传统的排队模型已经难以满足云计算中任务调度系统的高效和低成本目标要求,为缩短云任务调度过程任务等待时间及提高虚拟机任务调度系统的执行效率,提出一种云环境下基于 M/Geom/C/∞排队系统的任务调度模型,利用改进的排队模型提高云任务调度系统性能;对该模型中系统的嵌入马尔可夫链性质、平衡条件、稳态分布和条件随机分解结果进行了分析,给出该模型的稳态队长的随机分解和稳态等待时间等性能指标;结合数值例子,准确的找到服务率与期望队长、期望等待时间之间的关系及其它系统稳态性能指标;通过云任务调度系统的仿真,实验结果验证了该模型能够快速地完成云任务的调度,提高了虚拟机资源的平均利用率.%The traditional queuing model has been difficult to meet the requirements of high efficiency and low cost of task scheduling system in cloud computing,in order to shorten task waiting time and improve the execution efficiency of task scheduling system in the cloud task scheduling,a task scheduling model based on M/Geom/C/∞ queuing system in cloud environment is proposed,the improved queuing model is used to improve the performance of cloud task scheduling system.The properties of the embedded Markov chain,the equilibrium condition,the steady -state distribution and the conditional random decomposition are analyzed,the performance index of the model,such as stochastic decomposition and steady -state waiting time of the steady -state queue length,are given.A numerical example is given to accu-rately find the relationship between the service rate and expected queue length,expected waiting time,and other steady -state performance index of the system.Through the simulation of the cloud task scheduling system,the experimental results show that the model can quickly complete the scheduling of cloud tasks,and improve the average utilization of virtual machine resources.
展开▼
