作者: Xiaozhi Du , Yong Qi , Di Hou , Ying Chen , Xiao Zhong
关键词:
摘要: Software rejuvenation is a preventive and proactive technology to counteract the phenomenon of software aging system failures improve reliability. In this paper we present analyze three policies for an operational with multiple degradations, called preemptive rejuvenation, delayed mixed rejuvenation. These consider both history data current running state, action triggered on basis predetermined performance threshold interval respectively. Continuous-time Markov chains are used describe analytic models. To evaluate these polices expediently, utilize deterministic stochastic Petri nets solve Numerical results show that deployment in leads significant improvement availability throughput. better than standard policy, policy best one.