IPv6 (technical) SIG, Wednesday 20 Aug, 11:00-12:40 pm KAZU YAMAMOTO: Good morning, everybody. People sitting over there, can you hear me? If you can hear me, please raise your hand. Thank you very much. This is the IPv6 technical SIG. Thank you all for coming. We’ll explain about the sponsorship at first. All meetings are sponsored by KRNIC. That is our primary sponsor. This meeting is supported by JPNIC. Thank you for KRNIC and JPNIC. Thank you very much. This room is very large and I guess people sitting over there are very shy. It is very hard for you to look at the screen or hear my voice. So would those who are sitting in the back of the room, would you come closer to us please? OK? So this is describing the contents of this SIG. Today, we have five presenters. I'm sorry. My name is Kazu Yamamoto. I'm one of the co-chairs of the IPv6 SIG. Today, I will chair this session and, today, we have five presenters. The presenters are sitting in presentation order. The first speaker is talking about the IPv6 allocation status report. The next speaker talk about IPv6 and the last three speakers talk about what's going on in Japan. So let me call the first speaker. The first speaker is Guangliang Pan from APNIC. He will talk about the allocation status report. I will explain about translators. We have two translators between English and Korean. So, if you want to listen in Korean, please use the headsets. GUANGLIANG PAN: Good morning everyone. My name is Guangliang Pan. I'm a hostmaster of APNIC. Today, I would like to give you an update on IPv6 allocation statements. In my presentation, I will give you an overview of global IPv6 allocations, followed by APNIC allocation details. Then, we will look at the IPv6 routing table summary and APNIC Whois database registrations. First, let's look at all the blocks which I have received from IANA. I remember when I did a similar report in 2002, this table was very simple. Each IR only received one /23. Now, you can see the big change in the last one. As I mentioned before, the data is up to 30 June. Let's look at the IR allocations. So far, all IRs have about 400 IPv6 allocations in the world. RIPE NCC take more than half of the total allocations and APNIC has 113 allocations. It is nearly 30% of the total allocations. Of the allocations in the APNIC region…. This graph shows IPv6 allocations by year from each IR. IR started IPv6 allocation in August of 1999. You can see the growth. It is very fast. One thing here is, before this year, APNIC always make more allocations than ARIN in the same period. You see, always, APNIC is more than ARIN in the same period. However, from this year, ARIN has taken over APNIC. They make more allocations than us. It seems that ISPs in US are more interested in IPv6 now. Let's look at the global IPv6 allocations. So far, we have about 50 economies which have received IPv6 allocations in the world. Japan received the most allocations and closely followed by the US. Germany has the most of allocations in Europe. This is IPv6 Internet Exchange assignments made by all IRs - RIPE NCC also get the most of assignments. LACNIC has not made any assignments so far. OK. Let's look at IPv6 allocations in Asia Pacific region. We have 12 economies which have received IPv6 allocations in AP region. Japan, Korea and Taiwan are the biggest users, with Japan taking up more than half of the allocations. And you have 20 allocations this year. Japan still received more allocations than others. Indonesia started with IPv6 space this year and they got two allocations as well. If your organisation wishes to receive IPv6 address allocation from APNIC, please contact us. This is other Internet Exchange point assignments by APNIC. So far, we have made seven assignments. Five of them are /64. If your organisation received a /64 and would like to upgrade to /48, you can contact APNIC. However, you have to renumber your network. APNIC started making IPv6 critical infrastructure assignations since this year. So far, we have made four and they are for Australia, Taiwan and Japan. Let's look at the prefixes in the global IPv6 routing table. You can see the red bar which represents RIR allocation range. And the blue bar, it is from the other range. You can see most of them allow /32, which must be our minimum allocation size. And some of them allow /35, which was IR previous minimum allocation size. And the longest prefix is /48, which also must be the minimum assignment size. Regarding to the six-point range, most of them allow /24 and /28. As I mentioned before, APNIC has made 113 allocations. Nearly half of them have allowed the /32 to the Internet. Some of them still allow /35. This is including some organisations. They receive /35, but they are upgrading to a /32 and some organisations, they received /32 but still allow /35 or multiple /35. However, we still have above 