Julian Elischer <julian_at_elischer.org> writes: > I wish people would stop trying to deny that we have serious work in > front of us to get the VFS and disk IO figures back to where they were > before. A great way to help getting this fixed would be to write automated benchmarks and regression tests, or at least solid test protocols. The documentation for these benchmarks and tests should clearly explain which (part of a) real-life workload they are intended to simulate. These benchmarks and tests would serve a double purpose: as hard evidence to back your arguments in discussions about performance regressions, and as tools for developers who wish to do something about such regressions. DES -- Dag-Erling Smørgrav - des_at_des.noReceived on Mon Apr 25 2005 - 08:04:42 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:32 UTC