Analysis of the Internet Routing Registry
A method and an on-line service to automatically extract peering information from the IRR.
The Internet Routing Registry (IRR) is a large distributed repository of information, containing the routing policies of many of the networks that make up the Internet. It consists of several registries that are maintained on a voluntary basis. The routing policies are expressed in the Routing Policy Specification Language (RPSL). The IRR can be used by operators to look up peering agreements, to study optimal policies, and to (possibly automatically) configure routers.
The IRR is a unique and significant source of information to understand the Internet routing. However, since it is maintained on a voluntary basis, information contained inside it may be inconsistent or outdated. Moreover, routing policies can be expressed in RPSL by using complex constructions, and the level of accuracy of the descriptions may vary. All these factors make the task of automatically extracting useful information from the IRR a difficult one.
Within this project, we thoroughly analyze IRR data in order to extract BGP peering relationships from the registries. We believe that such information is useful both to take advantage of an alternative, possibly richer, source of topological information, and to assess the correspondence between registry information and actual routing data.
We obtained the following results: