You are here

IGF 2019 – Day 4 – Saal Europa – WS #403 IPv6 Independence Day: Rest in peace IPv4

The following are the outputs of the real-time captioning taken during the Fourteenth Annual Meeting of the Internet Governance Forum (IGF) in Berlin, Germany, from 25 to 29 November 2019. Although it is largely accurate, in some cases it may be incomplete or inaccurate due to inaudible passages or transcription errors. It is posted as an aid to understanding the proceedings at the event, but should not be treated as an authoritative record. 



>> MODERATOR: Good morning, everyone.  Thank you all for coming to our workshop, IPv6 Independence Day, Rest in Peace IPv4.  We are grateful to IGF for the opportunity to discuss the future of the internet.

Also I would like to thank the panelists for accepting our invitation and for the support in our research.  My name is Eduardo Barasal Morales, and I am the coordinator of the training area at  And here I will be the moderator for this roundtable.

We will also have my colleague, Tiago Jun Nakamura moderating the online participation.  So let me introduce our six panelists.

First we will listen to Ms. Burger from the German government.  And then Mr. Moreiras from the Brazilian technical community.  Followed by Mr. Souissi who will represent the French government.  Next, Mr. Tamon will bring us the African technical community view.  And two remote panelists, Mr. Leghris from the Morocco Civil Society, and Mr. Howard from the American private sector.

So this is how we have planned this roundtable.  I will start with a short introduction to the theme.  Then each panelist will have up to 10 minutes to present their points of view.  After that, we will have 20 minutes for the panelists to debate among themselves as well with the audience through the open microphone.

And we will finish with wrap-up and conclusions.  So let's start our introduction.  Currently, we have two versions of the internet protocol, version four or IPv4, which has address exhaustion issues.  And version 6, or IPv6, which is the newer version developed to substitute IPv4.

However, even though IPv6 has been around for 20 years, its use is still very low.  And that is what this graph made by Google shows.  The green line represents the percentage of users that access Google over IPv6.  As you can see, only 30% is running over IPv6.  This means that 70% is still using IPv4.  Looking at this graph and knowing that Google is a huge internet company, it would probably be safe to infer that most of the internet traffic is still running the old version IPv4.

However, IPv4 days are numbered as the other graph shows.  The different line shows the number of IPv4 addresses left in each region of the world.  As you can see, all of the regions are suffering from IPv4 exhaustion.  Probably in less than two years there will be no more IPv4 address left to be assigned.  So we will need to migrate to IPv6 soon if you want to guarantee the stability of the internet.  And we are here to discuss how and when the preparation for this transition can be done.

So here are three policy questions to guide our discussion.  First, when would the ideal time be to transition from IPv4 to IPv6?  Second, what role would each stakeholder play in that transition?  And third, how can we plan this transition without affecting internet governance principles taking into account security, stability and resilience of the internet.

Now we will hear our specialists and we will start with Ms. Burger.  So please.

>> CONSTANZE BURGER: Good morning, everyone.  I care about the deployment of IPv6 since 10, 12 years, the big issue in the Public Administration of Germany.

And let me mention one thing first.  In my opinion, we do need more attention for the theme for the topic of infrastructure development.  All applications, services need reliable infrastructure and networks.  And these topics need more awareness in the IGF and in the discussion and in the political discussion.

And so I want to introduce our IPv6 and IPv4 strategy of the Public Administration of Germany.  Our strategy for the use and of IPv6 and IPv4 in the public administration has many point of views.

Public administration has to provide its IT services for everyone who needs access to the service.  In Germany, themselves and abroad we have to implement and provide the services without discrimination against any part of the society.  Our goal is maximum accessibility.  And our next goal is to introduce IPv6 only and get rid of IPv4 but not everywhere.  IPv4 disabling must not lead to a digital divide.

After the German constitution, we grant federal independence for authorities.  Therefore, my opinion or our opinion is to implement IPv6 is not just to switch on IPv6 and to switch off IPv4.  We have to mention and to aware a lot of other themes we have to care about.

How to put the v6-v4 strategy.  The public administration has to lead by example.  Activities to support the introduction of v6 in the German public administration.  We run a local internet registry including special organization.  In case of the federal state, our constitution overrides, we set up sub LIR structure so is every organization able to do their own technical implementation after a framework.

We discuss the multi-stakeholder issue.  We have /23 IPv6 address space and we have corresponding address plan.  We set up public administration information network and we launch an IPv6 master plan for the federal public administration.

First, I can introduce short view to the LIR.  We are running this Local Internet Registry common with our provider, that is the BWK.  Our operative part of the LIR and the strategic part is in the ministry themselves.

