Preventive Maintenance for Virtualized Local Disaster Recovery Plan

Abstract

Preventive maintenance combined with disaster recovery planning will help to minimize the hardware and software problems if disaster occurs. Disaster recovery processes involve unplanned interruption of services. Unplanned downtime is mainly caused by software failure due to software aging rather than hardware failure. Preventive maintenance (software rejuvenation) is used to counteract the software aging phenomenon. In this paper, we propose preventive maintenance model for virtualized local disaster recovery plan through a stochastic Petri net model. In the proposed model, active-standby virtualized clustering architecture is employed. We analyze how preventive maintenance can improve the system availability of virtualized local disaster recovery plan. We perform the evaluation of the proposed model using SHARPE simulation tool.