Hi Paolo,
We have a Voldemort Perf Lab we use internally to identify performance regressions and test improvements.
I think it does not support changing the load pattern midway through a test. Rather, we configure many different settings we wish to test (in terms of load characteristics and/or in terms of client/server configs) and VPL then goes through every desired permutations
for 30 minutes or more each time, and gathers and plots the data.
I don't know if our perf lab internally re-uses the open-source perf tool and just adds extra tooling and automation around it, or if it's completely different. Xu might be able to give more details.
Is there something specific you're interested in when looking at changing load patterns, versus just doing a few different (completely separate) tests?