On 22 Sep 2017, at 13:24, David Farmer <farmer@umn.edu> wrote:
On Fri, Sep 22, 2017 at 5:04 AM, Tim Chown <tjc@ecs.soton.ac.uk> wrote: ... and ARIN are on a last /10 policy which sees applicants get a /28 to a /24, so presumably those /28’s are routed at some level; that’s been in place for some time, how is it working out? ...
This ARIN policy is in section 4.10 of ARIN's NRPM;
Hello WG, To shed some more light on routability of prefixes longer than /24, we started an experiment when the aforementioned policy change was being discussed at ARIN. From October 2014, we started announcing 6 prefixes from ARIN’s designated block 23.128/10. There are /24, /25 and /28 prefixes, with and without IRR objects. We then measured visibility and reachability of those prefixes both on the control plane (by looking at BGP) and on the data plane (using RIPE Atlas trace routes). You can find the original paper by Emile Aben at https://labs.ripe.net/Members/emileaben/propagation-of-longer-than-24-ipv4-p... as well as a followup in 2015 https://labs.ripe.net/Members/emileaben/has-the-routability-of-longer-than-2... and finally, a more recent followup by Stephen Strowes on 10th of July 2017: https://labs.ripe.net/Members/stephen_strowes/bgp-even-more-specifics-in-201... All the best, Kaveh. ——— Kaveh Ranjbar, Chief Information Officer, RIPE NCC.