Optimal Scheduling Model of Cloud Storage Resources Based on Double Threshold Load Balancing Control

International Conference on Manufacturing Technology, Materials and Chemical Engineering

机械制造;材料科学;化学工业

Zhan, Liuchun^1 ; Huang, Changjiang^2 ; Lin, Mei^3

Huali College Guangdong University of Technology, Guangzhou, Guangdong, China^1

Guangzhou University Sontan College, Guangzhou, Guangdong, China^2

Nanfang Vocational College, Jiangmen, Guangdong, China^3

关键词: Cloud resource allocations;    Cloud storage systems;    Load-balancing control;    Nonlinear time series;    Piecewise interpolation;    Resource optimal allocation;    Resource optimization;    Resource scheduling design;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/392/6/062205/pdf
DOI  :  10.1088/1757-899X/392/6/062205

学科分类：材料科学（综合）

来源: IOP

【 摘 要 】

Effective resource scheduling design in cloud storage system can integrate cloud computing central resources and improve resource optimal allocation ability. An optimal scheduling method for cloud storage resources is proposed based on two-threshold load balancing control. The spatial distribution of storage resources is optimized by using grid region collocation method, and the model of cloud storage resource distribution information flow is constructed by nonlinear time series reorganization method, and the auto-correlation feature matching of cloud storage resource information flow is carried out. The association rule feature quantity of cloud storage resource is mined, the phase space of cloud storage system is reconstructed according to the distribution attribute of association rule feature, the priority attribute list of resource scheduling is constructed, and segment interpolation method is adopted to allocate cloud resource. The threshold value of resource scheduling is determined by adaptive optimization of cloud resource allocation threshold using piecewise interpolation method, and load balancing control of cloud resource allocation is carried out by double threshold control method. The partition scheduling of cloud storage resources is implemented according to the priority list. The simulation results show that the proposed method has good load balance, high precision and high throughput, and it has good application value in resource optimization.