Towards fast profiling of storage devices regarding access sequentiality

This work presents a tool for storage device profiling named SeRRa. Our tool obtains the sequential to random throughput ratio for reads and writes of different sizes. Several optimizations aim at adapting applications' access patterns in order to generate contiguous accesses for improved performance when accessing storage devices like hard disks. However, when considering other storage options like RAID arrays and SSDs, the access time ratio between contiguous and non-contiguous accesses may not compensate for these optimizations' cost. In this scenario, the information provided by our tool could be used to dynamically decide if optimizations are beneficial for performance, an important task to provide I/O throughput. In order to provide this information efficiently, SeRRa employs benchmarks to obtain the values for a subset of the parameter space and estimates the remaining values through a linear model. Extensive test campaigns with our tool on nine storage devices have shown median errors of approximately 5% (up to 55%) while taking only 1/168 of the originally needed time.