Design And Implementation Of Unified Fault Management System

The traditional Network Management System (NMS) is mostly constructed as a single application on a single machine. Management jobs are constrained on some special workstations. This kind of system is too big and complex with worst migration. It requires a high-performance workstation. The spreading of WWW, Java and Corba technology offers us a most revolutionary solution. Web-based, distributed, layered architecture is an ideal management mode where there is browser there is management. Further, Java and Corba enable the management system to be totally platform-independent, easy-migratable and distributed. J2EE leads this to a much higher extent.Unified NMS has gradually become the goal of enterprises. It is to provide a common platform and reusable components and thus provide a framework for quick NMS development. Unified Fault Management (UFM) is one of the most important component. UFM has been the crucial weapon that device provider competes others. Automated process, quick fault detection, fault locating and recovering, fault correlation and etc. have been imporant aspects to telecom system's success. Developing high performance, totally automated fault managment system has been an agenda for most enterpises.This thesis describes firstly the general distributed architecture for NMS system. By comparison with other solution, it shows the virtues, feasibility and availability of this solution. Then it introduces UTStarcom UFM component, which is based on this architecture, from perspectives of requirement-analysis, system-design, system-implement and etc. in detail. It shows the general process of constructing a common component in a Web-based, distributed architecture for the network management system.