云环境下网络感知的虚拟机分配问题的求解方法

Network-aware virtual machine assignment method in cloud

提出了一种云环境下网络感知的虚拟机分配问题的求解方法,该方法能求解带有任务优先级和强制任务的最大虚拟机分配问题等5类问题。该方法通过将虚拟机分配问题编码成对应的SAT类问题,并调用现有的SAT类求解器进行求解,可以更有效地解决较大规模的虚拟机分配问题。最后,通过实验验证了本文算法的合理性和有效性。通过与现有算法在以上5类问题中进行对比,表明了本文算法具有更高的求解效率和更大的求解规模。

SAT-based methods for network-aware virtual machine assignment problem in cloud environment are proposed in this paper. These methods can solve five kinds of virtual machine problems including that with task priorities or with mandatory tasks. Each of the five kinds of virtual machine assignment problems is converted into a corresponding SAT-like problem and is solved by the existing SAT-like solver. It could solve the larger scale of the five kinds of virtual machine assignment problems more efficiently. Experimentally,the proposed methods are demonstrated to solve the five kinds of virtual machine assignment problems effectively. In contrast to the existing algorithms in the five scenarios mentioned above,the proposed methods can solve a larger scale of data with higher efficiency.