The use of real-time systems can differ from their initial design and requirements. New missions may take place in a more agressive environment and cause faults that are not managed by the initial fault tolerance mechanisms. Before attributing a new mission it is interesting to know the limits of a real-time system in order to avoid a new validation process. In this paper we intend to evaluate the fault resilience. We deal with the fault tolerant scheduling analysis under pseudo-periodic faults and fault bursts.