40% of our allocation not in the routing table. One organisation returned the IPv6 allocation to APNIC. If your organisation changes and you're not going to use your allocation, you can always return it to APNIC. According to the resource IPv6 policy, our members have to register all their assignments in APNIC Whois Database. This included your customer assignment and also your own infrastructure assignment as well. So far, we've found about 1,000 assignments registered in APNIC Whois database and most of them are /48. It is matched with the minimum assignment size. APNIC also provide IPv6 reverse DNS delegation service to our members. So far, we've found 63 ip6.int objects in our database. It is funny. We also found 63 ip6.arpa objects. Actually, they are not the same range and same prefixes. Of course, some of them are same. At this stage, you can register either ip6.int or ip6.arpa or even both. However, the ip6.arpa object is preferred, as you don't need to change it in the future. These are some useful URLs. The first one links to IANA website and gives you the information on the whole IPv6 address space. And the second one lists all the IR allocations. The third one gives you the resource IPv6 policy and how to request IPv6 address space from APNIC and the last one tells you the IPv6 reverse delegation service and how to change the .int object to .arpa object. Questions. Does anyone have any questions? Q. You've mentioned that I have mostly - most of this is registered in /48s. Are there also assignments made in other sizes as well? A. You refer to this graph? Q. Not the graph but I presume from your explanation, most of the assignments are in /48 so, from that, I thought maybe there are assignments made in other sizes. A. Yeah. Is it regarding to the Internet whois database registration? Most of the assignments is /48 and this, this assignment registration is encoding the customer assignment and also the own infrastructure. Like some of them registered /40 or /36 for the infrastructure. And this - some of them is the infrastructure and some of them is for the customer as well. Q. OK, so some are larger than /48? A. Yeah, the minimum assignment size is /48. So no-one can be longer than that. Q. OK, thank you very much. A. You're welcome. KAZU YAMAMOTO: Probably, you are thinking about IPv4 allocation but, unlike IPv4 allocation, IPv6 is supposed to, you know, fix to the customer in /48, so that's why the assignments are /48 is very large. OK, are there any questions? I have a couple of questions about your presentation. You use the critical infrastructure like that? A. Regarding to the critical infrastructure assignment, the criteria for organisations to receive critical infrastructure, and your organisation should be like IR, like APNIC, or NIR like KRNIC or JPNIC or, you know, organisation which provides a service like DNS service, top-level domain name service, and also the country code service like a DNS service, the top-level DNS service. You're either IR or NIR. Q. Thank you. Another one of my questions is, from your record, ISP in United States now obtaining many, you know, IPv6 address space. But I have not had any information that, you know, ISP in United States IPv6 much service. I'm wondering why. A. Yes. Most of them, they just received allocations standardly. Probably that's the reason. Maybe our staff maybe can answer your question but, yeah... Q. Thanks. Could I get some clarification on the question again. It was about the ISPs with IPv6 allocation. A. Allocation to, you know, ISP in United States is now increasing. Q. Yes. A. But I have no - I don't have here about the ISP who actually is using the service so I'm wondering why, you know, ISPs in United States are taking IPv6 address at the moment. Q. At the last … we had many IPv6 workshop and, at that workshop, we had some vendors visit to speak to the attendees of the ARIN meeting and we had some implementers. People there were doing things with IPv6 and talking about the services that they were offering. And there are Internet service providers in the region who are offering services in IPv6 right now and, in fact, ARIN itself has contacted several of them to find out if they have services available for some things that we're looking at doing and they've answered yes, that they have the service available. Also, there are some announcements in the ARIN region about one of the larger users of address space in the region, the US military, and the US Government has released a paper that states that contractors in the near future will need to be IPv6-capable, so I think perhaps there's a reaction there as well, but the registrations are definitely picking up. A. Thank you. Q. Yeah, I have a question to you. (The last questioner) Are you saying that the US military is taking the addresses from global space. Is that from global space for the US