網頁伺服器最佳客戶端連線數量之研究

近年來，隨著環保意識的抬頭，節能減碳的觀念逐漸受到重視，而電腦機房的能源消耗問題也成為重要的課題。因此，本篇論文在此提出取得最佳客戶端連線數量的架構設計，其內容包括CPU使用率與電流關係、連線數量與CPU使用率關係以及連線數量與反應時間關係，並採用ApacheBench進行壓力測試實驗，配合本論文所推導出的最佳連線數量定義式，找出網頁伺服器建構於不同硬體設備與系統架構上的最佳客戶端連線數量，作為啟動實體機器或虛擬系統的依據，使能源使用效率與可用性達到平衡。除此之外，針對電腦機房空調的能源使用效率，本論文提出時段空調能源使用效率(SCPUE)，並依照其定義結合論文中的實驗結果以及客戶端連線需求預測模組，推導出電腦機房空調電流之數學模型。提出電腦機房空調電流之數學模型的目的，是為了實現該數學模型於現實環境中的電腦機房(數據中心)，依據數學模型的計算結果動態調整電腦機房(數據中心)之空調設備的電流量，提升電腦機房(數據中心)空調之能源使用效率。

關鍵字

實體機器；最佳客戶端連線數量；電腦機房；虛擬系統

並列摘要

Recently, the concept of energy efficiency and carbon reduction progressively gets more attention. And the energy inefficiency problem of computer room is an important issue. For this reason, this paper provides a framework design to get the optimal connection number of a web server. This includes getting the relation between CPU utilization and current, the relation between connection number and CPU utilization, and the relation between connection number and response time. And using ApacheBench and optimal connection number analysis derives from this paper to get the optimal connection number of a web server in deferent hardware equipment or system architecture. The optimal connection number as the condition for turn on physical machine or virtual machine. Eventually finds the balance between energy efficiency and availability. Otherwise, this paper also provides Scheduled Cooling Power Usage Effectiveness (SCPUE). And according to the experiment result and forecasting models to get the math model of air-conditioner current in computer room. In the future, we can use the model to control the air-conditioner current to increase the energy efficiency of computer room.

並列關鍵字

Physical Machine ； Optimal Connection Number ； Computer Room ； Virtual Machine

參考文獻

[2]. Jim Gray, and Daniel P. Siewiorek,“High-availability computer systems,” Computer, September 1991.

[3]. Dara Kusic, Jeffrey O. Kephart, James E. Hanson, Nagarajan Kandasamy, and Guofei Jiang,“Power and Performance Management of Virtualized Computing Environments Via Lookahead Control,”Autonomic Computing, 2008. ICAC '08, pp. 3-12, June 2008.

[4]. Vinicius Petrucci, Orlando Loques, and Daniel Mossé,“Dynamic configuration support for power-aware virtualized server clusters,”WiP Session of the 21th Euromicro Conference on Real-Time Systems, 2009.

[5]. Takayuki Imada, Mitsuhisa Sato, Yoshihiko Hotta, and Hideaki Kimura,“Power management of distributed web savers by controlling server power state and traffic prediction for QoS,”Parallel and Distributed Processing, 2008. IPDPS 2008, pp. 1-8, April 2008.

[7]. Apache, http://httpd.apache.org/

國際替代計量

網頁伺服器最佳客戶端連線數量之研究

未授權

主題瀏覽