The following are the outputs of the real-time captioning taken during the Tenth Annual Meeting of the Internet Governance Forum (IGF) in João Pessoa, Brazil, from 10 to 13 November 2015. 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.
***
>> ANDRES PIAZZA: Good afternoon.
Good afternoon, everyone.
We are about to start the session, the workshop 194. Welcome to everyone here in the room, remote participants also. This workshop is organized by the Regional Internet Registration, the five RIRs. My name is Andres Piazza, and I'll be moderating this workshop. And I am the manager, the leader for the strategic relations in LACNIC, the local area, the one that serves the regional region and the rest of the Latin American Caribbean.
We are having an update on the IPv6 transition, and we are having -- we hope we have a roundtable where we can have a flat and open conversation. But we also have some panelists that, it's an interesting set of different views for IPv6, a diverse set of panelists. I want to introduce them. And to start with opening remarks for every one of the panelists. So I want you to welcome Sylvia Hagen. She is the owner and CEO of a consulting and communication company called Standard Connection. She is the author of "IPv6 Essential," a book well-known in this field.
Aaron Hughes is with us, CEO and CTO of Six Connect, a company working with Internet resources on these matters. They also work strongly with IPv6.
And we also have, Izumi Okutani. She is the policy -- -- maybe -- maybe you can hear me better, I hope.
Okay. Izumi Okutani is the policy liaison for JPNIC. The registry for Japan. And she is an active member of the Internet community and the RIR community.
Also, Nicolas Antoniello is with us. He has been one of the members for the LACNIC community and now he is an adviser for the industry ministry in his country.
And Antonio Moreiras is here on the right. He is the manager for development and projects at the CGI.BR. He is in charge -- part of the project that he is in charge of are the ones related to IPv6.
And also we have in the other corner, Mohamed Elbashir, special adviser for the Ministry of ICTs in Qatar.
So we have a diverse table and we have different experiences. And the challenge of the IPv6 transition has -- I hope we at least can feel the temperature about what are those challenges now.
So I want to give the floor to Sylvia to start her opening remarks.
>> SYLVIA HAGEN: Okay. Is that better? Okay. Well, thank you for the opportunity to be here to share my information. So I'm working out of Switzerland. And in the last couple years I consulted many different organizations in different industries. And there are several challenges that I see to IPv6 deployment. And let's say I'm more focused on the enterprise deployment, less than, let's say, ISPs.
So the things that I bring in are from an enterprise perspective. And there is usually one of the first hurdles is that the technical people, they see the need. So it's usually the people from the network that understand that they have to initiate an IPv6 project. And the first hurdle to take is to get the C level people on board. Because if you try to do an IPv6 project in an enterprise and you don't have C level people supporting it, you won't get anywhere. So it's usually the network people or in some cases the address people. Because large organizations even have the problem of running out of IPv4 space internally. And so that is a driver for deployment.
And then the second step is that many organizations, they, for a long time, did not consider IPv6 and now finally management understood, since the IANA pool is empty, that maybe they should do something. And they decide okay now let's get done with it, and try to do a quick and dirty approach. And so if they are thinking in a sustainable way, they should be taking a little more time than just trying to flash it out as quick as possible, because IPv6 has a different architecture from IPv4. And it takes everybody some time to understand where the differences are.
And only if you take the time will you have the opportunity to sort of take advantage of the advanced features of IPv6 and of the new architecture. And this will, in the end, save a lot of money in the way of -- it will save operational costs. Because if you have a clean and nice design and use the opportunity to clean out your network, then it will be a lot easier, less complex and less costly to maintain that network in the future.
>> ANDRES PIAZZA: Thanks very much, Sylvia.
I want to give the floor to you and start with your opening remarks.
>> AARON HUGHES: Hi. I'm Aaron Hughes I'm on the board of AIRON and actually several other roles. I live in California and our customer base is global, but we see primarily our customer base is in the US.
What we're seeing is that generally speaking, Internet Service Providers are, for the most part, complete with their IPv6 implementation, and that generally involved net stacking tools, integration with CRM, billing systems. Really all the basics that you need to provide service as a service provider. Their needs are generally focused around IPNS management and the automation of turn ups. What we see in the enterprise space, well, enterprises that look like service providers, so the public facing enterprises is that they are generally concentrating on things such as home grown provisioning systems for the back end of virtual machine implementations, and are running into a lot more challenges with vendor support. So people that are using things like Citrix or VM ware, open stack, their Administrators are generally holding them up. Not that the vendors are holding them up, but the home grown systems that back that are currently a stumbling block.
What we're seeing in the more traditional enterprise space right now is that people are either in the planning phases or have started to do work on IPv6 on internal infrastructure, where they are really learning how to deal with what was traditionally in the FC 1918 space now actually touches the global Internet, which creates fear in a lot of people. They realize that they actually have to have due consideration for things like security policies and more emphasis on firewalls and load balancers and where their end points or demarks are actually at.
Generally speaking, in the States and Canada and the Caribbean, I think there is a significant uptake. I don't see anybody who is saying we don't need v6. They are in some point of a transition plan, and I think that, you know, outreach and education are fairly well penetrated in the region. And we're seeing a lot of good progress.
One of the more interesting points is on the v4 side, I've seen quite a few times personally people now saying I'm sorry, we no longer have any IPv4 resources. So not only are we out from the IANA pool perspective, the IRI pool perspective, but seeing service providers and enterprises saying sorry, we no longer have resources. We are currently searching for resources or renumbering existing public numbered resources into either the 198 space or v6 only and reutilizing those. But this is a recent thing where I'm seeing people saying I'm sorry, we have no more.
>> ANDRES PIAZZA: Of course, Sylvia, go ahead.
>> SYLVIA HAGEN: Just to add something to this most recent story. I was hearing more than one story in Europe recently where providers even started to charge more for IPv4 services.
>> ANDRES PIAZZA: If anyone wants to jump in, it's totally welcome, so they can do this, even if they didn't do their own opening remarks. So feel free to do that the whole session.
