Re: Vote NO on R-L-G IP Address Allocation proposal

[Geoff Huston <[e-mail]>]

Peter and Ross,

> We have no good short term options. As I have said before, the IP internet is
> dying, it just doesn't know it yet. It will soon go the way of the research
> Decnets which outgrew the address space. RFC 1366 is an attempt to tide us
> over until a new approach is here, and it appears to be our best hope.

Tony has hit the nail firmly on the head. Sitting out on a far limb of the
Internet, with a single connecting lifeline things sure look differnet
from here. I guess that the distance adds a certain degree of perspective
and detached observation to what is a hot topic.

Yes, it appears that we are outgrowing our address space, and the big-internet
list is a graphic illustration of the fact that there is no simple
quick fix remedy that is within out immediate grasp.

I have said in the past that tinkering with address allocation strategies
is a very indirect way of trying to ease the burden of deployed networks
coping with growth, but it has to be remembered that once the B space 
is exhausted we will start to romp through the C space with gay abandon -
and the inevitable consequence is a future expansion of the number
of routable entities (assuming that we persist with the A / B / C 
routing structure). Broad geographic allocation mechanisms are at least
susceptible to aggregation strategies - which is a more positive position
over what we have right now. In my humble opinion RFC1366 attempts to postpone
the day of adoption of poor man's IP routing, and is worthy of considered
attention on those grounds.

Having argued personally in the past that provider-based address allocation
mechanisms is suboptimal, I must admit that there are few viable short
term alternatives which have a hope of flying within the current
Internet. An anarchic, or unordered address allocation strategy will
not be seen as being equitable within the broader international
domain, and this document at least attempts to set out a strategy
which is seen as being equally accessible for all parties. I'm
not sure that one can aim for any loftier objectives at this
late stage in the game.

I suppose the problem is that coping with exponential growth is something
that most people never really meet up with .. and the Internet is now
on the final stages of capability to service this explosive growth. What
next is anyone's guess.

My 2c anyway...

Geoff Huston
Australia