Generating realistic impressions for file-system benchmarking

Abstract

The performance of file systems and related software depends on characteristics of the underlying file-system image (i.e., file-system metadata and file contents). Unfortunately, rather than benchmarking with realistic file-system images, most system designers and evaluators rely on ad hoc assumptions and (often inaccurate) rules of thumb. Furthermore, the lack of standardization and reproducibility makes file-system benchmarking ineffective. To remedy these problems, we develop Impressions, a framework to generate statistically accurate file-system images with realistic metadata and content. Impressions is flexible, supporting user-specified constraints on various file-system parameters using a number of statistical techniques to generate consistent images. In this article, we present the design, implementation, and evaluation of Impressions and demonstrate its utility using desktop search as a case study. We believe Impressions will prove to be useful to system developers and users alike.