[arin-ppml] Proposal - Remove Initial Small Assignment Requirements for IPv6

[hostmaster at uneedus.com]

I forgot to mention that I am OPPOSED to this.

I see the router slots in the DFZ of IPv6 as the limiting factor, and we 
should not hand out PI addresses to those that are not actively using them 
to multihome on larger networks.  I agree with leaving the current 
standards that require a minimum number of IPv6 addresses in order to 
receive PI addresses.

By default, most providers give out more addresses than anyone needs.

Albert Erdmann
Network Administrator
Paradise On Line Inc.


On Mon, 13 Sep 2021, Larry R. Dockery wrote:

> 
> Thank you for your time in providing this information. This is the best argument against the proposal.
> 
>  
> 
> If this is a significant issue and holds true, then 1) most organizations do not and should not qualify for PI space. And 2) These orgs should, per the
> design recommendation in “IPv6 Address Planning”, use NTPv6 with ULA internally to avoid ISP lock-in inherent with IPV6 PA space.
> 
>  
> 
> This is far better than IPv4 NAT + RFC1918 in that it is stateless, but is an unfortunate workaround that somewhat inhibits the end-to-end principle.
> 
>  
> 
> That would be an unfortunate end-state for IPv6 that most SMB’s are still behind NAT, but may be the best technical way forward.
> 
>  
> 
>  
> 
> From: David Farmer <farmer at umn.edu>
> Sent: Monday, September 13, 2021 10:32 AM
> To: Larry R. Dockery <lrdocker at co.douglas.or.us>
> Cc: arin-ppml at arin.net
> Subject: Re: [arin-ppml] Proposal - Remove Initial Small Assignment Requirements for IPv6
> 
>  
> 
> The intent behind section 6.5.8.1 is not to conserve IPv6 address space but to conserve slots in the IPv6 route table, AKA the default-free zone. The
> abundance of /48s and /44s, or /40s, /36s, /32s for that matter, are irrelevant, there is only a finite number of routing slots available. BGP multihomed
> end-users will use a routing slot regardless of the source of the address space they use, so it is best if it comes from an RIR. However, single-homed
> end-users can be aggregated by their provider, yes this comes at a cost of renumbering for those end-users, but eliminating renumbering for those end-users
> comes at a cost of an IPv6 routing slot for the entire Internet. Therefore the cost of renumbering born by the end-user has to be balanced against the cost
> of a routing slot born by the entire Internet.
> 
>  
> 
> The IPv6 route table is currently growing quite quickly, see the following;
> 
> https://blog.apnic.net/2021/03/03/what-will-happen-when-the-routing-table-hits-1024k/
> 
> https://blog.apnic.net/2021/01/05/bgp-in-2020-the-bgp-table/
> 
> https://blog.apnic.net/2020/01/14/bgp-in-2019-the-bgp-table/
> 
> https://labs.ripe.net/author/stephen_strowes/visibility-of-ipv4-and-ipv6-prefix-lengths-in-2019/
> 
> https://bgp.potaroo.net/v6/as2.0/index.html
> 
> https://www.space.net/~gert/RIPE/weekly/2021/37/
> 
>  
> 
> The current 6.5.8.1.c was adapted from the IPv4 requirements when Draft Policy 2010-8: Rework of IPv6 assignment criteria was adopted.  At that time IPv4
> slow-start was still in effect, and there was still an IPv4 free pool. When the IPv4 transfer market became the primary source of IPv4 address space,
> slow-start was no longer practical or functional, and the initial allocation for IPv4 was changed. However, for IPv4 there is now the additional cost of
> obtaining the IPv4 block on the transfer market which somewhat offsets the removal to slow-start at least to some extent.
> 
>  
> 
> So, while I do not support the wholesale removal of section 6.5.8.1, I would support relaxing, possibly significantly relaxing, or otherwise
> modifying 6.5.8.1.c-e which are the current technical qualification for non-multihomed end-users. Fundamentally, it is not practical to have every business
> that could afford to pay ARIN's fees to avoid renumbering and to receive an IPv6 routing slot. It is not even entirely clear, there will be sufficient IPv6
> routing slots for every end-user that is willing to BGP multi-home.
> 
>  
> 
> Therefore, I believe there needs some kind of technical criteria that a non-multihomed end-user needs to meet to qualify to receive a Provider Independent
> IPv6 Allocation, and it needs to be more than just the ability to pay ARIN's fees.
> 
>  
> 
> And for clarity, I do not support the proposal as written.
> 
>  
> 
> Thanks.
> 
>  
> 
> On Mon, Sep 13, 2021 at 9:51 AM Larry R. Dockery <lrdocker at co.douglas.or.us> wrote:
>
>       https://www.arin.net/participate/policy/proposals/2021/ARIN_prop_301_orig/
>
>        
>
>       I would like to hear community feedback on this proposal. Thank you.
> 
> _______________________________________________
> 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.
> 
> 
>  
> 
> --
> 
> ===============================================
> David Farmer               Email:farmer at umn.edu
> Networking & Telecommunication Services
> Office of Information Technology
> University of Minnesota  
> 2218 University Ave SE        Phone: 612-626-0815
> Minneapolis, MN 55414-3029   Cell: 612-812-9952
> ===============================================
> 
> 
>