Replication control for fault-tolerance in distributed real-time database systems

In many distributed real-time systems, certain critical data need to be replicated to achieve fault-tolerance and improved performance. Schedulers for such distributed real-time systems must satisfy timing constraints of transactions and preserve data consistency. In this paper we presents a replication control algorithm, which integrates real-time scheduling and replication control to achieve the desired level of fault-tolerance. The algorithm adopts a majority consensus scheme for replication control and employs epsilon-serializability, a correctness criterion which is less stringent than conventional one-copy serializability. The performance of the algorithm is compared with that of a real-time token-based algorithm