______________________________________________________________________

                           DRAFT TRANSCRIPT

                          IPv6 technical SIG

                       Wednesday 7 September 2005

                               11.00am

______________________________________________________________________

KAZU YAMAMOTO:

This is IPv6 technical significant. This session is sponsored by
Nominum. And this session is organised by three persons. The chair is
me, Kazu Yamamoto. I am chair. And we have two co-chairs. One is
Tomohiro Fujisaki, NTT and the other is Tao Chen, CNNIC. Unfortunately,
Tao Chen has a presentation at NIR SIG so he is not coming here.

Anyway, let's get started. So this is the agenda for this session
(refers to screen). We have five presentations today.

First of all, I will review the action items. We have no action items,
so, if you have any opinion to IPv6 technical SIG, please speak out.
Speak up. (No response) OK, no opinions. So let's move on to the
presentations.

So let me call the first speaker. First speaker is from APNIC. His
presentation title is 'IPv6 allocation status report'. Sunny.

SUNNY CHENDI:

Good morning, everyone. I'm Srinivas (Sunny) Chendi from APNIC. From
those who are new to APNIC meetings, this is the regular update that
APNIC does in APNIC meetings about IPv6 status. In this, I'll be talking
about some RIR IPv6 statistics and APNIC allocation/assignments. Then,
I'll be looking at the routing table. After that, the APNIC
registrations in the APNIC Whois database.

As of July 2005, we have 213 allocations in the Asia Pacific region.
This includes various allocation sizes, not just /32s. Contrasting, as
you can see on the slide, RIPE NCC made 494 and ARIN 152. This is as of
July 2005, the statistics from each individual RIR – LACNIC 34 and
AfriNIC 10. By the way, AfriNIC is a new RIR, just established. Some of
the IP address ranges that have been allocated by other RIRs transferred
to AfriNIC.

Some of you requested to put a slide not just allocation numbers but the
size of allocations that each RIR is making in the world. (Refers to
slide) This is the slide. As you can see, RIPE NCC and APNIC are going a
long way, making IPv6 allocations in their individual regions. We're
close to 1,093 million RIPE NCC have made as of 21 August this year and
APNIC, 1,021 million.

IPv6 allocations received from IANA by APNIC – from the last update we
did, we received three allocations. The last one, as you can see - one
in May, a /19, and one in July, a /19, and recently, last month, a /21.

Allocations to APNIC (refers to slide). So far, as of August, we
received about 34. Compared to last year, the allocation received by
APNIC is growing. From the last presentation we did, the last update, we
made about 36 /23 s as of July 2005.

(Refers to slide). These are the allocations that APNIC made in various
economies. The new additions that we have from the last update made in
Kyoto Japan, is Pakistan and Macau. They each received one allocation
from APNIC. And India received two more allocations, this is mainly
because of initiatives in this region for ISPs to implement IPv6.

Now, if you compare the IPv6 allocations by year, in 2004, APNIC made 53
allocations for the entire year and, if you compare that against this
year, as of July, end of July, APNIC already made 30 allocations and we
have requests for other allocations as well.

Some of the large allocations that APNIC made – Vectantnet - we reported
that in the last APNIC meeting in Kyoto Japan. Apart from that, APNIC
made allocations to Softbank, Kornet and Telstra Clear in New Zealand.
As you can see, we made a /20 to Softbank, /20 to Kornet and /30 to
Telstra Clear. This is based on IPv6 infrastructure.

And there's no change in the expansion of the initial allocation space.

APNIC implemented a new fee structure for NIRs and confederations. This
is an EC decision. NIR members applying for IPv6 will receive a 90%
discount in their fee. This is based on the fact that they have already
evaluated IPv4 request so there's a discount applied for IPv6 request.
And also for confederation members, if they maintain a single pool, the
per-address fee will be waived. But they must maintain a single pool for
their members.