We provide every sub-LIR address space for running their IT infrastructures.  We separated our address plan after federal states and federal government, and you can see this just an overview how we implemented our address plan.

As a very interesting and new subject, in our digitization program, we are going to set up our public administration information network.  In our opinion, network infrastructures are essential for the public administration.  We need strong security and high availability for all use levels.  In the moment, we consolidate networks to respond to the threat level, to prevent future attacks, to create transparency and reduce complexity.

The next step, an important step is to enable the federal infrastructure and we set up an IPv6 master plan with ad hoc activities with migration strategy, with -- we figured out migration areas and we do a lot of things around security and procurement issues.

This master plan we set up the federal infrastructure is not only with the focus on infrastructure themes.  We also have to mention that the human resources, the skills, the knowledge, the training but also the contract and procurement issues are very important to set up the infrastructure issues.

These all is with our view to set up these new things.

And I want to hear your response to our way forward to implement IPv6 and I'm happy to hear your comments.  Thank you for your attention.

>> MODERATOR: Thank you, Ms. Burger.  Then we will have Mr. Moreiras from the Brazilian technical community.

>> ANTONIA MARCOS MOREIRAS: Good morning, greetings, everyone.  It is very good to see you here.  And it is good to see you interested in IPv6.  Not everyone can see the importance of IPv6 yet as you see.

I said that yesterday the IGF network, I mean the wi-fi network, we are using the internet connectivity that we have here is limited since we do not have IPv6 connectivity.

This is totally unexpected in the Internet Governance Forum and it is really, really a shame.  The IPv6 you can see in this room as far as I know is not provided by IGF organization but by our highly enthusiastic person in this same home.

I don't know who.  Yep, you?  No.


>> ANTONIA MARCOS MOREIRAS: Well, I want to talk about IPv6 transition.

I remember about 20 years ago there was a discussion regarding if we should talk about IPv6 transition or IPv6 deployment.  And then it was mostly agreed that we should call it IPv6 deployment because in those distant days the little of understanding about IPv6 was so low by network operators that some of them could think they should turn off IPv4 immediately at the same time they were deploying IPv6 and it would be a complete disaster.

So the word deployment starts to be used because it was better to teach the network operators what they should do in the short-term.  And I ask now, did we forget that we are really in a transition to IPv6?  Did we forget that keep IPv4 working, adding that, adding CGNET how this was part of a strategy in the way to the IPv6?  Did you forget that?

Or, are we not talking of a transition anymore?  IPv4 and IPv6 maybe are going to co-exist forever and ever.  Well, for the internet forward it seems to be clear in a statement made in 2016, IAB says the IAB expects that the ITF will stop requiring IPv4 compatibility in new or extended protocols.

Future IETF protocol work will then optimize for and depend on IPv6.  We recommend that our networking standards assume the use of IPv6 and be right so they -- be written so they do not require IPv4.  We edge courage the industry to -- encourage the industry to develop strategy for IPv6-only operation.

So according to internet architecture board, our goal is still an IPv6-only internet.  But I think it's not clear for how the people involved anymore.  In preparation to this panel, we organized a survey among ISPs in Brazil asking some questions about the transition.  We had 116 response.

And one of our questions was if IPv4 would be turned off in the internet, specifically in BGP routing system some day, about 40% said no.  The main justification is that legacy equipment, 40% of the Brazilian ISPs according to our survey do not think we will be able to solve the problem of legacy equipment.  So IPv4 will be present forever in the internet.

We also see a growing market for IPv4 addresses.  This market is succeeding in bring to the routing table addresses that before were not being used.  The governing process of the RIRs was not so successful in the same task.  But this marketing, this market is also lowering the pressure to IPv6 deploying.  It is slowing down IPv6 adoption, or it could be at least.

I think the IPv4 market can be good at short-term for ISPs, companies, the economy, whatever, but if we still have the purpose of getting an IPv6-only internet, it is very difficult to me to believe that the IPv4 market will help us with that.

I have heard some very complex arguments on the contrary, but I still -- but I'm not convinced.  In the last year, IPv6.Br completed 10 years.  IPv6.Br is the naming of the set of projects to foster IPv6 deployments in Brazil.  And we had a special technical conference to celebrate the anniversary.

And in this conference, we organized a panel called IPv6 deployment challenges and IPv4 shutdown.  It was supposed to be just a tease, a provocation.  We did not really expect that the speakers invited would have anything to say about IPv4 sunset.  But we were wrong.

