Virtual machine initial placement strategy method based on Rendezvous Hash algorithm

Abstract

The invention discloses a virtual machine initial placement strategy method based on Rendezvous Hash algorithm. The method comprises the following steps: step one, firstly defining a physical mainframe set PM existing in a data center, and a virtual machine set VM needing the initial placement; step two, defining parameter for a given virtual machine vmi; step four, defining that any virtual machine vmi is provided with a distribution weight score set Wi to correspond to each mainframe for any virtual machine vmi when the mainframe number n is less than 4; step five, establishing a virtual layered structure when the mainframe number n is greater than and equal to 4, wherein the virtual layered structure comprises two parts of virtual layered nodes and a real mainframe node cluster, the virtual layered nodes are distributed in a tree way, and the real mainframe node cluster is a set formed by multiple real mainframes and associated with the virtual leaf sub-nodes at the bottom of the virtual layered nodes. By comprehensively considering the load balancing, the virtual machine mainframe performance, the data center energy consumption and like factors, the reasonable scheduling of the physical resource is finally realized.

Description

technical field [0001] The invention relates to a virtual machine initial placement strategy method based on Rendezvous hash algorithm. Background technique [0002] Cloud computing is a new hot spot that has attracted the attention of the industry in recent years, and virtualization technology is an important means to realize cloud computing. Through virtualization technology, the cloud data center uses the virtual machine to share the physical machine resource pool, and uses the physical machine unit to carry the virtual machine and the virtual machine unit to carry the application service. The cloud data center has become a collection of high-performance computers with thousands of physical servers and network equipment. Due to the large amount of resources and strong heterogeneity, cloud data centers have more prominent requirements for on-demand services, dynamic and elastic resource management, and service quality. However, cloud data centers are now facing problems ...

AI Technical Summary

Problems solved by technology

[0007] In order to overcome the unbalanced load of the existing virtual machine placement strategy and the lack of comprehensive consideration of CPU and memory factors, the present invention proposes a virtual machine placement strategy method based on the Rendezvous hash algorithm, which is suitable for cloud computing environments. The initial placement of physical resource scheduling to the host, the establishment of the mapping relationship between the virtual machine and the host, the entire strategy comprehensively considers load balancing, virtual machine host performance, data center energy consumption and other factors, and finally realizes reasonable scheduling of physical resources

Images