Abstract: Current cloud computing providers allocate their Virtual Machine（VM） instances via fixed price-based or auction-
like mechanisms. However, most of these algorithms require static supply virtual machine, and are unable to accurately predict the user demand, lead to underutilization of resources. To this end, an auction-based algorithm for dynamic VM provisioning and allocation is proposed that takes into account the user demand for VMs when making VM provisioning decisions in this paper. It treats the set of available computing resource as “liquid” resources that can be configured into different numbers and types of VM instances depending on the requests of the users, and the proposed algorithm determines the allocation strategy based on the users’ valuations until all resources are allocated. The mechanism is evaluated by performing simulation experiments using traces of real workload from Parallel Workload Archive, the results show that the proposed method can guarantee to bring the higher income for cloud providers, and improve the resource utilization rate.

Key words: cloud computing, virtual machine instances, auction, allocation, cloud providers, resource utilization rate

摘要： 当前云计算供应商通过定价算法或类似拍卖的算法来分配他们的虚拟机（VM）实例。然而，这些算法大多要求虚拟机静态供应，无法准确预测用户需求，导致资源未得到充分利用。为此，提出了一种基于组合拍卖的虚拟机动态供应和分配算法，在做出虚拟机供应决策时考虑用户对虚拟机的需求。该算法将可用的计算资源看成是“流体”资源，且这些资源根据用户请求可分为不同数量、不同类型的虚拟机实例。然后可根据用户的估价决定分配策略，直到所有资源分配完毕。基于Parallel Workload Archive（并行工作负载存档）的真实工作负载数据进行了仿真实验，结果表明该方法可保证为云供应商带来更高收入，提高资源利用率。

关键词: 云计算, 虚拟机实例, 拍卖, 分配, 云供应商, 资源利用率