One of the speakers that was from one of the biggest ISPs and telecom operators in Brazil had a very solid and detailed plan of how it would be -- how its network would be evolving, how would be the evolving steps of its networks from the current state to a future of IPv6 only.  I was very surprised and I was very impressed.

But should I be impressed?  Probably no.  Our network operators should be doing this kind of planning.  So, bringing this to IGF is not provocation.  It is a call for action.' are talking about one net and then we must have one vision.  We must have one vision.  One shared vision about how our internet will evolve technically.

Are we still going to an IPv6-only internet?  Can we agree with that, that we are on a transition to IPv6 and not simply on an IPv6 deployment?  If so, then some day we will have an IPv4 sunset.  IPv4 will be turned off on the internet.

So why do we create efforts that could lead in other direction such as the IPv4 market.  So why we are not discussing the strategies for IPv4 sunset?  There are so many questions we could be asking and answering in the way to the IPv4 sunset.  Why IPv6 is not yet deployed?  It is still lack of separate -- it is still lack of knowledge or understanding.  Should it be enforced by laws or regulations?

Should we have coordinated shutdown, for example, of specific date or specific condition.  What should do without break anything on the internet?  I don't know the answers.

I'm not even sure about the questions.  But we should start talking about this and it's why we are here and why we organized this workshop.  Thank you.

>> MODERATOR: Thank you, Mr. Moreiras.  That was interesting.  Now Mr. Souissi will speak.

>> SAMIH SOUISSI: Thank you very much.  One of our missions is to accelerate the transition in France.  If we can start the presentation.  Okay.

So as you can see, and it is not a surprise, Europe is currently experiencing a shortage of IPv4 addresses.  It is not surprise.  But don't worry, the internet won't break and won't stop functioning.  But if you care about internet growth you should start worry.

So like here, most of the actors are thinking about keeping IPv4 and here I mentioned why keeping IPv4 is not a good idea.  The options to keep IPv4 scalable are mainly to it is either we buy IPv4 on gray market or rent or use carrier grade NAT to share IPv4 between several customers.

Both options me be harmful because for buying IPv4 and here and example in France.  For buying IPv4, certain services are being blocked.  The video streaming banking services due to a temporary wrong geolocation when the IPv4 addresses are acquired, located in Latin America or other place.  And then use them in France, the services may be locked temporarily and can be for the users and operators that acquired the IPv4.

It is also creating potentially a sizeable barrier for new entrants and small actors due to the increase of price of IPv4.  Now an IPv4 costs more than $20 and sometimes play prefer buying IPv4 instead of deploying IPv6 that is free.

One other reason that buying IPv4 may be harmful, it increases the risk of seeing an internet split in two at a certain time.  For example, when we have small actor that would obliged to have IPv6 only hosting, it won't be able to be accessed by end points that are IPv4 only.

For the second option using CGN.  It prevents certain types.  Controlling smart home systems and online gaming from functioning properly.  And we see this mainly with many gamers that are alerting about the fact that if they are behind the CGN they won't access all their ports, thus they won't be able to play their favorite game.

Second, it is difficult to identify suspects in judicial investigation was terrorism, cyber pedopornography.  They are have interested in IPv6 to have them detect with the people and communication within the pedepornographic forums, and the detection is low due to the CGN.

And the black listing issues.  If one address is black listed from a certain service for example, banking service due to security reasons it may lead to black listing of all machines behind the CGN.

As we see the issues as a regulator we think there is only one option that needs to be done to accelerate the transition to guarantee the openness and growth of the internet while avoiding the issues of IPv4.

Now I will speak mainly about the work we are doing actually in order to promote IPv6 in France either through publishing a barometer and increasing transparency about the transition or fed rating the community in a multi-stakeholder approach.

But before, let's talk about the transition on itself.  Like when we went from NCP to IPv4, it happened in one day.  This is not possible with IPv6 because of the complexity and the size of internet today.

So like we started the transition in 2003, but now we are still at the beginning of the cohabitation phase and we need to generalize the deployment of IPv4 with the IPv6 with -- IPv6 with IPv4 in all parts of value chains in order to start to speak about IPv4 finding out and the real transition.

So we are still at the beginning even if we are more than 15 years after.  And it is kind of alarming.  As you can see in the figure that you don't see that well because it is far, we show the state of the transition of the different stakeholders in France.  There are bottlenecks like mobile operators and hosting providers and gaps and information system for enterprise and administrations.

But also IoT.  Let's speak about the barometer.  I will go briefly.  It is a report of 30- pages that you can find online at our website, but I will bring highlights here in order not to spend too much time talking about it.  And we collect information directly from operators and also AFRINIC which manage the CCTLD of France that provides us information about the hosting part.