Aaron, please, if you want to continue. I don't know.
>> AARON HUGHES: That's certainly enough for an introduction. I'll hand it off to Izumi.
>> ANDRES PIAZZA: So Izumi Okutani. Her opening remarks. Also, if you can tell us a little about the whole integrated approach that you are having in Japan.
>> IZUMI OKUTANI: Sure. Hello. I'm Izumi Okutani. I'm from a national Internet registry in Japan, but what JPNIC does is not simply managing the resources including IPv6, but we also work on more community initiatives of trying to raise awareness on issues that really helps the Japanese Internet, including the promoting of IPv6. So that is where I come from. And so I just wanted to share a little bit of a picture of the initiatives that we have taken in Japan, comprehensively. What we learned from them and remaining issues that we observe.
In terms of where we are in terms of deployment rate, I think we're in the well top ten. So we can do better. But we're doing like relatively okay, I suppose.
And in trying to get there, there were three issues that we observed. One is the common issue of chicken and egg that, you know, we will do it if others do it.
And the second is that even if people genuinely are motivated to do it, you can't do it on your own.
And the third is that we need key business decision makers to make the decisions.
So we can't just, you know, individual organizations or single groups of community within Japan can't do it, we need to get different, like industry groups within Japan getting together and then working on this issue. So two bodies were set up. One is very much a community-based initiative called IPv4 exhaustion task force. It sounds backwards from where we're going. Exactly. But the idea was to prepare for the issue towards the exhaustion. So over 20 key organizations within Japan in the industry got involved. So -- it included the operators groups, such as JANOG. It included a cable TV providers groups, or Japan users group, and the Government ministry also joined, the Government layer. So having the groups of people, not just the core experts on IPv6, who always had interest in IPv6 anyways, having these groups of people really raised like awareness and the need on the national level. And they were able to reach out to the respective communities or members that they had.
So this has helped. And the issue of training that I think Sylvia has mentioned, this was again conducted jointly, identified an issue that we need to work together on the national scale. And so we provided trainings per different Sector, access line provider, content provider, ISP, so this was actually an initiative that started from this task force. So that was one thing.
And the other initiative was initiated by the MIC Government so they set up a Study Group. And they were being facilitator rather than trying to lead anything. So they invited academia experts in the industry and then as well as the key decision makers. And this was really important. Because each of the ISPs had to give a report of the progress that they were making, and it actually sort of created an environment of soft competition. Because if you were not making progress, you had to explain why. And it was a bit embarrassing in front of the Government and the other competitors.
So this was actually -- it wasn't intended that way, I suppose, but then the result of setting up this study report and producing a report actually had this sort of unintended but positive side effect to it. And it helped. If there are issues that you cannot resolve it in your organization, you can raise it. Okay. This is something that we're facing and we need help from the others. And there will be others in the table in that same place in the Study Group. So they were given the homework and they had to report back at the next meeting. So this helped in the progression. And that's where we have gotten into a relatively high deployment rate globally.
But things are not perfect. We still have remaining issues. And the stronger issues that we still have is the area of mobile. We actually don't have strong connections with the mobile phone industry, and then the Internet industry. So now the Government has set up a second phase of the Study Group, inviting the mobile phone providers, and trying to do a similar thing to motivate them.
And then also we still have the issue of having the contents to be IPv6 ready, I suppose. This is something that is commonly an issue throughout the world.
So that is an overview. And that's it from me for now as an introduction.
>> ANDRES PIAZZA: Thank you very much. We will, before giving the floor to the next panelist, I want to know if there are any questions in the audience and also to remind them that, as this is a conversation, questions are welcome at any time. Also, if there is anything, any questions, maybe not comments yet, but at the remote Moderator, I want to ask if there is anyone else who has a question in the remote area, also to know that.
So okay. Not yet. So I will give the floor to Nicolas Antoniello. And also, continuing with this, you will also be able to do your opening remarks of course, but also I wanted to ask you about this approach of integrated efforts and how do you see that applied in our region, in the Latin American region. Also with this role of the Government that Izumi was describing for the Government, a motivator for the outcomes to happen. That is the part that I wanted to see.
>> NICOLAS ANTONIELLO: Hello, everybody.
As Andres said, now I'm working as an adviser for the Ministry of Industry, which is also in charge of telecommunications. But in the past -- that is a recent, very, very actual role -- but in the past I've been working for more than ten years within an ISP, in different positions within an ISP in my country. And also, I've been working as -- kind of working with the community in the role of the Public Policy integration for the Latin American region.
And I will try to or would like to divide the discussion of the situation of IPv6 in the different regions into, maybe, into two aspects. One would be related to policy. Policies related to IPv6 and IPv4 exhaustion. And the other more related to the technical aspects of IPv6 development. Challenges, and the state-of-the-art, actually.
Regarding the the policy side, I also don't want to deeply talk about it, mainly because on Friday we will have a media special session on that. But in integration for the IPv4 exhaustion, there were the final kind of four phases, mainly in most regions were defined more or less the same instances. Mainly, the one phase differs from the other mainly in the amount of IPv4 addresses that remains in the pool of the RIR, of the RIR.
So in Latin migration we are in what we call phase 2. The last phase is phase 3. And probably between this month and next month, we will be going up to phase 3. That would be the last phase. That would mean that we won't be allocating anymore IPv4 blocks to existing providers. And there will only be a reserve block called the soft landing block reserved for new incomers. So if you are a new provider, you can request an IPv4 block just to make sure that you have enough IPv4 to face the transition from IPv4 to IPv6, unless you want to go IPv6 only, which at present probably from what I learned last week it won't be a very good idea.
And Regarding IPv6 in the region, in the LACNIC region, at the beginning the efforts were more concentrated on providers, on ISPs and providers, because they run the infrastructure, which is necessary to connect different devices and all that stuff.
