Excuse the off-topic question, but does anyone know if there is some sort of list anywhere of service providers who are running IPv6 in a production capacity, either to tunnel IPv4 or to offer native IPv6 services? I’m not looking for test or research networks but rather a list of IPv6 networks that are actually carrying customer traffic.
Thanks,
Irwin
NTT/Verio has been offering commercial IPv6 services since April 2002
I seem to recall.
http://www.v6.ntt.net/globe/index_e.html
http://www.soi.wide.ad.jp/ipv6_summit/2001/slides/03/2.html
I've gotten postcards from Hurrican Electric about their
Free IPv6 service as well. I'm sure many other people have gotten
them..
I seem to recall that C&W provided native IPv6 for the
Atlanta IETF.
I'm seeing more providers looking at offering native
IPv6 services with Juniper as well as Cisco routers since they
can converge the networks using ISIS. The OSPFv3 support is in
varying stages depending on your vendor.
I'm seeing more and more ipv6 hits in my web server logs
as time goes on (when i put the AAAA record in place and my host
stack is working ok). They mostly come from Europe and Asia but an
increasing amount are coming from locations within the US.
I honestly see most of the backbone providers offering
native IPv4 and IPv6 services in the next few years. Contact
your provider as you can probally get in on any beta service
offerings they currently have.
- Jared
Am I the only one that thinks IPv6 is a minimum of ten years out before
you see actual non-geek demand?
Maybe I have my head in the sand, but we have an abundance of available IP
space, and people who care encrypt their traffic anyways. Is IPv6 better
than IPv4? Yes. Enough for it to motivate everybody to switch? Debateable.
I just don't see how the snowball is going to get started. With
address-starved nations using IPv6, that only makes for more available
IPv4 space and less of a motivation to migrate for the rest of us. And in
the US, it's seriously driven by the geek early-adopters. I know that I
only tested it out to satisfy the geek in me, not to try to solve any real
problem that needs a solution. And when I present the issues to non-geeks,
they all suggest, in a round-about-way, every single last one of them,
that whatever benefits I'm talking about are insignifanct compared to the
fact that they no longer have a nice dotted quad to remember.
So, how does IPv6 go from the shores of Japan and the minds of geeks
across America to being the primary protocol used on the net?
Andy
I think the place you're going to see IPv6 adoption is
government networks - I've seen several RFPs from
various government offices which are requiring some
degree of IPv6 capability. This is going to drive the
various large carriers who have these governments as
customers to implement a generically working IPv6
solution.
Once it's a product, I think you'll see some people
buying it...
-David Barak
David Barak <thegameiam@yahoo.com> writes:
I think the place you're going to see IPv6 adoption is government
networks - I've seen several RFPs from various government offices
which are requiring some degree of IPv6 capability. This is going
to drive the various large carriers who have these governments as
customers to implement a generically working IPv6 solution.
This is the same phenomenon that drove the explosive adoption rates
of the ISO OSI protocol stack and the Ada programming language.
Jim Shankland
What I continuously remind myself is the transformation of the
internet from 10 years ago to now. When you look at what has happened
in comparison, I wouldn't rule this out at all. Obviously IPv4 is going
to be the primary for internetworks for some time but I do expect
traffic levels at the IPv6 exchanges to pick up. Personally, I find
some mirrors I connect to have a IPv6 address where they don't
rate-limit it, so when the next release of RH/FreeBSD come out, it's
quicker to download via IPv6 than IPv4 as there are no contention issues
for gaining access to the ftp server.
I see people doing IPv6 deployments but not quite as fast as
the IPv4 deployment speed of the past 10 years, but with most sites
enabled in the next 3-5 years at most.
- Jared
> I honestly see most of the backbone providers offering
> native IPv4 and IPv6 services in the next few years. Contact
> your provider as you can probally get in on any beta service
> offerings they currently have.Am I the only one that thinks IPv6 is a minimum of ten years out before
you see actual non-geek demand?
I don't know about timeframe, but I certainly don't see it soon.
Is IPv6 better than IPv4? Yes.
I'm not sure that is widely agreed on. Specifically, I am not sure
there are many features of v6 that are wanted and haven't been ported
back to v4.
Enough for it to motivate everybody to switch? Debateable.
Some of the new features in ipv6 have already been tried in other
protocols... protocols that ipv4 has already replaced in the market.
They won't be the adoption drivers people seem to think they are. v4
may not be as efficient as v6, but it is efficient enough, and until
it cannot handle the apps the mainstream users need, the mainstream
won't use it.
