[arin-discuss] /29 limit for ARIN SWIP whois

This discussion needs to move to the ppml as it concerns a policy and its merits and rationale.

Ray

> -----Original Message-----
> From: arin-discuss-bounces at arin.net [mailto:arin-discuss-
> bounces at arin.net] On Behalf Of Leo Bicknell
> Sent: Monday, January 07, 2008 10:14 PM
> To: arin-discuss at arin.net
> Subject: Re: [arin-discuss] /29 limit for ARIN SWIP whois
>
> In a message written on Sun, Jan 06, 2008 at 10:23:23AM -0500, Joe
> Maimon wrote:
> > Is there any overriding reason to limit ARIN swip to /29 or bigger?
>
> I will point out (in IPv4):
>
> /32 assignment (e.g. dial up, DSL, etc) is by definition 100% utilized.
>
> In terms of subnets, which only make sense if you have two more more
> devices (router + one or more hosts):
>
> /31 subnet by definition is 100% used.
> /30 subnet by definition is 100% used (router, host, network,
> broadcast).
> /29 subnet is at minimum 50% used (router, host, network, broadcast).
>     If we further assume this was done because a /30 was not large
>     enough (e.g. people are doing the right thing) there must be at
>     least 5/8's, or 62.5% in use.  Also, while the standard may be 80%
>     utilization, which would require 7 of the 8 IP's to be in use;
>     that leaves an interesting corner case where 5/8 and 6/8 can't
>     fit in a /30, but don't meet 80%.  Thus it makes sense to count
>     5/8 and 6/8 as fully utilized, making it all but impossible to
>     have an underutilized /29.
>
> Now, one of ARIN's primary uses for the data is to insure assignments
> were made in accordance with ARIN's rules when someone requests more
> space.  There's no reason to review a /30, /31, or 32, as there's no
> chance those assignments were under-utilized.
>
> --
>        Leo Bicknell - bicknell at ufp.org - CCIE 3440
>         PGP keys at http://www.ufp.org/~bicknell/ Read TMBG List -
> tmbg-list-request at tmbg.org, www.tmbg.org