Soldato
True but it means that they need to analyse that data and see what is needed to fix it.See that is the thing, before RMQ that 5-10 seconds was greater even then. They are still working and developing RMQ so that test was find out where, why and how it falls over. The fact that they went okay we have data at 500 player and can see where bottlenecks are so lets push to the 1000 and see what else that brings. That is literally the point of the test.
No doubt there are issues of scaling and using SM whilst in RMQ but saying it doesn't work and back to the drawing board isn't really what it means either.
It could be a one line change in the code or it could be a complete redoing of the RMQ ie back to the drawing board.
Dont forget that the last SM test, they gathered the data and based on that data HAD to redo the whole messaging que system!