military? A. The US military has not received their allocations yet and my understanding is that they will make a request soon but I think the reaction is the address space hasn't been given to the US military. Those numbers do not reflect in any way the address space that's been given to the United States military but there may be a reaction going on in the region to the announcement that they've made that vendors and providers need to be capable in the v6 area so they're looking for the address space themselves. KAZU YAMAMOTO: It's time to go to next speaker. If you have a question from this presentation, please ask at the end of this session. So let me call the next speaker. Next speaker is Tomohiro Fujisaki from the NTT corporation here to talk about the deprecation of site local addresses. TOMOHIRO FUJISAKI: Thank you. Good morning everybody. My name is Tomohiro Fujisaki. I want to speak about the deprecation of site local address. The IPv6 site local address was deprecated. In my presentation, I will summarise the discussion. Then, I will speak about why it was deprecated, with the problem of site-local address, and then I'll introduce the new local-address to substitute for site-local address. Site-local address is used with the unicast service. It is defined in RFC 3513. This address is designed to be used for addressing inside of a site without the need for a global prefix. The site-local address is a 'scoped' address. As a scoped address, the IPv6 is - the local address has link scope and a global address has global scope. The site- local address, the IPv6 address is a central address. It must not go to the global Internet. This is the format of site-local address. The site-local address has 10 bits. And it has 54 bits subnet ID. But the subnet address should be used with 16 … because the site- local address of IPv6 has 16 bits so it should be... with the site-local address and local address. OK. Change graphic about why the site-local address was deprecated. Many problems with the site-local address. First, the ambiguity of the address. The concept of site is very, very not well defined. So this is an example of the usage of site- local address and the link between two sites are called site A and site B. Each site has the same site- local address. It's hard to operate between the sites. This is an address ambiguity problem. The next problem is an address leaking problem. Of course, the problem with the site-local address... the global Internet. For example, the other address may contain the site-local address in its information. So each site-local address links to the global Internet using the site-local address. The next problem is the site-border router itself. The border belongs to much sites and so, each site treats it for its site and the graphics is the same but also the routing table for each site. Moreover, the protocol does not cater for this. So the site border and its routing protocol, so this may be linked to the problem of site border router. OK, the other problem is very simple. If a mobile need in site A with an address will move to the site with the same address, maybe miss local application. So the same program will be having if the nodes are using site-local address. Many programs use a site-local address. But the local address itself is very, very important. So last IETF, it was decided that the local address must use these things. This was for the site-local address. Each point described so please take a moment to look at it. One local address was proposed. This is a unique local IPv6 unicast address and this address meets all previous requirements and this address has global uniqueness and, of course, use for local communications inside of a site and this site is not expected to be routable on the global Internet because the address does not have the application for a global address. This draft will be published as a WG draft soon. This is the form of the unique local IPv6 unicast address. It comprises the prefix, the global ID. It has 10 bits in the prefix. This is proposed for the global ID. There are two types of global ID like these. The first is used to centrally assign and the second is used to locally assigned without any registries. The first is assigned by some registries. Each has a locally assigned global ID and this should have 40 bits of space. This is the replacement of the site-local address. For your information, this is the other space to be used for the documentation purpose. This is the documentation address. Anyone can use this address for technical books, articles and in training material, like this example right here. Of course, this address space can be used to represent any application -- without any application. This address space is also proposed at IETF with this Internet draft. OK, the site-local address was deprecated and the standardisation of a new