So what we do here is what we call data-driven regulation.  Name and shame or name and shine depending where you are on the podium.  As you can see like we have two operators that are more than 70% of client that are activated.  Here we are at the fixed network.

And something to mention that is interesting is like for free.  For example, they installed the new firmware on the vast majority of their boxes in May of 2019 and removed the ability of disactivating IPv6 which significantly increases the use of IPv6 in France.  And like we encourage this kind of approach.

In here, the four main operators on their fixed line, network, sorry, they have provided forecast for the percentage of customers that are IPv6-enabled.  And as you can see for some operators the evolution does not answer much to the IPv4 shortage.

And this year, in order to improve our monitoring, we enlarged the scope of the gathering to add operators that have between 5,000 and three million customers.

As you can see, we only have like four operators that are -- sorry, three operators that are doing IPv6 today for small operators.  And even though Europe is currently experiencing the shortage like some operators still have no plan to deploy IPv6 on fixed network, which is like as I said before, it seems problematic.

Now, if we focus on mobile, here we tried to divide by APN.  We did Android and Android tethering and data only and iPhone.

We only have two operators starting activating IPv6, mainly on Android and for iPhone it started like in September of 2019 and doesn't mention here because the figures are off the month of June.  So we have the brief telecom in orange that made the remarkable push on iPhone in 2016 with 68% and 30% of IPv6 ready at the end of October.

And they provide us also the forecast and same thing for the operators between 5,000 and three million.  And as you can see, even more than on the fixed lines, the pace on mobile network, future deployment is likely to slow down the transition.  And in order to better orient the customer, we added the -- in our barometer the devices that are IPv6 enabled per default by the operators in their last software update.

So the list is available on our barometers.  And something to add also on the mobile part, mainly to speak about 5G.  In the last procedure for awarding licenses to use frequencies in the 3.4, 3.8 gigahertz band in metropolitan France, we added that the license holder must make its mobile network compatible with the IPv6 protocol by December 31, 2020.

It is to get the intra-operability and not slow down the use of service only on IPv6.  Through the window of frequency attributes, we added the obligation of IPv6 compatibility which if we say it is for the 5G-APN it will be for full 4G and 3G and increase the compatibility on the mobile network.

If we go to the hosting services, which are one of the main bottlenecks in France, it continues to have a slow transition.  Only 27% of the top 730 of websites visited in France are IPv6-enabled.

And if we see the most visited web pages as Cisco mentions through Alexa top 730, the number grows to 60%, which shows it is mainly the bigger websites that are on IPv6.  If we focus on the French top level domain, we only have 15.5% in IPv6.

This is a schema at the top, the level of growth is quite slow.  And what we wanted also is to do some name and shame for hosting providers.  And as you can see, there is only one that has like more than 75% of IPv6 enabled website and all of the others of top 10 for French top level domains have less than 10%.  And in the barometer, we put figures for the top 30.

Same thing for mail hosting.  That like probably it is even slower than web hosting.  And the situation is quite alarming with only 5.8% in IPv6.  And the slow pace of the transition of this section of internet value could force IPv4 to be kept longer than expected, which will result in more costs et cetera, so the transition as we said before should be done on different level.  And in like, you know, harmonized way.

One more or less positive point is the DNS infrastructure with 78% activating IPv6 and with the same approach we differentiate with the different hosts.

So where does France stand?  Like if we take a look at the figure, like we started in 2003 and until 2016 thanks to the deployment of free we stayed at 5% of users and then when orange start deploying we started deploying and then at the mobile level.  Then on May of 2019, as I said before, right after the activation per default in all free boxes of IPv6 and removing the ability, removing the possibility of disabling IPv6 like France percentage of users grew to 36% today.

And then in the barometer, we compare also the different statistics either use or content and transit providers with the different countries.  And even if the situation is not that great, but we tried to find positive things like France today is like in the first position in Europe after Belgium, Germany, and Greece.

And but it is not ranking between countries.  It is even the first country is at more or less 50% which is not necessarily enough and I think that everyone in this room is agreeing with me.  So we need to accelerate and do more things.  Now I will talk about the IPv6 task force in France to finish my presentation.

In the task force we tried to gather all stakeholder in France in order to accelerate and solidify the transition in France.  So I don't see the slide there.  Okay.  So what happened in 2016 is that we gave a report to the government about the deployment in France.  And we proposed different number of actions.  One of the actions was setting up a barometer.  But another action was to set up appropriate place for discussion.

