Proceedings 11th Workshop on Parallel and Distributed Simulation (1997)
June 10, 1997 to June 13, 1997
E. Deelman , Dept. of Comput. Sci., Rensselaer Polytech. Inst., Troy, NY, USA
B.K. Szymanski , Dept. of Comput. Sci., Rensselaer Polytech. Inst., Troy, NY, USA
The efficiency of parallel discrete event simulations that use the optimistic protocol is strongly dependent on the overhead incurred by rollbacks. The paper introduces a novel approach to rollback processing which limits the number of events rolled back as a result of a straggler or antimessage. The method, called breadth-first rollback (BFR), is suitable for spatially explicit problems where the space is discretized and distributed among processes and simulation objects move freely in the space. BFR uses incremental state saving, allowing the recovery of causal relationships between events during rollback. These relationships are then used to determine which events need to be rolled back. The results demonstrate an almost linear speedup-a dramatic improvement over the traditional approach to rollback processing.
discrete event simulation; spatially explicit simulations; breadth-first rollback; parallel discrete event simulations; optimistic protocol; rollback overhead; rollback processing; straggler; antimessage; simulation objects; incremental state saving; causal relationship recovery; speedup
E. Deelman and B. Szymanski, "Breadth-first rollback in spatially explicit simulations," Proceedings 11th Workshop on Parallel and Distributed Simulation(PADS), Lockenhaus, AUSTRIA, 1997, pp. 124.