Processing Distributed Transactions in a Predefined Order

Consider distributed transactional memory systems where transactions residing at nodes of a communication graph operate on shared, mobile objects. A transaction requests the objects it needs, executes once those objects have been assembled, and then forwards those objects to other waiting transactions. We study the predefined order scheduling problem of committing transactions according to their priorities. This problem naturally arises in areas, such as loop parallelization and state-machine-based computing, where producing executions equivalent to a priority order is needed to satisfy certain properties. Specifically, we study predefined order scheduling considering two performance metrics fundamental to any distributed system: (i) execution time - total time to commit all the transactions and (ii) communication cost - the total distance messages travel. We design scheduling algorithms that are simultaneously efficient for both the metrics and rigorously evaluate them through several benchmarks on random and grid graphs, validating their efficiency. To the best of our knowledge, this is the first study of predefined order scheduling in distributed systems.