[arin-ppml] IPv6 policy: subsequent allocations?

[Chris Woodfield]

Speaking from personal experience, I’ve seen situations like this:

1. An organization knows they’ll need to adopt IPv6 eventually, and goes ahead and requests/receives a block without much thought to their allocation plan (/32 should be enough for anyone, right?)

2. Some time later, the org moves forward and starts building an address plan, and in the process realizes that given best practices (/48 minimum to end sites, aggregation on nibble boundaries, etc) that a /32 isn’t enough space for them.

3. Rightly or wrongly, the org expects they won’t qualify for a subsequent allocation under 6.5.3, so they break best practices to make do with that initial block. Or worse, they abandon the effort entirely.

So, I feel like it would be reasonable for policy to allow for an organization to have a “redo” of their initial allocation request. One condition that may make sense is that this be on a request-then-return basis, although that might be moot if original sparse allocation is still sparse enough.

-C

> On Jun 25, 2026, at 15:16, William Herrin <bill at herrin.us> wrote:
> 
> Howdy,
> 
> I didn't see any feedback on the draft policy rewriting section 6.5,
> so I want to step back and solicit your opinions on what ARIN's IPv6
> policies should become. I'm going to ask some questions and break them
> into separate message threads so that they can be followed separately
> according to your interest.
> 
> 
> The question for this thread is: How should ARIN handle second and
> subsequent allocations of IPv6 addresses?
> 
> Current policy says that your IPv6 addresses have to be employed and
> in use according to somewhat complicated definitions of in-use in
> order to qualify for additional addresses. ARIN will then try to
> expand your netmask. If they can't because someone else's allocation
> is in the way, they'll allocate a new, larger block and ask (but not
> require) you to renumber out of the old one.
> 
> This approach has a couple of odd artifacts. The amount of addresses
> you're qualified for under the subsequent allocation criteria aren't
> exactly the same as what you qualified for under the initial
> allocation criteria.
> 
> A different approach could be to have a single rule set to determine
> how many addresses you're qualified for. If you want more addresses,
> you apply the same rules you did originally to the current conditions
> on your network. This would avoid inconsistencies but it would also
> get rid of the "prove you're using IPv6 addresses efficiently"
> requirement.
> 
> Given that IPv6 addresses are plentiful and if we don't do anything
> too silly they'll remain plentiful well into the foreseeable future,
> do we need registrants to prove efficient use before resizing their
> address allocation? That's a lot of paperwork for the registrant to
> prepare and the analyst to examine, all of it blocking whatever
> improvement the registrant wants to make to their IPv6 network. We
> need it for IPv4 because folks have a financial motive to lie, but is
> that true of IPv6?
> 
> Anyway, what do you think? Continue to treat initial and subsequent
> IPv6 allocations differently? Try to merge their criteria into
> something more uniform? Your views are respectfully requested.
> 
> Regards,
> Bill Herrin
> _______________________________________________
> ARIN-PPML
> You are receiving this message because you are subscribed to
> the ARIN Public Policy Mailing List (ARIN-PPML at arin.net).
> Unsubscribe or manage your mailing list subscription at:
> https://lists.arin.net/mailman/listinfo/arin-ppml
> Please contact info at arin.net if you experience any issues.