APNIC 32 - Destination::IPv6

Transcript - Opening plenary


Due to the difficulties capturing a live speaker's words, it is possible this transcript may contain errors and mistranslations. APNIC accepts no liability for any event or action resulting from the transcripts.

Monday, 29 August 2011
09:00 (UTC +9)
Opening Plenary

Dr Eun-Jung Park: Welcome to the Opening Plenary.  We are going to invite the speakers for the keynote speech.  We are going to have the Chief Scientist, Geoff Huston, and Prof Chon Kilnam.

First, let's invite the Chief Scientist, Geoff Huston, for his keynote speech.  Please welcome Geoff Huston.

Geoff Huston: Good morning, all.  My name is Geoff Huston, I'm with APNIC and it's nice to see you all here in Busan.  Thank you to our Korean hosts for having us here, it's a delightful place.

This morning I would like to talk about something that has taken up my life for many, many, many years, and that's this thing that we kind of thought would happen way back in 1990, when the Internet was only just taking off.  Even then, we realized there were two problems out there -- the routing space might come into meltdown, and inevitably we were going to run out of addresses.

So I would like to look at that address exhaustion question again today and take a little more time and see where we have got to at the moment.

In this telecommunications industry we've been doing this gig for years.  If you go back far enough, you find that in the 1830s we figured out that if you banged electrical current down a copper wire, you could make something jiggle at the other end -- presto, up came telegraph.  We managed to span the world with wires and bring the world closer together.  Instead of taking six months to send a message from Europe to Asia, across an unreliable boat, we could do the same in a matter of days, with this new telegram.  The world got incredibly smaller very, very quickly.

Then we did it again with telephones.  After the first exposition of the telephone in the Philadelphia World Exposition in 1876, the rush was on.  Everyone who was anyone had a telephone handset within a mere decade.

The explosion of technology probably makes even the Internet's explosion of technology in our lifetime look dull and ponderous.  The mania of telephony was all-embracing.

The telephone history is rich.  They did so, so much.  They effectively created a fantastic, incredible system.  By the end of the 20th century, the telephone sector alone, the telephone companies were the largest employers in every single economy -- AT&T alone had more than half a million employees -- and they were the richest sector we had ever invented.  It was our major achievement of that century, as far as I can tell.  A very rich history.  But that doesn't mean they were right all the time.

Some of their mistakes were truly awesome.  I always liked ATM.  It was a brilliant idea but completely and totally flawed.  The idea that somehow I could make my time division multiplex slices self-identify and this would be the answer to how to share wires was a dismal failure.  Even their dream of how to make the telephone network handle the up and coming computer data revolution, the wonderful dreams of, "You packet guys have got it all wrong, the future is ISDN."  Yeah, right.

What worked?  What worked were things that they never even thought would ever take off.  SMS was brilliant.  SMS was never meant to happen.  You were never meant to have thumb arthritis from keying letters and words on a 10-digit number pad.  You weren't meant to do this, ladies and gentlemen, it was a mistake.

When they first brought it out, SMS networks all over the world collapsed.

IP was also a mistake.  It was never meant to happen.  It was merely an experiment.  You're very erroneous, you folk, you should be running open systems interconnect, the real way we were meant to do broadband.

Indeed, over the last couple of decades, most of the things that industry did deliberately were hideous mistakes.  Most things that have worked and shape our lives today, right the way through to Twitter, were mistakes.  They were never meant to have happened.  They were surprising.

Given the track record of this industry getting it hopelessly wrong deliberately, we are now faced with a really interesting decision.  How do we manage this transition as an industry?  Because this is a difficult problem.  This whole transition from v4 to v6 has many imponderables.  My suspicion is that the risks of getting this wrong are quite high -- higher than the probability of even getting it right.

Let's have look at this and see if we can figure out if this industry is indeed going to call this particular transition a success.

The first question is:  Surely you've heard and I've heard, for years, "I don't need to worry about this.

What are you worrying about?"  I'm saying, "You've run out of v4 addresses!  There is no choice, guys, you have to do v6."  The fact that the amount of v6 out there with clients is not 20 per cent, is not 3 per cent, at best on a good day, going downhill with the wind behind you, it's 0.3 per cent of clients choose v6 by preference.  "Oh, no, it's okay, Geoff, don't worry, it's just a matter of time.  It's going to happen, it's just inevitable.  When exhaustion really bites then this industry will move.  Stop panicking.  We've got this under control."

I worry about that kind of message.  I worry about the message I hear from the big providers: "It's okay, we've got plans.  You might not see us doing v6, we might be sending reset messages arbitrarily in our TCP resetter" -- I heard this morning -- "but it's OK, don't worry about it, we have a plan."  Maybe we should worry.