So in the region there were, besides the efforts of the -- of global organizations, besides the efforts of local organizations, the region through LACNIC and through LACNOG has and continues to do a lot of capacity building regarding IPv6, how to run an IPv6 network, how to deploy IPv6 in your network, which are the best common practices, the best ways to interconnect the networks, to exchange packets using IPv4 and exchange packets using IPv6 and all of that.
And also, on some certain technologies, like MPLS, BGP and a lot of technical stuff -- I don't know how many of you are from the technical side, so I don't want to run deeply into technical things.
And then one other effort that has been done, and it continues, is regarding the provider, the ISPs in the region, is the -- how to lower the difference between the number of blocks that are allocated to ISP -- I mean to IPv6 blocks allocated to providers, and the number of those blocks that are actually being published through Internet. I mean, that are being advertised to what we normally call the global routing table. Because in the region, there is a big difference yet between the kind of -- the amount of IPv6 that are allocated and the number of IPv6 licenses that you can see on the Internet.
Then what I believe are the -- what I believe are the challenges for the future. And I would like to ask and answer a question that I always like to do regarding IPv6. Who are the final users in Internet related to IPv6? We normally say okay, the person that uses this kind of devices, like computers, cell phones and all of this stuff. And I agree, that is wrong. The final users if you are talking about IPv6, for me at least, are the devices, not the persons. So when you ask how many of you use IPv6, one possible question is I don't care. Because if you -- if you're talking about IPv6, I believe it's a relation between, in an Internet of Things scenario, it will be a relation between the things, the cell phones, fridge, computers, light bulb, whatever. It has an IP address and can be connected to the Internet and infrastructure. It doesn't necessarily need the interaction of the -- of a person. What we need to connect doesn't carry through IPv4 or IPv6 or whatever Protocol.
So that leads to one of the main challenges for me, that is the IPv6 deploy in all of these massive use devices, mainly because one of the things I believe is that at first most of the enterprises that build these devices concentrate on transition mechanisms, which is some way of translating to or from IPv6, from IPv4, and all of this stuff, and not concentrating in what we call active IPv6 support. And active IPv6 support, it's also a way -- it's a very, very, very big issue. Because it involves not only the main specification of IPv6 Protocol, but a lot of other Protocols that need to be deployed so that IPv6 can work actively.
So that's, for me, that's one of the main challenges. In our region, it's how to get these devices, all the devices, running IPv6.
>> ANDRES PIAZZA: Thank you, Nicoslas. Maybe we will come back to the government. But a very interesting and disruptive approach. And also if anyone wants to react to this last part, especially the one that mentioned the role of the Internet users, it would be really welcome. If you want to. Also, in the audience.
And, well, we also have Antonio Moreiras here. So, Antonio, you can, apart from your opening remarks, you also are in charge of the operation and the projects related to IPv6 for, at least the more or the leader of the region, that has more influence in your country. It's not a small country. So you have a lot of challenges. And if you want to tell us the challenges for Brazil and what you see for the challenges for our region, the Latin American region, it would be interesting to hear from you.
>> ANTONIO MOREIRAS: Okay. Thank you. Very well.
I would like to talk about what we have been doing the last years regarding IPv6 adoption in Brazil.
We are a not-for-profit organization that works as an executive arm for the Brazil Internet Distributing Committee. We manage the ccTLD.br. We are the Brazilian NIR, the national registry. And we have other actions and initiatives. For instance, we manage 25 Internet exchange points in Brazil, and we are the local host of this meeting.
We perceive that we could help an IPv6 adoption work in information training and working with awareness coordination. So we tried to do it the last years. We have created and maintained a website about IPv6, with a lot of technical information, in Portuguese, since 2008. We have created an eLearning course introducing IPv6, also in 2008. We have hands-on training that is four or five days long that is about 36 hours, free of charge for the participants by them coming from their domain names.br. We have trained about 4,000 people since 2008 from more than 700 different autonomous systems.
We developed virtual experiments, a virtual teachers teach IPv6 based on CORE, the virtual experiments are using our training, and we published a book that you can buy in a bookstore in Brazil or download free of charge in PDF format.
We have organized and helped some big IPv6 events, called the Brazilian IPv6 forum. In a month we will have the Sixth Brazilian IPv6 forum. Maybe the last one. No.
And about coordination, we have organized and held meetings between ISPs, Telecom operators, our Telecom regulator, our Government, vendors, and some other players, since 2011. In these meetings, we described the IPv6 transition. We tried to reach consensus about some Brazilian scheduling for the transition. And we tried to create and sign a document, an agreement, with this scheduling.
We have failed in this. We could not have this document signed. But the process was very good and it gave good results recently.
For example, the Federal Government -- our Federal Government has created internally a schedule for IPv6 deployment in the Government agencies. Just in 2014. The end of 2014. Our Telecom operator organized a meeting between the main Telecom operators, which are also the main ISPs in Brazil, and they also agreed with a scheduling for IPv6 adoption, which also happened in 2014. And the schedule would start in 2015 this year.
The group of banks have addressed the question of IPv6 needs in institutions. They conducted a test, a trial, in 2014, and it was well succeeded. Now each institution is supposed to be working on an internal project to deploy IPv6.
So all that said, the exhaustion of IPv4 in the LACNIC region also happened in the last year, in 2014, the phase two. That is for all the practical matters, that is exhaustion.
So in this year, we have started to see some results. We started to see the trafficking growing. So now we have about maybe 7.5 percent of our old users using IPv6 in Brazil, according to APNIC and Google data. It was 0.1 percent one year ago. So it's good.
(Applause)
Oh, but it's not only Brazil. I think it's in Latin America as a whole. We have a big ground of IPv6 adoption because, I think, of the different countries followed a bit different paths, but the exhaustion happened in the last year and things started to happen everywhere. We have this 50 gig of IPv6 traffic at NIC.br, in Sao Paulo, that's our biggest IXP. We have 3,500 systems in Brazil and 3 thousand IPv6 allocations.
So things are happening. But there is -- but, I think I will let the "But" for outside for my initial remarks. But there is a lot.
>> ANDRES PIAZZA: I want to wait for the "But" and we will come back for that certainly, Antonio. Thank you very much.