Fortunately, unlike VHS vs Betamax, you can quietly include support
for both in a single device with no or little excess cost, and, like
cassettes vs CDs, you will be able to have one mass market for the
good enough right now, and one smaller market for the -philes, and
slowly shift from one to the other. First, though, you have to make
ipv6 perform. ipv4 has been worked on by some very competent people
for a long time; to pull out yet another comparison, like rotary vs
piston engines, even if ipv6 is theoretically better than v4, the
implementation and support have to catch up before the wide deployment
can really get started.
-Dave
Hi Irwin,
Excuse the off-topic question, but does anyone know if
there is some sort of list anywhere of service providers
who are running IPv6 in a production capacity, either to
tunnel IPv4 or to offer native IPv6 services? I'm not
looking for test or research networks but rather a list of
IPv6 networks that are actually carrying customer traffic.
funny you're asking today. Some hours ago I was browsing
www.sinet.ad.jp and while I did not understand a word but
the English! button there was a text saying I was accessing
that page with IPv6, and it just worked and v6 was in no way
slower at all than IPv4 when I actually compared these. It's
nice that it's just working fine your applications are ready
for it.
That being said, the network I work with has IPv6 customers
(natively) and is actually doing transit for some other
networks in Europe and also in the US to help IPv6 to spread
quicker. It's native IPv6 on all core nodes at the moment
and we only have local tunnels in Europe where nothing else
would work. Cool thing is 90% of all European exchanges are
already offering IPv6 services, and in the US of course
Equinix does IPv6 as of late, as does Telehouse and PAIX. So
it is easy to get started with IPv6 if you are at one of
these exchanges.
I'm happy enough more and more IPv6 network get the hang of
not simply (re-) announcing the full table but only their
own routes, asking their upstream provider to do IPv6
upstream for them. I may not want to receive the full table
in Europe by someone who has tunnels on a router where I am
native to the US.
Regards,
Alexander
PS: AS-TISCALI-V6PEERS in RIPE, if anyone cares, or
route-server.ip6.tiscali.net 
Am I the only one that thinks IPv6 is a minimum of ten years out before
you see actual non-geek demand?
Well, this probably doesn't fall completely under the non-geeek category,
but larger US Universities are starting to deploy v6 on their campuses as
well as using it for native transit over I2. One driver is the research
folks (a.k.a geeks). There is, however, another business driver. Those
wonderfully huge class B's that we have grandfathered to us for free
from so many years ago are starting to fill up with the explosion
in devices on campus - all our vending machines, card readers on every
door, pervasive wireless, every student with both a laptop and a desktop...
Getting a nice, free (from I2) chuck of v6 space and pushing the community
to use it is definitely preferable to paying ARIN for another /16.... If
Universities start offering pervasively, you might see this spur growth in
other sectors.
Eric 
Thus spake "David Barak" <thegameiam@yahoo.com>
I think the place you're going to see IPv6 adoption is
government networks - I've seen several RFPs from
various government offices which are requiring some
degree of IPv6 capability. This is going to drive the
various large carriers who have these governments as
customers to implement a generically working IPv6
solution.
Nearly every customer of mine has required IPv6 in their RFPs for over a
year, but not a single one has turned it on even for testing.
Once it's a product, I think you'll see some people
buying it...
You mean once Windows has it enabled by default, people will start using it.
IMHO, the only chance IPv6 has of widespread US deployment is if it can
happen without end users knowing they're using IPv6.
Unfortunately vendor C still ships nearly all of its L3 switches and core
routers with forwarding engines that don't grok IPv6 packets, even if said
vendor has supported IPv6 in software for several years now.
S
Nearly every customer of mine has required IPv6 in their RFPs for over a
year, but not a single one has turned it on even for testing.
Agreed. Similar experience over here.
> Once it's a product, I think you'll see some people
> buying it...You mean once Windows has it enabled by default, people will
start using it.
IMHO, the only chance IPv6 has of widespread US deployment is if it can
happen without end users knowing they're using IPv6.
Windows customers will not notice it once they can accept DHCP'd IPV6
addresses and their provider does 6to4 mapping and what not.
Unfortunately vendor C still ships nearly all of its L3 switches and core
routers with forwarding engines that don't grok IPv6 packets, even if said
vendor has supported IPv6 in software for several years now.
Vendor C wants you to upgrade to new hardware-level IPV6 ASICs once demand
is high enough. And then a new + verison once you need wirespeed.
Deepak Jain
AiNET
Right, but it means that more network providers are
having to offer some type of solution. This will
enable Windows (or whatever) to have it on by default
and actually have it work.
Vendor C's issues with v6 are a problem, but they're
not the only provider of core or edge gear...
