Network Fault Detection and Recovery in the Chaos Router

摘要: Chaotic routing, which allows packets to follow non-minimal routes, provides a basic level of fault-tolerance by allowing messages be routed around faults without requiring priori knowledge their locations. However, the mechanisms for doing this can slow and clumsy at times. We augment routing with limited amount hardware support fault, detection, identification, reconfiguration so that network automatically reconfigure itself when occur. present high-level design these mechanisms, driven goal achieving reasonable reliability exorbitant cost.