On Fri, Jan 2, 2009 at 5:15 PM, William Herrin <[email protected]> wrote: > On Fri, Jan 2, 2009 at 4:58 PM, Christopher Morrow > <[email protected]> wrote: >> On Fri, Jan 2, 2009 at 3:19 PM, William Herrin <[email protected]> wrote: >>> ARIN policy requires multihomed users to demonstrate a need to deploy >>> 500 machines on the Internet in order to qualify for end-use IPv4 >>> address assignments. This number was selected above the operations >>> minimum of 125 addresses based on the IETF and NANOG advice that their >>> assistance is required to conserve routing slots so as not to overload >>> the BGP system. >> >> I believe actually if you can demonstrate a 'unique routing policy' >> you can at least get a /24 from one provider, and an ASN from >> ARIN/RIPE/APNIC and enjoy a routing slot in the DFZ all your own. > > Hi Chris, > > Correct. Your ISP can justify giving you a /24 in their next request > if you request the /24 for multihoming purposes. And ARIN will then > give you an AS#. However, that /24 will often come from a /8 where > ARIN has declared the minimum allocation size to be /20. Though
that really depends on the ISP, their allocations and the person(s) doing the allocation at the time...There was no distinction like that where I was recently doing this. > fortunately few, some carriers treat all /24 cutouts in such blocks as > traffic engineering and filter them out. This results in an incomplete sure, there are examples of this, mostly those folks are limiting prefixes for financial reasons (don't want to upgrade 7500's or ...) > multihoming solution which still consumes a slot in most folks' > routing table. also true. My point, and I realized as I hit send I was splitting hairs, was that routing slot allocations come from a range of places, just noting that RIR's don't allocation without (hopefully) just cause isn't the only way a slot is won. Also, one point that I feel needs to be addressed is that internal route tables are most often quite a bit larger than the globally visible table. Ted and I both made this point at the RAWS meeting in AMS (and it's in the slides I pointed to previously). That growth is not as easy to quantify from the outside, unfortunately, and also not well documented. -Chris _______________________________________________ rrg mailing list [email protected] https://www.irtf.org/mailman/listinfo/rrg
