The asynchronous system model is widely used as a programming model mainly because of its ability to model most existing systems. Also, programming to this model is easy and results in portable applications due to its weak assumptions. However, many important practical problems are not solved in this model (for example, consensus). To circumvent this limitation, researchers have added synchronism assumptions to the asynchronous model. These assumptions may be added in the time domain (i.e. the system sometimes behave synchronously) or in the space domain (i.e. there is some portion of the system which always behaves synchronously). In this work, we take the space based approach and equip asynchronous systems with a small synchronous subsystem (i.e. a Wormhole). We then use good engineering to support our assumptions in this portion of the system based on cheap off-the-shelf components such as Switched Ethernet networks, message prioritization and built-in hardware clocks. Finally, we discuss how some services, like perfect failure detection, can be implemented in such subsystem and propose safety mechanisms to be applied when the synchronism assumptions do not hold.