传统的副本管理策略存在副本利用率偏低的问题,在设置副本时未考虑到节点的动态性和异构性,使系统中产生过多无用的副本并对热区不能有任何的缓解,影响系统的可扩展性。本文提出一种基于副本收益的副本复制算法。根据设置副本所减少的查询跳数的多少来决定是否在节点上设置副本。研究的内容包括:研究文件副本的收益模型,设计副本复制算法等。算法的目标是复制副本到选出的节点能以较小的开销获得最大的收益,即复制的节点所节省的查询跳数是最少的。 The traditional copy management strategy has the problem of the low copy utilization rate. When setting the copy, the dynamic and heterogeneous nodes are not taken into account, which results in excess useless copy in the system and can not alleviate the hot area. System scalability. In this paper, we propose a replica replication algorithm based on replica gains. According to set the copy to reduce the number of query hops to decide whether to set a copy on the node. The research contents include: income model of research paper copy, design copy copy algorithm and so on. The goal of the algorithm is to replicate the replicas to the selected node to get the maximum benefit with a small cost, that is, the number of query hops saved by the replicated node is the minimum.