[arin-ppml] An article of interest to the community....

[Owen DeLong]

On Sep 1, 2011, at 8:16 AM, Chris Engel wrote:

> 
>>> 
>> 
>> Absolutely. I just think that Chris underestimates what it is going to cost him
>> to try and prolong the life of IPv4 and that as he begins to realize those costs,
>> IPv6 will become radically more attractive in terms of cost/benefit ratio.
>> Unfortunately, he probably will not realize this until he has wasted relatively
>> vast sums of money on expensive equipment to try and keep IPv4 on life
>> support when that money would have been better spent moving towards
>> IPv6.
>> 
>> Such is the short-term thinking of a market.
>> 
> 
> Owen, 
> 
> Every usage case will be different, but for the record, I can't fathom what those costs would be in our case. We don't actually need more public IPv4 address then we currently have and even if we did due to growth I can't imagine it would be likely to be more then low double digits. We already, for purposes that have nothing to do with address conservation, make pretty extensive use of technologies that allow for sharing of public address space (NAT/PAT, Load Balancing, Host-Headers, etc) and would want to preserve similar functionality even under unlimited IP address availability.
> 

In the case of your particular situation, they will be much farther down the road. For access providers, OTOH, the costs will be pretty massive as their networks continue to grow and they try to squeeze more and more customers into the same address space.

> Private IPv4 address space is free and already works with all our existing hardware/software and applications. We don't make use of Peer to Peer applications as decentralized management of technology doesn't make sense for our business model. In other words, even if there was some Peer to Peer application which we found useful we would want to route all of our usage of it through some centralized gateway on our end so that we can uniformly apply policy and logging one place. So we would WANT to have ALG's in place for any Peer to Peer apps we might use, REGARDLESS of the availability public addresses.
> 

Again, this is a very limited perspective. As a member of the Advisory Council and in my day job, I must consider a much broader range of circumstances than just your particular network.

The extreme costs of IPv4 at the access provider level WILL drive access customers to IPv6. Even to the extent that access customers are able to maintain some level of IPv4 capability, it will be significantly degraded and they will begin to prefer sites and services that can deliver to them on native IPv6.

So, the cost that will be coming to you as a content provider network will be that of losing customers to others that have implemented IPv6 as the access eye-balls find IPv4 to provide a more and more degraded user experience.

The cost to an enterprise, OTOH, of failing to deploy IPv6 will simply be that they become increasingly disconnected from the internet as a higher and higher percentage of the internet moves to IPv6.

The cost to both types of organizations if they implement IPv6 but still try to support IPv4 will, of course, simply be the cost of doing everything twice.

> So I'm really struggling to see what sort of costs remaining on IPv4 and avoiding IPv6 would have for the foreseeable future? The only thing I can think of is not being accessible to IPv6 only traffic. Quite frankly I can't see anyone in the next decade (if ever) who got assigned an IPv6 only address NOT setting up the capability to tunnel or proxy onto the V4 network. So where is the loss, even there? What I see is alot of FUD being sold about the need to be on the IPv6 train now. That's SOP for vendors trying to sell new technology solutions. I think there are industries or verticals that it probably does make sense to be on the IPv6 bandwagon but for the rest of us, not so much. It's all cost and very little, if any, gain.

If you think that providing severely degraded service to those that can't get native IPv4 will be just fine, then, I encourage my competitors to try that.

If you're participating in the global internet, then, both ends of the connection matter. The access networks will have to go to IPv6 because there simply isn't enough address space in IPv4 to continue scaling them even with NAT and other service degrading techniques for address sharing.

> 
> About the only thing I would want IPv6 for these days is my own private playground....to get myself and our staff a little more experience on it, in case it does gain momentum......but quite frankly we're all too busy doing productive work that needs to be done today..to spare the time for that sort of thing. YMMV.
> 

It is gaining momentum. The before-and-after graphs from world IPv6 day show this in a pretty striking reality.

Owen