local address is now in progress. This will be an information presentation about this. Q. I have a question on this. How important - if you go back two slides, that's it, for FD 00/8, which is locally assigned, how important is it that you have uniqueness when you actually pick a number? A little bit important or very important? A. OK. This address was assigned by some registries, so, to use this address space, someone needs to apply to some bits, they need to pay for some... but, in the local address, can be used - Q. I appreciate that. My question was - how important is uniqueness - A. Oh, yes, yes. This address space? In the global ID bit of this address space, it's used locally so it's not guaranteed perfect uniqueness. KAZU YAMAMOTO: Broadly, if two sites combine to one site, the uniqueness is easier. Q. OK, then my observation is that the pool of possible global IDs in that /8 is 240. If people pick purely absolutely random numbers from that pool. After 1.246,000,000 attempts to pool numbers, the probability of collision is greater than a half. So, what I am saying is that, if uniqueness is important in locally assigned space, you cannot assure any useable uniqueness because 1.2,000,000 is a very small number using this … any more, so I would suggest that the draft is quite weak in splitting it in half if uniqueness is important. A. I think you are talking about collision for the entire Internet. But this is just, you know, combining only two sites, you know? Q. My question was though that, if uniqueness is important - and that's the big if - if uniqueness is important, then self-selecting a number, even from a pool of 40 bit, is a very, very poor mechanism. A. Yeah, yeah. Q. And, as long as the v 6 working group and you folk are aware that it's not very good, then go and do it, but it isn't very good. KAZU YAMAMOTO: Any other questions? OK. Q. You explained about the problems of site-local address. One of them was the leakage into the global Internet. What has caused such leakage? It's a quite basic and simple question but I would like to know if there is any identified mechanism for such leakage. A. OK, this problem is from an IPv6 … net 10 address has the same problem. KAZU YAMAMOTO: For example, looking at the routing table, if you are not - if you forget the private address, it can cause it to leak the routing information inside a site. Another example is EMS. Actually, a site with ... maintain two kinds of domain. One is for the outside. The other is for the inside. The inside information leaks. So there is no mechanism. Just by accident. RANDY BUSH: I think this shows the problems we're having with 1918 addresses. It's misconfiguration, either intentional or mistaken. Intentional misconfiguration happens in an operational list. There are continual arguments about ISPs who use RFC-18 address space internally and leak it externally. That's intentionally. "Oh, it's OK. We're doing it." Then tracers won't work, e- mail doesn't work. With site-local addresses, we're trying to solve a routing problem with an addressing solution. We're trying to say, "These addresses shouldn't be routed externally." Making magic address space is a bad thing. We learned this lesson over and over and over again. Q. Could you please simply remind us of what stage in the Internet draft or RSC process or discussion process is this topic? (Displayed on screen) Q. And the next steps are likely to be what do you think? RANDY BUSH: I'm involved in the IETF process. There is consensus agreed on the working group mailing list and in a working group meeting. Be aware that, for those who don't know the IETF, a mailing list is more important than a meeting because it allows everybody in the world to participate. But there is consensus to remove site- local from the RFC and the pending Internet draft. The process and text changes to do this are being discussed. At the same time, one or two people who disagree with the removal are making a loud fuss, but there is very clear consensus to remove it. (Question inaudible) KAZU YAMAMOTO: Could you repeat the question again. Q. I'm sorry. My question is whether the deprecation of site-local will necessarily be followed by the new mechanism for the distribution of /48s in the way that it's being proposed and at what stage is there a separation between deprecation of site-local and the substitution of this assignment mechanism. A. Are you saying the problem is that both are in the draft? A. I think, is one in addr and one is in unicast format? I think there's essentially very clear consensus and agreement between the IETF and IANA and the other RIRs that /48 is the minimum. The IETF is not responsible for anything to the left of /48 and it's policy boundary. I think there's clear agreement on that. I think, if you have problems in that not being in an RFC, a revised RFC at the moment, then you