Tracking Network Flows with P4

Open Access
Authors
Publication date 2018
Book title Proceedings of INDIS 2018: Innovating the Network for Data-Intensive Science
Book subtitle held in conjunction with SC18: The International Conference for High Performance Computing, Networking, Storage and Analysis, Dallas, Texas, November 11-16, 2018
ISBN
  • 9781728101958
ISBN (electronic)
  • 9781728101941
Event 2018 IEEE/ACM Innovating the Network for Data-Intensive Science
Pages (from-to) 23-32
Number of pages 10
Publisher Los Alamitos, California: IEEE Computer Society
Organisations
  • Faculty of Science (FNWI) - Informatics Institute (IVI)
Abstract
Tracking flows within a single device, as well as tracking the full path a flow takes in a network are core components in securing networks. Malicious traffic can be easily identified and its source blocked. Traditional methods have per- formance and precision shortcomings, while new programmable devices open up new possibilities. In this paper we present methods based on the P4 programming language that allow to track flows in a device, as well methods toward full path tracking. A core component of this work are Bloom filters, which we have implemented fully in P4. To validate our approach and implementation we have carried a study in a specific use case, namely the detection of SYN attacks.
Document type Conference contribution
Language English
Published at https://doi.org/10.1109/INDIS.2018.00006
Other links https://www.scopus.com/pages/publications/85063336369
Downloads
Tracking Network Flows with P4 (Final published version)
Permalink to this page
Back