40 Years of Computing at Newcastle

Department Technical Report Series No. 449

Building Fault-tolerant Distributed Computing Systems using Standard Components

P.A. Barrett
S.K. Shrivastava
N.A. Speirs
A. Waterworth

University of Newcastle upon Tyne. 1994

Abstract

The aerospace industry is making increasing use of computers in the implementation of safety-critical systems. New generations of airliners (Airbus A320/330/340, Boeing 777), for example, are using digital computers in their primary flight control systems - systems upon whose integrity and availability depends the safety of the aircraft and its passengers. Such systems are required to be fault-tolerant so that they can continue to function correctly in the presence of a finite number of component failures. Current generations of fault-tolerant computers for safety-critical applications tend to make extensive use of special-purpose hardware, and are thus expensive and inflexible. This paper investigates the possibility of constructing fault-tolerant computer systems using standard hardware components, replicated to an appropriate degree and communicating via special-purpose software protocols. The Voltan family of fail-controlled nodes is introduced and described, and ways of incorporating Voltan nodes into Integrated Modular Avionics (IMA) architectures are presented Means of overcoming the potential drawbacks of such nodes are discussed. In particular, possible extensions to IMA gateway modules in order to provide communications and data validation services in support of Voltan nodes are described.


Department Technical Report Series - 1994
Department Technical Report Series Index
Contents Page - 40 Years of Computing at Newcastle
Technical Report Abstract No. 449, 27 June 1997