摘要
针对LVS集群的加权最小连接数算法在反映实际负载和具体性能方面的缺陷,提出一种基于负载预测的自适应权值负载均衡算法。采用一种自适应AR算法计算负载预测值,使服务器权值能够根据预测结果做出调整,进一步提升负载均衡效果,更加有效地利用系统资源。实验采用WAS工具进行仿真,并对算法效果进行对比,实验结果表明,该算法能够更好地在LVS集群系统中实现负载均衡。
Aiming at the problem that the weighted least connection algorithm can not accurately reflect the actual load and processing capacity of the server,a load balancing algorithm based on load prediction and adaptive weight was proposed.An adaptive AR algorithm was used to calculate the load prediction value,so that the server weights were adjusted according to the prediction results,further improving the load balancing effect and making the most of the system resources.The experiment was simulated with WAS tool,and the effects of the algorithm were compared.Experimental results show that the proposed algorithm can achieve better load balancing effect in the LVS cluster system.
引文
[1]Zhang Y,Han Y,Sun G.Research on the construction of high available load balancing cluster based on LVS[C]//Workshop on Advanced Research&Technology in Industry Applications,2016.
[2]WANG Chunjuan.Research and analysis of LVS cluster algorithm[J].Computer Knowledge and Technology,2013,9(26):5963-5964(in Chinese).[王春娟.LVS集群算法研究与分析[J].电脑知识与技术,2013,9(26):5963-5964.]
[3]Sun D,Zhu Q.Comparison and improvement of load balance scheduling algorithm based on cluster technology[C]//International Conference on Electrical and Electronics Engineering,2014:67-74.
[4]Wu Y,Luo S,Li Q.An adaptive weighted least-load balancing algorithm based on server cluster[C]//International Conference on Intelligent Human-Machine Systems and Cybernetics,2013:224-227.
[5]Kanakala R,Reddy VK.Performance analysis of load balancing techniques in cloud computing environment[J].Telkomnika Indonesian Journal of Electrical Engineering,2015,9(18):1-6.
[6]Ren X,Lin R,Zou H.A dynamic load balancing strategy for cloud computing platform based on exponential smoothing forecast[C]//IEEE International Conference on Cloud Computing and Intelligence Systems.IEEE,2011:220-224.
[7]Hu Y,Zhu S.Load-balancing cluster based on Linux Virtual Server for internet-based laboratory[C]//Industrial Electronics and Applications.IEEE,2014:2181-2185.
[8]Bosque JL,Toharia P,Robles OD,et al.A load index and load balancing algorithm for heterogeneous clusters[J].Journal of Supercomputing,2013,65(3):1104-1113.
[9]Liu C,Hoi SCH,Zhao P,et al.Online ARIMA algorithms for time series prediction[C]//30th AAAI Conference on Artificial Intelligence.AAAI Press,2016:1867-1873.
[10]Hua ZQ,Xue DM.ARIMA based time series forecasting model[J].Recent Advances in Electrical&Electronic Engineering,2016,9(2):93-98.
[11]ZHANG Zonghua,ZHANG Haiquan,WEI Chi,et al.LVS cluster load balancing algorithm with adaptive weight leastload[J].Communications Technology,2016,24(3):248-251(in Chinese).[张宗华,张海全,魏驰,等.基于加权改进的AR模型的负载预测研究[J].计算机测量与控制,2016,24(3):248-251.]
[12]ZHAO Ming.Research on network bandwidth allocation algorithm based on AR(n)model[J].Microcomputer Applications,2017,33(6):61-63(in Chinese).[赵明.基于AR(n)模型的网络带宽流量动态分配预测算法研究[J].微型电脑应用,2017,33(6):61-63.]
[13]HUAN Qiuyun,QIU Xiaohui,LIU Xiaofei.Variable step LMS algorithm using norm of the hyperbolic tangent function[J].Journal of Signal Processing,2014,30(1):93-99(in Chinese).[还秋云,邱晓晖,刘晓飞.引用范数的双曲正切函数变步长LMS算法[J].信号处理,2014,30(1):93-99.]
[14]QIN Xia.The design and implementation on LVS load balancing cluster[D].Xi’an:Xidian University,2014(in Chinese).[秦霞.基于LVS负载均衡集群的设计和实现[D].西安:西安电子科技大学,2014.]
[15]YANG Ting,WAN Liang,MA Shaoju,et al.LVS cluster load balancing algorithm with adaptive weight leastload[J].Communications Technology,2017,50(4):741-745(in Chinese).[杨婷,万良,马绍菊,等.一种自适应权值最小负载的LVS集群负载均衡算法[J].通信技术,2017,50(4):741-745.]