Under the support of cloud computing,more and more applications choose cloud platform as their deployment platform.In order to respond to changing workloads,application scenarios,and user target,service providers need to dynamically adjust the existing resources in a scalable way.However,cloud platform resource management approach via virtual machine has many drawbacks.Virtual machines are heavyweight,it is not suitable for fine-grained and flexible allocation of resources.Meanwhile in the hybrid cloud background,the virtual machines can not migrate quickly.Container technology,to some extent,can make up for the lack of a virtual machine.However,traditional resource management method based on the virtual machine is not very suitable for container technology.To solve the above problems,this approach designed resources infrastructure solution and scheduling way which are more suitable to container.This approach models cloud resources precisely via nonlinear function,and could tune parameters using genetic algorithms,in order to optimize the performance of adjustment.This approach also computed rational allocation of container deployment location to improve physical resource utilization.In addition to that,this approach combines many bottom characteristics of container to adjust the procedure of load-balancing.We also conducted experiments to verify the effectiveness of the method.%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.
展开▼
