摘要
随着教学信息化的不断深化,校园云平台越来越普及,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校园应用的特征,从而导致负载不均和资源浪费。为了解决这一问题,提出面向校园云平台的虚拟机调度机制,定义课程需求模型、物理机负载模型,并在此基础上提出虚拟机部署方法。该机制可适应教学应用周期性、可预测性、批量性等特点,并实现节能及负载均衡的目标。实际平台验证表明,该机制可有效降低能耗并实现负载均衡。
With the continuous deepening of teaching informatisation, campus cloud platform is becoming increasingly popular, but theresource utilisation is still low in practical applications, and the key problem is that current scheduling mechanism of virtual machines doesn, ttake into account the characteristic of campus applications, which results in a waste of resources and load imbalancing. To tackle thisproblem, this paper puts forward a scheduling mechanism of virtual machines for campus cloud platform, defines the course requirementmodel and the physical machine load model, and proposes on this basis a virtual machine deployment algorithm. This mechanism can adapt tothe characteristics of teaching applications such as periodicity, predictability and batching, etc. , and achieves the goals of energy saving andload balancing. Actual platform verification shows that the mechanism can effectively reduce power consumption and realise load balancing.
作者
马倩
王菁
王岗
Ma Qian;Wang Jing;Wang Gang(Beijing Key Laboratory on Integration and Analysis of Large-Scale Stream D ata, North China University of Technology, Beijing 100144 , China)
出处
《计算机应用与软件》
CSCD
2016年第8期99-104,共6页
Computer Applications and Software
基金
国家自然科学基金项目(610330 06)
北京市自然科学基金项目(4131001)
北京市属高等学校创新团队建设与教师职业发展计划项目(IDHT20130502)
关键词
校园云
调度机制
虚拟机放置
降低能耗
负载均衡
Campus cloud Scheduling mechanism Virtual machine deployment Energy saving Load balancing