John Sonchack, University of Pennsylvania; Oliver Michel, University of Colorado Boulder; Adam J. Aviv, United States Naval Academy; Eric Keller, University of Colorado Boulder; Jonathan M. Smith, University of Pennsylvania
Measurement plays a key role in network operation and management. An important but unaddressed practical requirement in high speed networks is supporting concurrent applications with diverse and potentially dynamic measurement objectives. We introduce Flow, a switch accelerated telemetry system for efficient, concurrent, and dynamic measurement. The design insight is to carefully partition processing between switch ASICs and application software. In Flow, the switch ASIC implements a pipeline that exports telemetry data in a flexible format that allows applications to efficiently compute many different statistics. Applications can operate concurrently and dynamically on identical streams without impacting each other. We implement *Flow as a line rate P4 program for a 3.2 Tb/s commodity switch and evaluate it with four example monitoring applications. The applications can operate concurrently and dynamically, while scaling to measure terabit rate traffic with a single commodity server.
Open Access Media
USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are posted after the event are also free and open to everyone. Support USENIX and our commitment to Open Access.
author = {John Sonchack and Oliver Michel and Adam J. Aviv and Eric Keller and Jonathan M. Smith},
title = {Scaling Hardware Accelerated Network Monitoring to Concurrent and Dynamic Queries With {*Flow}},
booktitle = {2018 USENIX Annual Technical Conference (USENIX ATC 18)},
year = {2018},
isbn = {978-1-939133-01-4},
address = {Boston, MA},
pages = {823--835},
url = {https://www.usenix.org/conference/atc18/presentation/sonchack},
publisher = {USENIX Association},
month = jul
}