Providing resiliency is an important aspect of network service. Redundancy is one of the major solutions by which resiliency is offered, but it comes with an increase in Capex/Opex costs and usage of resources. This paper utilizes the emerging Software-Defined Networking (SDN) and Network Function Virtualization (NFV) paradigms to provide resilient services with minimal amount of redundancy. Two algorithms are proposed, one for the migration of the functions from the failed to the functioning server(s) and the other to improve the robustness of the failed part of Service Function Chain (SFC). The algorithms have been studied using a simulation model. The results show that the reactive handling of failure is practically feasible while honoring the Service Level Agreement (SLA) and an SFC can be made robust enough with respect to the parameter considered. © 2018 IEEE.