臺灣博碩士論文加值系統

隨著網際網路的發展，以及全球資訊網的建立，網際網路的使用者可以不出家門便知天下事。由於硬體設備的成長遠低於網路使用者的成長，在漫遊全球資訊網時已逐漸無法維持一定品質的需求。一個熱門網站的使用者會因為該網站服務效率低而對此網站的信任度與興趣下降，所以如何提昇網站服務品質的相關研究已經逐漸受到從事網路研究者的重視。一般來說，研究的目標通常從（1）伺服器傳輸能力與（2）伺服器工作效能這兩方面的改進著手。
在本論文中，我們提出一套方法對這兩點做改進的研究。在網路傳輸方面，利用服務重導的方式將使用者請求重新導向傳輸時間低與工作負載量低的Mirror伺服器；在伺服器工作效能方面，則是發掘使用者網頁瀏覽行為情節，再配合網頁預取技術降低單一伺服器負載量及減少使用者請求送至伺服器服務次數。統計分析的結果顯示，採用動態重導策略可在平衡群組伺服器負載量的前提下，為使用者選擇距離最近、工作量負荷最輕的Mirror伺服器；使用者網頁預取則可以降低使用者請求伺服器服務的次數，並且降低使用者等待網頁傳輸的時間。

Due to the fact that the hardware development never catches up with the growth of web users, the service quality of a web server is being unable to satisfy its clients’ requirement. A user will find that he must spend unendurable time on browsing a web page, or even the web server rejects his request. Such a condition will surely lower the interests and even the trust of a user to that site. Therefore, the research on raising the quality of service of a web site is getting more attention by the internet researchers. Most of previous work was concentrated on two research issues: (1) the transmitting ability of the web server and (2) the working performance of a web server.
In this paper, we designed a scheme to deal with these two issues. The proposed scheme includes two techniques: service redirection and web-page prefetching. The statistical results show that service rediretion dispatches users’ requests to workload-less and distance-closer web servers in the condition of the balance of these servers, and web-page prefetching indeed reduces the quantity of users’ requests to the sever and the subsequent processing time of requests, and the time of waiting a requested page returned.

第一章 簡介 3
1.1研究動機與目的 3
1.2論文架構 5
第二章 相關研究 6
2.1影響網站效能的因素分析 6
2.2網頁伺服器 8
2.3網頁探勘（WEB PAGE MINING） 11
2.4相關改良網站伺服器效能之研究 12
第三章 網頁伺服器效能之改良 14
3.1簡介 14
3.2建構群組伺服器 16
3.2.1 服務重導 17
3.2.2 動態服務重導 21
3.2.3 群組伺服器功能探討 25
3.3使用者瀏覽網頁之預取技術 26
3.3.1 常用行為情節分類─網頁探勘（Web Page Mining） 26
3.3.2 使用者瀏覽網頁預取 30
3.4動態網頁之處理 36
第四章 模擬測試 38
4.1動態服務重導測試 38
4.2使用者瀏覽網頁預取測試 42
第五章 實作 46
5.1實作環境 46
5.2服務重導 47
5.3網頁預取模組 50
第六章 結論及未來研究方向 52
參考文獻 54
附錄 56

[1]Ayrun Iyengar, Ed MacNair and Thao Nguyen, “An Analysis of Web Server Performance,” Global Telecommunications Conference, 1997. GLOBECOM ''97, IEEE, Volume: 3, p. 1943-1947.
[2]Ben Laurie and Peter Laurie, “Apache: the Definitive Guide.” O’REILLY 1997.
[3]Edith Cohen and Haim Kaplan, “Prefetching the Means for Document Transfe: A New Approach for Reducing Web Latency.” IEEE INFOCOM 2000, p. 854-863.
[4]Greg Holden, Nicolas Wells and Mattew Keller, “Apache Server Commentary.” The Coriolis Group 1999.
[5]Maurice Castro, Michael Dwyer and Michael Rumsewicz, “Load Balancing and Control for Distributed World Wide Web Servers,” Proceedings of the 1999 IEEE International Conference on, Volume: 2, 1999, p.1614-1619.
[6]Sandra G. Dykes, Kay A. Robbins and Clinton L. Jeffery, “An Empirical Evaluation of Client-side Server Selection Algorithms,” IEEE INFOCOM 2000, p.1361-1370.
[7]Sumeet Dua, S.S Iyengar and Eungchun Cho, “Discovery of Web Frequent Pattern and User Characteristics form Web Acces Logs: A framwork for Dynamic Web Personalization.”
[8]The Apache Team, “Apache HTTP server project.” http://www.apache.org/.
[9]Valeria Cardellini, Michele Colajanni and Philip S. Yu, “Dynamic Load Balance on Web-server System”, IEEE Internet Computing, May Jun 1999, p.28-39.
[10]Yiming Hu, Ashwini Nanda and Qing Yang, “Measurement, Analysis and Performance Improvement of the Apache Web Server,” Performance, Computing and Communications Conference, 1999 IEEE International, p. 261-267.
[11]Netcraft Web Server Survey, ”http://www.netcraft.co.uk/survey/