IaaS(infrastructure as a service,基础设施即服务)模式云训练是一种以云计算为基础的新型装备模拟训练.云训练中,虚拟机的优化放置是提高资源利用效率、降低运行时资源调度工作量的基础.阐述了云训练的内涵,并对云训练中的虚拟机放置...IaaS(infrastructure as a service,基础设施即服务)模式云训练是一种以云计算为基础的新型装备模拟训练.云训练中,虚拟机的优化放置是提高资源利用效率、降低运行时资源调度工作量的基础.阐述了云训练的内涵,并对云训练中的虚拟机放置进行了数学描述.提出了一种改进的免疫克隆优化算法(MICOA),采用反向优化算法对初始抗体进行优化,通过变异概率与范围的自适应控制,保证算法演化初期抗体群的多样性与搜索空间的完备性,以及演化后期的局部寻优与最优解质量.引入抗体-抗体亲和度筛选最优抗体,保证抗体群的多样性.通过对虚拟机放置进行仿真实验,表明该方法可以有效提升资源利用率,实现系统综合优化目标.展开更多
提出了一种云数据中心基于数据依赖的虚拟机选择算法DDBS(data dependency based VM selection).参考Cloudsim项目中方法,将虚拟机迁移过程划分为虚拟机选择操作(VM selection)和虚拟机放置(VM placement)操作.DDBS在虚拟机选择过程中...提出了一种云数据中心基于数据依赖的虚拟机选择算法DDBS(data dependency based VM selection).参考Cloudsim项目中方法,将虚拟机迁移过程划分为虚拟机选择操作(VM selection)和虚拟机放置(VM placement)操作.DDBS在虚拟机选择过程中考虑虚拟机之间的数据依赖关系,把选择与迁移代价值比较小的虚拟机形成侯选虚拟机列表,配合后续的虚拟机放置策略最终完成虚拟机的迁移过程.以Cloudsim云计算模拟器中的虚拟机选择及放置策略作为性能比较对象.实验结果表明:DDBS与Cloudsim中已有能量感知的算法比较起来,在虚拟机迁移次数和能量消耗方面都比较少,可用性比较高.展开更多
Nowadays,virtual machine migration(VMM)is a trending research since it helps in balancing the load of the Cloud effectively.Several VMM-based strategies defined in the literature have considered various metrics,such a...Nowadays,virtual machine migration(VMM)is a trending research since it helps in balancing the load of the Cloud effectively.Several VMM-based strategies defined in the literature have considered various metrics,such as load,energy,and migration cost for balancing the load of the model.This paper introduces a novel VMM strategy by considering the load of the Cloud network.Two important aspects of the proposed scheme are the load prediction through the support vector regression(SVR)and the optimal VM placement through the proposed dragonfly-based crow(D-Crow)optimization algorithm.The proposed D-Crow optimization algorithm is developed by incorporating crow search algorithm(CSA)into dragonfly algorithm(DA).Also,the proposed VMM strategy defines a load balancing model based on the energy consumption,load,and the migration cost to achieve the energy-aware VMM.The simulation of the proposed VMM strategy is done based on the metrics such as load,energy consumption,and the migration cost.From the results,it can be shown that the proposed VMM strategy surpassed other comparative models by achieving the minimum values of 7.3719%,10.0368%,and 11.0639%for the load,energy consumption,and migration cost,respectively.展开更多
Cloud data centers, such as Amazon EC2, host myriad big data applications using Virtual Machines(VMs). As these applications are communication-intensive, optimizing network transfer between VMs is critical to the perf...Cloud data centers, such as Amazon EC2, host myriad big data applications using Virtual Machines(VMs). As these applications are communication-intensive, optimizing network transfer between VMs is critical to the performance of these applications and network utilization of data centers. Previous studies have addressed this issue by scheduling network flows with coflow semantics or optimizing VM placement with traffic considerations.However, coflow scheduling and VM placement have been conducted orthogonally. In fact, these two mechanisms are mutually dependent, and optimizing these two complementary degrees of freedom independently turns out to be suboptimal. In this paper, we present VirtCO, a practical framework that jointly schedules coflows and places VMs ahead of VM launch to optimize the overall performance of data center applications. We model the joint coflow scheduling and VM placement optimization problem, and propose effective heuristics for solving it. We further implement VirtCO with OpenStack and deploy it in a testbed environment. Extensive evaluation of real-world traces shows that compared with state-of-the-art solutions, VirtCO greatly reduces the average coflow completion time by up to 36.5%. This new framework is also compatible with and readily deployable within existing data center architectures.展开更多
The drastic increase in engineering system complexity has spurred the development of highly efficient optimization techniques.Many real-world optimization problems have been identified as bilevel/multilevel as well as...The drastic increase in engineering system complexity has spurred the development of highly efficient optimization techniques.Many real-world optimization problems have been identified as bilevel/multilevel as well as multiobjective.The primary aim of this work is to present a framework to tackle the bilevel virtual machine(VM)placement problem in cloud systems.This is done using the coupled map lattice(CML)approach in conjunction with the Stackelberg game theory and weighted-sum frameworks.The VM placement problem was modified from the original multiobjective(MO)problem to an MO bilevel formulation to make it more realistic albeit more complicated.Additionally comparative analysis on the performance of the CML approach was carried out against the particle swarm optimization method.A new bilevel metric called the cascaded hypervolume indicator is introduced and applied to measure the dominance of the solutions produced by both methods.Detailed analysis on the computational results is presented.展开更多
文摘IaaS(infrastructure as a service,基础设施即服务)模式云训练是一种以云计算为基础的新型装备模拟训练.云训练中,虚拟机的优化放置是提高资源利用效率、降低运行时资源调度工作量的基础.阐述了云训练的内涵,并对云训练中的虚拟机放置进行了数学描述.提出了一种改进的免疫克隆优化算法(MICOA),采用反向优化算法对初始抗体进行优化,通过变异概率与范围的自适应控制,保证算法演化初期抗体群的多样性与搜索空间的完备性,以及演化后期的局部寻优与最优解质量.引入抗体-抗体亲和度筛选最优抗体,保证抗体群的多样性.通过对虚拟机放置进行仿真实验,表明该方法可以有效提升资源利用率,实现系统综合优化目标.
