摘要
网站云是一种多租户的云部署架构.研究了基于Eucalyptus实现的网站云资源综合调度策略,提出了以HTTP请求为细粒度任务的贪婪调度算法,并引入了带权机器学习算法模型,使系统能主动感知请求高峰并进行负载均衡.实验证明达到了较理想效果.
Website cloud is a multi-tenant cloud deployment architecture. This thesis studies the resources scheduling policy of the Eucalyptus-based website cloud, and the HTTP request as f＇me-grained task for the greedy scheduling algorithm. It also discusses the model of weighted machine learning algorithms, that the system can take the initiative to perceive the request peak and load balancing. Experiments show that the ideal.
出处
《计算机系统应用》
2012年第10期174-178,共5页
Computer Systems & Applications
