Network Survivability Simulation of a Commercially Deployed Dynamic Routing System Protocol

With the ever-increasing demands on server applications, many new server services are distributed in nature. We evaluated one hundred deployed systems and found that over a one-year period, thirteen percent of the hardware failures were network related. To provide end-user services, the server clusters must guarantee server-to-server communication in the presence of network failures. In prior work, we described a protocol to provide proactive dynamic routing for server clusters architectures. We now present a network survivability simulation of the Dynamic Routing System (DRS) protocol. We show that with the DRS the probability of success for server-to-server communication converges to 1 as N grows for a fixed number of failures. The DRS’s proactive routing policy performs better than traditional routing systems by fixing network problems before they effect application communication.