And, finally, Mohamed Elbashir. And you will help us, Mohamed, to cover this, the world, not only because you work in the Middle East region, but because you also come from the African region and you've been an activist. And we know each other from those times, activists in the Africa region for Internet usage. So you can help us understand a bit more the environment for the two regions. It's a lot of challenge. But whatever contribution you make is welcome. So thank you.
>> MOHAMED ELBASHIR: Thank you very much. And I'm very happy to share with you some experience regarding the role Government and regulators could play in terms of facilitating the adoption of IPv6. And maybe at least pushing for that.
I think just to build on what Izumi has said, Government can play a major role and regulators could play a major role in terms of bringing all the stakeholders together in one table and, basically, build a forum.
So I'll share with you an experience recently done in Qatar, where I currently work, which is basically developing a national IPv6 strategy and the national IPv6 migration plan. I hope you can hear me very well. So I'll give you experience about establishing a national IPv6 strategy and a migration plan for the State of Qatar. And the objective is that although Qatar is a small country, in 2022 it will be hosting the World Cup. And there are major requirements to having IPv6 readiness. I think it's important.
So we need to study what other countries are doing in terms of national, let's say, strategies or implementation plans. And a good model out there was a Singapore model, where they also have a national IPv6 strategy and immigration plan. So three or more details about maybe how you can replicate those models.
It's number one important to have a Government agency which is playing that leading role. It could be the Telecom regulator or the ICT ministry. And that Government entity should be able to bring everyone to the table. So calling for, let's say, for a formation of IPv6 task force, a national task force, which is basically a multistakeholder group of mobile operators, academia, Civil Society, ISOC chapters, Government network operators, application providers, basically, even banks. I mean, so entities who really have major networks, and we need them to be sitting at the table and understanding the importance of IPv6. So I think that's an important milestone.
And then after that, we did a survey of the current status of what is exactly available. What IP ranges they have, from where, what AS numbers. So we did a study of the current status of the country. And based on that, a draft of a national IPv6 strategy was developed with the support of those stakeholders. So, basically, the components of that strategy is, number one, education and awareness. And it focuses on having a platform where they can share experiences and also best practices. And showcasing IPv6 implementations as well, and workshops, and organizing conferences as well.
Part of the awareness was establishing a test lab with the University. So that was also important.
The other component of the strategy was to build demand. So we were helping as well private entities from different sectors. For example, oil and gas or big corporates, tried to negotiate and talk between them, and Telecom operators. Incumbent operators, unfortunately sometimes they don't have incentive to provide IPv6. We have legacy incumbent operators who have enough pool of IPv4 and sometimes they don't see that there is a need to introduce or talk about IPv6. But those corporates who have a few networks, they are feel the reality that they need to start speaking. So part of this was introducing the stakeholders to the RIRs. In the case of Qatar, it was RIPE NCC and having the support of RIPE NCC for the different stakeholders.
The groups worked on providing night lines for the Government working network procurements. So, for example, what type of requirements or compatibility is required for Government tenders when you are procuring network equipment? As well as provisioning from Telecom operators. And that was important, the Government has huge networks and it's important for them to understand the importance of IPv6. For the migration plan -- so that was a strategy.
The migration plan basically was to try to develop a template of proposed migration plans for their stakeholder or Sector. So, for example, we did look at let's say financial Sector banks. What are the steps required for a bank to, let's say, to start planning for IPv6 adoption? And we, I think, identified four major areas. Starting from internal auditing. I think the first step is to assess what is there in the corporate. That's number one. What systems are available. What type of networks are currently available.
And then start planning for the implementation of IPv6. And that could include training. That could include reviewing network equipment procurements.
And the third step is installation and commissioning in terms of doing test labs and testings and launching services, which include the go live activities. So a set of agreed migration plans has been discussed within that forum and agreed with the stakeholders. And the timeframe for those migration plans is between two to three years. And actually, as Izumi said, it's useful to have a forum where the stakeholders are reporting their progress. So every month the same task force sits together and everyone actually shares their progress along the line, which has enabled them to evaluate how they are doing.
So I'm happy to give anymore details about this model.
>> ANDRES PIAZZA: Thank you very much, Mohamed.
I believe there is a question or a comment in the remote channel. So if the remote Moderator wants to jump in.
>> REMOTE MODERATOR: Yes, there is. Thank you for the floor. There is a comment in the remote participation hub. It is from -- it is written by Alajandre from Venezuela. And he says I just want to add a comment based on Nicolas' comments. In LACNIC we have had, in every event during the last two years, trainings oriented to reach IPv6 in the end-user. I just wanted to point this out since I share his thought regarding we are missing IPv6 in the massive user infrastructure.
>> ANDRES PIAZZA: Thank you very much. I just want to point out that Alajandre is also a staff of LACNIC. I just want to raise that.
Any other comment? Yes. Chris.
>> AUDIENCE: Hi. Thank you. Chris Buckridge from RIPE NCC. I wanted to take this opportunity as RIPE NCC to spin off of that from Qatar. But the road show project which has been something that we have deployed around the Middle East region, the Arab States particularly, and which has been a bit of a new initiative -- well, since 2011, but something that we have been been building since 2011 to provide practical training in a lab-based environment, and it has hit milestones recently, which I thought was useful to point out. This last weekend we had the 30th of these events in countries around the region. This one was held in Lebanon. And this was held in conjunction with the public Sector in that country. So we found that particularly in the Arab region, working closely with Governments to provide that training to regulator staff or to use the Governments to get in contact with the technical staff and some of the incumbent providers there, has been a really useful way for us to reach people that we might not otherwise.
The other key point there about the road show is that we are using community members, rather than necessarily RIPE NCC staff as the trainers. And so we have worked with trainers from Europe, from the State, from some other parts of the world and brought them in, funding that, and providing the opportunity for them to be there.