Maybe we should be very, very concerned.

Because I'm not sure that technology really works the way we're told.  You can look back at some of these things and go, "I'm sorry.  It was inevitable.  There was no choice.  It was just going to happen.  Don't worry."

The original telephone networks couldn't share.  If I wanted to have two conversations happening into my building I needed two wires; three, I needed three wires.  If this picture of New York is anything to go by, by the time everyone wanted a telephone handset, there was way more traffic up in the sky than was happening down there on the ground.  And when it snowed and it all came tumbling down, everyone must have taken a holiday until they rigged all the wires back up.

The progression from wires to virtual circuits, when we look back 100 years later, goes, "That was just easy."  All of a sudden we figured that we could do frequency division multiplexy and time division multiplexy and share the wires and, instead of having a forest of wires, we could do it all with one wire and virtualize the circuit.

I would like to have a cute photo of a virtual circuit, but it's white.  Just imagine it.  Virtual circuits have their problems, too, as we have lived through.  The world of circuit switching is an expensive world.  As the world gets bigger, the costs don't magically decrease.  Volume based industries don't work very well because circuit setups take time and effort.

When you are switching time, you need hideously accurate clocks.  As you try to switch time in finer and finer detail, your clocks have to oscillate at megahertz, at gigahertz, at terahertz.  As you get bigger, it just all gets harder.

At some point you give up.  It seems so inevitable now in 2011 that the switch to packets was a no-brainer, it just had to happen, because all of a sudden the packets had to tell the networks what they were doing, because it was a waste of everyone's time to make the network more complicated.  It was not a winning proposition.

So technology was inevitable, wasn't it?  Oh, yes.

So the transition to IPv6 is inevitable, isn't it?  Oh, yes.  Or maybe no.

Because whenever you do these jumps -- I'm on one surfboard in the middle of the wave and I want to get on to the one beside me -- that jump is a leap of faith.

That transition gets awfully messy, it's not clean, particularly when the new technology isn't seamlessly backward compatible with the old one.

Because to get from here to there, we have made life hard, harder than the original protocol designers ever thought.  We were meant to have finished this transition.  This room was meant to be IPv6 connected.

We were meant to be talking about IPv4 and museums.

We're not.  You are all running v4, every last one of you is sending v4 packets.  Some of you, and only some, are sending v6.

We are going to wait to up the challenge level to degree of difficulty 10.  Because we're not going to transition when there are heaps of v4 addresses around, we are going to wait until there are no more addresses.

We are going to make this transition mind-bogglingly hard.  We are going to stress an industry that makes us money by mindless process, by forcing it to be creative.

You don't want this.

Because all of a sudden we are going to make an excursion in transition by, instead of making our networks simpler, clearer, cleaner, cheaper, we are going to re-equip them with the same paraphernalia as we had in virtual circuit worlds.  We are going to start equipping our networks with a whole bunch of novel technologies that we only ever tried before at the edge.

Carrier grade NATs are certainly wonderful things, aren't they?  We've never really had any experience at how they load, we have no real idea of what the compression factors are going to be; not now, but in four or five years time when life gets serious.

All of a sudden, we are going to start erecting stuff in the network.  That stuff is actually rationing devices, scarcity devices, apertures that dim your vision of the network, that reduce the clarity of that conversation between one user and another.

Then, just to make it worse, folk are going to say, "That's not good enough, I want to get my content near to those users."  The whole world of content distribution networks is going to start to take an added impetus, as folks say, "Look, I can't sit there in a datacentre, because folk can't see me, I have to replicate my content inside every single network."  All of a sudden, networks will change shape.

If we still persist and still want to increase the degree of difficulty, we have an instant answer way back from the 1980s, yet again; application level gateways will rear their ugly little heads, because at some point the carrier grade NATs won't scale, true, and if we are still in a world of denial we will head into application level gateways.

All of this should send phenomenal warning bells, because we are not one industry, we are a lot of separate components.  There is a carriage industry, there's a content industry and, even inside there, there are folk who do search, there are folk who do social media, there are folk who distribute large amounts of data, folk who just make instant on-time services.

What about this transition from the point of view of the network provider?  Over the last few years, they have been told, "Get out of the way, don't adorn the packets, just get rid of them.  Make your network cheaper.  Stop investing money into the network, invest less, reduce the price, users will love you, content will love you, everyone will love you, get out of the way."

