本文選題：Web服務(wù)器集群 + 負(fù)載均衡��； 參考：《計(jì)算機(jī)工程與應(yīng)用》2017年10期

【摘要】：大數(shù)據(jù)時(shí)代的快速發(fā)展和大數(shù)據(jù)戰(zhàn)略的明確提出,使得Web服務(wù)器集群將面臨更加復(fù)雜和嚴(yán)峻的負(fù)載挑戰(zhàn)。傳統(tǒng)的負(fù)載均衡算法存在著明顯的局限性。提出了一種基于強(qiáng)掛起弱預(yù)測機(jī)制的負(fù)載均衡模型,該模型利用強(qiáng)掛起機(jī)制和基于層次分析的三次指數(shù)平滑預(yù)測算法進(jìn)行負(fù)載均衡動態(tài)調(diào)度。實(shí)驗(yàn)結(jié)果表明該模型在系統(tǒng)瞬時(shí)性能異常、高并發(fā)和重負(fù)載交互情況下的負(fù)載均衡效果優(yōu)于傳統(tǒng)負(fù)載均衡算法。
[Abstract]:With the rapid development of big data era and the clear proposal of big data strategy, Web server cluster will face more complex and severe load challenges. The traditional load balancing algorithm has obvious limitations. A load balancing model based on strong hang and weak prediction mechanism is proposed. The model uses strong hang mechanism and cubic exponential smoothing prediction algorithm based on hierarchical analysis to carry out dynamic load balancing scheduling. The experimental results show that the load balancing effect of this model is better than that of the traditional load balancing algorithm under the condition of abnormal transient performance, high concurrency and heavy load interaction.
【作者單位】： 中國傳媒大學(xué);安防大數(shù)據(jù)處理與應(yīng)用北京市重點(diǎn)實(shí)驗(yàn)室;
【基金】：國家自然科學(xué)基金項(xiàng)目(No.6137006)
【分類號】：TP301.6;TP393.09
，

本文編號：1897739