[1542] in North American Network Operators' Group
Re: Policy Statement on Address Space Allocations
daemon@ATHENA.MIT.EDU (Eric Kozowski)
Fri Jan 26 02:47:37 1996
Date: Thu, 25 Jan 1996 23:37:22 -0800
From: Eric Kozowski <kozowski@structured.net>
To: dennis@Ipsilon.COM
Cc: nanog@merit.edu, cidrd@iepg.org, iab@isi.edu, iesg@isi.edu, iana@isi.edu
>> Here is a memory refresher re: IPv4 address space usage:
>[...]
>>Tony> CIDRD Minutes
>
>I think you're confusing the number of routes with the amount of address
>space routed (the latter is the only thing the note you included commented
>on). They're only weakly related.
No, go back and read that section again. It specifically talks about IPv4
address _usage_, NOT number of routes. Later on in that email from Tony Li,
It does talk about number of routes.
>The "CIDR/Address Allocation/etc." thing has hardly been a "debacle" in
>any sense of the word, and I have the existence of a still-operating,
>still-growing Internet to prove that.
It has been less than optimal. Inconsistant allocation policies and
inconsistant routing policies have caused this. User confusion as to
"how big of a cidr block is routable" and "where do I get IP addresses
from" are excellent examples of this. IMO, the IANA needs to setp in and
use the "Authority" part of their name to set a common allocation and
routing policy.
>You may not have the historical
>perspective to understand just how hard and dangerous a problem this
>looked to be 5 or so years ago.
Actually I've been involved in Internet routing, in one form or another,
for many, many years. I remember flag day, when the chage from NCP to
TCP/IP was made. I've actually laid my hands on a working LSI-11 running
Fuzzball. I think I have a half way decent historical perspective.
>Yet the process to get where we are,
>from figuring out what one might do about the problem, through the
>technology development and deployment, and the policy and reorganization
>work to take advantage of what was deployed, was better done than just
>about anyone could have expected. This was good work.
I'll agree. The technical side of making the Internet work has been
extrordinary. The policy side, after the NSFNet AUP started being
ignored, has left a lot to be desired.
>... while engineers can occasionally be competent at technology they're
>hardly ever good at policy.
Amen.
>I think we just need to all agree about which
>page we should be on, and I think we'll do just fine at this if we can
>talk about policy less in terms of adjectives, and more in terms of numbers
>we can measure, since we are mostly all just engineers.
>
>> We need to decide which is a bigger problem - number of routes or remaining
>> address space. We can't have it both ways.
>
>I don't think we should touch the question of "bigger" at all, nor do I
>think these two goals are in any way mutually exclusive. I just think
>there needs to be agreement to be consistent on the means to both ends.
This was precisely what I was trying to get at, in an oh so unelegant way.
What I should have said is that the current polocies make it such that it
will be very difficult to both preserve address space and keep routing tables
small (comparatively).
Eric
--
Eric Kozowski Structured Network Systems, Inc.
kozowski@structured.net Better, Cheaper, Faster -- pick any two.
(503)656-3530 Voice "Providing High Quality, Reliable Internet Service"
(800)881-0962 Voice 56k to DS1