Experiment allocations - we haven't made any allocations to what we
reported in the last meeting but there is a document published now for
this experiment at that URL (refers to slide) In the last update, we
didn't have that document so we only received it after the meeting and
we published it now. So, if anyone is interested, please go and look at
the document and see what experiment they are doing.

IPv6 IX assignments - most of them are made in the economies that there
is an NIR.

Critical infrastructure - again, most of them in the regions where the
NIR is, including one for APNIC. And there's one in Hong Kong. The
routing table (refers to slide). There's a new route that received a
/19, a large allocation, they're advertising a /19. This is obtained
from AS4608 as of 4 August 2005.

The assignments in the Whois database - these are only the allocations.
We have an increase in /48 assignments. They've gone up by 325
registrations in the database. And all the others are registered.

Any questions?

TOMOHIRO FUJISAKI:

Could you please show the slide of IX assignments? Could you please -
the number of assignments within KR and JP, because number of
assignments in JP is 3 but KR is 2.

(Note: JP received 3 assignments and KR 2 but, on the pie chart on the
slide, JP’s portion of the chart is smaller than KR’s)

SPEAKER FROM THE FLOOR:

The proportioning of the pies.

TOMOHIRO FUJISAKI:

Why is JP smaller than KR?

SUNNY CHENDI:

Oh, sorry. That's just a typing error. Thanks for pointing it out.

TOMOHIRO FUJISAKI:

Could you show the slide - routing table, IPv6 routing table. Actually,
I think the graphics - if possible, I would like to know the detail of
the route.

SUNNY CHENDI:

Sorry, I didn't get that.

TOMOHIRO FUJISAKI:

Which graphics... actually, the global routing table should not have the
/35 – it’s illegal.

GEORGE MICHAELSON:

There is no routing restriction that says prefixes cannot be longer than
/35. That is not a globally adopted policy.

TOMOHIRO FUJISAKI:

I want to say, which - the detail of the graphics -

SUNNY CHENDI:

You want to see the IP prefix of each route?

GEORGE MICHAELSON:

If you would like to see a routing table, we are happy to provide you
with a routing table so you can see these prefixes and get more
information about these longer prefixes. OK?

TOMOHIRO FUJISAKI:

Yes, yes. Thank you.

SUNNY CHENDI:

Thank you, George.

KAZU YAMAMOTO:

Any other questions? OK, so, thank you, Sunny. Let me call the next
speaker. This speaker is Sawabe-san, from the IPv6 Promotion Council of
Japan. He is will talk about current activity on research and
development of and deployment of IPv6 products in Japan.

NAOTA SAWABE:

Good morning. My name is Naota Sawabe from the IPv6 Promotion Council of
Japan. And today I will talk about current activity of research and
development of IPv6 products in Japan.

So, first of all, I will talk about IPv6 Promotion Council. And the next
- I will introduce the IPv6 products in our showroom. And thirdly, I
will talk about recent research and development in Japan and there are
five topics. And finally, I will conclude my presentation.

To show you the overview of IPv6 Promotion Council of Japan:  (refers to
slide). So, our promotion council was established in October 2000 and it
is non-profit and non-governmental organisation. It has various members
from worldwide major companies, such as ISPs, and so on. And the number
of members are about 360 now and this will rise.

(Refers to slide) And this shows the organisation of our council. The
chairperson is Jun Murai from Keio University. And our council has some
working groups. There is a network working group and an application
working group, a security working group and certification working group
and so on. And our council has many cooperating organisations, such as
JPNIC, IAjapan and so on. And the Japanese Government, the Ministry of
Internal Affairs and Communications, is an observer.

And this is a chart of activity area of our Promotion Council. And the
application areas are dispersed to solve the problems of real society.

In the next, I will introduce the IPv6 product in our showroom. (Refers
to slide) Our showroom is named Galleriav6. And there are many IPv6
products in our showroom and you can try these products. So please visit
our showroom, if you come to Japan.

And this list is IPv6 products in our showroom. I will introduce these
products briefly.