That's why last year we had the IPv6 workshop and had several problems explaining the delay to IPv6 like lack of training, quality of service issues and security issues et cetera.  You can find the summaries available on our website but there were suggested courses of action in order to accelerate.

That is why we set up an IPv6 task force that was launched two weeks ago.  It is in partnership with internet society and we involve different stakeholders.  We have ISPs, big ISPs, small ISPs, hosting providers, IXPs.  Equipment suppliers.  And it is important for us to gather different stakeholders that are involved in transition.

And the objective, as I said before, is to accelerate the transition through, one, facilitating information and best practice sharing between different stakeholders and two, working on concrete deliverables to help different actors in their transition.

In addition to physical meeting also there is a project to put online platform to synchronize the work and prepare the different deliverables.  It is not the separate telecom task force.  It is common task force.  So it is defined in consultation with the community of participants.  Thank you very much.

>> MODERATOR: Thank you, Mr. Souissi.  Now we will listen to Mr. Tamon from the African technical community.

>> MUKOM AKONG TAMON: Good morning, everyone.  So I'm going to start my presentation by sharing a proverb in Africa that says unless a man is willing to plant a seed knowing full well that he will never sit on or eat of its fruit, the village will never grow.

Unless a man is willing to plant a seed knowing that you will never sit on or eat of its root the village will never grow.  I want you to have this at the back of your mind as I speak because we will circle back at the end.

What does it take for us to accelerate IPv6 deployment and I speak from the perspective of you could -- if Constanze is an admiral in the Army think of me as the intelligence officer who has been in the battlefield, right.  More than 55 different countries, training more than 10,000 engineers.

So I know that what it takes is just not more training because people like to say oh, training.  It is just not more training.  In AFRINIC, for example, we trained more than 10,000 people.  So I can pretty much guarantee in every single network operator on the continent we have at least four people that we trained.  I'm talking about hands-on training.

We know that is just not more conferences.  We have got IPv6 in France and the business summit in Zurich and several sessions at IGF.

It is just not that.  It is definitely not more government policies or mandates which have failed spectacularly as the U.S. government.  It is definitely not more incentives which don't exist for now.

It is definitely not more search for the elusive business case which is the management which is for me is the mother of all management excuses.  Right?  And definitely it is just not another wild goose chase of something we call killer applications.  That is just plain lazy.

What it takes is decision makers who are willing to think beyond the short-term, right?  It is decision makers.  Every single one of us works in a company.  You know that engineers don't initiate change on a massive level.  That is on the decision makers.  Right?  So the problem we know is the skill exists, right?

We know that engineers are motivated because they are always going to want this thing, I want to do this thing or they have proof of concepts in all of the labs but the thing required to scale it to company level is outside the circles of influence.

That is on the decision makers, right?  We know that is not a lack of, you know, other operators from Africa to free in France to T-mobile, there is lots and lots of different companies in every sector that have taken the plunge and have demonstrated that it can be done.

The problem is simply that so many managers, so many executives or CIOs do not -- they don't have the vision and are not willing to back the engineers for IPv6 deployment projects.  And because we know the number one reason for project failure in every kind of organization is what?

Lack of management support.  Because management support or executive support unblocks a trinity of things without which any project is doomed from the beginning.  The first one is time.  Engineers need time to work on IPv6 deployment projects at scale.  Who controls the time?  The decision makers.

The second one is money.  So if you are in Africa you get training for free but AFRINIC isn't going to buy or replace equipment for you.  But free training, you still need to sponsor yourself in the flight and hotels.  All of that money is controlled, the purse strings is controlled by the decision makers.  And sometimes manager change equipment.

And finally, in several workshops, two weeks later they say I have gone back and deployed IPv6 and I have it perfect, and we asked why haven't you pushed to the customers.  They say it requires more resources and I don't have the authorization to pursue that on my own.  So these are the three things that executive support which can only come from decision makers unlocks time, money, and other people to work on the project.

So I think we all know here it is just not a matter of debate that IPv4 is ill-suited to the continued growth of the internet.  We know that IPv6 like a tree is guaranteed to be good for the internet and what is good for the internet is good for business.  So the same exaltation that the mentioned at the beginning applies to the decision makers.

Until the decision makers are willing to back profits with benefits to the organization and realized the next quarter or the next semester, only then can technology that we have outgrown like IPv4 begin to wither and only then can IPv6 get a chance.  And this perspective is what informs our strategy at AFRINIC, right?  So we bring training and focusing on managers and policy makers on the strategic impacts through actionary workshops.

