摘要
现有的负载均衡算法大都基于Web服务,并不适用于视频点播(VOD)。而基于VOD的负载均衡算法没有考虑用户操作以及一个节目不同片段的访问概率差异等因素对系统负载的影响。因此,设计一种基于统计学模型的负载均衡设计方法,从用户的随机性操作出发,统计分析其规律,根据不同节目不同片段的不同点播率对片源进行分块备份存储。通过仿真实验证实了算法较其他算法的优越性,提高了各节点服务器资源利用率,达到了负载均衡的目的。
Most load balancing algorithms are based on Web service and unfit for Video on Demand（VOD）system.They all neglect the effect of user＇s behavior and the differences among different segments in one program.Therefore,the load balancing algorithm based on statistical models was proposed.It commenced from random user behavior,analyzed the implied rules,divided the programs into pieces,copied and stored them,according to different order probability of diverse segments in diverse programs.The results of simulation testify its superiority to other algorithms.It improves the utilization rate of server resource and achieves load balance.
出处
《计算机应用》
CSCD
北大核心
2010年第3期765-767,共3页
journal of Computer Applications
关键词
负载均衡
视频点播
统计模型
用户操作
load balance
Video on Demand（VOD）
statistical model
user operation
