In this paper, we propose a recovery model to enhance system survivability through critical service resource reconfiguration. That is, the model focuses on how to preserve the system and resume its critical service while incident occurs by reconfiguring the system based on the remaining available resources without affecting the stability of the system. There are three motivating factors of importance in this recovery model namely, the response time of reconfiguration, the cost of reconfiguration, and the number of pre-empted non-critical service resources. The adoption of fault-tolerance using redundancy in our model is discussed from a new perspective.