This is IPv6 digital camera system and home gateway, made by Sanyo. And
images are remotely caught by IPv6 digital camera and images are
September to the home gateway via wireless LAN. And using a TV remote
controller, you can see and share the images and store them in the home
gateway.

And this is a web camera system, made by Canon. And it can be controlled
remotely.

And this is a remote-control-based node, made by Yokogawa. And it
provides to make a reservation of TV programmes from remote PC or
cellular phone.

This is a voice navigator, and it's also made by Yokogawa. It's a kind
of Internet tunnel to provide detailed information by voice which is
stored in the voice server.

This is a field information server, FIS, and it's also made by Yokogawa.
And FIS is a kind of sensor to monitor the weather conditions.

And this is a network solution controller, named Xancia. It's also made
by Yokogawa. And Xancia is a kind of gateway to control and monitor the
status of equipment in the distributed area from allocation. We
demonstrate the fire engine management.

And this is a DVTS made by Dentsu. It is a high-quality DV image used on
the IPv6 network. And it is used in videoconferences, live event relay
and so on.

And this is a videoconference system made by NEC. It uses MotionJPEG and
enables you to see remote images on your web browser.

And this is a printer made by Panasonic. It's using IPP and it's in
IPv6.

And this is a web camera and IP phone system made by Panasonic. The
camera can connect to the network and it is already a commercial product
used in various applications, such as monitoring. Visual communication
can be done using web camera and IP phone.

This is a video chat system made by Toshiba. It is using m2m-x protocol.

And this is also a digital system made by HIC. This system is based on
Playstation2, made by Sony, and it's also using m2m-x protocol.

And this is an IP phone terminal based on IP Centrex system made by
FreeBit. This phone is already small to more than 300 dormitories for
20,000 subscribers for ordinary use. It works on IPv6-based IP Centrex
system.

And now, I introduce the products outside our showroom.

This is an intelligent mobile IP terminal made by Fujitsu. It has a web
browser, videophone and digital television. The CPU is - FR-V is
installed in it. And the size and weight are same as Playstation
portable. And it connects to the Internet - full browser is installed
and it's got SIP, MPEG-4 and H.264.

Now, in the next, I will introduce recent research and development in
Japan.

First of all, the facility networking. Facility networking special
working group is one promotion council. This working group’s goal is to
reduce the life cycle cost of buildings by employing IPv6 network
technologies in an open and multi-vendor building system. And the
working group will standardise the building automation specification and
propose it as an international standard. And the proposal is not
designed to compete with the LonWorks or BACnet specifications. The
resulting specification will be proposed to the LonMark and BACnet.

This is the working group's recent activities: At Interop Tokyo, we
demonstrated the interconnection experiment. This media shows the
demonstration - I'm sorry, it's in Japanese characters. Many
IPv6-enabled nodes is connected using BACnet and SOAX/XML protocol, such
as web cameras, weather information sensors and remote control and so
on. Our second one is VoIP system interoperability task force
interoperability task force.

The current problem in Japan is that IP telephony service is becoming
popular rapidly in anywhere like homes, enterprises and ISP networks.
Although there is less establishment of interoperability because IP
telephony terminal vendors and IP telephony service providers tend to
develop their products or services on their own.

We believe that basic interoperability is not only among vendors but
also between the vendors and the providers will be necessary. In this
situation, VoIP system interoperability task force will be established.
It will be established to prepare the environment to test and evaluate
interoperability and to cooperate for aligns with local and global
standards bodies and to contribute to business activities.

And the task force aims to verify the interoperability at a technical
viewpoint and aims to provide the effort to the standardisation
organisations.

These are current activities of the task force. Major ISP and terminal
vendor joined in Japan and interoperability test has been done four
times. And cooperate with VoiceCon, SIPit and so on and VoiceCon will
take place in Japan. And the contract - the MoU about cooperation with
Multiservice Switching Forum in February 2005. And start the arrangement
with Taiwan about cooperation works.

And the next - the third one is Live E!  Project Live E!  Is an approach
that aims at the achievement of the infrastructure construction that can
use, process and share environmental information. Environmental
information is collected by digital weather station, IP camera, and so
on. They are set up by the individual or the organisation voluntarily.

And digital weather station acquires the weather information with low
cost and it sends the information to the data server via the Internet or
the Japan Gigabit Network. And by the installation of a lot of digital
weather stations, the environmental information can be utilised much
more.

And the development of a new activity is assumed in education, public
service and the business field.

This media shows the experiments in the Live E! Project. And digital
weather station is made up using LonWorks protocol. And it sends the
weather data to the data server using SOAP/XML protocol. And the weather
data is stored in the data server and utilised as - at the school,
public service and office.

And the fourth one is IPv6 Ready Logo Program. This is a certification
program for IPv6 products. It's run by the IPv6 Forum. Certification
working group in our promotion council supports this program, especially
at the technical area, such as test scenarios. Phase 1 started in
September 2003 and phase 2 started in February 2005.

Phase-1 meets the minimum requirement and it aims to encourage every
organisation that will deploy IPv6. It is concerned with core protocol
only. And Phase-2 meets complete requirements. So it is concerned with
core, IPSec, MIPv6, MLD, 6to4, NAT, IKE and NEMO.

And this is the current status of the Ready Logo Program. About 180
products have got the Phase-1 logo. So Phase-2 program started and now
public review of IKE has started. And this is the transition field trial
in Japan in this year. This is the experiment for practical use of IPv6
with products. For example, this is consultation services for residents
(refers to slide). And this is a video distribution service between
educational facilities (refers to slide)

So now, I'll conclude my presentation.

For the consumer, IPv6 telephone and information appliances will spread
gradually. On the other hand, in the field, like the building
maintenance and the sensor network for disaster prevention, the system
using IPv6 will spread rapidly.

We will contribute to spread IPv6 at worldwide scale through the
activity, such as IPv6 Ready Logo and VoIP System task force and so on.

Thank you for your attention. That's all.

APPLAUSE

KAZU YAMAMOTO:

So, any questions?

TOMOHIRO FUJISAKI:

You showed a lot of IPv6 products. Is it commercially available now?

NAOTA SAWABE:

Many of the products are a prototype but some products are commercial.
So, like, the web camera and the VoIP system, the VoIP system.

TOMOHIRO FUJISAKI:

Thank you. Can you let me know which products can I buy? Can you let me
know which products I can buy?

NAOTA SAWABE:

Yes.

KAZU YAMAMOTO:

So, do you have gear which you introduced in this presentation here
today, like the web camera? Do you have it here today?

NAOTA SAWABE:

No.

KAZU YAMAMOTO:

I'm wondering, you know, I'm Japanese, so I can go to the showroom, but
I'm wondering if other country's people here - how other country's
people can know what the demonstration of your kind of products or
services.

NAOTA SAWABE:

Now we have no application can use remotely, but, in our showroom, we
demonstrate many products. Only in our showroom.

KAZU YAMAMOTO:

I think there is a showroom in Korea also.

NAOTA SAWABE:

Yes. We have showrooms in Korea, China and Taiwan and the web camera and
IP telephone. So this is progress.

KAZU YAMAMOTO:

So, it would be nice if people can watch that kind of demo remotely

NAOTA SAWABE:

If you go to showroom in Korea and you can see the IPv6 products in the
showroom in Japan.

KAZU YAMAMOTO:

Any other questions?

(No further questions)

OK. Thank you, Sawabe-san.

KAZU YAMAMOTO:

The next speaker is Jordi Palet and he will speak about security and
auto-transition.

JORDI PALET:

The two presentations I am going to do now are related to one of the
European commission programs that has been working actively until June.
This project finished in June but we are going to continue the
activities because we have work and ideas related to one of them. The
first one is the concept of the distributed security. Actually we are
incorporating other people including people from the security working
group and the basic idea of this work is what happens when we deploy
IPv6 related to the security of the network. We are trying to convince
everyone that which IPv6 we want end to end connectivity all the time
and that it's IPv6 is better in the sense of the security because it has
IPsec already in the stack.