But also, then, to scale that, and this is the other milestone, we have been trying to build the train-the-trainer programm. So this last event in Lebanon was one of those where we had one of the graduates of the train-the-trainer program as one of the road trainers in the road show itself. So we have been building that for a while. Getting the train-the-trainer off the ground is harder than the straightforward trainings. But hopefully with that milestone reached, that will allow us to build out the program more and provide that kind of practical training where it's needed. So I think that is useful.
I'd be happy to talk to anyone who would be more interested in the details of the program. It might be useful or applicable elsewhere. Thanks.
>> ANDRES PIAZZA: Thanks very much. Paul Wilson?
>> PAUL WILSON. I'm Paul Wilson, the head of APNIC. And I spent time talking about IPv6 trying to explain the transition to people and also in all sorts of walks of life. One of the analogies that I use is that the transition from IPv4 to IPv6 is like the transition from gasoline to electric powered vehicles on our roads. And there are quite a few analogies, a lot of parts to that analogy that I could go into. A car mechanic has a very different job to do from the driver of a car. Between an electric -- a gasoline vehicle and an electric vehicle, the cars actually look the same and they function the same and they drive on the same roads, much like we have traffic on v4 and v6, driving on the same roads and functioning the same way at some particular level. Someone was asking about the users before. And the thing about the user in this analogy is that they don't really need to know anything. They can just sit there and enjoy the road and trust that every other part of the car, the fuel, the driver, is sort of doing the job for them properly. And it's quite interesting here, because many of you won't know it, but IPv6 is running on the network that you're using at the moment. And without knowing it, your computer is likely to be used IPv6 now. You can -- for those who are curious, and many passengers won't be curious about what fuel the car is using, but if you are curious, you can go to a website like the APNIC site, APNIC.net and it will tell you that you're using an IPv6 address on your computer at the moment. So I think that's an interesting thing about the user is that they may not need or want to know. They may want to know, though. Just like someone as a passenger who is very interested in getting somewhere, who has a business dependency on their car, they start to worry about what sort of car they are driving and whether the car that they are driving will be good for the future. Whether it will run out of gas or it will be okay. So as a user, your dependency on the Internet depends on what you are using the Internet for. And it may be well worth your while to find out whether IPv6 is being provided to you by your service provider.
The whole Internet is a complex Web of services and service providers. So no matter what your relationship with the Internet is, that service is being provided to you by one or several or quite a few different service providers. And if you're very invested in the Internet, if you are dependent on the Internet, then the thing that you probably want to do is to talk to your service provider, the people that you are relying on to get -- to keep you connected to the Internet, and ask them what are their plans for delivering you IPv6 services? So that could be an ISP, it could be a person who sells you a computer. It could be your own staff, if you have people that are doing that work for you, data centers and Web hosting, and the whole range. You depend on them and you want to ask them, are they going to provide you an IPv6 service? And you need to sort of trust the answer and work out whether it's the answer that you really want.
But that's really where we all are at, at the moment. We may not need to transition to IPv6 today, just like we don't need to drive electric cars today, but we kind of are starting to need to know when it's going to happen. And again as I said, if you have a high dependency on the Internet, as a Government does in terms of the national infrastructure, then you have to take a high degree of interest in the IPv6 and when v6 will be serving you for your independent needs.
Thanks.
>> ANDRES PIAZZA: Thanks very much, Paul, for this insight related to the users. And if there is any reaction to this, I want to give the floor to -- Izumi, are you reacting to this? Please. Thanks.
>> IZUMI OKUTANI: Thank you for sharing your relevance as a regular Internet user and what you can possibly do. And I want to add one more point about what you can possibly do. If you are working in an organization, maybe it's worth asking your system Administrator if your company's website is IPv6 ready. And if not, then maybe it's something that you can actually try to, well, it might be a bit difficult if you don't have technical knowledge to persuade. But at least in the case of Japan, there are guidelines provided for all regular networks, how to actually have your network be IPv6 ready. So it might be something that is worth asking a question within your company.
>> ANDRES PIAZZA: Thank you very much. Is there any comment at the remote channel? Or question?
>> REMOTE MODERATOR: Yes, there is. Alajandre from Carracos, Venezuela, he has a question for Mohamed. He asked if he could mention something about the financial aspect investment, KPEX, OPEX,
Thank you.
>> MOHAMED ELBASHIR: Thank you. I don't have really a figure I can quote. But I think it's -- there is theories about delays in implementation that means higher costs at, let's say, as you go.
So, for example, if your corporation is -- did not start any planning or considerations for IPv6 maybe now, you end up after two years maybe investing a big amount of money because we have a critical application maybe to launch or you have something, some service to roll out and there is no IPv4 available. Or you need to do, to purchase equipment. So it's better, really, to start preparing from now and studying your options and reviewing your network and how you can start adopting IPv6.
The more you delay it, the more costly it will become. Maybe others can jump in.
>> ANDRES PIAZZA: Thanks very much, Mohamed. Any reactions on the floor? Oscar.
>> AUDIENCE: Oscar Robeilles, CEO of LACNIC. Just on the line of what Antonio Moreiras said, that Brazil has been doing for the last months a remarkable job in the deployment of IPv6. And if you happen to be an end user, like using the analogy that was mentioned by Paul, if you happen to be an interested user in the IPv6 or you are the manager or a technical manager of a network that is willing to deploy IPv6 in the Latin American region, just to remind that we have three online courses. Some of them are already in Spanish and English. One is for the basic knowledge of IPv6. The other one is the advanced course. And the third one, it's a methodology to test current software infrastructure to make sure that it is compatible with IPv6.
So whether you are an end-user or a manager of your infrastructure, we have relevant information, online information, online courses for you. Just connect to LACNIC.net and you'll have those three options.
Thank you.
>> ANDRES PIAZZA: Thank you very much, Oscar. And I heard about delays on IPv6 and the economic impact. I heard from Sylvia about how she referred to saving money in the operation costs as one of the things that should be considered. So I want to come back to you, Sylvia, regarding this economic impact. But Sylvia, first if you want to. So if you can go deeper on that. Thanks.
