在大规模P2P网络中,为了提高数据的可用性,可以通过在系统中多个节点上建立数据副本的方法来增加数据的冗余度。为了减小副本建立的开销,可采用一种基于文件分片的副本建立策略,将需要复制的数据文件先进行分片,然后把这些分片在一组选取好的候选节点间进行分发复制。该方法能够充分利用网络带宽,明显减小建立副本所需的时间开销。
In large scale of P2P network, in order to get high data availability, to actively replicate data to other nodes is used to increase data redundance. To reduce replication cost, it was better to use a file-partition based replication strategy. Data file could be partitioned into pieces which would then be distributed and copied within a collection of node selected beforehand. It can sufficiently use the network bandwidth and obviously reduce the replication time cost .