should bring that up with the addressing director of the IETF and we'll see that you get what you need. Is that helpful? A. Yes. KAZU YAMAMOTO: Looking at the site-local, we couldn't remove it for political reasons so we keep it in mind but we make it deprecated. So the deprecation of site-local address is large consensus. So, we think just we should think about unique address at this moment. So, it's time to go to next presentation. Next presentation is from me. OK. Hello again. My name is Kazu Yamamoto and, today, I am talking about the IPv6 show case. This is the N+I 2003, Tokyo. So, first of all, let me explain what N+I is. N+I Tokyo is the largest network show in Japan. It's located in Makuhari, Tokyo, from July 2 to July 4. And you can find information on this website - www.interop.jp. This year, many people visited to the show and, to much surprise, before visiting - people visiting the, you know, network N+I, they checked website of this and the key word search is summarised like this. And the number one key word is IPv6. So this is evidence that many people are interested in IPv6 in Japan. Then, let me talk about IPv6 showcase in N+I. This year, we have, you know, full-time IPv6 show case and we have 10 sponsors and 22 enterprises joining in and today's theme of this year is 'beyond the net'. And this is a picture of the IPv6 show case, a part of IPv6 show case. This is a very, very large booth this year. And we have eight zones, including home network, ubiquitous zone, and so on. Well, today, I will focus on the home network zone and the ubiquitous zone. There is one speaker talking about Feel6 from FreeBit. FreeBit is in the home network zone. Then, Mr Nagashima will talk about Chiervo from, you know, Japan Telecom, joining to the office network zone. So, I'm concentrating on the home network zone and ubiquitous zone only because the rest will be covered by two other speakers. Before I will explain about the home network zone, I will direct you share the picture, the IPv6 Internet with the audience. My picture about the IPv6 Internet is like this, you know. There is graph space Internet and any kind of physical device joining, connect to the Internet, you know. For IPv4 Internet, only computers are connected at this moment. But, thanks to large address space of IPv6, any kind of physical device can join. And because IPv6 Internet is not limited so any device can communicate with each other. You can enjoy that by directional communication. So with the example here that we can control video recorder from cellphone. So, in the IPv6 world, many new abbreviations and many new services will appear, I think. And this is the model of home networks. Here is the IPv6 part of it. This is the IPv6 network. And if this is IPv4 Internet, the intermediate node connecting the home network to the Internet is something. So we can enjoy the bidirectional communication but, in the IPv6 world, this is just a router. So, you know, normal site network can provide information to the outside and the Internet can give access to the normal inside of the network. This is a big advantage of IPv6. So globally, in the near future, we can take a picture outside and directly send the picture to, you know, a printer in our home network and et cetera. OK. Let's go down to the network zone of IPv6 show case. So this is a digital camera. And this digital camera, the memory of this camera is …. This camera can, you know, send a picture taken directly to, you know, over wireless to, you know, image terminal or something. And, of course, this camera supports both IPv4 and IPv6, so you can, you know, typical usage is storing picture to image server or sending a picture direct to a printer over wireless network and, because this is a still camera with microphone and a speaker, if we use a couple of these cameras, we can, you know, enjoy video chat and so on. Next good application is controlling an air conditioner. This is air conditioner made by National which is one brand of Panasonic. This air conditioner, of course, speaks in IPv6 and this is a cellular phone provided by KDDI. Unfortunately, this cellular phone does not speak IPv6 at this moment, nor IPv4. So controlling the air conditioner by this cellular phone is, you know, like this - cellular phone speaking proprietary protocol. KDDI maintain the translator between the IPv4 and the IPv6. So, using the cellular phone, we can access to the air conditioner in our network and the typical usage is, you know, powering on the power switch of the air conditioner just before we arrive my home, to keep, you know, the air comfortable to us, and such and such. There. So, controlling the home network by cellular phone is not a dream any more. We can enjoy that. But there is, you know, the weakest link in the Japanese IPv6 network. So let me explain the show-stopper to the deployment of IPv6. A