We call them action hack-a-thons.  Shorten the time from learning six months or six years in some cases to five days.  Thank you very much.

>> MODERATOR: Thank you, Mr. Tamon.  Now we will hear the remote panelist Mr. Leghris from the Moroccan Civil Society.

>> CHERKAOUI LEGHRIS: Thank you, Eduardo.  Can you hear me?


>> CHERKAOUI LEGHRIS: Good morning.  Good morning for everyone.  I am Cherkaoui Leghris from Morocco.  I'm a member of the Moroccan society, the ISP structure in Morocco.  And also a member of the task force IPv6 in Morocco and also ITA professor.

So sorry, I am (?) there is no IPv6.  I think that this meeting will be the first to encourage the IPv6 initiative.  I hope that next I can take this point of importance.

So when we speak about IPv6 and IPv4, I think there is no incentive to move to IPv6.  We have issues and we have IPv4 attrition.  This is a big issues that should impose the movement or the transition in IPv6.

So the important portion what is the right time to stop IPv4.  I think that we should have new services.  We are waiting to the 5G, for example, for the services which informs to be connected in IPv6, I think that is the most important is to have the new services just with IPv6.

Also, I think that chosen the object to be connected in internet should be also almost reason to make stop IPv4 because we could not address all objects.  So now how we -- how do we prepare technically?  I think that the first thing to signal is there is no fixed day, there is no days to move to IPv6.  IPv6 is a big project.  It is not an update of IPv4.  We should take to try to pass to move to IPv6.

All the stakeholders in every country should work together to success this big project.  For example all government in every time in every country should take action plans with the deadline with the action and with everyone agrees all of the stakeholders to migrate to IPv6.

Also on the ISP side, on the operator side, they should do IPv6 the new IPv6 services.  The users is not a good idea.  We should to initiate the need in the users we should create a user basis and why users are to use IPv6 services.

In the academic side, I think that we should initiate a road to have more skills in the IPv6 users.  We have more academic models in university in training organization.

In the community technical, I propose also to have a more workshops in IPv6.  More technology in IPv6.  More technology built with IPv6 in the internet infrastructure.

Now to the transition.  I think that this project we will have more time to this transition.  We should take more time to this transition.  IPv6 is a long and a big project.  We should work together to facilitate this.

Another point is we should have more consensus, we should create more consensus with all stakeholders in each country.  And we should work together to go ahead in IPv6.  And we should satisfy all users in the services and so on.

In Morocco, I value the innovative who will make the project to plan the migration at three four years ago.  We have found with operators of mobile and all actors throughout the society and technical community and academic and so on.

This will go to specify only in IPv6 and to give more idea or to plan the migration in our country.  Thank you very much.

>> MODERATOR: Thank you, Mr. Leghris.  We had a problem with the other remote panelist, Lee Howard.  Let's start the microphone session.  Any comments from the audience, the speakers?  Please?

>> AUDIENCE: Rajish from India.  Largest percentage wise IPv6 in India.

Germany is way ahead to India but in the user wise, India is the larger user base of IPv6 deployment.  India is having around 65% Google user base and the pain capacity is very low.

That is the reason the lifeline of the equipment which is being installed in anywhere into the country on the broadband or mobile phone the lifeline is very large.

And in those phone still IPv4 is working, IPv6 is not working.  They cannot replace the equipment just to take the IPv6.  That's one of the major challenge.  Why we are right now limited to the user who are buying the new equipment.

Secondly, the technical guy of corporates are not so encouraged, in the major problem they are facing is the address is so long that they cannot remember.  They are simply remembering IPv4, they cannot remember the IPv6.  And they have to change the equipment of the corporate network.

As well as their skill capacity they want to use what they have.  They don't want to use into the new.  While a lot of trainings, a lot of workshops has been done by ISP association to which I am heading.  But some discouragement is there until the time the IPv4 is not getting completely exhausted and they have not left out to just switch to IPv6 finally.

Otherwise, until that time, it will work.  For the startup, they require a small pool of /23 minimum.  As far as I remember, in 2008 I moved the proposal for reducing the allocation size from /21 to/22.

In the exhaustion just into the discussion, not in the serious talks.  But still the Whatsapp people are requiring a small pool to start up and create the network and to understand the thing.

We are welcoming the IPv6, we are deploying the IPv6 but at the same time, rest in peace IPv4 I doubt.

>> MODERATOR: Thank you very much for that perspective.  I want to answer to one specific part about the concern of the network engineers and the length of the addresses.

