Hi, I just had a conference call with Samantha Lo (student) and Rocky Chang (associate professor) from Hong Kong Polytechnic on the prepending research. This email is a status update on what we are doing and what we want to do. Comments/suggestions welcome. Current status -------------- There seems to be good communication and agreement on where to go. Today we discussed which RRCs to announce the test prefixes from. We decided that RRC07 (Stockholm), RRC10 (Milan), and RRC14 (Palo Alto) are the best candidates because they have a realistic number of upstreams and a more balanced AS graph (shown in attached PDF file). We obtain BGP data from the RIS (using my scripts) and from ORV and looking glasses (using scripts by Samantha). In total we have about 150 views of the Internet. This might further be enriched by TTM, but I don't know if I'll have time to implement that. Goals ----- - Submit a paper to IMC 2006 conference (deadline end of May). Author list might be something like Lo, Chang, and Colitti - They also want to present this at nanog in June Proposed setup -------------- Everything will be in the NCC network and under our control: - We have a machine here that peers with the 2 or 3 RRCs involved - The machine runs the Python BGP announcer I wrote and announces communities to the RRCs - The RRCs announce varying degrees of prepending to our transit peers based on the communities they receive. A route-map will be used for this purpose. Next steps ---------- - Samantha will set up a test-lab, write the route-maps and test the setup (this week) - Once we know how long that is going to take, we inform our peers that we will be doing the experiments. - Target: begin experiments soon after RIPE meeting Cheers, Lorenzo -- Lorenzo Colitti lorenzo@ripe.net Network Engineer +31-20-5354471 RIPE NCC www.ripe.net