[arin-ppml] Opposed to 2010-9 and 2010-12

Owen DeLong owen at delong.com
Thu Oct 7 16:21:26 EDT 2010


On Oct 7, 2010, at 1:12 PM, Christopher Morrow wrote:

> On Thu, Oct 7, 2010 at 3:59 PM, Owen DeLong <owen at delong.com> wrote:
>> Probably a /24. That allows a /56 for end-sites which is suboptimal
>> (end sites should be at least a /48), but, hopefully doesn't consume
>> too vast a swath of IPv6 in the process (roughly a  /8).
> 
> can't we let the ISP decide what makes sense? it seems (to me) that a
> /48 for a business-type link (your traditional T1/T3 customer type,
> and office, etc.) is perfectly rational. It seems, to me, that a /56
> for a consumer (dsl/cable/etc) is also quite fine.
> 
If the ISP wants to give /48s out in 6rd, that's a /16 for each ISP.
I am opposed to giving out IPv6 /16s to ISPs for 6rd. I think that
is really really bad stewardship of the address space.

> There are, I'm sure, ISP folks who'd decide to just assign a /48
> across the board... I'm not sure that guidance (aside from general
> scoping) is required from ARIN to the members/users.
> 
/48 across the board for native IPv6 is fine. That's not the
topic of this discussion. The topic of this discussion is the
upper bound for how much IPv6 space we would give to an ISP
to enable 6rd. Giving /20s or /16s strikes me as a particularly
bad idea.

Owen




More information about the ARIN-PPML mailing list