I usually mention that cellphones and perhaps some captive users
with no alternatives in China are likely to use IPv6 in large
numbers. For brevity, I didn't mention it in the message you quote.
Since the cellphone has its own inbuilt applications and typically
local sources of stuff the mobile carrier is trying to sell to the
customer, there is no major barrier to using IPv6.
I suspect that many carriers will want to give their customers a
public IP address, IPv4 or IPv6. If they do that, the end-user can
run their own VoIP software and bypass the carrier's voice system.
The end-user might also be more able to purchase services, video
etc. (AKA "content") from other companies than the carrier or its
affiliates.
I don't have any direct knowledge of IPv6 in China - its just
something I heard of, which sounds likely. I guess those people
would not have a choice of an IPv4 service, because there is only
one ISP in their area.
All these are new users.
For the next 5 or 10 or whatever years, I doubt that there will be
an IPv6 routing scaling problem. I think that the predicted masses
of cellphone customers and likewise lots of users behind some
massive ISPs in China will be connected via large ISPs, I guess with
a small number of BGP advertised prefixes.
I think that there would only be unsustainable growth in the number
of IPv6 BGP routes if there was a widespread general uptake of IPv6
by end users - including those who want to sell things to the
legions of predicted mobile users. A routing scaling problem for
IPv6 would emerge if there was a few hundred thousand of these
organisations who decided they need to be on IPv6, and that they
want multihomed and/or portable space so badly that they get their
own PI space. Maybe this will happen.
Who says there needs to be growth in the number of IPv6
BGP routes? If we map/encaps the entire IPv6 space as an
overlay over the existing IPv4 Internet, we keep IPv6
prefixes out of the BGP routing tables and we get to
scale through mapping w/o affecting routing scaling.
Fred
fred.l.templin@boeing.com
There is only one routing scaling problem at present - IPv4. The
only way IPv6 will help with that is if enough ordinary end-users
like present-day end-users, with desktop PCs, applications, servers,
server farms, hosting companies, office networks etc. decide that
they want to pay for a service which is IPv6-only, and at the most
share a single IPv4 address via something like Dual Stack Lite.
It is really hard for me to imagine this happening - search for
"lite" at: http://psg.com/lists/rrg/2008/maillist.html.
Other folks have no trouble imagining this sort of thing. But they
have been imagining this for over a decade, and it hasn't happened
yet.
They tend to think of the IPv4 sky falling in ca. 2011 when the
fresh blocks of unused address space are snapped up, but I think
this will be the start of a long period of using the space more
intensively. I think there is a lot of scope for doing this, and
that for a very long time it will be cheaper and better to keep
looking after customers with connectivity to the Internet they want
and need: IPv4, rather than trying to sell them a service to some
other Internet which only connects to a fraction of users.
I am not saying that there won't be a significant number of captive
home/office customers and cell-phone users on IPv6 in by 2018 or
perhaps 2013 - they may well be quite a few. I am not saying that
IPv4 will go forever, or that IPv4 NAT is a good thing. I am not
yet convinced by the various arguments for why large numbers of
current end-users and their like will adopt IPv6-only services
inside 10 years.
- Robin
--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg
--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg