Fault-tolerant systems

Two mishaps early this year forcefully reminded the authors how important computer systems' reliability has become to human well-being in this last decade of the 20th century. In January, large parts of AT&T's long-distance telephone switching networks failed because of a software problem. Then, in February, a 'fly by wire' Airbus A320 aircraft crashed in Bangalore, India. While a definitive cause for the accident has yet to be established (preliminary investigations suggest that the control computers might well have performed to specifications), the fact that the A320 employs a fully electronic cockpit with no mechanical backup for control has raised much speculation about the computers' role in the crash. The authors stress that the basic idea behind building in a fault tolerant capability is to provide the system with extra (redundant) resources.<
>