We need to consider we have many more devices and not necessarily PCs as
Geoff introduced this morning. And also that most of these devices
probably will be moving around, will not just be connected to single
network. So looking into this situation we would probably need to find a
new global security model or a new security architecture.

When we talk about security, we are looking at the very, very biggest
scope of security so we are also looking not just into securing the
information but also for example attacks or even virus, spam - all these
kind of things so we are looking far beyond what usually security means
in the sense of pure IP security.

So what we have today is what we call network-based security model. In
this model we have - in the picture here in the screen, we have an
example of two networks connected to the Internet and we have a firewall
in the networks and each of the firewall has a different security policy
defined by the network or security administrator. The firewall actually
is behaving as the policy enforcement point. It is the only machine
providing the security to the network. So if there is any kind of
attack, maybe one of the security policies is protecting against that
one but it could be the situation that another security policy on
another network, it is failing to protect against that attack.

So what happens with this today network security architecture? Well, we
assume that attacks come only or mainly from outside and we assume also
that everyone in all the - in the same LAN segment, is already trusted.
We are assuming that the nodes are not moving outside the network. And
there are not any kind of backdoors. There are no modems or ADSL lines
in this network so attacks will not happen from this way and of course
we assume - that's the main problem - that the host will not be directly
accessed from outside the network so we don't enable this end-to-end
connectivity.

Of course this security model is simple, it is easy to manage. There are
minimum points of configuration and there are almost no protocols that
really need to implement security because we're not allowing any
end-to-end security, just network to network, border to border. However,
this model has a lot of drawbacks. Its essential point of failure in
terms of performance and availability - do not address the threats that
are coming from inside the network even if those threats could be more
dangerous than from outside. Usually the firewall is acting also as a
NAT or proxy and not allowing the end-to-end communication. There is a
need for a special solution for transport mode secured communications.
Certain types of technologies which are being as described today mostly
in scientific communities but probably in the future in business sectors
also like GRIDs - they don't work or don't work in a simple way. We are
preventing innovation with the lack of end-to-end communication.

What we are proposing is the host-based security model. The idea is we
move all the security to every one of the hosts in the network or at
least to those capable of doing the security because processing
capabilities - and then we get the firewall or maybe a different
matching on every organisation decision as the main contour if the
policy of the - the security policy of the network. Again we have an
example where we have two networks and we have also a policy decision
point. So we provide a policy to each of the networks. It could be still
different, even for different host in the network. For example, when
there is an attack, it can fail to attack the network and in some
conditions it can attack one of the clients but this client is at the
same time advising the network that there is a threat and providing
alternative measures to protect against that.

So we have a situation where we have pieces moving around and we have
for example someone in his office which is moving to a hotter spot and
obviously there is not a firewall which is complying with the security
in his office but instead if he's his own firewall, his own security
device, he's able to continue with the same protection and at the same
time move the laptop to the home and still keep going with the security
or even, for example, the ISP can provide a new service which has been
the policy decision point because when new users don't have the
knowledge, they're usually not comfortable with security so you can have
the security configuration but it might be better if you have this
service coming from the ISP or an alternative company. I think I don't
have too much time so I'll skip some of the slides but basically during
the graphics I have already described what we -- described what we want
to do with this new distributed or end to end security model and all the
advantages.


Maybe we will go to the next presentation and make questions together.
Yes, I think this is better. OK, let's go to the second one. In this
work, what we are trying to do is to - OK, looking into one of the main
features of IPv6 is getting IPv6 configured on every device. However, we
don't have - in most of the cases, IPv6 collectively in the
organisations, in the enterprises and at the same time not in the home
networks, in the networks the ISPs are providing today. Sometimes you
have transition mechanisms but we're not able sometimes to use the
existing transition mechanisms easily and the users don't have the
knowledge to dramatically configure these mechanisms. We're trying to
facilitate the users. So following the approach of configuration with
IPv6, we're trying to do an auto- configuration.

