Virtual Machine Scheduling and Rescheduling Method Based on Network Awareness

Abstract

The present invention provides a virtual machine scheduling and rescheduling method based on network perception, including the virtual machine's initial scheduling and rescheduling. The initial scheduling first obtains the status of the current data center, judges whether there is an initial communication matrix, and chooses whether to perform initial clustering according to whether there is an initial communication matrix, and then selects the shortest subtree for the entire group of virtual machines or some virtual machines after clustering. The algorithm deploys it to a virtual machine. Then compare the current state of the data center with the historical state, and trigger rescheduling when the network communication matrix changes significantly. In the process of rescheduling, first calculate the priority of each group of virtual machines, first reschedule the virtual machines with higher priority, and try to move the virtual machines in this group to the physical machine in the same rack. If it fails, use the rack-oriented virtual machine The steps of machine greedy aggregation are rescheduled. The present invention improves the performance of tasks deployed on the virtual machine and the overall network communication efficiency of the data center through proper virtual machine migration.

Description

technical field [0001] The invention relates to a virtual machine scheduling method in a data center, in particular to a virtual machine scheduling and rescheduling method based on network perception. Background technique [0002] With the development of virtualization technology and the popularization of the concept of cloud computing, more and more users choose to complete their work tasks by renting computing resources from the data center instead of purchasing physical devices themselves. In order to improve resource utilization, how to efficiently allocate resources requested by users to each physical server is one of the important issues that a data center manager needs to solve. An efficient virtual machine scheduling strategy can improve the resource utilization of the data center and reduce energy consumption during operation. Existing scheduling algorithms comprehensively consider the CPU, memory and network requirements of virtual machines, and deploy virtual mac...

AI Technical Summary

Problems solved by technology

However, the current virtual machine scheduling strategy does not dynamically consider the changes in network communication traffic between virtual machines, nor does it consider the problem of resource utilization reduction and network delay increase caused by virtual machine exits.

Images