On 10-apr-2006, at 7:43, Tony Hain wrote:
Instead of dissecting the numbers into chunks that give you the
answer you
want, how about looking at the big picture?
[...]
The real issue is that Geoff's linear projections against the
current .75
/8's per month going out from the RIRs to hit a 2012 date don't
match the
historical growth.
The problem is that nothing matches historical growth, because it
contains elements that have proven resistant against modeling. (Note
that Geoff has three different projections and the linear one doesn't
hit the ceiling in 2012, with 0.75 /8s per month = 151 million/year
this date would be 2015.)
Also, taking a very short term view creates misleading
windows that lead to claims like yours that we are now on a slower
pace than
last year, so not to worry. While the graph does show that we were
above the
projection curve last year and below it so far this year, the
overall trend
since Jan 2000 is tracking the projection very tightly.
I don't see it. If I use a formula of deltayear(n) = deltayear(n-1) *
x and start in 2000, the best fit (where the yearly differences
between the projection and reality as a percentage added up equals
zero) is a an x of 1.09. This is obviously ridiculous because both
2002 and 2005 are off by around 30% (93 vs 69 million and 120 vs 168
million). If I ignore 2002 and 2003 it comes out to 15%. This lands
us in 2010 as the year IPv4 runs out, by the way, with the projection
for this year at 180 million addresses. At 9% this would be early
2014 with a projection of 131 million addresses used this year.
The only way I can fit the projections closely to reality is by
picking 2002 as my start date and assuming 34% growth. This way,
we're out very close to the turn of the decade, but it does mean
we'll be using up 222 million IPv4 addresses this year. And that's
something that the current figures just don't seem to support, even
though 2006 so far as increased from 35 million when I wrote my
earlier message to 45 million now. However, for 222 million it would
have been something like 61 million by now. (But looking at the data
this closely doesn't do much good.)
The good news is that at the end of the year, we'll have a much
better picture: either the mini-trend of around 34% growth in yearly
address use that started after 2002 will turn out to have continued
more or less, or it turns out it wasn't a trend after all, just like
the dip in 2002 wasn't a new trend. Until that time, I'll continue to
assume 2010 - 2015 with 2012 as the most likely moment for IPv4 to
run out.
Changing the RIR policy is a hopeless cause. This would have to be a
simultaneous global change and the process for getting global
agreement
takes at least 2 years (as shown by the only global agreement they
have;
IPv6 policy, and the much longer time it is taking to debate
changes to it).
By the time anything could be done there wouldn't be enough left to
worry
about.
I don't think the actual changing is the hard part, but coming up
with a new policy that is better than what we have now, is. We'll
never really run out of /24s and blocks that aren't much larger
because even though <= /18 blocks make up 90% of all allocations they
make up less than 10% of the total address space used, i.e., less
than a /8 a year. So we only have to reclaim a single /8 per year to
accommodate those requests. For the really big blocks that ISPs are
burning through so fast these days, I don't see a reasonable policy
that can slow this down without basically making IPv4 effectively run
out for them at the time of the policy change rather than when we're
really out. Either you give those ISPs what they need or they'll have
to start putting more than one customer behind a single address.
So a policy change to make the IPv4 space last longer for the big
users would be impossible. There are two things we can do, however:
- try to avoid destructive end-game behavior, for instance by
imposing a maximum block size at one point
- set aside a limited amount of IPv4 addresses (like the last 100
million addresses) for smaller address users rather than give the
last bit to the large users
There is however and interesting policy question: should we allow
IPv4 addresses to be sold? Some people are in favor of this, but I
don't see the upside of formally allowing it. (People are going to do
it to some degree anyway.) The down side is that you can forget about
getting back sizable legacy space chunks and people have even more of
an incentive to hoard address space rather than use it. And this will
break up the address space in much smaller fragments which doesn't
help the routing table.
With the advent of RIR-anchored address space certificates the RIRs
must decide whether they allow trading or sub-delegation of address
space or prohibit it.
You are correct that we don't know what will happen in the future.
Unrealistically long projections don't serve anyone except those
who look
for solace in the fantasy land where nothing changes. The world
needs the
wake up call that reality is about to hit them in the face and they
will
need all the time there is left to develop a managed IPv6
deployment plan.
We only get to cry wolf so many times. You and Geoff came up with
predictions last year, I suggest you both do the same this year and
the following ones. But accept that if it really turns out to be 2009
or even 2008 the transition will be a painful one for many people
anyway and the numbers just aren't hard enough yet to make people
want to start experiencing that pain now when it's still possible to
rationalize the problem away.
If they don't start now they will be forced into a crash deployment
when
they try to get more space and find out the pool had long ago run
dry. The
IETF as a whole needs to wake up as well and stop developing for a
dead end
technology. By the time any new work items make it through the
process there
will not be any new IPv4 space to deploy it.
When we run out of IPv4 space obviously very many people will have
IPv4 addresses and they'll want to keep using them. This won't be so
much the end date for IPv4 as the start date for IPv6.
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf