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

Owen DeLong owen at delong.com
Thu Oct 7 16:27:51 EDT 2010

On Oct 7, 2010, at 1:20 PM, Joe Maimon wrote:

> Owen DeLong 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).
> Does it provide enough space only for whom a /32 for native was enough or for all?
By definition a /24 is enough for any ISP to do /56s for all of their IPv4 customers because
there are only 32 bits in IPv4. Since 6rd contains a mapping of the IPv4 /32 into the
IPv6 address, this is pretty basic.

Perhaps I don't understand your question.

> Would ARIN need to get another /12 to be able to address this from a distinct prefix? Would we expect all the other RiR's to do the same, and could IANA be kind enough to do it on a /8?
Ideally I would like to see IANA set aside a block for this, such as 3f00/8. The RIRs might need more
than one /12 each to do this if they are giving out /24s.

> Could we expect all and sundry to deploy 6rd under these guidelines and how many bits would that actually consume?
Don't know how many would deploy it. However, if EVERYONE using IPv4 today deployed in this manner and received a /24, it would consume a total (worldwide) of a bit less than a /8 (~/10+).

> And all even before any real use by real people.
Not sure I understand this statement. That amount of space would facilitate 6rd to every existing customer of every ISP currently receiving IPv4 services and all future IPv4 deployments as well.


More information about the ARIN-PPML mailing list