【作者】 赵竞； 余宏亮； 张堃； 郑纬民；

【Author】 ZHAO Jing~1 YU Hongliang~1 ZHANG Kun~1 ZHENG Weimin~1 (1.Department of Computer Science and Technology,Tsinghua Univ.,Beijing,100084)

【机构】 清华大学计算机科学与技术系；

【摘要】 如何维护广域网分布式哈希表存储系统中副本的可用性和一致性一直是一个难题。本文提出的解决策略记录了副本产生、修改及维护过程中的失败信息,并且根据这些信息对副本进行修复迁移。这一策略不仅能维护副本的最终一致,而且能加速副本修改信息的扩散。该策略使用一种适应性的算法来减小额外的带宽消耗带来的影响。经过模拟程序以及实际系统的验证,和传统的扩散算法相比,本文提出的策略能够把副本修复的速度提高10倍以上,同时有效避免了带宽波峰给系统带来的影响。更多还原

【Abstract】 It is a challenge to design and implement a wide-area distributed hash table(DHT)which provides a storage service with high availability and consistency of the replicas.This paper explores this problem and provides a possible solution,which records failing information during the process of generating,updating and moving replicas,and employs the information for recovery work.This solution,including an adaptive recovery strategy not only keeps eventual consistency of replicas but also quickens the spread of updates,and also saves bandwidth.This paper provides simulation results of the strategy.Compared with traditional epidemic algorithms,our strategy can provide an over 10 times’ quicker replica recovery speed,and can avoid bandwidth burst effectively.更多还原