Cloud computing dynamic resource scheduling system and method

Abstract

The invention relates to the dynamic resource scheduling technique in the cloud computing field, and provides a cloud computing dynamic resource scheduling method based on the feedback and a prediction mechanism. The method aims to overcome the defects of the cloud computing resource distribution and scheduling technique in the prior art, and can achieve balance use of various computer resources in a cloud computing environment, obtain satisfactory load balance under small pay expenses and improve comprehensive efficiency of system scheduling. According to the scheme, various performance indexes of virtual machines are monitored in real time in operation processes of the virtual machines, operation circumstances of all the current virtual machines within a next short period of time are predicted according to the monitored various current performance indexes of the virtual machines and a virtual machine state prediction model based on state feedback when a task request comes, and the most adaptive virtual machine is selected by combining a prediction result and a required task type and the required task is distributed. In addition, the invention further discloses a corresponding cloud computing dynamic resource scheduling system applicable to dynamic resource scheduling.

Description

technical field [0001] The invention relates to a dynamic resource scheduling technology in the field of cloud computing, in particular to a cloud computing dynamic resource scheduling system based on a feedback and prediction mechanism and a corresponding scheduling method. Background technique [0002] In recent years, with the rapid development of information technology and the increasing scale of the Internet, the amount of business and data handled by the Internet is also growing rapidly. In order to effectively process these massive data and services and optimize the user experience of using Internet services, cloud computing technology has emerged as the times require. Cloud computing provides users with convenient, fast and secure data storage and network services through computer and network technologies such as distributed computing technology, parallel computing technology, virtualization technology and load balancing, bringing new developments to the development ...

AI Technical Summary

Problems solved by technology

Moreover, in the existing cloud computing resource scheduling technology, the one-sidedness of the existing scheduling strategy may lead to the occurrence of blind transfer of physical resources, which needs to be avoided as much as possible

In addition, in the process of task execution, since the execution of the task will change the resource consumption type of the task, the type of the task will also change accordingly, which makes the static resource allocation and scheduling often cause insufficient or wasteful resources , and manual dynamic resource adjustment has obvious lag, so it is necessary to monitor the execution status of the task in real time, and intelligently judge the resource consumption type of the task through the resource consumption type of the task. If the resource consumption type of the task is different from the initial If the task type is inconsistent, and will not be consistent for a long enough time, and this inconsistency seriously affects the health value of the virtual machine, it is necessary to dynamically reschedule the task at an appropriate time to achieve a balanced use of each resources (CPU, memory, disk, I / O, network), avoid single resource bottlenecks, increase the density of virtual machines, eliminate hot spots, and improve business processing capabilities, but the existing resource scheduling schemes do not take this into account

Images