结构化P2P网络的负载均衡技术研究

Research on Loading Balance in Structured P2P Network

在结构化对等网络中,负载失衡现象的产生将会造成单点失效、网络拥塞、请求延迟甚至是系统瘫痪等系统应用性能问题。本文针对传统复制算法中存在的无有效的预防热点策略、忽视节点性能差异性和副本维护机制等问题,提出基于节点性能分类和数据,节点分析的自适应负载均衡算法。通过实验模拟证明本文算法能够有效地实现结构化对等网络的负载均衡,降低数据请求报文的丢失率,从而有效地提高系统节点的利用率。

In structured P2P network,loading imbalance results in the problem of single failure,network congestion,request delay and even worst problems.This paper proposes an effective loading balance algorithm to prevent hot spots based on the analysis of requested datas and nodes processing difference.Through experiments and simultaed results,the proposed algorithm can effectively improve loading unbalance in structured P2P network.