All of a sudden, we are saying, "We didn't mean that.  Spend money, spend large amounts of money.  Put more gunk in the network because we have run out of v4 addresses, spend enormous amounts of money."  Then next year, or maybe the year after if this IPv6 transition works, we'll say, "We didn't mean it.  Write all that money off.  It was a bad mistake.  We've done IPv6 now, we don't need your investment any more."

Are they going to do that?  Oh, no.  This is a large industry and we are talking billions of users.  They don't make decisions lightly and their investment life cycle for equipment is measured in decades.

If they start investing in network infrastructure, it's not just for tomorrow or even next week; this is a long-term strategy they're embarking on and they won't come out.  Maybe they will never come out.

Maybe once they have made that investment and maybe once the content industry has been given no choice but to give up on end-to-end coherency, to give up on the model of network neutrality, maybe once we have all given up we will head in a completely different direction and there will be no more v6, because no one will have remembered that was the aim.

The risk is that we head in an entirely different direction.  I don't want that.  I'm sorry, that's not what I've been working for for the last 20 years.  I'm not sure all of you want that either.  Some of you might.  If you're working for a large carriage provider, maybe this is the way ultimately you will make back all that money Google stole from you.  Maybe.  But I don't want it.  And I suspect most of you don't want it either.

How can we manage this transition and ensure that collectively all of us head towards v6 and we keep that objective in mind as we go?  How do we ensure we don't get distracted by optimizing gunk in the network?

Because we are inventive people.  We have a short attention span but we are very inventive.  "I can do a better CGN than you, my one's bigger, faster, better."

"Oops, he's done an even bigger one, I have to study CGNs now."  All of a sudden, as an industry, we lose sight of the long term for the price of optimizing the short term.  All of a sudden, we're there perfecting stupidity in the network.  Because we're like that as people, we're easily distracted.

How do we ensure we don't get distracted by these transitional things?  This is a hard question.  I would like to point out just how hard it really, really is.

We have a number of challenges in keeping our eyes on v6, and the first challenge is, this is no longer a world of monopoly telcos telling us what the plan is.

There is no one in control.  This is not a single industry with a single voice confronted with, will we do v6 or not?  No more is there one person out there in any country saying, "I'm in charge of the telco, it's my call."  Oh, no.  We deregulated.

All of a sudden, all of us are stakeholders in this industry.  There are many different players.  What Google wants is not necessarily what Verizon wants or in this context what Yahoo! might want is not what Korea Telecom may want.  All of them would like my money, I'm the consumer.  They are all competing for that money.

They are not necessarily friends in that competition to see the inside of my wallet.

So, as we explore uncertainty, each of us in our industry will head off in different directions at the same time -- otherwise known as chaos.

What we are doing is exploring all possible solutions simultaneously.  That's a challenge.  Because as we deregulated this global telecommunications industry, we lost our plan.  There is no plan.  There's no direction, there is no script.  The future is no longer inevitable.  There's just the interplay of market forces.

If you think those market forces are predictable, if you truly believe that, go and play the markets and make your billion dollars.  The rest of us haven't quite figured it out yet, because market forces are weird, unpredictable and erratic.  Consult any financial analyst and tell them what happened over the last few years.  Financial markets and market forces are strange.

And we are subjecting the Internet to them.  And the outcome is erratic.

But let's make it harder.  Let's up the pressure.

We are not all going to exhaust at the same time.  APNIC exhausted effectively around 20 April.  That's it, the end.  We are now down to the last /8, we have now flatlined at the bottom.  The other RIRs haven't.

Around the other regions, as you see, they are all coming down at different paces.

If you extrapolate this forward, it seems highly likely that the RIPE NCC will come down to its last /8 in February next year, 2012.  Fascinatingly, the other really large community in this planet, North America, at this level won't reach its last /8 until two years later -- two years later -- oddly enough, at much the same time as AfriNIC and LACNIC.

Even if I do see some certainty there, things look strange.  RIPE next year, LACNIC, AfriNIC and ARIN two years later.  ARIN changed its policies earlier this year.  At the time IANA ran out in February, ARIN started the ARIN community, put forward a policy that said, "We will use three-month allocation windows," which sounded fine.  But if you take just that data and push that forward, if the last few months is an indicator of the future, that new trend line actually says that ARIN will eke out an existence until it gets to the last /8, ooh, some time in late 2017.  It's bizarre, but that's the way the numbers look.

So I come to this kind of model of human behaviour.

If you talk about climate change, you get to much the same kind of conclusion, that sometimes accepting what you see as reality is really, really, really hard, because the general reaction of almost everyone who is human is the ability to truly empathize: it's not happening unless it happens to me.  Until it happens to me, it doesn't happen.