typical home network is like this. You know, ISP backbone is now IPv6-ready and ISP router is IPv6-ready and home network, including Windows XP or Macintosh now support IPv6. But, you know, home router - very few home routers support IPv6 and ADSL carrier, unfortunately, broadband access server, operated by ADSL carrier, does not support IPv6. So, typical use of camera make IPv6 functionality at this moment in Japan butif you should buy home router which supports both IPv4 and IPv6, they can make use of tunnel to the ISP. Because we have the computer, you know, IPv6 or IPv4 tunnel by manually, it is very struggle for non-technical users. So Alex Yeoh will talk about Feel6 to do this tunnel automatically. So we will listen to his presentation and, if ADSL server has BAS system support both IPv6 and IPv4, ISP can assign IPv6 address to your home router automatically, so we can enjoy functionality of IPv6. So the weakest link, at this moment, is ADSL and home router. This is another example of application of, you know, home network. This is hard links to -- a recorder provided by Sony. Of course, this hard disc recorder is speaking IPv6 and using cellular phone provided by NTT DoCoMo, we can control the hard disc recorder. So we set, you know, to record a TV program from outside home network. And talking about another application in home network zone, there is a cute web camera. So both web cameras is provided by Panasonic. And also Panasonic/NEC is operating web camera and using this kind of camera, you know, we can control this camera using IPv6 from outside home network. So this is very good application for IPv6. OK. Before I go to ubiquitous zone, I would like to explain about mobile IPv6. Because ubiquitous zone mainly focuses on mobile IPv6, I will talk about the advantage of you know, mobile IPv6. In our ubiquitous zone, we are demonstrating the cellular phone in the future. Because cellular phones at the present is using very limited frequency range and the number of users is increasing. So range is very limited. The number of users is increasing. So a cellular phone is - we can use cellular phones for broadband purpose. But considering the wires - both networks have been widely deployed and we can enjoy high-speed and broadband service with PHS and WLAN using cellular phone. So a cellular phone in the near future should choose best communication media for its location. If you use cellular phone, for example, in this room, cellular phone should use wireless instead of cellular phone frequency range. So, you know, because IP is independent of communication media, IP is the natural choice for the cellular phone in the near future. And, because there are many, many users of cellular phones, we have to provide a large space. So we should choose IPv6 naturally because IPv6 does not provide mobility functionality. So a cellular phone in the future would use mobile IPv6 then, in ubiquitous zone of IPv6 showcase, we demonstrate the mobile IPv6. This is a PDA-providing … with Linux software running on it. That will support mobile IPv6. And this is two laptops with a camera-operating system running on them. What is this demonstrating? … First of all, media independency. One is WLAN and one is 100 BaseTX. We are switching the media to show that video streaming is continuing. And we demonstrated in the session that it is speaking voiceover, you know. We demonstrated that, you know, this can accept the session even with this moving around without the DNS and such and such. And if, you know, if you look at its demonstration with the picture cellular phone in the near future, that this is a very interesting demonstration I guess. I think many people in Japan are believing the concept of IPv6, so I think this network opens the door to any kind of devices and new operations. But there is a show-stopper to deploying IPv6 at this moment. The weakest link is, you know, the router, the carrier. So we have to counter that. That's my presentation. Any questions. I would like to go do next speaker. Next presentation is by Alex. ALEX YEOH: Hello. My name is Alex. First of all, I'd like to talk about this particular solution that can be made to tackle the weakest link. Can you hear me now? OK, my name is Alex, representing FreeBit, which is a telco company. We are looking at tackling what I just mentioned. First of all I want to talk about who made it and why make it and what is it? Who makes the solutions? This company like FreeBit and it has been regarded as one of the best ISPs in Japan and the range of services that they provide is for building blocks. They provide services at a minimum cost. They provide website designs, content and provide local services, software and also have a customer support centre to help, which is non-expensive, and consultations. So, FreeBit has been working with over 150 ISPs, so we understand the