Performability evaluation: where it is and what lies ahead

The concept of performability emerged from a need to assess a system's ability to perform when performance degrades as a consequence of faults. After almost 20 years of effort concerning its theory, techniques, and applications, performability evaluation is currently well understood by the many people responsible for its development. On the other hand, the utility of combined performance-dependability measures has yet to be appreciably recognized by the designers of contemporary computer systems. Following a review of what performability means, we discuss its present state with respect to both scientific and engineering contributions. In view of current practice and the potential design applicability of performability evaluation, we then point to some advances that are called for if this potential is indeed to be realized.<
>