I see articles in Network World, even as recently as a few weeks ago, talking about IPv4 exhaustion as a possible future.  I sit here in APNIC land in this community, going, "I'm sorry, 'possible future'?  How about 'definite past'?  It happened."  "Oh, no, it didn't happen to me, therefore it hasn't happened at all."  So, "No, it's not happening until it's happening to me," appears to be the problem out there.

The fact we are all going to hit the wall at various times means we are not working together.  That's a problem.  Exhaustion will occur variously over many years.  And that is the cause of yet another challenge here, around regional diversity.

Let's look at this transition again, as little balls on a track, little trains running around.  APNIC has entered the weird dark world of transition and, instead of leaping straight across to v6, because most of the players in this region feel they can't run an IPv6 network and damn the consequences, all of them are forced to continue running v4 networks.  So they have to enter transition land and have to start working around CGNs and ultimately where that leads.

The RIPE NCC is getting hideously close to exactly the same position for their community as the NCC itself runs out of addresses to get European and Middle East communities in the same boat.

Let's run it a bit further forward, because in early 2012, LACNIC, ARIN and AfriNIC, by the current projections, aren't there.  Their reality is different.

Run it forward further and even up into 2013, all of a sudden we are living different lives.  The realities of exhaustion are occurring differently in different regions.

So I'm the manufacturer of equipment and I look at who is ordering what kind of gear from me.

Fascinatingly, the kind of equipment that I see being asked for in the Asia Pacific market is not the equipment being asked for in, say, the North American market or not even the same equipment as the European market.  It's no longer one Internet, it's no longer one market, it's a whole bunch of differing markets, all because of the case that exhaustion is running at different time scales.

What's the risk we no longer do the same thing any more, we no longer head to the same objective?  What's the risk of a single Internet inside all of this; the coherency of one single architecture that binds all of this stuff together?

You know, long-term planning is great.  I remember the whole five-year plan, and I keep on hearing, "Transition will take many years, five years easily, 10 years, yes."  What plans did you have back in 2006?

I don't remember mine.  Honestly.  I ask my favourite telco, "What were your plans?"  "That was a different CEO.  That was a different something.  Those are old plans.  Those aren't plans any more.  Our new plans are terrific."

Nobody remembers the old plans.  We are off in a different world, and this is always the case.  This industry does not long-term plan, and the longer the period of transition, the higher the risk the original plan is completely forgotten.  We will be heading off perfecting new realities because we are all experiencing the world in different ways.

Oddly enough, each part of the globe has a distinct risk of building on an IPv4 platform and perfecting their world in different ways.  Up there in the top right somewhere is the vision of v6, receding and dimming.  Because we have lost the plan.  We are all heading off in different ways.

Market pressures are terrible because each of us react differently to these differing pressures and there is no longer one Internet any more.  The longer the transition, the more probable we fracture.  I don't want that.  I don't.  I'm pretty sure you don't either.

So what can I do and what can you do to help things through?  How can we avoid at least making this worse and possibly make it better?

I would like to leave you with just three thoughts here.

This is one Internet, not five.  This is one community, not five regional communities.

As you consider various policies around addressing and distribution of addresses, don't think about what's good for my company or even what's good for my country or even my region, think about how you can align the larger interests with your interests.  If we can continue with that alignment, we have a chance of one Internet at the end of this.  If we fracture, if we diverge to the point where we see ourselves in an adversarial system, we are in a terrible place, and I'm not sure how we get out of it.

We have to think about common interest and we have to determine to find ways for local interests to converge with common interests, or we won't have an Internet at the end of this.

People react to scarcity in very different ways.  In general, if I say that all the food shops are going to lock up tomorrow, you would all run down right now to the supermarket and get as much as you can.  Scarcity and hoarding go together as natural reactors, natural couplets.  It's the worst possible thing we can do right now.

Don't hoard addresses.  Don't try and prolong the pain.  Because if you eke out addresses and give everyone not what they need but just a little bit, and extend the pain not just for two years but for 20 years, we will perfect, as humans, pain.  What we will end up with is a network that will have no resemblance to v6 because we are not committed any more.  Extending scarcity simply prolongs the pain and, increases the uncertainty of the entire process.

Use the addresses, get them out there, get this through with.  Because the best chance of getting to IPv6 is the shortest possible transition we can engineer.  Because the more time and money folk invest in employing more and more grandiose bits of their network to compress addresses and ration connectivity, the more they will enjoy it, the more they will want to see returns on that investment and the more reluctant they will be to take that equipment and junk it next year.  Don't make us perfect bad solutions.

Make this fast.  If we are truly committed to a single and coherent v6 Internet then we really have to make this transition as fast as possible.

As an objective, just think: I don't remember what five years ago really was in terms of planning in this industry.  In five years, you won't remember this.  If you want v6, you've really got to make transition happen, even well within that timeframe.  That's your challenge.  Thank you.


Normally I would be tempted to take one or two questions, but I am getting into Prof Chon's time, so I will hand it over to him.  Thank you.

Dr Eun-Jung Park: Please welcome Prof Chon Kilnam, who will give us a second keynote speech.  Please give him a warm round of applause.


Prof Kilnam Chon: Good morning.  Welcome to Busan.  This is one of the best beaches in South Korea, if not in the world, and I hope you enjoy outside, on the beach.  The hosts took some trouble to set up the Open Office instead of a Microsoft Office, I have a reason, and if we have time, I will let you know.

Today I am going to talk on the future Internet and Asia, including IPv6.

A challenge we are having in this millennium for the next 10 or 20 years is connecting 1 million schools in China -- China has about 1.2 million schools -- or India, or Africa, using broadband.  That's what we have here in South Korea today, or in the case of China, if you talk about Beijing, Shanghai, the metropolitan area, yes, you do have today.  But throughout China or throughout Africa, no, we don't have today.  Is it possible or how can we do it?  It is a challenge for us in this future Internet.

Before I get into this topic, talking about APNIC and me, I chaired the first APNG formal meeting in 1991, 20 years ago.  Then, one of the agenda was the spinoff of APNIC.  Would you believe, this was the first APNG meeting and we discussed spinning off the APNIC, and we approved.  And here you are, 20 years later, it is still in good shape.  And if you look at it some more, this is what we did in roughly about 25 or 30 years, starting from the ANW, if you remember, in 1984, or the PCCS, 1985, then in 1991 we created APNG, then we called it the APCCIRN.

Then we made a bunch of working groups and each working group was on its own, like APIA, APRICOT, in the case of security, APCERT, and the satellite is APIII.

Yes, we made many of those organizations in Asia, mostly in the 1990s, and that was the contribution to APNG, including APNIC.

The Internet research I was involved in since the 1960s, and in the 1970s and 1980s we have what was called the Arpanet, and then we looked at how to bring this Arpanet technology into Asia.  You may call it the other thousand, because at that time there was 1,000 or 10,000 Arpanet users in the USA.

Then in the 1990s was the Internet for all mediums, and today we have about 2 billion, and now we are looking, for the next 20 or 30 years, what kind of Internet we are looking into, so that we are doing some research.

To be more specific, American NSS came up with this idea about future Internet research in the mid 2000s, 2005, 2006, and they came up with a lot of issues -- security, the current Internet doesn't have the security the US Federal Government needs in this century; and the mobility is something we didn't think about when we created the Arpanet in the 1970s; and the scalability, as you see, and IPv6 is one of the issues for scalability; and we are looking into the billions and if you look into the Internet of Things then we are looking into the trillions.

Then the Europeans, in particular AFP6 and AFP7 flagship projects, came up with applications like the future Internet use for eHealth or eGovernment, all those e or smart applications for using the future Internet.  Also, in the last about five or six years we are coming up with new technology, like cloud computing or Internet of Things or the Green IT, what you may call the Green Internet, or the Internet for the green society.  That is the second group of requirements.

Then some of us are looking at the new users.  When we started this exercise five years ago, we had 1 billion Internet users, which we called the first billion.  Let's look at the other billions, those remaining.  Today we have 2 billion users and the implication is that there will be 5 billion Internet users.  Within 10 years, we will have about 5 billion users, means about two-thirds of the world's population, and it means the majority, a super-majority of the world population will have Internet access.

If we are looking at how to access the Internet, this is particularly important in Asia Pacific.  We have a paradigm shift.  A long time ago we used a desktop, which doesn't move, either using a CRT terminal or desktop PC.  In the last 20 years we have been using a mobile version, typically the one you have in front of you, or a notebook computer or laptop computer.

Then fortunately -- I should say fortunately -- in the last four or five years we started having a phone which is capable of handling the Internet, including the smartphones, and the tablets.  It is very likely the majority of the other billions, those new Internet users, will use smartphones only.

The reason is typically people cannot spend about one month's income for the buying of the device.  So if you are American or Japanese, probably you could spend $10,000, but if I ask you, "Can you spend one year's salary, $100,000, for Internet access?", probably no.About $10,000 is painful but you could pay.  But for developing countries, one month's salary is $100, so that's all they can spend.  What can you buy?  No chance you can buy a smartphone and a tablet or a laptop computer.  You can have only one.  Then the choice is very obvious; a smartphone.  So we really appreciate this telecom industry or what this community can come up with smartphones.

