https://scholars.lib.ntu.edu.tw/handle/123456789/151170
標題: | 網路醫學資訊系統之可擴充性與容錯機制 Scalability and Fault Tolerance Mechanisms for a Network Medical Information System |
作者: | 郭斯彥 | 關鍵字: | 元件物件模式;分散式物件元件模式;物件導向;可靠 度;可用度;可擴充性;容錯機制;COM;DCOM;object oriented;reliability;availability;scalability;fault tolerance mechanism | 公開日期: | 31-七月-1999 | 出版社: | 臺北市:國立臺灣大學電機工程學系暨研究所 | 摘要: | 隨著寬頻網路技術日新月異的發 展,醫學資訊系統的使用不再侷限於醫 院本身,將醫學資訊系統架構於網路上 已是一個不可避免的趨勢。利用分散式 元件物件模式(DCOM),可讓醫學資訊系 統的程式碼發展具備物件導向及可重複 使用的特性,配合新世代的寬頻網路環 境,使得醫學資訊的傳遞更具即時性。 分散式元件物件模式是由微軟公司 所提出, 它延伸自元件物件模式 (COM),用來支援在各種網路環境上不 同電腦間物件的通訊。藉著DCOM,我 們可以將已經在以COM 為基礎的應用 程式、元件、工具或知識輕易的轉移到 標準化的分散式計算環境上。 由國外的研究論文中發現,隨著 客戶端請求的數目增多,或是在伺服 器上的物件數目增多,都會使延遲增 長。所以本計畫擬將系統的可擴充性 作深入研究,並提出與實現可改進效 能的方案。其次,我們也將容錯的機 制加以考慮來改進系統的可靠度,並 增加可用性。我們將先建立容錯模 型,找出處理錯誤的機制,提供一個 以分散式元件物件模式為基礎的高可 靠網路服務。 未來的網際網路應用將以分散式物 件為基礎,工業界目前在這方面的進展 非常快速,所以我們學術研究必須在這 個熱門的領域投入人力與心力,而這個 子計畫以可擴充性及容錯機制為課題, 將大幅提高分散式計算環境的效能與可 靠度。 It is due to the rapid development on broadband network technology that the medical information system is no longer just conformed to a medical center or a hospital itself. It is an inevitable trend to construct the medical information system on top of the Internet. With Distributed Component Object Model (DCOM), we make the source codes of medical information system become object oriented and reusable. Together with the next generation broadband network environment, the medical information spreads widely and will serve in a real-time manner. DCOM is proposed by Microsoft. It extends Component Object Model (COM) to support communication among objects in different computers on the Internet. Because DCOM is a seamless evolution of COM, we can leverage our existing investment in DCOM-based applications, components, tools, and know-ledge to move into the world of standards-based distributed computing. It has been shown that the latency increases when the number of requests from clients or objects on the server increases. Thus we will focus on the scalability of the system, and propose and implement new mechanisms for better system performance. Then we will add fault tolerance mechanisms in the system to improve both the reliability and availability. We'll first setup proper fault model and then find out mechanisms of fault-handling to provide a DCOM-based, highly reliable network service. It is foreseen that Internet applications will be based on distributed components in the near future. The progress on this topic is very fast in the industry, and we have to keep pace with it in the academia. This project focuses on the topics of scalability and fault tolerance mechanisms. We are sure that it will be of great help in increasing the performance and reliability of a distributed computing environment. |
URI: | http://ntur.lib.ntu.edu.tw//handle/246246/7725 | 其他識別: | 882219E002006 | Rights: | 國立臺灣大學電機工程學系暨研究所 |
顯示於: | 電機工程學系 |
檔案 | 描述 | 大小 | 格式 | |
---|---|---|---|---|
882219E002006.pdf | 45.07 kB | Adobe PDF | 檢視/開啟 |
在 IR 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。