A theory for observational fault tolerance

Abstract

One reason for the study of programs in the presence of faults, i.e. defects at the lowest level of abstractions [2], is to be able to construct more dependable systems, meaning systems exhibiting a high probability of behaving according to their specification [18]. System dependability is often expressed through attributes like maintainability, availability, safety and reliability, the latter of which is defined as a measure of the continuous delivery of correct behaviour, [18]. There are a number of approaches for achieving system dependability in the presence of faults, ranging from fault removal, fault prevention and fault tolerance.