To summarize, the Internet users will be about 7 billion, and the Internet users, which I call the first billion, are us, 2 billion today.  90 per cent of non-Internet users are in Asia and Africa, and within 10 years we will have 5 billion Internet users, and naturally most of those will be new users, either in Asia or Africa.  That's the homework or challenge we are going to have.

Looking at the future Internet technology more carefully -- in particular for the other billions, the remaining 5 billion people -- current future Internet research which is being carried out in North America, Europe and this part of north-east Asia is focusing on the issues raised among the developed countries or for the first billion.  Are they doing enough research for the other billions?  Not much.  Until they have to use up their tax money for their people, they are not doing future Internet technology for the other billion.  Who is going to sponsor?  Who is going to fund?  That is the issue. This is particularly for North Asia, including Africa.

Looking to the issues for the other billions, in a typically emerging economy, first of all, their infrastructure is very weak, their fibre optics backbone is not a given.  For developed countries it's a given.

But for the developing countries, in the big cities you can have fibre, but otherwise you can't.  The implication is you should be able to have a backbone based on wireless technology.  That's not too easy.

Human resources -- again, we have an extreme shortage of human resources among the emerging economies.  Suppose if there was a system in a remote village, over 5,000 people, how can we find the network engineer who can manage the server and the network?

Access capability: as I told you, in particular these are either smartphones or laptop computers, because they can spend about $100, not the $1,000 or the $3,000, they cannot have both.

Can you write a paper, a 10-page term report, using a smartphone?  It's a challenge, not easy.  So we have to develop some technology in such a way that, even if you have only the smartphone, you can handle whatever you need.

Then we have the local language and the local culture support.  This is something that each country, each culture has to solve by themselves, because otherwise what will happen is -- which is happening in many parts of North Asia and Africa -- you just access the website of the USA or Europe.  But that doesn't really help.  You should be able to access your own culture, including your own language.

In this area, China, Japan and Korea is a sort of the exception.  Once I did a survey, China, Japan and Korea is the only place, like search engine and other social networking, American made Google, Facebook, are not the majority.  If you go outside China, Japan and Korea, it's all the same, Google, Facebook and Twitter, possibly because of the culture and language issues.

Can we transfer or generalize this technology into the rest of Asia Pacific and Africa?

If I summarize again, the first billions, the current 1 or 2 billion users, yes, the developing countries are addressing these issues.  The other billions, I guess it's probably our turn for Asia.

I guess we really appreciate in the last 40 years where the Americans and to a lesser amount the Europeans contribute mostly to the Internet development, and we sort of enjoy their development, which is the open technology, so they share.

Probably it is our turn to contribute, in particular the technology for the developing countries.  So this is a tremendous challenge, including IPv6.

The future looks pretty nice.  I attended the APRICOT meeting last February in Hong Kong, and most of the APRICOT activities are organized and handled by the next generation -- people in their 20s, 30s and 40s.

They are very active, very kind of youthful who did these activities, just like what we did back in the 1980s.  I guess it will be very good for Asia and Africa.  This is a general description.

Then what can we do, or what can you do?  First of all, in the Internet technology area, you work towards excellence.  There are a lot of problems we have to solve, so you have to come up with some good and excellent ideas as a group.  Next is to connect each of the islands of excellence by network, which we call the network of excellence.  Then we need many networks of excellence in many areas, and collectively we work on this future Internet technology development under deploy.

In such a way, 10 years from today, 5 billion Internet users, including many new Internet users in Asia Pacific and Africa, can enjoy the Internet as much as western civilization.

Thank you very much.


If you want to see some activities we are doing, Africa Asia Forum on Network Research and Engineering, our website is africaasia.net, and through this one we are collaborating between Africa and Asia on the Internet area, including the future Internet.

Then we have the Asia Future Internet Forum, website is asiafi.net.  Currently, this one almost exclusively is China, Japan and South Korea, and with collaboration with North America and Europe, and we try to expand this activity into South East Asia and South Asia.

Any questions or comments?  Otherwise, thank you very much.


Dr Eun-Jung Park: With this, we are going to wrap up the keynote speeches and we are going to present a performance here.  We are celebrating the Opening Plenary session and opening of APNIC, and we are going to present a performance.

Here we are going to present a boy dance group which is very famous and is an icon of Korea, so let's meet Ecstasy, the bboy group.  Please welcome them with a big round of applause.