>> MUKOM AKONG TAMON: I want you to think about the world of typist.  The first time he stop using a typewriter and got a computer.  For so long as she kept or he kept using the mentality of a type writer and the computer she was never going to be getting value out of a computer.

The exact same thing happened here.  I think it is important that the people who do training understand that we are just not trying to, you know, take every single bad habit that we have learned because lots of engineers all over the world have learned some terrible habits around IPv4.

And every smart engineer I know understands that the half life of the knowledge in IT is two years.  Which means if you don't learn new skills and upgrade your skills in two years you become irrelevant.

At AFRINIC we do and we demonstrated this over and over.  The length of the IP address is a $5 problem.  IPS system and basic skills, that is a $5,000 problem.  Please don't solve $5 problems.

>> AUDIENCE: You have answered your own question and my question.  Smart technician.  Some technicians who work as usual into the journal, we don't want to change.  But the smart who really think of development and who really want to deploy the new technology into the network, they will definitely change and they are changing.

That is the issue.  That few engineers who are not so smart, they think it is going, let them to and they work according to that.  That is the problem.

The management of that corporate has to come forward and why not deploy the new equipment and why not deploy the new technology into the network for the betterment.  And we are encouraging, we are now directly talking to the corporate head for deploying the IPv6 into their network.

When the service providers are not allocating for the IPv4 to their demand automatically they are changing to IPv6.

>> AUDIENCE: I would like to comment about the equipment.  In the point of the companies it really is just solve because the equipments have obsolescence.  So ISPs all the time come to us and say well, I have a network and our CPs don't have IPv6 support.

What can I do, it is very expensive to exchange out this equipment.  So forget it.  Forget your current network.  Forget your current equipment.  The equipment will be obsolescence in five years.  In ten years maximum.

Then you change it because of other reasons.  You have to -- you have to worry about the new networks.  The ISPs are growing.  Are connecting new people.

So let's not do new networks with IPv4 only.  Let's do all of the new networks, all of the new implementations with IPv6.  We started to do it in 1998.  If we had done is, it would be already solved.  It's not.  So it is simple.

>> AUDIENCE: Just correcting, all of the ISPs in India are IPv6 capable and their network is IPv6 enabled.  They are giving the IPv6 network who their customer.  It is the discouragement from the customer side.  Either the customers are not accepting IPv6 immediately until the same, the requirement is not increasing.

Like I have given the pool of /24 to one of my customers.  The moment he will demand /23 I will say that you have to come to the IPv6 as I will give you as number of IPv6 to you for your network.  Then they are changing.  Until the time their network is small and they their requirement is small, they are not so encouraged to change the network.

All of the service provider is there.  One more problem is there.  The wish of the government should also come forward.  The moment they meet mandate three that all of the payment gateway, all of the bank railway, airlines has to be IPv6, drastically there will be a change into the network.

Then the user will not be able to access through IPv4 so they have -- for they have to forcefully change to IPv6.  It is the government initiatives that should come by mandating them.  All this application will be on IPv6 only, not IPv4.  Still a the lot of banks and payment gateways are there accepting IPv4.

>> MUKOM AKONG TAMON: Within Africa, 15 governments have gone down the route of mandates and like the U.S. government failed spectacularly.

That is because government has two weapons in the toolkit, regulation and mandates.  The other one that I don't think most governments understand is the power of economic incentive.  Let me finish.

If there is one thing every operator will tell you the government is not a stakeholder of my network.  The government does not help pay customers.  The government does not buy me equipment.  The government cannot tell me what is in my commercial best interest.  This is something they are trying to get governments in Africa to do.  In huge countries, the government is the buyer of IT services.  Maybe the government can have the power of incentive.  The same governments deploy IPv6.

If they themselves don't have IPv6 on their network.  That brings the problem of credibility.  Why are you trying to put a mandate to deploy technology that you yourself are not.

If you find an example in the world where mandates I have worked I will be glad to know.  I guarantee you all those that tried maybe stand to try something else and the power of economic incentive.  Imagine what would happen in the Indian government said X date?  What if the Indian government said in five years' time any contractor who does not have IPv6 will not get an Indian government contract.

You are not forcing anyone.  You are going to make people to do the writing for themselves.  I think that might be more effective and cost a lot less.

>> AUDIENCE: We have to understand one thing in India, the literacy level is very low and the new user base coming from the rural side is no doubt illiterate in education but they are still illiterate.

One more thing we have to understand the politicians who are running the government are not at all a technology guy.  They don't understand.  And they don't want to compel anything so politically.

>> MODERATOR: Please, Ms. Burger.

