Carnegie Mellon University Parallel Data Lab Technical Report CMU-PDL-16-105. October 2016.
Saurabh Kadekodi, Vaishnavh Nagarajan, Garth A. Gibson
Carnegie Mellon University
File system aging has been advocated for thorough analysis of any design, but it is cumbersome and often bypassed. Our aging study re-evaluates published file systems after aging using the same benchmarks originally used. We see significant performance degradation on HDDs and SSDs. With performance of aged file systems on SSDs dropping by as much as 80% relative to the recreated results of prior papers, aging is even more necessary in the era of SSDs. Still more concerning, the rank ordering of compared file systems can change versus published results.
We offer Geriatrix, a simple-to-use aging suite with built-in aging profiles with the goal of making it easier to age, and harder to justify ignoring file system aging in storage research.
KEYWORDS: File System, Aging, Disrtribution Fitting, Modeling, Tool
FULL TR: pdf