Full-stage load sharing comprehensive optimization method for cloud computing management platform

Abstract

The invention provides a full-stage load sharing comprehensive optimization method of a cloud computing management platform. All resource utilization conditions of the platform virtual machine are integrated, the full-stage resource utilization efficiency is improved, and compared with the prior art, two obvious progresses exist: firstly, when the virtual machine is created, an original algorithmonly selects a physical host according to the size of a memory, yet the algorithm of the invention comprehensively considers a plurality of factors such as a central processing unit, the memory and ahard disk, so that the result is more accurate; and 2, the original algorithm only plays a role in the deployment stage of the virtual machine, yet the algorithm of the invention can play a role in any stage, and automatic load adjustment can be carried out according to actual conditions during system operation. Through experimental argumentation and comparison, after the method provided by the invention is used, the load difference among the central processing unit, the memory and the hard disk among the physical hosts is obviously reduced, and the overall performance of the cloud computing management platform is obviously improved when the load is relatively large.

Description

technical field [0001] The invention relates to a comprehensive optimization method for full-stage load sharing, in particular to a comprehensive optimization method for full-stage load sharing of a cloud computing management platform, and belongs to the technical field of cloud platform load optimization. Background technique [0002] In recent years, the cloud computing industry has developed rapidly, and cloud computing-related technologies have become more and more advanced. Under the traditional model, if an enterprise wants to build an information technology platform, it needs to purchase infrastructure such as hardware and software by itself. This method is too costly and is not conducive to the expansion of the platform. The emergence of cloud computing allows service providers to configure infrastructure and build a cloud computing management platform, and users only need to purchase corresponding services from service providers. The cloud computing management plat...

AI Technical Summary

Problems solved by technology

On the whole, the load sharing method of the cloud computing management platform in the prior art mainly has the following defects: First, when receiving a request to create a virtual machine instance, the load balancing module of the cloud computing management platform only selects a remaining virtual machine instance according to the size of the memory The physical host with the largest memory is used to create a virtual machine instance, but the resource utilization of the cloud platform is closely related to many indicators, in addition to the memory size, there are also the number of CPUs, disk capacity, network bandwidth, etc. The cloud computing management platform only It is obviously unreasonable to use memory as the only consideration index. The physical host selected in this way may meet the requirements for creating virtual machine instances, but it is difficult to be the most suitable host. The cost of creating a new virtual machine instance on an inappropriate physical host The sum of various resources far exceeds the sum of the resources of creating a virtual machine instance on a suitable physical host. In some cases, the methods of the prior art not only fail to achieve the effect of load optimization, but may have counterproductive effects; The load balancing strategy of a technical cloud computing management platform only works in the deployment phase, that is, when Nova receives a virtual machine creation request, the Scheduler module will play a role and select a suitable physical host to create a virtual machine instance, but if a certain The physical host is overloaded due to the network environment or other factors, and the scheduler cannot move the virtual machine instance to reduce the load. This requires the cloud computing management platform to schedule virtual machines based on dynamic data, and the monitor needs to be used to monitor the resource utilization of each physical host in real time state, also needs the support of dynamic migration technology, which can better improve resource utilization, but the existing technology has no related improvement design; third, the complexity of the algorithm of the partial load sharing optimization method in the existing technology is relatively high, and there is no connection Real platform resource utilization is limited by the high computational complexity and cannot be extended to practical applications. It has almost no practical value, and in many cases it will interfere with the normal work of the platform.

Images