Modeling Software Rejuvenation on a Redundant System Using Monte Carlo Simulation

In this paper, rejuvenation is modeled in a redundant computer system in order to counteract software aging. The evolution of such a system in time could be modeled by a Markov or a semi-Markov process. Nevertheless, due to generally distributed repair times of system components, the entire system is non -- Markovian. Thus, Monte Carlo simulation methods are used for studying system's evolution in time. Through this approach, the main dependability and performance measures are computed in order to ascertain the efficiency of software rejuvenation.