mechanism proposed by us.The experiment shows that QCMR-DG can provide data grid applications good data access speed(better than the required QoS for each application respectively) with comparatively low overhead.
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2005年第z1期11-14,共4页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家高技术研究发展计划资助项目(2002AA444140)
国家重点基础研究专项经费资助项目(G1999032705)
关键词
分布式系统
QoS-簇
数据副本
有偏分布
QoS-覆盖
按簇复制
副本一致性维护
distributed system
QoS-cluster
data replica
skewed-distribution
QoS-overlay
replication on clusters
replica consistency maintenance