>> SYLVIA HAGEN: Okay. I will try. It's a pretty complex question. And it depends on the scope and the depths of how companies look at that. So one thing I find when I'm working in enterprises is that very often people don't differentiate between investment and operation. So they just say well, IPv6 costs a lot of money. But you have to look at it.
So one thing is that every network that we run since ever we run networks has some annual cost in just continuing to evolve the network. And we have done several things. So in the '90s, we introduced the HTTP and then we started to implement IP Second and we started to do that, but we had to learn all of the technologies, we had to implement them. It was just part of evolving the network. And now in 2015 or maybe soon we will just implement IPv6. And it's one other aspect of how we deinvolve the network.
So you can say IPv6 is a lot of high cost that we wouldn't have if we won't care about IPv6. It just means that you evolve your infrastructure. And why do we do that? We do that because we want to run services on top of it. So we don't drive around with a Ferrari or a Tesla on a street that we use for horses. Because the Ferrari wouldn't look good on that street. So we have to build our street to run our services. So that is number one.
And then we have to make the differentiation of how much do we have to invest to build the roads and to update our services, and how much we would cost later on to maintain that. So if you look at Gardener's figure, Gardener says if you deploy IPv6, it will cost you approximately 6 percent of your annual IT budget divided by the number of years it takes you to get there. You're not doing this in six months. After that, it will cost you approximately 1 percent of your annual operational costs to maintain that IPv6 integrated network. And so that is a very general figure, but in some cases where we looked at it more closely, we could say yes, it's meaningful.
And so what I think is that in many companies they try to assess the cost right in the beginning, before they even have a real plan of what they want to do. Right? So actually, what you need to do is first you do a strategy and you define a target architecture. Then you start to look at what does it take and what does the roadmap look like? And after that, you can sort of define your RFC requirements, what equipment, hardware and software upgrades that you need. Only at that point you can quantify the cost that it will take you to get any further.
So when you were talking about your experience in Africa with the banks, I was just thinking, you know, many customers just look at what they have and they assess their equipment and they say what can this equipment do, and then they try to build an IPv6 strategy based on what their equipment can do. And I don't think that's the right way to go. You will limit yourself.
So the way to go is to find a strategy, the target architecture, and then assess what you have, and then you can sort of work on how much will it cost and what is a good roadmap to deploy that.
>> ANDRES PIAZZA: I saw you nodding in some parts of the answer so if you want to react here.
>> AARON HUGHES: Sure.
You know, I think when people look at implementing IPv6, particularly if they look at an enterprise network, the biggest challenge is that the tangibles are difficult to see. Historically when we upgrade things, we may look at something like say a PBX running D4 and investigate moving to a VoIP system. And we can count up the number of phones that we know what our reoccurring costs will change to. We can see what hardware we have to replace. Because it's a small thingm. Or we look at something like we have 50 dot matrix printers and we will replace them with 50 laser printers. We can say that is a tangible capX investment and it will add these features and it will change our operating by X or Y over this period of time. When you talk about the underlying protocol, it's not familiar. It's been a long time where we had to replace these things. Where we are going from thin net and token ring to IP, it's less common in this world.
So I urge people to look at this in a way that is more tangible to them. Find the things that are important, such as public facing services. And address them first, as an element. And look at things in your back office, in sort of a step by step, evaluate the needs that are important for your organization. Where you're going to feel pain first, that is where you're going to run out of resources that you need, say, to continue to provide services over IPv4 and attack those next.
And then the things that are less painful to survive, whether they are on 1918 or they are just not growing, and then attack them in a tertiary manner. And those kinds of approaches are a little easier to swallow. You'll always be upgrading infrastructure. So there is always going to be capX, a depreciation schedule, and your OPX will be what it needs to be to support your infrastructure. Those things always go up and to the right. It's not that IPv6 costs more. It's just a change in technology that you'll have to figure out how to spend.
>> ANDRES PIAZZA: Thanks very much, Aaron. I see several reactions to this topic. So we will go with these reactions, and after that, I have you Nicolas and then you in the room. And then I won't forget to go back with Antonio after this. Nicolas, you wanted to react?
>> NICOLAS ANTONIELLO: Thanks. I was thinking about an example that came to my mind. Suppose you are -- you want to run or you're running a provider -- a content provider business like delivering video. Like online video club. And you have to buy one million set top boxes to deliver to one million, which is, you know, a public -- you are offering this service. You say okay, I'm going to buy one million set top boxes and give one to each of my customers. All this runs over an infrastructure which is not provided by me, it's provided by -- I made a contract with a service provider to, you know, you need Internet in your home. I deliver the video through Internet and you receive it on your set top box, which is also given to you by me.
And my service provider actually doesn't provide IPv6 support in the home users. So I don't care and I buy one million set top boxes which I don't know if they support IPv6. And then my service provider, which has run out of IPv4 addresses, decides to go towards IPv6 and starts giving IPv6 or starts giving IPv6, or they start giving IPv6 only in their home.
So I would have to see if my set top boxes support IPv6. And what about -- what if I ask the, you know, the one that made those set top boxes and they said oh, no, you should have told us that you want to run over IPv6 before, because for that you will have to change the set top box. In the best case you have to change the figurer and that's a little cost, but if you have a change one million set top boxes. For a service provider like that, the highest cost is the box that they give to each user. It's higher than any other investment that they have to do in the business. So if you have to change all of your set top boxes, you are basically your business ends there.
So -- and why your business ends there, because you don't care about IPv6, because you were focused on video delivery, quality of emission and sound and all that stuff. And something which now should be very easy to specifying in a, you know, in a call for specification for a vendor of set top boxes, which is support for IPv6. You don't care about it and you, you know, you miss. That is one example that we should -- we may think about it in terms of the cost of not thinking about IPv6 support in the things we use within our companies.
>> ANDRES PIAZZA: The time is starting to be as scarce as the addresses. But I want to turn to Antonio, but I'll go back.
So you had something you wanted to go back to and then you had a question from the audience. So thank you.
