缓存是解决流媒体网络瓶颈和磁盘I/O瓶颈的重要方法.用户在访问流媒体系统的时候呈现出非常复杂的行为,这些行为对于设计有效的流媒体缓存方法至关重要.现有的算法往往以泊松分布,zipf分布来表示用户行为,这往往就失之精确.本文根据用户的点播行为和操作行为,设计了符合实际用户行为的大容量缓存算法,并进行实验,实验结果表明效果良好. Caching is an important way to solve bottlenecks in streaming media and bottlenecks in disk I / O. Users are experiencing very complex behavior when accessing streaming media systems, which are crucial for designing effective streaming media caching approaches. Existing algorithms Often Poisson distribution, zipf distribution to represent the user behavior, which often lose accuracy.In this paper, according to the user’s on-demand behavior and operational behavior, designed in line with the actual user behavior of large-capacity cache algorithm, and the experimental results show that the effect good.