twest I still haven't done this test, but it was already on my list. This is the last method I know that should work to resolve this problem.
But I intend to know if there would be any easier way to resolve this, or maybe the Enhance Team could notice this and prepare a solution in future updates. @Adam
This could be a very practical situation: Let's say I have two web server roles in my cluster, one dedicated to openlitespeed clients, and the other for clients who need nginx. Then one day, the server with openlitespeed goes down suddenly, and I move all the backups of its websites onto the nginx server. At that time, I can't change the web server because there are still other active nginx websites running.