隨著網路、硬體、軟體等相關技術的快速進步，一些醫學資訊系統的發展環境，已到達分散、複雜、龐大且多樣。在一個這樣的環境下，常因一個小環節的問題，導致整體系統無法正常運作，而系統管理者也不知如何快速下手找出真正的問題點。我們希望能有一套系統，協助管理者正確、快速的判斷並解決問題，進一步，我們將容錯機制整合在醫學資訊系統。

　　本實驗室發展了影像管理系統與教學模擬系統。影像管理系統主要的功能是管理醫院影像檔的收集與派送；教學模擬系統，是一套線上學習系統，藉由與系統的互動模擬中完成醫學知識的學習、探討及考試等事項。但因兩系統分散於醫院各地，導致系統異常時，我們往往無法快速得知並即時處理。

　　本論文中，我們將整個監控系統中各節點，在垂直上以邏輯分網路層、作業系統層與應用層；在水平上，各層的監控，我們分主控端監控模組與被控端監控模組。資訊的收集除各主機資源使用狀況與物件效能的資訊外，也收集各主機事件發生的紀錄檔與應用程式紀錄檔。達到全面性監控網路上各節點的運轉情況。

　As the related technologies of internet, hardware, and software progress rapidly, some developments of the medical information system have been distributed, complex, huge, and diverse. In such environment, a little problem will cause the whole system to fail and the system manager can not find the problem immediately. We hope that there is a system which can help system manager to solve the problem correctly and quickly. Furthermore, we integrate the fault tolerance into the medical information system.

　Our laboratory develops a picture management system and a simulation system for teaching. The main function of the picture management system is to manage the collection and distribution. The simulation system for teaching is an online learning system. By means of interaction with the system, the users study and discuss the medical knowledge. The users also can be tested in the system. However, because these two systems are distributed in the whole hospital, we can not perceive and deal with the problem immediately when the systems fail.

　In this thesis, we monitor each node in the system. Vertically, the system is divided into network layer, operating system layer, and application layer by logic. Horizontally, the monitor of each layer is divided into master monitoring module and slave monitoring module. Not only the resource usage of each server and the performance of each object, but also the Log files of each server and application are collected. Therefore, monitoring the operating situation of each node in the whole network is achieved.

[1] Zhou, W. ,“Detecting and tolerating failures in a loosely integrated heterogeneous database system”, July 15, 1999,Computer Communications Volume: 22, Issue: 11, Page: 1056-1067

[2 ] King, A.; Hunt, R. ,“Protocols and architecture for managing TCP/IP network infrastructures”, September 30, 2000, Computer Communications Volume: 23, Issue: 16, Page: 1558-1572

[3] Deri, Luca; Mattei, Eugenio,” An object-oriented approach to the implementation of OSI management “, August, 1995,Computer Networks and ISDN Systems Volume: 27, Issue: 9, Page: 1367-1385

[4] Gavalas, D.; Greenwood, D.; Ghanbari, M.; O'Mahony, M., “Advanced network monitoring applications based on mobile/intelligent agent technology ” , April 1, 2000, ComputerCommunications Volume: 23, Issue: 8, Page: 720-730

[5] O. Wolfson, S. Sengupta, Y. Yemini, “Managing Communication Networks by
Monitoring Databases”, 1996, IEEE Transactions on oftware Engineering.

[6] G. Jakobson, M.D. Weissman, “Alarm Correlation”, 1993, IEEE Network, Nov.,
page: 52-59.

[7] RFC 1155: Structure and Identification of Management Information for
TCP/IP-based Internets.

[8] RFC 1157: Simple Network Management Protocol (SNMP).

[9] RFC 1213: Management Information Base for Network Management of TCP/IP-based
Internets. MIB-II.

[10] Stalling W. “SNMP, SNMPv2, SNMPv3, and Rmon 1 and 2”,
Addison-Wesley: Reading, MA, 1999

[11] Douglas R. Mauro & Kevin J. Schmidt, “Essential SNMP ”,
O’REILLY, US , Mar, 2003

[12] Microsoft Operation Manager,Micorosoft Corporation,
http://www.microsoft.com/mom

[13] Tivoli, IBM Corporation ,http://www-306.ibm.com/software/tivoli