Communication mechanisms are most of the time implemented efficiently and therefore lead to good performances as long as the data transfer is done in an ``undisturbed environment'', i.e. the resources (CPU, memory bus, network interface) are fully available. On the other hand, communication reactivity during computing phases, i.e. the ability to react quickly to the network, is hard to ensure. However, by centralizing the network accesses, it is possible to maintain a good reactivity even during computing phases. In that context, we describe a new server, interacting with both a communication library and a thread scheduler, able to provide a good reactivity whatever the execution context is.