[address-policy-wg] 2011-04, "Extension of the Minimum Size for IPv6 Initial Allocation"

On 14 November 2011 16:18, Tom Hodgson <tom at someaddress.net> wrote:
> One thing that I do wonder is whether the (limited) overhead of the NCC
> processing these requests mean it is more desirable to limit this
> "re-request" to a single shot per-LIR (which will likely push people to
> request the whole /29), or whether permitting multiple requests grabbing a
> /32 at a time (up to the /29) is desirable. Personally I assume that most
> networks using this policy extension would go for the /29 straight off, but
> then there maybe further interaction with the charging scheme which
> dissuades this.

The logic was to try and remove the guesswork from the process. For
someone adjudicating the request its a fairly simple binary logic (in
fact its almost programable) and therefore (I assume) a limited amount
of impact. It also gets round the "oh bugger" moment where someone who
could have gone for a /29 but only went for a /30 because their maths
was wrong can't get the space, even though its effectively reserved
for their future use, due to the HD rule.

As I said if we can deal with the HD rule then I think we can maybe revisit this

J
-- 

James Blessing
07989 039 476