In this paper we present a novel framework for disaster resilience, called FRADIR, which incorporates reliable network design, disaster failure modeling and protection routing in order to improve the availability of mission-critical applications. To the best of our knowledge, this is the first comprehensive framework which utilizes tools from all these fields in a joint design of disaster resilient connections. In particular, we introduce a new probabilistic regional failure model, which does not only take into account the distance from the epicenter of the failure, but includes the (improved) availability values of the network components into the model, too. Based on the failure list generated as the result of the availability-aware disaster failure model, dedicated protection approaches are used to route the connection requests. We demonstrate the concept and benefits of FRADIR through experimental results in two real-like network topologies. Our proof-of-concept implementation shows that with the interplay between protection routing, failure modeling and network update procedure the network performance in terms of blocking probability and average resource consumption can be significantly improved, which makes FRADIR a possible competitor to provide disaster resiliency in critical infrastructures.