[ppml] getting converts to V6

[John Paul Morrison]

This all gets at the question of  "why must PI always be the answer?".

At the beginning of the Internet, things worked well, but we soon found 
out the address space and routing weren't scalable.
PI was the only thing around, and people allocated globally unique IP's 
to everything (hey, I might be able to afford a full time 9600 slip link 
to the 256kbps backbone
so I can read netnews at home one day!). There's nothing wrong with 
those assumptions if it scales, which it didn't with IPv4 and the 
routers of the day.
(The irony: I have my globally unique addresses and a lot more 
bandwidth, but the market has been so conditioned by lowered 
expectations and an un-even playing field that I can't route them for 
the price I'm willing to pay)

The growth of the Internet was unprecedented and took everyone off 
guard.  People made some bad decisions - look at the stock market 
bubble, and people made technical fixes to IPv4 and the routing 
architecture of the time. Some of those technical and policy decisions 
were justifiable to keep things going, but they lowered expectations. 
They created second-class net citizens. The quick fixes got us used to 
kludges and hacks to keep things working, and we're so used to second 
best that people ask "why must PI always be the answer?"

IPv6 solved the address scalability problem.

BGPv4 and Moore's law took the urgency off of the routing problem for 
IPv4, and hopefully IPv6 for a while. I make BGP and networks run for a 
living, I think routing scalability is still an issue, but I don't see 
the same sense of panic any more. People in the industry now expect the 
growth, so maybe they aren't as shocked by it any more. I don't know 
what the answer is, but I think there are more tools and people working 
on sensible solutions.

Quoting from IETF 68 "5-10 years of growth" and "No need for panic" 
(they conclude there's no firm answer and more research required). I 
prefer to think that we'll just reach a plateau or natural limit on the 
Internet - and I hope in most areas as well - we live in a world with 
finite resources and humans with finite capabilities. We don't expect 
cars to double in speed every few years. Keyboards don't need to get new 
keys every generation etc. (keyboards will be replaced by something 
altogether before humans evolve extra (fewer?) digits!)

Back to Policy:

We have the luxury of being able to do the "Right Thing".  We're not 
going to run out of (v6) addresses and the routing infrastructure is not 
going to collapse in six months or a year, so we don't need knee-jerk 
solutions. My impression is that we're on the right track with respect 
to PI in IPv6. I definitely don't think we should use the shortcomings 
of the current routing architecture to justify making any new bad or 
unfair policy decisions.

Soft Landing:  this proposal is reasonable in principle and it or 
something very much like it needs to be the roadmap for getting from 
IPv4 to v6.

If this proposal doesn't gain consensus, then surely the consensus is 
"there is no policy on transitioning from IPv4 to IPv6". if that's the 
real consensus it would be nice to have it expressed. I would be happy 
to help construct such a policy, since it's a win either way.


Heather Schiller wrote:
>
> So, why isn't your complaint that no one has built a good, reasonable, 
> easy method or tool for renumbering or developed an assignment scheme 
> or way of routing that would allow you to not have to renumber or 
> renumber entirely?   There's more than one way to handle a problem.  
> Why must PI always be the answer?  (because the other is harder? takes 
> more time? more collaboration? __?)

Owen DeLong wrote:
>>> in the IPng wars, many folks asserted that without a solution to the
>>> routing problem, adding more address bits would both fail to solve
>>> existing problems and also make existing problems worse and also cause
>>> new problems.  as you can see, those folks lost the war.  ("too
>>> little,
>>> too soon." --tli)
>>
>> Sorry, what problem?  I've been told "just build bigger routers" so
>> many times now, maybe a million lemmings might not be wrong.
>>
> The lemmings are wrong.  There is no feasible way to scale a routing 
> solution
> that continues to overload the end system identifier and routing 
> locator onto
> the same number. In the end, even the telcos have abandoned this idea,
> and it is long past the time when IP should have in the IDR world.