On Evaluating the Reliability of Multiprocessor Systems
This paper is concerned with a class of dependable computing systems, where the requirements on reliability figures go together with specific requirements on performance and cost. In a common approach, the reliability requirements are fulfilled by introducing redundancy into the system, managed by a fault tolerance technique, namely error processing and fault treatment. Extensive studies on the development of a wide number of fault tolerance mechanisms have appeared in the literature, together with evaluation of the system benefits deriving from the application of each single mechanism. The contribution of this paper is an analysis of integrated fault tolerance strategy on the proposed multiprocessor system.