need. And that's why the design is for solutions and it's to help them transfer their existing IPv4 network into IPv6 solutions. Why we make the solutions? Currently in Japan there's been a change in the network situation due to two factors. The rapid growth factors. The nature and the need for change. Traditionally, with PCs, but recently, a lot of the changes have been made, like the IP telephone with electronic devices. So if you look at the numbers, the numbers for subscribers back in May, and the numbers are approaching to 11,000,000. And the nature of the access has changed. Traditionally, like I said, on this diagram here is only PCs accessing to the Internet, but with the changes right now a lot of people have a home network because of the availability of broad band and the network's always on and people are accessing into the home network from other places such as office, Internet cafes and now device locations. For example, IP Phone, very good example. In Japan right now, they have functions for VoIP. This is the forecast by the year 2007 which is getting to 25,000,000 IP Phone users. Another example is consumer electronic devices. In Japan right now, the popularity with the recorder and with brands like Panasonic and Toshiba, they make these through mobile cell phones. A lot of people from outside the world are able to access the network with these controlling devices. Look at the numbers again. Another forecast for the year 2007, we believe that it will be close to 5.5,000,000. So mobile phones by then will be accessed by remote control, accessed into their home. Like the air conditioner, which was just mentioned. It has proven to be very popular right now. So in Japan, a lot of the Internet enabled phones are common. We have close to 65,000,000 units of mobile phones with Internet access. So why make the solutions? To supply IP Phones, such as the recorder and other devices which will come in the future with global IP, we’ll realise it's not enough. It's why we need the solutions, like Feel6. So what is Feel6? OK, we're working with over 150 ISPs and frequently understand the local ISP needs in the business world. We understand the weakest link and how this problem is changing the pattern. We need solutions to overcome this. This particular solution, Feel6, enables existing IPv4 network to act as an IPv6 by using technology. So, number one is to satisfy the needs of the IPv6. And also it is a practical solution for people to use right now from today. So how does it work? Feel6 technology is based on five standard technologies. DTCP and software based on 64-bit implementation, database as well. So this solution is designed to overcome those hurdles. Upgrading from IPv4 network to IPv6 with minimal investment. Scarcity of IP addresses. Security issues. Hurdle for users to switch to IPv6. In the security issues which with IPv6 is now built in. And also another hurdle for users in Japan is to switch from IPv4 to IPv6 is a contents problem. We'll discuss that later. So how does it work? At the bottom here, there are two elements. Number one is the server - it acts as the Internet gateway. And the client, which acts as the gateway to the home network. So with the server acting as the gateway, this allows existing IPv4 network to obtain access into the v6 Internet world and the other way around is allowing people to access into the IPv6 home network, so they made the structure. In the middle is the IPv4 network and on the top is the Internet. By placing Feel6 solutions, with the server acting as the gateway to the Internet, it's always the client between the current IPv4 network and the area. This network immediately becomes an IPv6 network. So how does it work? Feel6 Client is an installation on the home PC or a router. There are two solutions I will discuss later. The client will access to a server via existing IPv4 network and once this server here will set up a tunnel, allowing the tunneling to go through. So the IPv6 is read from the top of the IPv4, allowing it to go through the existing areas and to the IPv6 Internet. So in effect, what we have right in the middle, nothing's changed. It's still the existing IPv4 network without any change to hardware or any adjustment. Here’s the IPv6 network and on the top of gaining access. So Feel6 is more than just a concept. Like the tunnel server, made by Hitachi, to unwrap packets coming from the home router. Home router made from Yamaha, which sends it through to the tunnel server. In between are the existing IPv4 carriers. There are two solutions. We can install BitBasket6 or we can use the Yamaha router which I mentioned earlier on. With this BitBasket6 it is built in. So the PC would use this. They -t will contact with the DHCP server. And then with this home router is the name which connects to the server. And all devices behind this router, inside your