Linux Disk Scheduler Benchmarking

[Dr. Peter Chubb, Project Research Officer – Gelato] Over the last six months, Google has sponsored Gelato to take a close look at the disk chedulers in Linux, particularly when combined with RAID.

We benchmarked the four standard Linux disk schedulers using several different tools (see our wiki for full details) and lots of different workloads, on single SCSI and SATA disks, and on hardware and software RAID arrays from two to eight
spindles (hardware raid) and up to twenty spindles (software raid), trying RAID levels 0 through 6.

We had to fix some of the benchmarking tools (the fixes are now upstream), and we developed a new one: a Markov Chain based replay tool, which allows a workload to be characterised and then a similar workload generated.

Original version is always available here at

Share this post

One thought on “Linux Disk Scheduler Benchmarking

  1. The BFS (Brain Fuck Scheduler) is hot recently, which is still under heavy development. Keep a note if you’re interested in this topic.

Post Comment