Prediction association iterative allocation method and system for cloud computing task resources

Abstract

The invention provides a prediction association iterative allocation method for cloud computing task resources. The prediction association iterative allocation method comprises the following steps: step 1, according to resources allocated in a resource allocation mode, executing subtasks according to the plurality of subtasks and a serial and / or parallel structural relationship between the subtasks; step 2, dynamically optimizing and allocating resources through the task execution state and the resource consumption state in the subtask execution process, and controlling the execution process of the subtasks so as to avoid triggering task scene evaluation scores of an event mechanism and an optimization evaluation mechanism; and step 3, carrying out overall prediction of the subtasks in stages in the subtask execution process to realize overall optimization of resource allocation. According to the method, resources can be prevented from being allocated again, overall optimal matching can be maintained, and the execution process of subtasks in each stage and optimization of resource allocation in subtasks are guaranteed.

Description

technical field [0001] The present invention relates to the technical field of resource allocation, in particular to a method and system for predictive association iterative allocation of cloud computing task resources. Background technique [0002] In order to meet the actual needs of analytical computing such as big data and artificial intelligence, cloud computing technology has developed rapidly in recent years. Cloud computing is to use the server cluster located in the cloud as the execution carrier of the analysis operation. The server cluster obtains the analysis operation task submitted by the client through the network, and executes the calculation in response to the demand of the analysis operation task, so as to feedback the analysis operation task to the client. Compute the execution result. [0003] Cloud computing takes advantage of the scale advantages and centralized management advantages of cloud server clusters, can undertake massive, high-concurrency and...

AI Technical Summary

Problems solved by technology

Moreover, due to the influence of various factors and the characteristics of the task itself, there are obvious dynamic uncertainties in the process of task execution. Therefore, the optimal matching relationship between tasks and resources is unstable, and it will often be broken after a period of operation. Causes the need to allocate resources frequently, but resource allocation itself is a kind of resource consumption, if it is too repeated and too frequent, it will obviously affect the execution of cloud computing

Images