為因應使用者對網頁服務的需求，現有的單一網頁主機的架構，早已不敷使用，因此分散式代理伺服器系統因應使用者龐大的需求而產生，但是其所運用的負載平衡機制，常常會因為切換代理伺服器的技術，而造成使用者需要重新進行認證的問題發生。
分散式代理伺服器系統所遇到最大的問題在於網頁重新連線時，伺服器之間連線的轉移；在連線轉移的同時，又要能解決客戶端的連線正常，並且不需要重新認證。所以，本篇論文主要擬解決的問題就是我們要設計出一種整合型之分散式代理伺服器系統，讓使用者在登入需要認證的網頁過程之中，不會因為代理伺服器之間的資料交換，而導致使用者需要重新登入。

In order to meet users’ needs for web services, the existing structure of single web host is no loner adequate. Therefore, in order to meet enormous user needs, the Distributed Proxy System was born. However, due to the Load Balancing Mechanism applied to it, the problem of user re-authentication often occurs, because of the technique of proxy server switching. 
The biggest problem encountered by a Distributed Proxy System is the connection shift among servers when a webpage is reconnected again, the connection of the Client End should be kept normal at the same time when the connection is shifted, without any re-authentication. Therefore, the main problem that this paper intends to solve is: We will propose Single Logon Distributed Proxy System; thus, when entering a web page that requires authentication, a user will not have to re-login due to the data exchanges among proxy servers.

第一章 緒論...................................................1 
　　1.1 研究動機..............................................1 
　　1.2 研究內容..............................................2 
　　1.3 論文架構..............................................2 
第二章 相關研究...............................................4 
　　2.1 叢集伺服器............................................4 
　　2.2 Linux虛擬伺服器(Linux Virtual Server).................5 
　　　2.2.1 VS/NAT模式........................................7 
　　　2.2.2 VS/TUN模式........................................9 
　　　2.2.3 VS/DR模式........................................10 
　　2.3 負載平衡機制.........................................12 
　　　2.3.1 現今商業有關網頁負載平衡.........................12 
　　　　2.3.1.1 軟體的負載平衡...............................13 
　　　　2.3.1.2 硬體的負載平衡...............................14 
　　　2.3.2 分散式代理伺服器.................................14 
　　2.4負載平衡機制的優缺點..................................16 
第三章 系統架構..............................................17 
　　3.1 Single Logon Distributed Proxy System的網路系統架構..17 
　　3.2 Single Logon Distributed Proxy System的連線運作方式..19 
　　　3.2.1 SLDPS的連線步驟..................................19 
　　　3.2.2 TDPS切換代理伺服器時的異常問題...................20 
　　3.3 SLDPS的連線需求對應表................................22 
　　　3.3.1 SLDPS的對應表定義與簡介..........................23 
　　　3.3.2 SLDPS對應表之範例................................24 
　　　3.3.3 時間區塊.........................................25 
　　3.4 資料的同步...........................................27 
　　　3.4.1 連線資料的同步狀態...............................27 
　　　3.4.2 資料同步的方式...................................28 
　　3.5 TDPS連線時的異常狀況.................................30 
　　　3.5.1 連線時的異常處理步驟.............................31 
　　　3.5.2 判斷連線資料回覆步驟.............................33 
　　　3.5.3 Proxy Gateway的容錯處理..........................33 
　　　3.5.4 Proxy Gateway的異常處理步驟......................33 
第四章 系統模擬..............................................35 
　　4.1 模擬環境.............................................35 
　　4.2 模擬結果.............................................38 
第五章 結論與未來發展........................................49 
　　5.1 結論.................................................49 
　　5.2 未來發展.............................................49 
參考文獻.....................................................51 
附錄-英文論文................................................54 

圖目錄

圖2-1 叢集伺服器..............................................5
圖2-2虛擬伺服器...............................................7
圖2-3 VS/NAT模式架構圖........................................8
圖2-4 VS/TUN模式架構圖.......................................10
圖2-5 VS/DR模式架構圖........................................11
圖2-6 分散式代理伺服器.......................................15
圖3-1 SLDPS網路系統架構圖....................................17
圖3-2 SLDPS與TDPS連線流程圖..................................20
圖3-3 切換伺服器異常流程圖...................................21
圖3-4 SLDPS連線需求流程圖....................................22
圖3-5 時間區塊同步狀態示意圖.................................28
圖3-6 時間區塊同步示意圖.....................................29
圖3-7 傳統的資料回覆異常流程圖...............................30
圖3-8 SLDPS資料回覆流程圖....................................32
圖4-1(a) Siege壓力測試模擬過程與結果.........................37
圖4-1(b) AB壓力測試模擬過程..................................37
圖4-1(c) AB壓力測試模擬結果..................................38
圖4-1(d) Web Server上之執行緒................................38
圖4-2(a) AB在2台伺服器下之需要重新登入之次數.................40
圖4-2(b) Siege在2台伺服器下之需要重新登入之次數..............41
圖4-2(c) AB在4台伺服器下之需要重新登入之次數.................41
圖4-2(d) Siege在4台伺服器下之需要重新登入之次數..............42
圖4-2(e) AB在6台伺服器下之需要重新登入之次數.................42
圖4-2(f) Siege在6台伺服器下之需要重新登入之次數..............43
圖4-3(a) AB在2台伺服器下之平均回應時間.......................45
圖4-3(b) Siege在2台伺服器下之平均回應時間....................45
圖4-3(c) AB在4台伺服器下之平均回應時間.......................46
圖4-3(d) Siege在4台伺服器下之平均回應時間....................46
圖4-3(e) AB在6台伺服器下之平均回應時間.......................47
圖4-3(f) Siege在6台伺服器下之平均回應時間....................47

