In this paper, the idea of omitting the recovery time when it is too short that it can be considered as negligible is incorporated in a software rejuvenation model. Two new models are introduced, in which the instantaneous availability is defined when firstly the rejuvenation time and secondly both rejuvenation and repair times can be omitted. The main purpose consists in regarding system software as operational when the time spent in a non-operational state, like the rejuvenation and the failure states, does not exceed a predefined time threshold. By modeling the evolution of such a software rejuvenation model using continuous-time Markov chains, we define its availability depending on whether the rejuvenation and the repair critical times are constant or random variables. CDF of the critical rejuvenation time threshold h(t):PDF of the critical rejuvenation time threshold G(t):CDF of the critical repair time threshold g(t):PDF of the critical repair time threshold A(t):availability of the system in the original model à 1 (t): availability of the system for the new model 1 à 2 (t):availability for the new model 2 π(t): state probability distribution Q:i n finitesimal generator matrix p ij (Á,Á): transition probability from state i to state j W:rejuvenation time (r.v.)