文摘提出了一种云数据中心基于数据依赖的虚拟机选择算法DDBS(data dependency based VM selection).参考Cloudsim项目中方法,将虚拟机迁移过程划分为虚拟机选择操作(VM selection)和虚拟机放置(VM placement)操作.DDBS在虚拟机选择过程中考虑虚拟机之间的数据依赖关系,把选择与迁移代价值比较小的虚拟机形成侯选虚拟机列表,配合后续的虚拟机放置策略最终完成虚拟机的迁移过程.以Cloudsim云计算模拟器中的虚拟机选择及放置策略作为性能比较对象.实验结果表明:DDBS与Cloudsim中已有能量感知的算法比较起来,在虚拟机迁移次数和能量消耗方面都比较少,可用性比较高.
文摘Nowadays,virtual machine migration(VMM)is a trending research since it helps in balancing the load of the Cloud effectively.Several VMM-based strategies defined in the literature have considered various metrics,such as load,energy,and migration cost for balancing the load of the model.This paper introduces a novel VMM strategy by considering the load of the Cloud network.Two important aspects of the proposed scheme are the load prediction through the support vector regression(SVR)and the optimal VM placement through the proposed dragonfly-based crow(D-Crow)optimization algorithm.The proposed D-Crow optimization algorithm is developed by incorporating crow search algorithm(CSA)into dragonfly algorithm(DA).Also,the proposed VMM strategy defines a load balancing model based on the energy consumption,load,and the migration cost to achieve the energy-aware VMM.The simulation of the proposed VMM strategy is done based on the metrics such as load,energy consumption,and the migration cost.From the results,it can be shown that the proposed VMM strategy surpassed other comparative models by achieving the minimum values of 7.3719%,10.0368%,and 11.0639%for the load,energy consumption,and migration cost,respectively.
基金supported by the National Key R&D Program of China(No.2017YFB1003000)the National Natural Science Foundation of China(Nos.61572129,61602112,61502097,61702096,61320106007,and 61632008)+4 种基金the International S&T Cooperation Program of China(No.2015DFA10490)the National Science Foundation of Jiangsu Province(Nos.BK20160695 and BK20170689)the Jiangsu Provincial Key Laboratory of Network and Information Security(No.BM2003201)the Key Laboratory of Computer Network and InformationIntegration of Ministry of Education of China(No.93K-9)supported by the Collaborative Innovation Center of Novel Software Technology and Industrialization and Collaborative Innovation Center of Wireless Communications Technology
文摘Cloud data centers, such as Amazon EC2, host myriad big data applications using Virtual Machines(VMs). As these applications are communication-intensive, optimizing network transfer between VMs is critical to the performance of these applications and network utilization of data centers. Previous studies have addressed this issue by scheduling network flows with coflow semantics or optimizing VM placement with traffic considerations.However, coflow scheduling and VM placement have been conducted orthogonally. In fact, these two mechanisms are mutually dependent, and optimizing these two complementary degrees of freedom independently turns out to be suboptimal. In this paper, we present VirtCO, a practical framework that jointly schedules coflows and places VMs ahead of VM launch to optimize the overall performance of data center applications. We model the joint coflow scheduling and VM placement optimization problem, and propose effective heuristics for solving it. We further implement VirtCO with OpenStack and deploy it in a testbed environment. Extensive evaluation of real-world traces shows that compared with state-of-the-art solutions, VirtCO greatly reduces the average coflow completion time by up to 36.5%. This new framework is also compatible with and readily deployable within existing data center architectures.
文摘The drastic increase in engineering system complexity has spurred the development of highly efficient optimization techniques.Many real-world optimization problems have been identified as bilevel/multilevel as well as multiobjective.The primary aim of this work is to present a framework to tackle the bilevel virtual machine(VM)placement problem in cloud systems.This is done using the coupled map lattice(CML)approach in conjunction with the Stackelberg game theory and weighted-sum frameworks.The VM placement problem was modified from the original multiobjective(MO)problem to an MO bilevel formulation to make it more realistic albeit more complicated.Additionally comparative analysis on the performance of the CML approach was carried out against the particle swarm optimization method.A new bilevel metric called the cascaded hypervolume indicator is introduced and applied to measure the dominance of the solutions produced by both methods.Detailed analysis on the computational results is presented.
