Measurement and Monitoring Framework for the Evolving Internet
Nolan Rudolph
Committee: Ramakrishnan Durairajan
Honors Bachelor Thesis(Jun 2020)
Keywords: Networks, Measurement, Monitoring, Telemetry, Internet, eBPF

Granted the annual trends in increasing internet usage, the demand for scalable and high-performance networks continues to rise. Furthermore, as networks have evolved in new and profound ways, the tools and practices used to measure them have not always kept pace with their evolution. In light of both these events, this paper aims to provide a deeper insight on the measurement and monitoring frameworks for the evolving Internet and the design they should endow. We begin by producing a light-weight and scalable network flow to packet synthesizer designed to be a telemetry system for new measurement frameworks. Next, we examine the Extended Berkeley Packet Filter and its applicability to active and passive network measurements by designing an Express Data Path flow collector. Lastly, we present MicroMon, a multi-dimensional monitoring framework for geo-distributed applications using heterogeneous hardware and receive 10-50% throughput gains when applied to third party software.