I have been measuring some method execution times in a docker libra swarm setup.
What I found was that methods(such as process proposal) take longer to execute as the #nodes in the network grows. I am making sure that blocks are fully saturated. There are two possibilities: 1. Some method execution times strongly depend on the #nodes in the network. 2. As the #nodes in the network goes up, so does the required CPU power. Due to resource sharing methods take longer to execute.
I expect that there is no dependency on the #nodes in the network and that resource sharing is to blame.