>> ANTONIO MOREIRAS: Yes. Well, I think I'm not worried anymore about ISPs things. It seems to be -- address seems to be working, seems to be happening. And even the end-users, the companies, I mean, a lot of them, the technical people at least, it's not an "if," if we are doing IPv6 and maybe it's a "when," and maybe it's a "how" we are doing.
So people sometimes are asking things for us at NIC.br and I don't know how to answer them. Technical things. I'll give some examples. Maybe you can help me and give some ideas. For example, besides Google, gmail, almost nobody supports IPv6 in e-mail. Maybe some universities, but people from companies say that the -- it's a different world. It's a different thing. I want to see some big mail provider doing this. And they ask how can I do it securely? How filtering works? I don't know. I cannot answer that. Filtering, I don't know what are threats.
Other question. Companies keep asking how to deploy IPv6 securely. How to protect themselves against the DDOS attacks. Sometimes they say that the vendor doesn't support IPv6 yet. I don't know how to answer them. What to tell them to do now.
The question you talked about some minutes ago about the devices. I really don't believe that the end-user will be able to change that. But it's real, there are a lot of mobile phones, smart TV, video game, wireless hubs, access points in the market being sold, but being still manufactured. So they don't support IPv6. I'm not sure if convincing the end-user to ask for IPv6 will be the solution. But I don't know what is the solution. How to convince these vendors that IPv6 is needed now, at least for the new products. How to do it. I don't know.
One more, the last one. I promise you. I don't know an eCommerce or eBank website with IPv6, not only Brazil, worldwide. I don't know one. So if you know, please point it to me. You know? Okay.
But in general, why are not they deploying IPv6? Is it too early? Is it security? What is the problem? I don't know even the problem. They say that they will do it. Sometime they are planning for it. But I don't see one of them doing it. Why? There must be a reason for that. I don't know. Well, those are the questions.
>> ANDRES PIAZZA: Yes. I thank you, Antonio. I see Sylvia wanted to react. Anyone else wanting to react here? So I can let Sylvia react and then we will go with Olivia.
>> SYLVIA HAGEN: I wanted to answer a question. I know a few. I don't have the overall statistics, but I know a few examples. So first of all if you look on the Web site of the launch day, you see that about 16 percent of the top 1 thousand Alexa websites in the world are IPv6 enabled. So that's not bad. It's 16 percent.
>> Not eCommerce.
>> SYLVIA HAGEN: One eCommerce site that I know, it's not a good example.
>> AUDIENCE: That's Cisco, everybody expects Cisco to do that, so it's not the same. But this is a couple years they are enabled -- and they do billions of turnover over IPv6.
And another case study that we showed in June, last June, at our conference in Switzerland and where you can download the slides, it's not activated yet, but Morofski is dual stacking their website. So management, you know, the gemstone seller, they do more than 50 percent of their global turnover over the Internet. And they have three main Internet points all over the world. And management decided in 2012 that there was a business case to dual stack the website for those business reasons. So they are working on it. It's not activated yet. They are waiting for some upgrades for their online shop software but they will be IPv6 soon and you can check the slides. So it's just an example.
And then we have a very traditional enterprise in Switzerland. The Swiss Post, which you would expect versus Cisco to be last to go IPv6. And they are going. This website is dual stacked for over a year and they are looking at financing, which will do eBanking over dual stack. So it's not a ton, but there are some.
>> Thank you.
>>OLIVER: I'm the Chair of ISOC UK England. I've been advocating the rollout of IPv6 for ten years or perhaps more, and put together our company servers and got them to run IPv6 since I can't remember the date. But it was long ago.
I've been an advocate of pushing for IPv6 in the UK. The UK position has been terrible. It's been trailing in IPv6 adoption. We have a Government that did not want to take the lead into rolling out IPv6 or into promoting IPv6, because the belief is it should be market driven. It's one of the first countries in Europe that deregulated telecommunications that ended up with a flurry of telecommunication providers that then fought with each other. Either you were a good provider and you would win and be able to purchase other Telcos, or you were a bad one and you went out of business or you'd be purchased by somebody else. So there was a whole split in the market and then the market coming back together.
With regards to IPv6, it's been despairing. And in September I went to another forum, the IPv6 Council, because the IPv6 forum died, and the Council took over. We had several things that tried to launch IPv6 in the same way that we're discussing it here at the meeting. While I was ready to go to another one of these depressing meetings and choose a building to jump from, finally, because you know that it's never gone to happen. On this occasion I was surprised. And the discussion we had were different to the discussions that we had here. We were thinking about the past. The future is the IoT, it's today, it's not tomorrow. It's happening today. Anybody thinking that they will be about able to run out of IoT on IPv4 is kidding themselves. The devices that are out there is making a real case for IPv6 to be deployed. It's a disrupting factor.
B Sky B or Sky as they call themselves, Sky Broadband announced during the second quarter of 2015 that they were rolling out IPv6 to all of their broadband subscribers. That was a bit of a shock for many. Everyone thought maybe they just thought about it now. But, in fact, during the forums, the Councils presentation, they presented us with a presentation that showed a three-year work to reach the point where they are at today. They started in early 2012 -- sorry, 2013. There is a whole list of things which they had to do across the company in order to be able to reach that, including about a year and a half of staff trials, internally. And all of that being done without really telling the rest of the world that they're doing this, so as to be able to have the defining factor to be the first ones to be really out there. BT has for many, many years known about IPv6. They were one of the early people that were doing some test networks before the IETF finished the work that they were doing the different RFCs for IPv6. And BT for many, many years said oh, there are no plans to rollout IPv6. Guess what? Because Sky has now rolled this out and Sky is among the five largest ISPs in the UK, BT announced in September that in April 2016, that's only a few months ago, in April 2016, 50 percent of their customer base will have access to IPv6. By December 2016, everyone will be on IPv6. We are now seeing a change in the UK market that is very, very different from something that was just about a year ago the same sort of discussion as we had here. Now we're seeing competition on this.
