Abstract: In order to comprehensively and objectively analyse the performance for cloud computing systems center,this paper presents a performance analysis model of cloud computing center based on queuing theory.The characteristics of the user service request which reaches the center are described and a the system model for batch arrivals by using queuing theory is established,and the length of steady-state probability is solved in probability space.Finally,the simulation experiments are carried out to test the performance by using blocking probability,immediate service probability,etc.Simulation results show that,with the increase of the number of batch arrival request,average length of cloud computing center increases,and increasing the buffer length can reduce the blocking probability.

Key words: cloud computing system, performance analysis, queuing theory, mean waiting queue size, batch arrivals

摘要： 为全面、客观分析云计算系统中心的性能,提出一种改进的云计算中心系统分析模型。利用泊松分布描述批量到达云计算中心的用户服务请求变化特点,建立基于排队论的批量到达系统模型,在概率空间上求解队长的稳态概率,采用仿真实验对阻塞概率、立即服务概率等性能指标进行分析。仿真结果表明,随着批量到达请求数增加,云计算中心的平均队长相应增加,降低了系统的阻塞概率。

关键词: 云计算系统, 性能分析, 排队论, 平均等待队长, 批量到达

CLC Number: 

• TP393