We have a lot of transition mechanisms but while we have this auto-
configuration, we're not getting easily IPv6 at all. It is a kind of
contradiction. Even we have sometimes - not just the users which will be
able to config IPv6, we have users where IPv6 won't work because you
don't have connectivity in your local network. We're trying to break
this contradiction.

Basically, the idea is to look into the network in every scenario in an
automatic way when a device is connected and making easy this search for
which of the transition mechanisms will work. So at the end the user or
the device - maybe it is better to say the user will be able to not
worry anymore about a transition and the device at the same time will do
it automatically. If there is any way to get 95% connectivity, that will
be the way but if not to look for a selection criteria and base it on
the connection performance. For example, sometimes we are able to use 64
and this works automatically but maybe the 64 relay is far away from our
ISP network and then not very useful. It would be better to use
something nearby the ISP network but most users don't have knowledge to
look for Google for the best broker we have nearby and go to the website
and configure the connection and so on.

We are trying to make it as easy as possible. I'm not going to explain
all the letters but basically it is we have connectivity so go for it or
alternatively check for other transition mechanisms, which one is the
best one and then configure without any use of intervention.

Today we have already a small implementation looking for different
transition mechanisms like protocol 41 and we have also a server with
IPv6 over UDP and ISATAP, STEP, 6to4. We are looking at domain mechanism
and seeing which is the best in every situation and defining some
parameters for if the user is moving what he prefers. He prefers to take
the same prefix if possible, connecting to the same broker or he prefers
performance even if the prefix is a different one. These type of things
need to be taken into account when doing this also.

I think just to cover this, it is enough. Just to mention one important
thing that we also provided a solution and it has also worked that we're
following in IETF, it is how the device will discover in case of not
native connectivity, who is the tunneling point for doing this
transition? So we have looked into different possibilities for
discovering and we reached a conclusion that our candidate solution is
using DNS  SRV maybe associated with shared unicast. This is our
candidate solution. You can find all the information in the IETF
repository. I think it's better probably to go for questions about the
presentations. No questions.


SURESH RAMASUBRAMANIAN:

You were talking about host-based security model as opposed to
information-based security model as in that's something new that came
from IPv6. It has been around for years as far as I know in security and
knowing that an end to end connection just doesn't matter, it is much
more useful to look at the security of what hosts are connected on your
network. Those concepts are basic network security since probably the
dawn of time. It is not something that came along with IPv6 and I wonder
where you got the model or the idea that it's something new.

JORDI PALET:

We don't say it is something new. We say something that's not been
deployed globally. With the deployment of IPv6, we have a new
opportunity to go into that direction. The point is that we also believe
there is still a need for some kind of broader security. Probably a
combination between both architectures is probably a good way to
approach the situation where today probably 80% of attacks are coming
from inside the network. We really need a change in our today - separate
the architecture.

RANDY BUSH:

Let me clarify Suresh’s question. What's IPv6 gut do with it? It doesn't
solve the problem or do anything different? Just like anything else
about IPv6, it is just the same. You get a couple more bits - maybe.
IPv6 is contributing nothing to this problem.

JORDI PALET:  Well, I don't really agree in the sense that today you
have still IPv4 stacks where you don't have IPv6. It is a problem for
the manufacturers that don't comply with IETF specifications. The real
situation is having more bits as someone pointed out. It means you can
have enough addresses for any number of devices and allow always end to
end communication which also if you -- give you end to end security. I
think that's clear. We cannot do that with IPv4 for every device that we
want to connect to the Internet.  There is no way we can have enough
addresses.


SURESH RAMASUBRAMANIAN:

I consider that a feature. Now when you've got a cell phone and a
toaster and god knows what else connected over v6, APs were designed for
connectivity of the sort you imagine and security is something - end to
end connection secured by the connection doesn't matter when the hosts
are compromised. What's the point?

