Edison: Event-driven Distributed System of Network Measurement

Network measurement is critical in network management such as performance monitoring, diagnosis, and traffic engineering. However, conventional network measurement software tools are inadequate in high-speed network at scale. They also lack an event-driven mechanism to automate complex network measurement process. In this paper, we present Edi-son which contains an high performance measurement backend (Ares) to collect flow metrics, driven with an expressive frontend (Equery) to enable the composition of complex measurement tasks. Finally, we evaluate the effectiveness of our Edison framework on a distributed measurement platform deployed for 100Gbps international research networks.