SIMULATION OF SYSTEM LEVEL DIAGNOSIS IN DISTRIBUTED ARBITRARY NETWORK

The problem of distributed diagnosis in arbitrary network failures and repairs is considered in this paper. The basic idea behind incorporating a fault tolerance capability to a distributed system is to provide the system with extra (redundant) resources. A number of investigations have been attempted to extend traditional notions of “fault-tolerant computing”, to deal with the problem of failures, which affect the facilities of distributed systems and computer networks. as distinguish of diagnostic responsibility requires the flow of diagnostic information through the network, and the faulty facilities themselves may participate in this flow and may alter, destroy, or generate erroneous diagnostic information in the process, the whole diagnostic procedure itself becomes quite complex. The purpose of this study is to simulate a distributed system and carry out fault diagnosis under Arbitrary Network topologies. The distributed system level diagnosis algorithms discussed in this paper give a comprehensive idea about various issues one must keep track of while going to develop such fault tolerance algorithms. Since “system level diagnosis” is one of the steps in the process of building “distributed fault-tolerant systems”, reliability of such a system depends heavily on proper functioning of the diagnosis algorithm. The distributed system has been simulated in JAVA using the Console. Java program to create multiple windows, each one representing a different system node. Afterwards, distributed diagnosis algorithms have been simulated using this environment.