数据复制技术是一种提高P2P系统中数据可靠性和可用性的常用策略.现有复制方法大多只考虑副本数量,副本数量越多就越能提高资源访问效率,但采用这样的数据复制方法将会带来高昂的副本一致性维护代价.为平衡副本一致性维护的开销和多副本带来的访问性能提升之间的关系,该文提出了动态副本分布方法.文中首先给出了副本目录的设计和副本信息的获取方法,能够获得某一逻辑资源的所有副本信息.然后,根据逻辑资源的全局副本信息,对访问频率高且平均响应时间长的数据资源进行复制,并给出副本数量的计算方法.最后,根据用户访问特征和节点实时带宽等信息计算放置副本的最佳地点,使副本分布能够适应数据访问请求和网络带宽的动态变化.模拟实验结果显示,该方法能够实现全局优化的副本分布,以少量数据副本提升资源访问的性能.
Replication is a common method used to improve the performance of data access in P2P systems.It improves not only data access efficiency,but also data availability and fault tolerance.The larger number of replicas,the better performance it can obtain.However,a large number of replicas may lead to high overhead for unnecessary data replication and consistency maintenance in case of updates.The optimal replication strategy MACR (Minimum Access Cost based Replication strategy) is presented which takes into account the access frequency,the status of the network connection and average response time to perform optimal replication.Firstly,the design of the replica catalog is described.To each logical data resource,the replica catalogs have all the information of the replicas and the optimal replication strategy can benefit from it.Secondly,we use the access frequency and the average response time to decide which data resource should be replicated.Then,the method of calculating the number of the replicas is proposed.At the last,MACR strategy finds the appropriate site and executes replication.The simulation results show that MACR offers the benefits of shortening the response time with fewer replicas.