>> CONSTANZE BURGER: You are all right.  We have an interest to guarantee national sovereignty.  Therefore, we are going to run our own network infrastructures.  We are not just a buyer for IT.  We are running IT ourselves.  And therefore we are interested in to implement IPv6.

The problem is we are a federalist state and after constitutional rights the country's municipalities can do after the constitution their own decision.

And what we can do, we can bring in knowledge and training and convince them to integrate IPv6 and we can have that discussion to implement IPv6 is not just running IPv6, it is a question of implementing new structures.  The chance of having new ideas and new -- more transparency to bring in national sovereignty.

These are more political than buying just one infrastructure element or a device or to order a new provider.  These are more features.  That is our point of view.

>> AUDIENCE: Several points to start with the compatibility issue et cetera.  We don't care about compatibility we care about activation and we don't wait for clients to ask for activation like in our perception should be done by operators or hosting providers.  That is the first one.

>> AUDIENCE: All of the service providers are compatible and doing the services in IPv6.  Even their own network running on IPv6 and the new customers coming to the network they are going to the IPv6 only.  They are only much earn canned with the few old customers who are not changing.  Until the time they will not change.  Speaking about the mobile network --

>> AUDIENCE: Mobile network.  100% IPv6.  And the Vodafone is almost 70 to 80% IPv6 network.

>> AUDIENCE: In general it's just a matter of time that the device will no longer be in the network.

>> AUDIENCE: Only the rural portion of the country who are not buying -- they are buying the refurbished equipment.  By default they have to go with the IPv4 and the service provider cannot overlook those customers in the present time when we are struggling.

>> ANTONIA MARCOS MOREIRAS: And to come to the second part, the transition or even adding incentives, et cetera in general like incentives may be nice to have, but it is not like mandatory in order to at least do 90% of the transition.

It is more the end of the transition where maybe this incentive may lead to finish the transition.  But in general, for example, as I take the example of France because that is what we are doing, other than what we did for 5G where we asked for compatibility what we are doing is data driven regulation showing transparency about what actors are, one.

Second, to show them to every stakeholder where is the interest of transitioning and by talking there is like a psychological thing where people are afraid of something they don't know yet.  Let's put it this way and it can be directly linked to the training issues and the fact that as soon as they are reassured about quality of service issues, security issues, how they should handle that and that aspect filtering aspect et cetera, it may lead like those operators to put up in place their transition.

>> AUDIENCE: Compare the size of the population, use of the user base of the country, changing all of the user and size of the population will always be a challenge for implementing anything.

>> MUKOM AKONG TAMON: I had a similar argument.  We have to deploy IPv6.  It is completely different from we must get IPv6 to every single citizen on the same day.  Every project manager understands the importance of a phased approach.

Phase one, let us get IPv6 in government infrastructure for the next two years.  No operator needs to deploy IPv6 every single part.  Do the data center while leaving broadband.  Do the mobile network while leaving the fixed network.  It happens today.

My experience has been that most of the managers who say I can't do it because every single device on my network is not ready, every single user is just looking for an excuse.  People who want to do it will find something that you can do.  The small wins rather than thinking you can do every single thing everywhere at the same time.  For me that is the mother of all excuses.  Let's find something that we can do.  It is possible to find that in every single case.

>> MODERATOR: I think Mr. Leghris wants to say something.  Mr. Leghris.

>> CHERKAOUI LEGHRIS: I want to (?) to develop more skills and basis.

In my part of government they do not have the task in IPv6 (?) but the operator have it.  The operators just says that they have IPv6 but there is not a the lot of need on user side.

I think that all of the logic because the operators should create the need in users.  The operator expect that users should request IPv6 services.

In our logical -- in our previous commerce logic, the user nowadays is waiting to propose for him to new services and the operators should push in this orientation to implement and propose the new service in IPv6.  I think that is finally it is clear that if the operators want it, we should have the user and (?) begins tomorrow.  Thanks.

>> MODERATOR: Thank you.  So we don't have more time to debate.

So to conclude our session, again, I would like to thank you all for coming.  When we proposed this discussion, we didn't expect to set a specific date for the transition, but we wanted to raise awareness of all of the problems that we might have if we don't start to plan to think about this transition.

So as you can see from all of the panelists' speech there is a lot of work to be done.  We hope you have enjoyed, we hope to meet you again next year.  Thank you.  Rest in peace IPv4.



Contact Information

United Nations
Secretariat of the Internet Governance Forum (IGF)

Villa Le Bocage
Palais des Nations,
CH-1211 Geneva 10

igf [at] un [dot] org
+41 (0) 229 173 411