home, will be a sign of the IPv6 trace. So, what we have, right now is the contents problem, like I said before. People are not switching. To overcome that we provide services to facilitate the transition. To speed up the transition. So we've handled the weak link and we also wanted to handle from the users' perspective and that's why we’re coming up with a lot of solutions facilitating people to use it. One of the functions to access PC from PC. This is a browser-based application. By using a browser remotely, it can access to your home network. Another solution, in partnership with Sony, is the recorder here, mobile phone and the PC. Through the Feel6 network, people can actually control their home through the mobile. For example you can pick when you want to record TV programs. The information will be displayed on the mobile, making the access into home network possible. Allowing mobile telephones to send mail into received mail access from the PC. And to make the Feel6 solutions even more practical, we have launched a test bed from March 3. So far more than 50 ISPs have joined. Big names there. Besides ISPs actually, like I mentioned before, Sony with the recorder and Yamaha and Hitachi with the routers. So with the contents, we also look after the users, giving them the facility to switch from the IPv4 platform to the v6. So the numbers here are more than 80% of the Internet users in Japan now can use IPv6 because of Feel6 technology. Right now, over 80% coverage in Japan. So, Japan is ready for IPv6. In the future, more mail will be provided with the IPv6. You guys can use the Feel6 today and by dropping us a line and trying to use it. Any questions? How to control whole networks? If you're interested, please contact me. So I think we should go to the next speaker. TOMOHIDE NAGASHIMA: Hello, my name is Tomohide Nagashima, I work for Japan Telecom. My presentation will be very simple. My company is Japanese ISPs. We also provide many facilities, and from 2001 we provide trial services. In the next year, we provide commercial service, and we also are trying DSL service this year. We provide 100 users with routers. The problem is we can provide the access but we decided to go by the application services. This is the overview of application trial. This application which has a nickname of Chiervo. We provide instant message. The aim is by using... the user will use IPv6. The user will have IPv6. Chiervo also provides IPv6 with applications and get used to accessing the Internet from anywhere online PC is able. Requirement for trial users is windows XP+SP1,with IPv6 connectivity, provided by another trial service or by Feel6. Also compliance with specific rules. I want to talk about the technology trends. We can access only the client to the server. We can access from the server to the client. In this time client means in the home. This is one of the other services - the instant message. We also provide the meeting, using SIP/SIMPLE, the program. This is a view of the instant message. PC 1 and PC 2 communicate using IPv6, so there is no connectivity for the that one. Service to provide, directory service for each person. Easy access ranging from work place to the home. We have a plan to provide this service having no access restriction for all users participated. This is an overview. Also user A with PC2 of user A, they can transfer. In this case, network is private network. Without Global, cannot upload. Only download. This is our media coverage and PR event. From the first of May to October, it is in Japanese only. Wide media exposure and numerous inquiries. Several dozen of major technology websites in Japan. Several media of Korea and China. Interview from vsix-start.net. This is our business showcase. This was explained earlier. Feel6 will be expanding. The reason why with Feel6, we have trial service user. But this application can be used. So if you want to have access to other users, this can use also do using Feel6. In the future, we want to extend to more users. Not only for our users, but more IPv6 users. We also want to do group file sharing. Thank you. YAMAMOTO: Any questions. At the end of this session, you can ask any questions to the speakers. Are there any questions? OK. So I think we should discuss about, but because time is running out, I think we should discuss on the mailing list, if you think the current charter is good enough. We don't want to keep you much longer. Lunch is next door. I just received an announcement that there's also an IP policy lunch. That's started at 12:30 across the foyer. If any of you would like to attend lunch, you're more than welcome to attend the policy meeting. The cruise is on tonight. Please don't forget to be gathered at the lobby at 7pm. And also any help. Enjoy your lunch. Thanks to you who joined this session. Let's close this session with big hands. APPLAUSE