Alternate fault diagnosis strategy in a distributed environment
S.-C. Chu · 2002
All approach that uses a hardware device to perform fault diagnosis in a fully distributed multiprocessor system is proposed. Both the complexities of the diagnostic hardcore and interprocessor communication are reduced. Fault diagnosis in a fully distributed system with permanent faults is considered. Standard notation, definitions, and a framework for discussion are provided. A brief description of the general fault diagnosis process follows. The structure of the diagnostic hardware for each component in the system, called a component diagnosis element (CDE), is defined. The input dependency of each CDE is presented along with a discussion of how this diagnosis device carries out fault diagnosis in a distributed environment. The necessary conditions are provided for a class of fully distributed systems where faulty processors can be diagnosed independently within each processor provided that the number of faulty processors does not exceed a set limit.>