How does Trace Explorer work?

Trace Explorer digs through the published list of highest-traffic domains from Quantcast and Alexa and a list of all links from the English Wikipedia to pick a few hundred thousand web sites that seem notable. Over the last two weeks, it has re-run a TCP traceroute to each, and built a searchable tree in a Mysql database out of them.


The Internet is so big that this tool can never be complete or up to date, and it can only show you Cyberverse's view of the Internet. It also makes a bunch of simplifying assumptions about how routing and DNS work, and may occasionally show inaccurate or impossible paths. But you might still find it interesting.

Turning a bunch of traceroutes into a tree discards a lot of detail. It also can't properly represent multilink paths or different paths for different routes that end up passing through the same router; it only ends up keeping the most popular path. But it is still surprisingly useful for looking at the edge of the network, where there's a lot less path diversity.