We pick ports at the beginning of the MCSFinder simulation, but not during intermediate replays. This could interfere with (long-running) experiments, where another process could swoop in and take the port we chose at the beginning. This would then cause the MCS results to be highly inflated.
We pick ports at the beginning of the MCSFinder simulation, but not during intermediate replays. This could interfere with (long-running) experiments, where another process could swoop in and take the port we chose at the beginning. This would then cause the MCS results to be highly inflated.