Queue-based multi-target scheduling strategy for heterogeneous resources

Abstract

The present invention relates to a strategy for scheduling resources at a heterogeneous cloud data center, wherein the strategy is provided based on the queuing theory to fully consider the heterogeneity of resources and improve the utilization rate of resources. In this way, the service quality (QoS) of tasks and the load balancing of the cloud data center are ensured. Meanwhile, the energy consumption of the cloud data center is greatly reduced. The strategy comprises the following steps of 1, establishing a two-stage scheduling framework for the data center by means of a queuing model, and dividing the execution of all tasks into two stages, namely task allocating and task scheduling; 2, during the task allocating period, considering the heterogeneity of resources, and evenly allocating an acquired classification task set W(t) to all servers (PW) by means of the HPAC algorithm; 3, during the task scheduling period, considering the resource utilization rate and the service quality (QoS), establishing a virtual machine based on tasks queued in all the PMs by means of the MIUS algorithm, and executing the tasks.

Description

technical field [0001] The invention belongs to the field of cloud computing IaaS, and in particular relates to a resource scheduling strategy in a heterogeneous cloud data center. Background technique [0002] The resource scheduling of cloud computing IaaS layer is the key technology to realize cloud computing applications. The cloud data center is composed of a large number of heterogeneous servers (PM), and heterogeneity is common, including resource heterogeneity, task heterogeneity and virtual machine heterogeneity. Heterogeneity will directly affect the performance of the scheduling strategy, thereby affecting the service quality of the entire cloud data center. Therefore, research and scheduling for heterogeneity can make more rational use of resources and ensure the interests of cloud providers. [0003] Multi-objective scheduling is to achieve multiple target values ​​in a resource scheduling strategy at the same time, and use the idea of ​​multi-objective optimiz...

AI Technical Summary

Problems solved by technology

However, this method is severely restricted by the setting of weights and the given order of goals. At the same time, the public functions and constraints may be non-differentiable or non-linear, which also increases the difficulty of the linear weighted solution.

[0004] The current multi-objective scheduling strategies mainly focus on the research of homogeneous resources. Although some scholars have begun to study multi-objective scheduling under heterogeneous resources in recent years, this research is still in its infancy, and the proposed related strategies have resource utilization problems. Low efficiency, high energy consumption and other issues

Images