面向高校教学云服务的虚拟机动态调度方法

A dynamic scheduling method of virtual machines for campus cloud teaching services

随着教学信息化的不断深化,高校云平台越来越普及,但是实际应用中资源利用率仍然较低,核心问题在于当前的虚拟机调度机制未考虑高校教学应用的特征,从而导致负载不均和资源浪费。为了解决这一问题,从高校教学应用需求出发,提出了一种虚拟机动态调度算法(CRS),定义了课程虚拟机模型和物理机负载模型,并实现了基于OpenStack开源云平台的可对虚拟机进行动态调度的校园云平台。实验表明,提出的虚拟机动态调度方法达到了降低能耗及实现负载均衡的目标。

With the continuous deepening of teaching informatization, the campus cloud platform is becoming increasingly popular, but the resource utilization in practical applications is still low. The key problem is that the current scheduling mechanism of virtual machines does not take into account the characteristics of campus teaching applications, resulting in a waste of resources and load unbalancing. To tackle this problem, according to the requirements of teaching applications in universities, we put forward a dynamic scheduling algorithm of course virtual machines（CRS）, define the course requirement model and the physical machine load model, and implement a campus cloud platform based on the Open-Stack. Experimental results show that the platform achieves energy consumption reduction and load balancing.