Heuristic optimization of speedup and benefit/cost for parallel database scans on shared-memory multiprocessors

Previous work on parallel database systems has paid little attention to the interaction of asynchronous disk prefetching and processor parallelism. The authors investigate this issue for scan operations on shared-memory multiprocessors. Two heuristic methods are developed for the allocation of processors and memory to optimize either the speedup or the benefit/cost ratio of database scan operations. The speedup optimization balances the data production rate of the disks and the data consumption rate of the processors, aiming at optimal speedup while ensuring that resources are not allocated unnecessarily. The benefit/cost optimization considers explicitly the resource consumption of a scan operation and aims to allocate processors and memory so that the ratio of the speedup attained to the operation´s resource-time product is maximized. Such an awareness of resource consumption is crucial for intelligent resource management in parallel multi-user database systems, for example, to ensure adequate resource limits for operations that exhibit only small marginal gains in speedup. Both developed heuristics are computationally low-cost and thus suitable for dynamic optimization at runtime