Generated by GPT-5-mini| Paris Traceroute | |
|---|---|
| Name | Paris Traceroute |
| Developer | Olivier Bonaventure, INRIA, University of Louvain |
| Released | 2006 |
| Operating system | Unix-like |
| License | Open source |
Paris Traceroute is a network measurement tool designed to improve path discovery and topology inference in the presence of load balancing and routing asymmetry. Developed in academic and research contexts, it addresses limitations of classic traceroute by issuing flow-consistent probes and combining techniques from packet measurement, active probing, and topology analysis.
Paris Traceroute was created to reveal Internet path diversity and per-flow load balancing behavior by researchers affiliated with INRIA, Université catholique de Louvain, and collaborators from institutions such as Université Pierre et Marie Curie, Columbia University, ETH Zurich, Princeton University, University of California, Berkeley, Massachusetts Institute of Technology, Stanford University, Harvard University, University of Cambridge, University of Oxford, University of Maryland, College Park, University of Illinois Urbana-Champaign, Carnegie Mellon University, University of Washington, University of Toronto, National Institute of Standards and Technology, Cisco Systems, Juniper Networks, Google, Amazon (company), Microsoft, Facebook, Akamai Technologies, RIPE NCC, APNIC, ARIN, IETF, ITU, European Commission, National Science Foundation, DARPA, Cisco Research, Bell Labs, and Bellcore. The tool informed studies published in conferences like ACM SIGCOMM, USENIX NSDI, IEEE INFOCOM, and ACM IMC and journals such as ACM Transactions on Internet Technology and IEEE/ACM Transactions on Networking.
Paris Traceroute's methodology centers on maintaining consistent flow identifiers to avoid per-flow load balancing artifacts observed with traditional traceroute implementations from projects like Van Jacobson's traceroute and tools used by MCI Communications and Sprint Corporation. It leverages concepts from packet header manipulation used in research at IETF working groups and techniques similar to those developed in Scamper, Pathneck, hping, and nmap. The design draws on routing research from work by Vern Paxson, Lixia Zhang, Jennifer Rexford, Randy Bush, Geoff Huston, Mark Allman, Kannan Varadhan, Ramesh Govindan, Deborah Estrin, Scott Shenker, Kevin Fall, Sally Floyd, Van Jacobson, Peter Kirstein, Jon Postel, Radia Perlman, Barbara Liskov, Tim Berners-Lee, Vinton Cerf, Robert Kahn, Paul Mockapetris, Stephen Deering, David Clark, Arpanet pioneers, and others who influenced Internet measurement principles.
The implementation is a C-language userspace program for Unix-like systems integrating packet crafting and timestamping similar to tools used by tcpdump, Wireshark, libpcap, PF_RING, Netmap, DPDK, and Linux Kernel packet APIs. It interoperates with measurement infrastructures such as PlanetLab, RIPE Atlas, perfSONAR, CAIDA Ark, MAWI, Open Observatory of Network Interference, Netalyzr, SamKnows, Measurement Lab, and iPlane. The tool produces outputs consumed by topology mapping projects like CAIDA, DIMES, Skitter, RouteViews, BGPmon, Hurricane Electric, Team Cymru, Route Views Project, and visualization utilities used in Gephi and Graphviz.
Paris Traceroute has been used in research on interdomain routing, load balancing, and topology discovery for networks operated by AT&T, Deutsche Telekom, NTT Communications, Orange S.A., Telefónica, Vodafone Group, Verizon Communications, T-Mobile, China Telecom, China Unicom, KPN, Telstra, BT Group, Swisscom, Rogers Communications, Time Warner Cable, and content providers like Netflix, YouTube, Dropbox, Cloudflare, Fastly, and Akamai Technologies. Use cases include studying path asymmetry highlighted in studies by Laura Chappell, Nick Feamster, Arnaud Legout, Andrei Broder, Christos Papadopoulos, Lorenzo Colitti, Ethan Katz-Bassett, Matthew Luckie, Emile Aben, kc claffy, and Alessandro Finamore; validating routing across backbone links analyzed in Olivier Bonaventure's work; and informing routing policy discussions at IETF and regional registries like RIPE NCC and APNIC.
Critiques of Paris Traceroute focus on operational constraints in large-scale deployments, comparisons with measurement platforms such as RIPE Atlas and CAIDA Ark, and the challenges posed by middleboxes like firewalls and Network Address Translation appliances used by Cisco Systems and Juniper Networks devices. Researchers including David Levin, Nick Feamster, Bruce Maggs, Mark Allman, Bram Cohen, Dan Woods, Arvind Krishnamurthy, Nick McKeown, Scott Shenker, Jennifer Rexford, Randy Bush, and Geoff Huston have debated trade-offs between probe intrusiveness, measurement accuracy, and scalability in forums such as ACM SIGCOMM, USENIX, IEEE INFOCOM, ICNP, and IMC.
Evaluations reported in publications compared Paris Traceroute against legacy traceroute, Scamper, and active probing suites using datasets from CAIDA, RIPE NCC, PlanetLab, MAWI, Measurement Lab, and service provider telemetry. Metrics included path discovery rate, false-alarm incidence, measurement overhead, and ability to detect per-flow load balancing on links operated by providers like Level 3 Communications, Cogent Communications, GTT Communications, Zayo Group, and NTT Communications. Results demonstrated improved path stability in the face of per-flow load balancing and informed follow-on tools and standards adopted by communities around IETF and measurement platforms including RIPE Atlas and CAIDA Ark.
Category:Network measurement