Also, even though their forwarding mechanisa are not
completely functional, they do pass packets, so it'll
work, just not be optimized.
Unfortunately vendor C still ships nearly all of its L3 switches and core
routers with forwarding engines that don't grok IPv6 packets, even if said
vendor has supported IPv6 in software for several years now.
The inventors of tag-switching^H^H^H^H^H^H^H^H^H MPLS seem to be
firm believers in that if they don�t deliver the goods, the world will stand still
and wait.
Pete
I believe GX sells commercial IPv6 at select locations as well. Hurricane
Electric is probably the leader in the market though, as everyone else
seems to still be implementing v6 with dedicated low-end devices and
tunnels.
MAI will be offering IPV6 for their web services hosting.
> Nearly every customer of mine has required IPv6 in
> their RFPs for over a
> year, but not a single one has turned it on even for
> testing.
Lets say I have a lab with a few machines and I enable native v6 on
my primary v4 network connection. The 300kbps that my lab generates
is fine. What do I do if someone DOS's me with v6 packets and
my router dies taking out both v4 and v6? The routers we spec
can usually handle v4 traffic at levels on par with the link rates
so its not as big a deal for v4 DOS traffic (link saturation is more
the problem). When the v6 rate is substantially less than the link's
bandwidth, I get worried that a low-level DOS will take out our entire
network...
Eric
Thus spake "David Barak" <thegameiam@yahoo.com>
> Nearly every customer of mine has required IPv6 in
> their RFPs for over a
> year, but not a single one has turned it on even for
> testing.Right, but it means that more network providers are
having to offer some type of solution. This will
enable Windows (or whatever) to have it on by default
and actually have it work.
We can hope.
Vendor C's issues with v6 are a problem, but they're
not the only provider of core or edge gear...
Also, even though their forwarding mechanisa are not
completely functional, they do pass packets, so it'll
work, just not be optimized.
When a 30Mpps IPv4 box falls back to <200kpps for IPv6, I don't think "not
completely functional" is an adequate description. To me, that falls into
the "not supported" category.
S
> Vendor C's issues with v6 are a problem, but they're
> not the only provider of core or edge gear...
> Also, even though their forwarding mechanisa are not
> completely functional, they do pass packets, so it'll
> work, just not be optimized.When a 30Mpps IPv4 box falls back to <200kpps for IPv6, I don't think "not
completely functional" is an adequate description. To me, that falls into
the "not supported" category.
Clearly, you wouldn't deploy this box for a native-IPV6 app. I am guessing
Cisco is betting this box will have an upgrade available or be obsolete by
the time the majority of their customers want to pass 30Mpps IPV6.
Heck, a PC-IPV6 router will move more than 200Kpps, but I don't want to get
on that horse.
Deepak Jain
AiNET
Date: Thu, 12 Jun 2003 19:22:38 -0500
From: Stephen Sprunk
When a 30Mpps IPv4 box falls back to <200kpps for IPv6, I
don't think "not completely functional" is an adequate
description. To me, that falls into the "not supported"
category.
Okay, I'll make a fool of myself on-list -- certainly not the
first time.
Why not use the highest-order 32 bits of an IPv6 address for
interdomain routing... i.e., "overlay" them on IPv4 addresses
and/or a 32-bit ASN? Yes, it smells of classful routing. Call
me shortsighted, but how many billion interdomain routing
policies do we really need?
Probably OT, but seems semi-fitting for the thread.
Eddy
Well, i'll try to steer the conversation in a different
direction.
I think that some of the hardware vendors need to seriously look
at their design policies for their new linecards, platforms, processors
and continue to leverage their existing software so that we can get the
necessary solutions to operate our networks. What am I talking about?
Well, we need to insure that not only the platform can forward
at linerate with all the necessary features turned on. You need
to place rate-limits, acls, ipv4, ipv6, unicast-rpf, load-sharing,
mac accounting, received mac address acl logging (at least one "core" vendor
seems to be missing this still) and more. The platform needs to boot
in ~30-60 seconds. Yeah, NSF/HA will help things, but nobody ever needs
to do a cold start because there's never a power outage ...
there need to be sufficent processing power that there aren't
any problems (or percieved problems - eg: customers actually do expect
your routers to respond to icmp promptly otherwise they'll claim packetloss;
this isn't the case most of the time, but any percieved problem can possibly
cause you to lose customers) handling BGP updates and providing good
[interactive] response time.
I truly think that in order to provide all the necessary
features needed in the core we need the vendors to go through at least
2 more hardware generations to provide the features necessary if they
do not make too many mistakes. Otherwise we'll be chasing how to look
into the mpls packets to do DoS tracking for years to come.
- jared