Dr Eun-Jung Park: In a few minutes, the VIPs will come to the Conference Hall.  I hope this iconic meeting will be just as powerful and dynamic as the performance.

Now we are going to start the Opening Ceremony for the APNIC 32 Meeting.  In a few minutes, we will welcome the VIPs on their arrival.  Please welcome them with a big round of applause.


Dr Eun-Jung Park: We have the VIPs arriving, so with their presence this meeting is even more meaningful.

First I would like to invite Paul Wilson, the Director General of APNIC, and a representative from the Korea Communications Information and we have the Director from the same organization.  Thank you so much.

In a few minutes we are going to invite them to the stage.  Please give a warm round of applause, because we have other VIPs entering the hall.  Please welcome them with a big round of applause.

Director General Paul Wilson is entering the hall, and other VIPs are with us here on this meaningful occasion.

From now on, we are going to start the Opening Ceremony for the 32nd APNIC Meeting.  First we are going to invite the Director General, Paul Wilson, to give his opening remarks.  Please welcome him with a big round of applause.


Paul Wilson: Thank you very much, everyone.  Thanks for the kind introduction as well.  Welcome, everyone, to the APNIC 32 Conference to all of you who have come to Busan from near and far, and also to those of you who are joining us online.

These APNIC Conferences are open events, they are truly multi-stakeholder events, and anyone who is interested in what we have to discuss here is welcome to participate.

It is really great to see people here, not only members from ISPs, who may be our traditional audience, but also from government, academia and others, because I think this increasing diversity really shows the importance and the success of what we are doing here.

These are business meetings, however, of course, and in this beautiful location, in spite of the beautiful location, we all have some serious business to do.

APNIC 32 is the first Conference we have had since we reached the final stage of IPv4 consumption in this region, and the Asia Pacific in fact will be the first IPv6 region.  As the Conference byline says, our destination is IPv6.

Tomorrow, we have a full day of content on v6, on the transition, on technologies, on experiences.  I hope it will be a unique and very valuable experience as a learning and sharing event for everyone who attends.

We also have a lot of work to do on policy this week, that is on the rules and the procedures for managing IP address space in the region, v6 and v4, and this is a major part of the multi-stakeholder process of Internet management.

We have proposals to consider here that have come from China, Japan, India and also from the United States.  You see the breadth of the capture and the participation that we are seeing here in this meeting.

With those proposals, we are going to have some diverse and interesting discussions, for sure, possibly robust and challenging.  But with the understanding and mutual respect that I think we share here, we can get a lot done.  I think we need to remember the genuine intent of any proposal or contribution, but the genuine intent can be sometimes hidden in technical detail, it might be misunderstood and it might actually be satisfied by something different; by a compromise or even a completely different solution.

I think we need to remember that today the global impact of the work we do here is actually greater than ever.  The decisions we make in this Conference are going to influence others in real ways and on a global scale, and some compromises may also be needed in order to move ahead.

I think a compromise by definition should be a reciprocal thing.  I think we can rightly expect to see the effects of what we do here reflected by changes that occur in other parts of the world in future, and that's the way this regional system should work.

We have a really important week coming up for, for learning and networking, decision-making.  I really do look forward to your active participation, as I said, both here in this room and online through the webcast.

Of course, APNIC is here to serve you.  If you have anything you need at any time during the week, please don't hesitate to approach any of the APNIC staff who are here.  We are dedicated here to making your experience at APNIC 32 the best that it possibly can be.

Finally, very sincere thanks to KISA.  I was very pleased to meet the president, Mr Suh, and officials from KCC this morning.  I think they have done an amazing job so far in hosting the event, so thank you to KISA for sharing their country with us, for giving the Korean Internet community as well a unique chance to participate directly with the APNIC community in this joint endeavour.  Thank you all very much.


Dr Eun-Jung Park: What will follow are the words of welcome delivered by President Hee Jung Kim of the Korean Internet Security Agency.

Good afternoon, ladies and gentlemen.  I am the President of the Korean Internet Security Agency.  First of all, I would like to express my sincere thanks to all the participants for being with us today in this 32nd APNIC Meeting.  My special thanks go to Director General Paul Wilson of APNIC and the Director of Korea Communications Commission and Mr Yong Sup Shin of the same organization.  Thank you so much.

It is my great pleasure to host this meeting in the beautiful port city of Busan.  Ever since this event in 1969, the Internet has shown dramatic growth.

Worldwide, 2 billion population is using the Internet.

The industry size is about $2 trillion.

Backed by the boom of smart devices, such as smartphone, tablet PCs, the traditional Internet has evolved into mobile Internet and the smart Internet.

