Performability modelling and evaluation brings together twodisciplines that have long been treated separately in differentcommunities: computer and communication system performanceevaluation and system reliability and availabilityevaluation.
* Provides a single-source reference covering of all aspects ofperformability evaluation.
* Introduces a new approach to assessing the performance anddependability of systems thus revealing system characteristics thatcannot be uncovered separately by performance or reliabilityevaluation.
* Explains how systems can operate partially in the presence offailures by discussing:
-Capacity of failure prone clusters of transaction processingsystems.
-Network throughput when network nodes and links fail.
-Task completion times on degradable multiprocessor systems.
-A variety of modelling and evaluation techniques.
* Features software tool descriptions to support performabilityevaluations as well as an extensive bibliography.
Computer scientists and engineers will find this excellent overviewcompulsive reading as will postgraduates and researchers incomputer science and related fields.
The uniformization method in performability analysis; queues with breakdowns; techniques for fast simulation of models of highly dependable systems; closed-form solutions for performability; the task completion time in degradable systems; Markov-reward models and hyperbolic systems; specification and construction of performability models; a survey of performability modelling tools.