基于时间的软件恢复策略的建模与分析

针对软件在连续运行过程中普遍发生的老化现象，提出了一种嵌套的基于时间的软件恢复策略，对恢复过程的Petri网模型分析求解，并最终得到了最优恢复时间间隔序列和最优应用级恢复次数。该策略同时考虑应用级和系统级的恢复，从而进一步减少了恢复时间，降低了恢复成本和周期性应用级恢复策略预测失败的风险，提高了系统的可靠性。对于更复杂的系统，策略还可进一步嵌套进程级的恢复，从而具有一定的可扩展性。

Modeling and Analysis of Time-based Software Rejuvenation Policy

A new nested time-based software rejuvenation policy is proposed to prevent software from failure due to software aging. We analyzed quantitatively petri net model of rejuvenation sub-process and finally obtain a series of the optimal rejuvenation intervals and the optimal times of application-level rejuvenation approximately. Comparing to the conventional rejuvenation policies, the nested time-based software rejuvenation policy takes into account the application-level and system-level rejuvenation simultaneously. The results demonstrate that it is possible to consume less time and lower cost, to reduce the risk of failure to predict, to enhance the software reliability and availability with this policy. In addition, this policy is extendable for complicated system because process-level rejuvenation can be nested in application-level rejuvenation.