The problem of predicting the reliability of a distributed system based on the principles of Byzantine agreement is addressed. The system is considered inoperable or failed if Byzantine agreement cannot be guaranteed. The reliability models depend on a unified model of interactive consistency, which is based on a unique fault taxonomy appropriate for distributed systems. The unified model takes advantage of the fact that some faults may not be of an arbitrary nature, while still allowing for the fact that some faults may be arbitrary. A closed-form expression for the reliability and the mean time to failure of systems base on the unified model is derived. Each processor is allowed to have multiple failure modes, and the contribution of the interactive consistency algorithm is explicitly taken into account. The practical value of this unified model in designing ultrareliable systems is demonstrated by several examples.>

A novel technique that allows a single system to guarantee the execution of both periodic and aperiodic tasks within hard deadlines is presented. The approach is based on dynamically changing the replication factor of periodic tasks in response to aperiodic tasks. The technique is directly applicable to the problem of transient overloads that are time-critical. Preliminary results that show the applicability and the tradeoffs involved are presented.>

The authors address the problem of reaching Byzantine agreement in a distributed system in the presence of different types of faults and show that significant improvements in reliability and performance are possible if faults can be partitioned into disjoint classes. They show that, in a distributed system, to guarantee Byzantine agreement requires N>2a+2s+b+r where N is the total number of processors, a is the number of malicious asymmetric faults (a>

A description is given of the multicomputer architecture for fault tolerance (MAFT), a distributed system designed to provide extremely reliable computation in real-time control systems. MAFT is based on the physical and functional partitioning of executive functions from applications functions. The implementation of the executive functions in a special-purpose hardware processor allows the fault-tolerance functions to be transparent to the application programs and minimizes overhead. Byzantine agreement and approximate agreement algorithms are used for critical system parameters. MAFT supports the use of multiversion hardware and software to tolerate built-in or generic faults. Graceful degradation and restoration of the application workload is permitted in response to the exclusion and readmission of nodes, respectively.>