About the BGP peering sessions.
Hi, I am a new member on this list. I had a question regarding the nature of the BGP peering sessions of the RRCs. Is there any place I can find accurate information about the way the different RRCs maintain their peering sessions; as in how many of the sessions are multi-hop BGP sessions and how many are direct link sessions etc? I was wondering about this since, there has been a paper (see below) published in 2002 that said that most of the RRCs peering sessions were multi-hop eBGP sessions and as a result the BGP data collected here cannnot be an accurate refelction of what goes on the actual Internet/backbone routers. "Observation and Analysis of BGP Behavior under Stress", Proceedings of the Second ACM SIGCOMM Internet Measurement Workshop (IMW '02), pp. 183-195, Nov. 2002 Any inputs that I can get regarding this will be very helpful. Shivani.
Shivani Joshi wrote:
I had a question regarding the nature of the BGP peering sessions of the RRCs. Is there any place I can find accurate information about the way the different RRCs maintain their peering sessions; as in how many of the sessions are multi-hop BGP sessions and how many are direct link sessions etc?
Hi, apart from RRC00 in Amsterdam, which only has multihop peerings, all the RRCs in the RIS project are located at internet exchange points and are directly connected to their peers over Ethernet.
I was wondering about this since, there has been a paper (see below) published in 2002 that said that most of the RRCs peering sessions were multi-hop eBGP sessions and as a result the BGP data collected here cannnot be an accurate refelction of what goes on the actual Internet/backbone routers.
Actually, if you look at the peering statistics available on the DB status page: http://www.ris.ripe.net/cgi-bin/rrcstatus.cgi you will see that the multihop sessions on RRC00 only make up a small minority of RIS peerings. Cheers, Lorenzo
I was wondering about this since, there has been a paper (see below) published in 2002 that said that most of the RRCs peering sessions were multi-hop eBGP sessions and as a result the BGP data collected here cannnot be an accurate refelction of what goes on the actual Internet/backbone routers.
Actually, if you look at the peering statistics available on the DB status page:
http://www.ris.ripe.net/cgi-bin/rrcstatus.cgi
you will see that the multihop sessions on RRC00 only make up a small minority of RIS peerings.
but, in 2001-2, was this the case? randy
you will see that the multihop sessions on RRC00 only make up a small minority of RIS peerings.
but, in 2001-2, was this the case?
No, when we still had only RRC00 (back in 1999), all but 1 session were multihop. So, both the statement in this old paper and Lorenzo's observation are correct. Henk ------------------------------------------------------------------------------ Henk Uijterwaal Email: henk.uijterwaal(at)ripe.net RIPE Network Coordination Centre http://www.amsterdamned.org/~henk P.O.Box 10096 Singel 258 Phone: +31.20.5354414 1001 EB Amsterdam 1016 AB Amsterdam Fax: +31.20.5354445 The Netherlands The Netherlands Mobile: +31.6.55861746 ------------------------------------------------------------------------------ Look here junior, don't you be so happy. And for Heaven's sake, don't you be so sad. (Tom Verlaine)
participants (4)
-
Henk Uijterwaal -
Lorenzo Colitti -
Randy Bush -
Shivani Joshi