Automatic resource distribution method in cloud calculation environment

Abstract

The invention relates to an automatic resource distribution method in a cloud calculation environment. The method comprises the following steps: A. a "resource pool" in the cloud calculation environment is defined and managed; B. a virtual machine is pre-established; C. when a resource application is put forward by a user, resources in the resource pool is automatically matched by a management system, and the resources are listed to an optional queue if types and configuration are both matched; if the number of the optional queue surpasses a requirement of the user, the resources are selected from the optional queue to be distributed to the user according to a presorting dynamic resource distribution algorithm; D. when the pre-established virtual machine cannot meet the requirement of the user, insufficient number of the virtual machines are established after the resource application of the user is submitted. E. the resources are automatically applied and returned, and user data are confidentially stored or selectively deleted according to the requirement; and F. statics and analysis are performed on the requirement of the user, and a resource distribution report form is automatically generated. The process of the method is rapid and safe and automation is realized so that complexity and management cost of resource distribution in the cloud calculation environment are reduced, working efficiency is enhanced and resource utilization rate is also scientifically enhanced.

Description

technical field [0001] The invention relates to a resource allocation method, in particular to an automatic resource allocation method in a cloud computing environment. Background technique [0002] In the cloud computing environment, resources (such as virtual machines, servers, etc.) are huge in scale and have many customers, and most of them are virtual machine resources, which need to be created before they can be allocated. If the system creates virtual machines for allocation after receiving user applications, it will inevitably cause customers to wait, affecting system efficiency and customer satisfaction; but if a large number of virtual machines are pre-created in the system waiting for user resource applications, not only will the system’s own disk and other resources At the same time, due to the ever-changing configuration of virtual machines in user needs (for example, the configuration of virtual machine disk space may be 1G, 2G, 3G, 4G, 5G, 8G, 10G, 16G... the ...

AI Technical Summary

Problems solved by technology

[0004] a. Pre-create virtual machines with limited types of "high-demand" configurations or pre-create virtual machines with all configuration types: according to experience or individual user needs, pre-create virtual machines with certain "high-demand" configurations, although it is possible It saves system resource consumption to a certain extent and meets the resource application needs of some users to a certain extent, but it is not a perfect solution. The "high demand" configuration analyzed by experience or individual user needs may not be true for all The configuration that users demand the most, in this way, the created virtual machine may not be allocated for a long time, thereby wasting system resources, and at the same time prolonging the time for users to wait for resource allocation, reducing allocation efficiency

[0005] Pre-creating virtual machines of all configuration types obviously cannot meet user needs

The number of user requirements for virtual machines with different configurations is not the same, which not only causes low-frequency application for configuration of virtual machines to waste system resources, but at the same time, most user needs will not be met, prolonging the time for users to wait for resource allocation and reducing allocation efficiency

[0006] b. Allocate low-performance resources to users and reduce user satisfaction. If things go on like this, it will affect the benefits of cloud computing service providers: automatically match resources that meet the needs according to user needs, and allocate them when they are found. Although the allocation speed is improved, but at the same time Sacrificing customer satisfaction is obviously not conducive to the long-term development of cloud computing service providers

[0007] c. Consider the resource configuration in user demand in isolation, ignoring factors such as the relationship between all resource applications of the user and index performance: Although this method can improve the allocation speed, for example, if the user applies for a server and a database, if the search is performed according to the configuration , find any two servers and databases that do not have a relationship to the user, which will inevitably affect the data access speed during the user's use

Images