JORDI PALET:

The point is allowing innovation in the network, allowing new kinds of
services and applications that regard communication. We cannot go for
extending the network beyond the client server model efficiently without
end to end communication.


BILL WOODCOCK:

Suresh, you have to give Jordi this one. Granted IPv6 per se doesn't
actually have any impact on the security issue but he's right that
having more addresses does theoretically make it easy to do real address
assignment which does help deal with the end-end problem and I
personally don't want a session border controller for my toaster.


KAZU YAMAMOTO:

Just an idea, you know.

JORDI PALET:

You mean if we implemented the model - we have some preliminary
implementation but not ready for testing at the time.  There is actually
some implementations from the security - from the wide security angle. I
think they're working on that direction.

KAZU YAMAMOTO:

Any other questions?

JORDI PALET:

Just to add something, we will have a - most probably a BoF which this
topic, distributed security, in the next IETF. It is almost sure we will
have this meeting trying to go for a new working - if anyone is
interested in this, should try to join us and participate into that work
and also I want to say it is a kind of advertisement but not for
anything commercial. We have two publications which are free and I
wanted to carry a lot of them so you can get your own physical or
printed copy. One is IPv6 broadband and the other one is legal aspects
of IPv6, what are their implications from the legal point of view for
example from privacy and all kinds of things for data protection. Any
way you can download the PDF for this presentation because I said I was
already with too much baggage so I was not allowed to carry extra books
with me but you can download the books in PDF at the IPv6 www.
ipv6tf.org. Thank you.

KAZU YAMAMOTO:

Let me call the last speaker. The last speaker is Toshiyuki Hosaka.

TOSHYUKI HOSAKA:

My name is Hosaka from JPNIC. I will present JPNIC IPv6 registry
service, from May this year. This is brief history. In 1999 APNIC
launched IPv6 allocation service and one year later JPNIC started IPv6
request agent service for our members. This year, 2005, May 16, JPNIC
launched our own IPv6 registry system. So what is JPNIC IPv6 service? It
is completely the same as APNIC is providing to its members. We are
receiving requests from JPNIC members for allocation requests and
subsequent allocation. We also receive database registration including
allocation, suballocation, assignments and contact information
registration. We also received reverse DNS request from our members and
we are also providing Whois database services to public. The most
important point for the Japanese community is that these requests can be
done in local or Japanese language. And our host network information is
registered both in Japanese and English so from overseas you can view or
you can query the English information regarding certain Internet
information. So don't worry about that.

So before we launch our JPNIC IPv6 registry service, all the information
of our members is registered in APNIC Whois database so we have to
transfer the objects from APNIC Whois to the JPNIC Whois database. In
April this year, we send an email to our member to ask whether to choose
JPNIC management or

(Approximately four minutes’ worth of text are missing here, owing to a
server crash)

You can see the IPv6 allocation and administrative contact features
admin-c in JPNIC. This also shows less specific info and more specific
information in the port. We have not registered those informations so
the slide here shows no match.

So this is the result when you query the group handle, JPNIC group
handle. You can see the contact e-mail address and telephone number and
fax number. So you can reach them.

So I will show some statistics after we implement our own IPv6
registration system. So this is the allocation statistics.

(Refers to slide) This shows the number of allocations to our members.
We don't have significant - we implemented IPv6 registration system in
these statistics but we had significant growth regarding the assignment
records before we implemented our own IPv6 registration system, we
normally have over 7,000 /48 registrations in the APNIC database. But,
after we implemented our own IPv6 registration system, some specific ISP
started registration and we have over 230,000 /48 registrations just
now. This has been a significant growth.

So that's it. Any questions?

TOMOHIRO FUJISAKI:

I have a question. This shows the assignments in the database. Is all
the data related to JP transferred to the JPNIC database?

TOSHIYUKI HOSAKA:

Yes, it is.