表目錄

表2-1 軟硬體負載平衡方法比較表...............................14
表3-1 Proxy Gateway對應表欄位定義............................23
表3-2 Proxy Server 對應表/同步對應表欄位定義.................23
表3-3 Proxy Gateway對應表範例................................25
表3-4 Proxy Server 對應表範例................................25
表3-5 時間單位名稱說明表.....................................26
表4-1(a) 1500~4500 hits/15mins需要重新登入之次數表...........39
表4-1(b) 6000~7500 hits/15mins需要重新登入之次數表...........40
表4-2(a) 100~300 hits/min平均回應時間數據表..................44
表4-2(b) 400~500 hits/min平均回應時間數據表..................44

[包蒼龍 04]..............包蒼龍, 池至欽, 陳建伯, 李明龍, "突波式需求之網頁伺服器負載平衡架構," TANET, Oct. 2004, pp. 280-285.
[吳宜興 05]..............吳宜興, "動態網頁伺服器彈性負載平衡架構之研究," 大同大學資訊工程研究所碩士論文, 2005.
[阮東憲 06]..............阮東憲, "叢集網路檔案服務系統之設計與分析," 南台科技大學資訊管理研究所碩士論文, 2006.
[Chen 09]................Hsin-Hung Chen, “Performance Comparison of Web Server Load Balancing Frameworks,” Thesis of Master, Science Department of Computer Science and Engineering Tatung University, Jan 2009.
[Chen 08]................Jian-Bo Chen, “A Study of Web Server Load Balancing Architecture for Burst Mode Traffic,” Ph. D. Dissertation, Department of Computer Science and Engineering, Tatung University, Jul 2008.
[Guo 06].................Jiani Guo and Laxmi N. Bhuyan, “Load Balancing in a Cluster-Based Web Server for Multimedia Applications,” IEEE Trans. Parallel and Distributed Systems, vol. 17, no. 11, Nov 2006, pp. 1321-1334.
[Karger 03]..............David Karger and Matthias Ruhl, “New Algorithms for Load Balancing in Peer-to-Peer Systems,” Tech. Rep. MIT-LCS-TR-911, MIT LCS, Jul 2003.
[LVS 04].................Linux Virtual Server Available at: http://www.linuxvirtualserver.org/
[Liao 07]................Kuo-Chung Liao “A Fault Tolerance Mechanism on Web Service,” Thesis of Master, Department of Computer Science and Information Engineering, Tamkang University, Jun 2007.
[Markatos 02]............E. P. Markatos, D. N. Pnevmatikatos, M. D. Flouris etc. “Web-Conscious Storage Management for Web Proxies,” IEEE/ACM TRANSACTIONS ON NETWORKING, vol. 10, no. 6, Dec 2002, pp. 735-748.
[Microsoft 03]...........Microsoft Network Load Balancing. Available at: http://technet2.microsoft.com/WindowsServer/en/library/
[Pan 04].................Chi-Chien Pan, Kai-Hsiang Yang, and Tzao-Lin Lee, “Secure Document Access Architecture Based On Distributed Firewall AND Proxy Technologies,” Bulletin of the College of Engineering, N.T.U., no. 90, Feb 2004, pp. 127–136.
[Ramaswamy 05]...........Lakshmish Ramaswamy, Bugra Gedik, and Ling Liu, "A Distributed Approach to Node Clustering in Decentralized Peer-to-Peer Networks", IEEE Transactions on Parallel and Distributed Systems, vol. 16, no. 9, Sep 2005, pp 814-829.
[Sharma 08]..............S. Sharma, S. Singh, and M. Sharma, “Performance Analysis of Load Balancing Algorithms,” World Academy of Science, Engineering and Technology, vol. 28, Apr 2008, pp. 269-272.
[Squid 03]...............Squid-cache Available at: http://www.squid-cache.org/
[Xu 05]..................Yue Xu “Hybrid clustering with application to Web mining,” International Conference on Active Media Technology, May 2005, pp. 574-578.
[Yiu 05].................Fai-Sit Yiu, Francis C.M. Lau, and Cho-Li Wang, “On The Cooperation of Web Clients and Proxy Caches,“ International Conference on Parallel and Distributed Systems, vol. 1, Jul 2005, pp. 264-270.
[Zhang 00]...............Wensong Zhang, Shiyao Jin, and Quanyuan Wu, “Linux Virtual Server: Server Clustering for Scalable Network Services,” Ottawa Linux Symposium 2000.
[Zhou 03]................Bing Zhou, Jun-Yi Shen, and Qin-Ke Peng, “A parallel clustering algorithm for cluster system,” Second International Conference on Machine Learning and Cybernetics, vol. 1, Nov 2003, pp. 4-8.