Dual stack IPv6 for IPv4 depletion

Owen DeLong owen at delong.com
Fri Jul 10 16:12:50 UTC 2015


> On Jul 9, 2015, at 23:33 , Matthew Kaufman <matthew at matthew.at> wrote:
> 
> On 7/9/2015 3:07 PM, Owen DeLong wrote:
>> 
>> Can you offer one valid reason not to give residential users /48s? Any benefit whatsoever?
>> 
> 
> Sure. To avoid having to go back and deal with ARIN yet again for more IPv6 space.
> 
> One of the hopeful outcomes of IPv6 adoption was that an ISP could get enough to last "forever" in a single transaction. But "forever" isn't very long at one /48 (or more) per customer.
> 
> Matthew Kaufman

I don’t understand how that shortens forever if you ask for the right size block the first time.

I’ll be surprised if HE hands out enough /48s to empty their /24 any time short of something approximating forever. It’s been at least 3 years since I got that for them.

They’re definitely handing out a /48 per end site with the exception of free end-sites that don’t bother to click the “give me a /48” checkbox.

Getting IPv6 from ARIN is really easy. Getting more IPv6 from ARIN is really easy if you get anywhere near filling up your IPv6 block.

MUCH MUCH easier than IPv4.

As an example, I bet if they wanted to, Comcast could easily qualify for a /12 under current ARIN policy.

The latest figures I could find show them at just over 22.4 million broadband subscribers. Let’s assume they have 40 million just to be conservative.

If you packed those in, you could fit them in 3 /24s (actually, about 2.3 /24s technically). A /12 is 4096 /24s.

Please tell me again how you can’t hand out /48s per end-site forever with a single ARIN allocation?

Owen




More information about the NANOG mailing list