MULTI-QUEUE AND MULTI-CLUSTER TASK SCHEDULING METHOD AND SYSTEM

摘要

A multi-queue and multi-cluster task scheduling method and system, which relate to the technical field of cloud computing. The method comprises: constructing a training data set, wherein the training data set comprises state spaces and action decisions that correspond to each other on a one-to-one basis, the state space comprises a plurality of task attribute groups in a plurality of queues arranged in sequence, and the task attribute group comprises a task data amount and the number of CPU cycles required for tasks (S1); training and optimizing a plurality of parallel deep neural networks by using the training data set to obtain a plurality of parallel trained and optimized deep neural networks (S2); setting a return function, wherein the return function minimizes the sum of the task delay and the energy consumption by means of adjusting the return value proportion of the task delay and the return value proportion of the energy consumption (S3); inputting a state space to be scheduled into the plurality of parallel trained and optimized deep neural networks to obtain a plurality of action decisions to be scheduled (S4); according to the return function, determining an optimal action decision from among the plurality of action decisions to be scheduled and outputting the optimal action decision (S5); and scheduling the plurality of task attribute groups to a plurality of clusters according to the optimal action decision (S6). According to the method, an optimal scheduling policy can be generated by taking minimization of task delay and energy consumption as an optimization objective of a cloud system.