Reverse Traceroute

The internet is important to all of us. Albeit its importance, it is quite difficult to pinpoint problems or gather information about it in case the infrastructure is not under ones own control. For the rest of us, there are only two tools available that usually work on the public internet, which are ping and traceroute, which allow us to collect some information about end points on the internet.

ping is quite limited in what it can actually tell us. An answer to an echo request allows us to deduct information about interface reachability and the tool gives us a round trip time measurement for said interface.

traceroute gives us a lot more information. What you get out of traceroute is a list of routers between the host issuing the traceroute and the target host on the internet. For each router and the target host itself you also get round trip time measurements. The list of router however only covers the forward path. That might seem irrelevant at first, but the internet is highly asymmetric, which means, more often than not the forward path and the reverse path differ. So in case a potential problem is on the reverse path, traceroute is not helping all that much to identify it correctly.

The goal of our work is to define a new protocol, or more precisely protocol extentions, and to implement related tools that allow us to measure the reverse path between end points on the internet: reverse traceroute.

General overview (Talk at DENOG14, 15.11.2022)

Elevator pitch (Talk at IETF115, HotRFC session, 06.11.2022)

Links

1. Code on Github
2. Internet Draft

Timeline

1. 05.09.2022 - Published the 00 version of our Internet draft
2. 06.11.2022 - Presented reverse traceroute at the HotRFC session at IETF 115
3. 15.11.2022 - Presented reverse traceroute at DENOG14