Dear Jon regarding IPv6 kits, the cost depends on the business model. I talked to a Comcast guy some years ago about IPv4/IPv6, and they found that the IPv6 cost was less than managing IPv4 and NAT. As said this depends if the company just sells "pure access" (i.e. no device investment) how many routers you have to change, how many CPE (if you have some) you have to upgrade etc. Anyway, this was some years ago, I don't know what Comcast is doing in these days or what other providers are doing now. What I can say is that dual stack costs nothing on the CPE side (it costs on the access and core, but I guess nowadays most of routers do IPv6) My point is very simple: I am not asking anybody to switch off IPv4 tomorrow, I am asking to large ISP/Telco/Content to provide dual stack NOW. We already have a LARGE number of devices on the market which are IPv6 capable, which will be more and more pervasive and which are IP-Vsomething hungry. I don't know the exact numbers, but with iPad, iPhone, Androids we have a lot of Ipv6 capable devices each of them wanting an IP (if connected on the mobile operator's network) and wanting to stay online much more that in the past. Differently from the CPE they are upgraded regurarly, their number is higher tha the home CPEs, the integration is "seamless" in the user experience. By the way, the customer does not care if he/she is using IPv4 or IPv6 they care if google is working, if facebook is working if youtube is working... The IPv4/IPv6 thing is good only for us, the operators the customers at 99% do not care about it. So regarding the number of "devices" I would say that the mobile market has a great potential for IPv6 deployment much more than the fixed market. The second point: who said that we must not run dual stack? Let's start running dual stack and then let's see the IPv6 traffic grow day by day. The third point: the IPv6 contents. Most of the Internet (in terms of traffic) is IPv6 ready, I mean google, youtube, facebook, etc. The rest is slowly migrating and the real reason why "it's slowly migrating" it's because large contents providers are not providing IPv6 or customers are not using it. Again, the lack of IPv6 traffic depends on the fact that the site is not available on IPv6 or that the customer is not running any dual stack. Let's start with the top 20 sites for each nation, let's deploy IPv6-dualstack in the mobile market and let's see if the IPv6 traffic is going to grow or not. Summarizing: IPv6 traffic is not growing because nobody is giving out IPv6 addresses on the terminals (e.g. mobile phones), nor moving the content on IPv6 nor doing anything on that. Even if tomorrow all my customers will be on dual stack, most of them will still "request" traffic from IPv4 networks simply because the content is not available on dual stack (except google, youtube, facebook) How to "push" people to start deploying IPv6 NOW also depends on RIPE: do not give more IPv4 addresses to large IPv4 onwers till they provide NOW dual stack. I know that RIPE cannot charge for IPv4 address space consumption but at least can say "no more IPv4 to you" to those large AS who are not deploying IPv6 (i.e. dual stack). Let's start with those companies who have mobile terminals and the content providers and let's see IPv6 grow. As I said, what I see from my point of view, is just a market push to NOT deploy IPv6 because it would break a lot of "barriers" Just my 2 Euro cents Paolo
The main thing holding “large” providers back from rolling out IPv6 across the board is cost … you’ll have to increase the effective costs of IPv4 to $100 per IP before you’ll be at a point where the cost savings of switching to IPv6 outweigh the costs of depreciating the current IPv4 capable kit and going through natural expansion / upgrades to kit capable of IPv6
Vendors have been slow to produce kit which is IPv6 capable .. they’re starting to do this, but on the whole the big players have been holding off on the $xx million capex spend until the kit they have in place has depreciated sufficiently.
Some of the bigger players here in the UK were almost at the point of rolling out kit which is v6 capable when they found that this kit also did CGNAT and rolling out CGN is easier / cheaper than rolling IPv6 “in the short term”
Then you have to factor in CPE … from Broadband (and VoIP) perspective not all kit does IPv6 yet
Rather than forcing an exponential cost increase in v4 address space, work to roll out IPv6 now and get a _competitive advantage_ against the big boys - then make hay while their accountants tell them they can’t spend money on the v6 kit.
When they finally do we might find their land grab pricing has to disappear and they start to cover the costs of the new infrastructure they’ve had to roll out ….
All you’ll do by increasing the costs of IPv4 is make the smaller guys less competitive and drive a few more out of business whilst the big boys go “oh well another 20c per user on our cost base, c’est la vie
IPv4 is dead. IPv6 is the future. Roll out IPv6 .. push the vendors to provide IPv6 compatible kit … harder I know when the big boys aren’t pushing .. but push never the less .. if enough of us small guys push hard enough then we stand a chance of getting somewhere and maybe getting a competitive advantage over the behemoth again .. for a short while
Jon
On 18 Dec 2013, at 15:17, Vlad Dascalu <vladd@beetux.com <mailto:vladd@beetux.com>> wrote:
IP Market = resource < demand (charging won’t help) That's simply not true, it's economy 101 basic stuff: demand adjusts based on price. If the price is higher, the demand will fall until it's equal to demand.
Consider the case of large telecom that have an unique IP for every telephone line. A higher price is the only motivation that will encourage them to switch their private VOIP network to IPv6. At that higher price-point their transition will be feasible for them, why still being easily affordable by businesses which need a IPv4 for their website or other critical resources.
Resource depletion shouldn't prevent an efficient allocation of scarcity. Vlad
---- If you don't want to receive emails from the RIPE NCC members-discuss mailing list, please log in to your LIR Portal account and go to the general page: https://lirportal.ripe.net/general/
Click on "Edit my LIR details", under "Subscribed Mailing Lists". From here, you can add or remove addresses.
-- Ing. Paolo Di Francesco Level7 s.r.l. unipersonale Sede operativa: Largo Montalto, 5 - 90144 Palermo C.F. e P.IVA 05940050825 Fax : +39-091-8772072 assistenza: (+39) 091-8776432 web: http://www.level7.it