Fault-tolerant approach for deploying Server Agent-based Active Network Management (SAAM) server in Windows NT environment to provide uninterrupted services to routers in case of server failure(s)/ Efraim Kati

Abstract

With the explosive growth of the Internet and high demand on real- time network applications, the need for integrated service networks has emerged. Therefore, the Next Generation Internet project and as a part of this project the Server Agent-based Active network Management (SAAM) project was initiated. SAAM is a server based hierarchical routing architecture designed to provide Quality of Service routing services for network resource intensive applications. In the SAAM architecture, a small number of dedicated SAAM servers perform most of the network management tasks on behalf of the routers. The SAAM server has a great responsibility in the SAAM architecture and failure of the SAAM server can have a devastating effect on the performance of the entire network. In order to tolerate the failure of the SAAM server, this thesis examines the fault tolerance (ft.) for the SAAM server in two phases: local area ft., and remote area ft. For the local area ft., after a survey of the literature and commercial offerings, a recommended solution is proposed. For the remote area ft., a backup server model is designed and prototyped. The prototyped model provides robust error detection and a fast recovery from the failure of the primary SAAM server