This thesis investigates constructing test-bed environments suitable for testing enterprise software systems. Test-beds act as a kind of laboratory in which the quality of software systems can be studied. For credibility, it is important that laboratory conditions be as similar as possible to the real world. In the enterprise context, this means that test-beds must mimic interaction behaviour of tens-of-thousands of individual systems. Cameron's thesis introduces a method for creating suchlarge-scale test-beds. By modelling real system interactions and emulating those models test-beds are provided which mimic the behaviour and scale of real enterprise environments. This work enables a class of scalability and robustness testing not previously practical and ultimately enables development of more reliable software systems.