基于容器技术的云计算资源自适应管理方法

发布时间：2018-03-24 16:41

本文选题：云计算　切入点：Linux容器　出处：《计算机科学》2017年07期

【摘要】：云计算的发展使得越来越多的软件应用选择云平台作为部署平台。为了应对动态变化的工作负载、应用场景和服务质量目标,应用提供商希望能以一种可伸缩的方式对云计算资源进行动态调整。基于虚拟机的资源管理较为重载,难以实现细粒度的资源动态调整与混合云中跨平台的服务快速迁移。容器技术在一定程度上弥补了虚拟机的不足,然而传统的资源管理方法在诸多方面并不十分适用于容器技术。针对这一问题,提出了基于容器技术的云计算资源自适应管理方法,设计了更适用于容器的资源架构方案与资源之间的调度方式。与传统的线性建模方法不同,所提方法使用非线性函数对云计算资源进行更加精确的建模,同时用遗传算法进行参数调优,使得自适应调整响应更快、总体性能更好。所提方法还针对不同容器多维度的异构性,合理分配容器部署位置,提高物理资源利用率。此外,所提方法结合了容器技术多方面的底层特性,在分配负载等方面进行适应性调整。最后通过实验分析初步确认了所提方法的有效性。
[Abstract]:With the development of cloud computing, more and more software applications choose cloud platform as deployment platform. Application providers want to dynamically adjust cloud computing resources in a scalable way. Resource management based on virtual machines is overloaded. It is difficult to realize the fine-grained dynamic adjustment of resources and the rapid migration of cross-platform services in the mixed cloud. Container technology makes up for the deficiency of virtual machine to some extent. However, the traditional resource management method is not suitable for container technology in many aspects. In order to solve this problem, an adaptive management method of cloud computing resources based on container technology is proposed. Different from the traditional linear modeling method, the proposed method uses nonlinear functions to model cloud computing resources more accurately. At the same time, the genetic algorithm is used to optimize the parameters, so that the adaptive adjustment response is faster and the overall performance is better. The proposed method also reasonably allocates the container deployment location to the multi-dimensional heterogeneity of different containers, and improves the utilization ratio of physical resources. The proposed method combines various underlying characteristics of container technology and adaptively adjusts load distribution. Finally, the effectiveness of the proposed method is preliminarily confirmed by experimental analysis.
【作者单位】：复旦大学软件学院;上海市数据科学重点实验室(复旦大学);
【基金】：国家高技术研究发展计划(863)(2015AA01A203)资助
【分类号】：TP393.09

【参考文献】