Today, the Internet industry shows a growth of 7.7 per cent, which is $2.2 trillion in 2013 and it is expected by the year 2020 about 5 billion population will use the Internet.

The Internet is no longer just a near network communication technology, the Internet presents a new paradigm to individuals, the Internet provides rich lives, without any restrictions imposed by time and space.  For companies the Internet provides a great opportunity to create added values based on small ideas and today the Internet provides new opportunities in order to develop new driving engines, regardless of the size of territory and resources.

Internet is designed to work through the send out protocol which is TCP/IP and IP addresses are located on TCP protocols and IP address is the key resource that enables Internet access and data exchanges.  Mr Vint Cerf, the father of the Internet, declared contributions for the birth of Internet and TCP and IP and based on the understanding of importance of IP address, he said IP address is a fundamental of Internet development.

In a sense, I believe the stable supply of IP addresses can be very fundamental and very basic for the development of the Internet.

The Internet access, the Asia Pacific region that APNIC belongs to, has about 60 per cent of the world population and 40 per cent of Internet users, and has 30 per cent of world GDP.  The Asia Pacific area is a dynamic and promising market.  I believe in the sense this 32nd APNIC Meeting that has the APNIC members from 56 countries is very meaningful.  As we are all aware, IANA, Internet address resource management organization, said they are going to stop the allocation of IPv4 addresses.  We heard the announcement on 3 February.  So IPv4 centred Internet will close down and we are going to see IPv6 new paradigm for the future Internet.

I hope this venue can be the wonderful opportunity to discuss the future policies for effective operation of the Internet and practical issues and operation measures for the Internet technology.

Taking this opportunity, I would like to express my sincere thanks to the Director General, Paul Wilson, and other staff for making this event possible.  I hope you can enjoy the scenic beauty of Busan, the excellent city of Korea, during your stay here.

Finally, I would ask for your continuing affection and support and I wish you health and happiness.

Thank you.


Dr Eun-Jung Park: Thank you so much.  What will follow is a word of congratulations delivered by Mr Shin Yong Sup, a member of the Korean Communications Commission.

Yong Sup Shin: Good afternoon, ladies and gentlemen.  I am a member of the standing committee, Shin Yong Sup, of the Korean Communications Commission.  First of all, I would like to express my sincere thanks to Director General Paul Wilson of APNIC for spending time out of your precious schedule, and my special thanks go to the President of KISA and distinguished guests.

Today, in attendance with the Internet experts, we have the 32nd APNIC Meeting and it is my great honour to deliver this word of congratulations.

Recently we saw the advent of new technologies and new smart devices.  At the very centre of these changes we can see IP addresses and APNIC is playing a central role in these changes.  As Busan is emerging as the key IP city in Korea, in the sense, I believe, it is very meaningful.

In our daily life, the Internet has become the core part, and at this point the Internet is evolving itself.

We have the smart mobile, smart grid and smart buildings and we can see a new network of convergent service environments.

So the Internet environment can be expanded in an endless manner because of IP addresses.  The demand for IP addresses is rapidly increasing that cannot be compared with any time before, so we are going to see big changes in the future and I can say we are standing at a very important juncture.

This 32nd APNIC Meeting is taking place on the theme of the advent of IPv6, and we are going to have discussions on the frequencies and we are going to have discussions on the collaborations among Asian member countries.  In order to expand our effort for introducing IPv6, we are going to have discussions on this occasion.

Ladies and gentlemen, I believe this is the very time to announce the importance of IPv6 and also we should deal with all the necessary problems we can possibly face in the introduction of IPv6.  In order to do this we should have very close collaboration among member countries.

Korean Communications Commission, we are going to do our maximum effort in order to better deal with the new era of IPv6 and last October we provided policies in order to swiftly move to IPv6, and also we set priorities to have a stable transition to IPv6.

In addition, in order to better develop the vulnerable social users, we are going to provide full policy measures and provide greater support for these people and I hope the 32nd APNIC Meeting will be a very meaningful venue, where the Internet experts can have active discussions about IPv6 and also I hope this will provide a wonderful opportunity to move into the IPv6 in a solid manner.

In conclusion, I would like to express my sincere thanks to all the participants for showing great effort.

Thank you so much.


Dr Eun-Jung Park: We are going to conclude the Opening Ceremony of the 32nd APNIC Meeting.  The meeting will take place until 1 September and I hope you can enjoy the meeting and enjoy your stay in this wonderful Busan city.  Thank you very much.


Key Info


Paradise Hotel,
Busan, South Korea


28 August -
1 September 2011

Program included:

AMM, Policy SIG, IPv6 plenary, APOPS

Please wait...