Filebench

What Filebench is

Filebench is a Linux tool for benchmarking file systems and storage by generating workloads that represent application I/O behavior. It supports micro- and macro-benchmarking, with random and sequential tests, IOPS measurement, and latency measurement. Users can define custom workloads in Workload Model Language (WML), specifying filesets, processes, threads, and flow operations. Built-in workload personalities cover mail, web, file, and database server scenarios, and the documentation names webserver, fileserver, and mailserver examples. Shipped workloads may not match the size of a particular system, so the documentation recommends copying and adjusting them. Filebench is free, with a listed price of $0 and no usage limits stated. It provides a command-line interface. The README describes compiling and installing it with configure, make, and sudo make install; yacc and lex are required. The project directs questions, bug reports, and feature requests to GitHub Issues. Its repository license is the Common Development and Distribution License, Version 1.0.

Who it suits

Filebench suits Linux users who need to generate or customize storage workloads for benchmarking. Its command-line installation and workload definitions are aimed at users comfortable with build tools and configuration.

Getting started

1. Open the Filebench project on GitHub.
2. Install yacc and lex.
3. Compile the project with configure and make.
4. Install it with sudo make install.
5. Choose a predefined workload or describe one in WML, adjusting shipped workloads for your system as needed.

Verdict

Filebench offers configurable storage workloads and several benchmark types at no cost. Account for its Linux-only support, build requirements, and the need to adjust shipped workloads to fit a system.