TOMOHIRO FUJISAKI:

Thank you.

KAZU YAMAMOTO:

Any other questions? Do you have any recommendations to other NIRs from
your experience?

TOSHIYUKI HOSAKA:

This is the each NIR decision so we have not any recommendation to each
NIR. But according to these statistics, apparently we have a lot of
registrations to the database so, if each NIR believes their own IPv6
database is good for the registration matter, it could be recommended, I
think.

KAZU YAMAMOTO:

Any other questions? OK, thank you very much.

APPLAUSE

SURESH RAMASUBRAMANIAN:

It is not a very critical thing but, if the Whois released by JPNIC can
be standardised so that we can, for example, get it to be possible - I
think JPNIC had a... what I am actually trying to do is to build
something that would get bulk Whois admission so that we get contact for
our various reporting systems. So, if the Whois record output could be
standardised across different NIRs or at least made very similar to what
the other RIRs, it would be a little easy.

It's just a request, not a question.

TOSHIYUKI HOSAKA:

OK, the comment is the - we somehow should be standardise the output of
Whois results, across RIRs or NIRs? Is this correct?

SURESH RAMASUBRAMANIAN:

Yes

TOSHIYUKI HOSAKA:

I think that is a good comment, because everybody in this community uses
Whois for a network query, so that is worth considering.

SHIN YAMASAKI:

I'm from JPNIC as well. Let me just add several comments for Suresh,
your question. So, let me ask you - what actually would you like to have
standardised please? I think I know what you are saying, but -

SURESH RAMASUBRAMANIAN:

If we have a standard -

KAZU YAMAMOTO:

Use the microphone please.

SURESH RAMASUBRAMANIAN:

If there is some initiative towards building a more standard format for
the Whois - I don't know how to customise our scripts for passing Whois
data on RIR and NIR basis, it makes it easier for my scripts, I can read
it using Whois But, again, it's much easier for my scripts if it is
standardised, the various Whois.

SHIN YAMASAKI:

OK, thank you. Yeah, that's a good point. I guess quite a few people
have to change their script when we released the new Whois screen so I
understand what you are saying. So actually, people in this community is
working for CRISP, new standard, so I believe that technology can be
used for that purpose so we are keep doing - keep looking for that kind
of solution for future. I hope that will meet your requirement for
consistency and standardised Whois. Thank you.

KAZU YAMAMOTO:

Thank you, Hosaka. So, all presentations are finished. Before finishing
this session, I would like to provide housekeeping information, OK?

Again, the sponsor is Nominum and the important thing is Nominum lucky
draw is iPod, OK? So, if you want to get it, you need to fill in a form
at the APNIC help desk on the second floor. You need to fill it out by
12:00 Friday. The winner will be announced at the member meeting on
Friday. So please fill in the form.

The next thing is the APOPS BoF. The room has changed. The room was
Function Room 2 but the current room is Function Room 7, so please take
care. And the social events - the social event is Vietnam Museum of
Ethnology. You can watch water puppet show and you can have dinner. It
will be held on Wednesday. The first bus leaves at 1830 and the last bus
leaves at 1910. So you should not miss buses. It's sponsored by VNNIC. A
ticket is required. You can find the ticket in your tag. It's in your
name tag, OK? And this is open-air event, so APNIC recommend that you
should wear a long-sleeved shirt and long pants, OK.

And do you know MyAPNIC? Do you have MyAPNIC accounts? If you don't know
about MyAPNIC, you can watch demonstration that is available all day at
the help desk. Again, help desk is located on the second floor and that
is available at break times.

So we need to go to lunch. Lunch is on the Grand Ballroom 2. It's on the
first floor. This is Grand Ballroom 1. It's on the same level. This is
not El Patio, where we had lunch yesterday, OK. So we can eat the lunch
in Grand Ballroom 2 from today to Friday, OK, that's it. So any other
issues?

OK, I would like to close this session with big hands. Thank you very
much.

APPLAUSE