Towards Simulation of Parallel File System Scheduling Algorithms With PFSsim
Source: University of Florida
Many High-End Computing (HEC) centers and commercial data centers adopt parallel file systems (PFSs) as their storage solutions. As the number of applications concurrently accessing a PFS grows in both quantity and variety, it is expected that scheduling algorithms for data access will play an increasingly important role in PFS service quality. However, it is costly and disruptive to thoroughly research scheduling mechanisms in deployed peta- or exa-scale systems, compounded by the complexity in scheduling policy implementation and experimental data gathering. While a few parallel file system simulation frameworks have been proposed, their goals have not been in the scheduling algorithm evaluation. In this paper, the authors propose PFSsim, a simulator designed for the purpose of evaluating I/O scheduling algorithms in PFS.