During the performance tests of 3.9, the results were different among developers. Dev team has to repeat a test of 3.8 in order to compare those versions because it was depended on the developer’s hardware.
In order to save time and make performance tests more reliable, we should improve the process of conducting them.
create a forum topic to ask the community
gather team experience from 3.9 performance tests
the output of this ticket should be updated wiki page