The documents distributed by this server have been provided by the contributing authors as a means to ensure timely dissemination of scholarly and technical work on a non-commercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, not withstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.
|Author:||Ralf Kundel, Jeremias Blendin, Tobias Viernickel, Boris Koldehofe, Ralf Steinmetz|
|Kind:||In proceedings - use for conference & workshop papers|
|Book title:||2018 IEEE Conference on Network Function Virtualization and Software Defined Networks (NFV-SDN)|
|Keywords:||P4, Active Queue Management, Bufferbloat, CoDel, Low Latency, Queueing, Traffic Shaping|
|Research Area(s):||Network Mechanisms & QoS|
|Abstract:||Today, networks are still vulnerable to high laten- cies. An important reason for that is the Bufferbloat problem caused by big packet buffers as part of the forwarding equipment of modern networks. Even if these buffer sizes are configured in a reasonable way, they still have a fixed size which is always a compromise. Indeed, the required buffer size strongly depends on the RTT of the end-to-end network connection. In order to support dynamic buffer sizes and to reduce the negative impact of Bufferbloat, different Active Queue Management (AQM) algorithms have been proposed recently, such as CoDel and PIE. However, these algorithms are typically not available in state of the art network equipment. In this paper we show how recent AQM algorithms can be implemented using P4 programmable network equipment. In consequence, communication networks can be easily enhanced to benefit from state-of-the-art AQM algorithms. To this end, we provide an open-source available implementation of CoDel, one of the most prominent queue management algorithms, in P4. The implementation of such AQMs in P4 data plane hardware enables a massive reduction of latency in many use cases, such as traffic shaping in ISP access networks.|
|Full paper (pdf)|