And just to finish this off, the overall theme, and we had several other presentations by EE as well, which is the Orange and T-Mobile partnership, or company that put together, they are also rolling it out in the next year or so. They are serious about this. And the overall theme was if your company is not now looking at rolling out IPv6 and we are looking at ISP, mobile and land line, if your company is not having plans about this, look for another job because that company is going to disappear.
Thank you.
>> ANDRES PIAZZA: Thank you very much. I want to see any reactions to this or other comments. Kevin?
>> AUDEINCE: Hi. Thanks very much. Kevin from LACNIC.
I wanted to provide a partial comment and consideration for the communication discussion. And it deals with the region I'm from, which is the Caribbean, and the fact that if anyone is familiar with the entire landscape of Telecom and ICT, Internet providers in the Caribbean, you realize it's unique in terms of market driven growth and these sorts of things.
So in my own observation, I have come across situations where you do find some of the smaller ISPs saying that they are not thinking about IPv6. And where you consider situations where you have like in some instances tens, tens of ISPs, some of them with a client base, sometimes it's 20 or 30 people, simply because the Internet providers are not providing service to their communities. You see that there are interesting comments that come from that.
One of the things I noticed as well with some of the dominant providers is that sometimes they link v6. They are not thinking CAPEX, OPEX. They are thinking simply about the link to the client base and they realize the client base to grow opportunity is not high. And you see some unique comments coming out from those particular factors that speak for the telecommunications and Internet market within the Caribbean.
So it's just something that I wanted to put forward. Not necessarily as a personal view, but simply to say that there are situations where you do, indeed, find in micropopulations, and in microcommunities, the prevalence of comments that probably v6 isn't something that someone is considering at this point, despite the technological arguments. Despite the logic that we in the room here are quite familiar with.
Thank you.
>> ANDRES PIAZZA: Thank you very much. I want to know if there are any other remarks on the remote channel or here in the room, because we have two extra minutes before wrapping up.
Yes. Over there.
>> AUDIENCE: Hi. My name is Omar. I'm from Kabul, Afghanistan. Our country is new to these concepts. I've been on a mail list from the Minister of Communication and there was discussion with the local ISPs on the discussion about IPv6, but that never happened, because people did not know what it was. Awareness is one of the major issues we have to address in Afghanistan.
Other issues, like training, and technical know how.
The third one is the technical support from the providers or people who know how to help with the IPv6 implementation in Developing Countries. These are the three challenges.
My question now is: What are the support mechanisms available and who are the people in situations we have to talk to to help us with addressing these three matters? Thank you.
>> ANDRES PIAZZA: With the last seconds available, if anyone wants to jump and answer this question.
Okay.
>> AUDIENCE: I'm Sanjo and I came here in the Youth IGF program. I want to congratulate you all for the hard work on the factors to connect the next billion. And I want to ask you what is currently the most important step to prepare people to go towards the IPv6, or if it is the cost of the hybrid transition.
Thank you..
>> ANDRES PIAZZA: Thanks very much.
The screen is showing that the time for this workshop is over.
To the colleague from Afghanistan that asked a question, I want to just refer to the five RIRs, the five organizations, but we have representatives of five entities here. And we have a booth there with the marvelous coffee machine. So if you visit us, it will be interesting and we will follow-up the discussion.
If anyone wants to jump on the last question we will have final remarks and wrap it up.
Yes, please.
>> AUDIENCE: It's not a question, but just a comment. So with regard to the IPv6 transition up to date, we have been actually hearing about the stories from the supply side. But recently, this is probably known to most of you in the room, but from the demand side of the world, there is a very remarkable development that we are actually seeing. And there was actually a need for the IPv6 from adoption of the -- from the alternative things.
But one thing that I wanted to mention is the movement from Apple, that the IOS 9 requires IPv6 as a default. Otherwise, the application will get a 25 millisecond delay. So, actually, the demand side is going to demand IPv6 more going forward. And we are sure that it's not only the movement from Apple, but it's the content providers who actually hosts the whole world to move and adopt the IPv6 going forward. I wanted to mention it's not only the supply side has been actually working, putting a great deal of effort in terms of adopting IPv6. But now we do see that the demand, the user side, we are definitely going to make adoption of IPv6.
>> ANDRES PIAZZA: A brief comment, Nicolas.
>> NICOLAS ANTONIELLO: Adding to both questions. Regarding where to go to ask for help or for advice or for cooperation regarding IPv6 deployment, you may -- I would go to the registry. All five registries are doing lots of things. Lots of users regarding capacity building at all levels. Go to all technical communities, there are at least five technical communities, called the XNOGS. The one, the cooperation groups, are developed within each of the five regions. They have a lot of courses. They have also given a lot of capacity building, you know, in all technical aspects.
And also to lots of RIR institutions -- other institutions that care and do things and push with IPv6 deployment and they give services to help in running IPv6, and all of them are listed in many of the public access in the Internet.
And regarding the last question, I should, you know, refer to what Paul said earlier about that this is more for the mechanics. But the interested drivers may jump in and push forward. That is the main, for me, that is the main challenge for the future. To get all of the -- these millions of things running on IPv6 correctly.
>> ANDRES PIAZZA: Thanks very much.
If there is one more comment, I could allow. If not, I just want to also thank everyone for coming. And remind you that the technical community organizations, not only the RIRs, but others like the -- well, the rest of the organizations that are also here, are hosting a reception this evening. And that reception will take place later and there are coaches who will leave from the Northsky Hotel and the North Atlantic hotel. So if you are close to those hotels, you can go by bus to the reception. If not, we have more information at our booth, also.
Thank you very much.
Yes. Do you want to jump Izumi.
>> IZUMI OKUTANI: Just quickly. So we had the best practices forum on IPv6 environment, and I think there were a lot of common learnings from each of the cases shared. And this is actually documented. And so feel free to find it from the IGF microsite. And, well, I think how many of us have contributed in the content. So I hope that will be useful. Thanks.
>> ANDRES PIAZZA: Thanks very much.
(Applause)
(Over at 5: 35, 17:35)