- 學位論文
基於權重式分配之可擴充性網站叢集負載平衡機制
Load Balancing Scheme based on Weighted Distribution for Scalable Web Server Clusters
摘要
由於網路的蓬勃發展，網路人口爆炸性的增加，熱門網站提供的服務趨於多元化。像是一些各大入口網站、搜尋引擎、訂票購物網站，他們所會面臨的問題就是當大量用戶在同時間訪問網站，造成網站過大流量，使後端的網頁伺服器無法負荷如此龐大的需求。對於管理者而言，我們所要思考的是該如何分散伺服器負載以及流量並提升反應時間，使得網路服務更優異，因此為了解決這些問題，大部份伺服器均使用叢集式的網路架構，利用多台網頁伺服器來分散其負載及網路流量。所以使用正確的負載平衡演算法，平衡負載及流量以及反應時間即顯得更重要。在此論文中，我們提出了基於HTTP Redirection架構下的權重式排程法來達成降低反應時間的目標，而實驗結果顯示我們的方法確實能使叢集中的網頁伺服器負載分配得較為平均。
並列摘要
Due to the explosive growth of Internet, the population of Internet increases rapidly. The most common problem faced by the Internet web service provider is to resolve the congestion problem caused by the large number of on-line users at a very short time interval, so that the backend webserver can’t overloaded with such a huge work. For a system administrator, the most often encountered problem is how to make web services respond quickly and provide better quality of services for the users. Therefore, in order to solve these problems, our research toward use web cluster architecture, web server cluster uses many web servers to share the web loading. In addition, we must use the correct load balancing algorithm to reduce the workload and fasten response time. In this thesis , we proposed load balancing scheme based on weighted distribution for scalable web server clusters to balance the workload and reduce the response time. Experiment result shows our proposed scheme can really make the backend web server load more evenly in the cluster.