A replication technique based on a functional and attribute grammar computation model
Adel Cherif, Masato Suzuki, Takeshi Katayama · 2002
Presents a replication technique based on the FTAG (fault-tolerant attribute grammar) computation model, where instances of a replicated application are active on different groups of processors called replicas. FTAG is a functional and attribute-based model. The developed replication technique implements "active parallel replication", i.e. all replicas are active and concurrently compute a different piece of the application's parallel code. In our model, replicas cooperate not only to detect and mask failures but also to perform parallel computation. The replication mechanisms are supported by the FTAG run-time system and are fully application-transparent. Different novel mechanisms for checkpointing and recovery are developed. Rollback is achieved only if the system experiences multiple failures, otherwise forward recovery is performed. The replication technique takes full advantage of parallel computation to reduce the computation time.