Abstract : The author defines a 'fault' as the identified or hypothesized cause of an error. An error is the manifestation of a fault, an undesired state either at the boundary or at an internal point in the system. A failure is the inability of the system or component to provide the specified service caused by an error. This report surveys of fault taxonomies used in classifying faults, and enumerates many commonly used fault types. Faults can be classified according to many attributes including cause, origin, persistence, duration, nature, extent, value, activity, symmetry and malice. After defining these and other fault attributes, we classify some fault types, and discuss the difficulties encountered in classifying many faults according to these taxonomies without sufficient knowledge of the characteristics of the environments, systems, and applications in which the faults arise. We enumerate and define faults in three classes: physical, environmental and operational faults. Finally, we present a total fault classification to be used within systems to define the classes of faults of interest and to specify the measures taken to tolerate those faults.