Adaptive parallel I/O scheduling algorithm for multiprogrammed systems

摘要

As the rate at which disk drives read and write data is improving at a much slower pace than the speed of processors, I/O has risen to become the bottleneck in high-performance computing for many applications. A possible approach to address this problem is to schedule parallel I/O operations explicitly. To this end, we propose two new I/O scheduling algorithms and evaluate the relative performance of the proposed policies against two baseline policies. Simulation results show that the proposed policies outperform the baseline policies.