[153302] in North American Network Operators' Group
IPv6 evolution
daemon@ATHENA.MIT.EDU (Matthew Luckie)
Mon Jun 4 18:49:41 2012
Date: Mon, 04 Jun 2012 15:49:01 -0700
From: Matthew Luckie <mjl@caida.org>
To: nanog@nanog.org
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
IPv6 paths that are the same as an IPv4-level path are correlated with
better IPv6 performance according to: "Assessing IPv6 Through Web
Access - A Measurement Study and Its Findings"
http://repository.upenn.edu/ese_papers/602/
At the Feb NANOG I gave a lightning talk on trends involving dual-stack
ASes, beginning with the fraction of AS-level paths that are the same in
IPv4 and IPv6. As of Jan 2012, 40-50% of AS-level paths are the same in
v4 and v6 for dual-stacked origin ASes.
http://www.nanog.org/meetings/nanog54/presentations/Monday/Luckie_LT.pdf
We've looked further into the evolution of IPv6 since then. In
particular we looked at "what could be". We find that 95% of AS-level
paths could be the same today: where an IPv4 path contains ASes that are
not found in an IPv6 path to the same dual-stacked origin, we check to
see if the AS is observed in an IPv6 BGP path, and thus the AS is IPv6
capable at a minimum. A brief blog post on this is at:
http://blog.caida.org/best_available_data/2012/06/04/ipv6-what-could-be-but-isnt-yet/
Matthew
