The article considers one of the MTD strategies in which the defenders can continually reconfigure the system to increase the attackers’ uncertainty and consequently, their probability of success. The paper proposes a probabilistic. (Q, r)-strategy that involves the renewal and successive activation of preassembled and reconfigured cyber defense resources. The accumulation and application of operational resources are based on two levels of provisioning, which are randomly distributed. The model assumes that Q is not less than r. It also assumes the repeatability of cyber-attacks. The proposed stochastic model assumes continuous probability distributions of the resources’ lifetime as well as attackers’ activity time. The basic probabilistic characteristics of the model, such as stationary resource availability and emergency system shutdown probabilities, average level of resource reserve, and time to first shutdown, are derived based on renewal and regenerative stochastic processes.