[ppml] Re: [address-policy-wg] Is the time for conservation over?

[Owen DeLong]

Returning to list per Michael's request (and my original intent)...

--On Monday, October 27, 2003 12:36 -0800 Michel Py 
<michel at arneill-py.sacramento.ca.us> wrote:

> Owen,
>
> [I would rather post this to the ML]
>
>>> Michel Py wrote:
>>> This topic has been extensively discussed on ipv6mh in the past,
>>> and it's half-true and half-false. The issue is that "If every
>>> ASN had a single IPv6 /32" does not register. Large multihomers
>>> need either multiple /32s or to announce subsets (such as /36)
>>> of their own /32, which for all practical purposes does not
>>> change much in terms of number of entries. There is potential
>>> for much more than 12%.
>
>> Owen DeLong wrote:
>> This simply does not make sense.  Bear with me:
>> If you have a /32, you have more addresses available within
>> your autonomous system than ALL of the current IPv4 space
>> combined.  Ergo, I propose that there is currently no single
>> autonomous system that could in any way justify more than a
>> 32 of space.  Heck, a /32 contains as many /64s as the entire
>> IPv4 space contains hosts.  There's just no legitimate reason
>> to issue any autonomous system (at least any in existence today)
>> more than a /32.
>> If you are a single autonomous system, then, you have one routing
>> policy.  If you have one routing policy, then, your routes can
>> be aggregated as long as they are contiguous and bit-aligned.
>> If you are given a contiguous bit-aligned /32, there is no reason
>> any "large multihomer" in existence today needs more than this
>> to replace their current IPv4 space.
>> Perhaps I'm missing something in the (unnecessary) complexity of
>> V6 addressing.
>
> No, it has nothing to do with IPv6 and everything to do with
> geographical spread. If you are a multinational organization, you don't
> want to announce your address space in one piece. You have allocated
> some addresses to east coast, some to west coast, some to Europe, some
> to china, some to Australia, etc. So you announce more than one block to
> achieve global load balancing and avoid transporting traffic internally.
> Today, large organizations that would want to deploy IPv6 have only one
> option: lie to their RIRs pretending they are LIRs and obtain multiple
> /32s.
>
Sorry... If you want to do that, you have multiple autonomous systems.
The definition of an autonomous system is a collection of routes with
a consistent routing policy.  If you are announcing some from east and
some from west, etc. and don't want to provide a backbone between, then,
you are dealing with an autonomous system for east and an autonomous
system for west, etc.

This also assumes that the majority of large organizations don't want
to transport traffic internally (I don't know how true/false this assumption
is for various organizations).

I thought we had pretty well established that geographical allocation
policies are generally fairly broken.

>
> Indeed. I said "enabling", not "actually using".
>
Fair enough.

>
>> The only entities I've seen with a compelling desire to push
>> IPv6 are cellular phone companies (it's hard to address all
>> cellphones in IPv4 space) and Micr0$0ft with their desire to
>> provide an IPv6 only Teenager platform.
>
> Unfortunately it's legal, which means it probably won't work! Some
> parents would say that since teens would something illegal no matter
> what, it's better than they pirate mp3s instead of smoking pot and
> drinking underage while driving to the Saturday night disco where they
> have unprotected sex with another minor in a car.
> Sex, drugs, rock'n roll. Haven't you been a teen in a past life?
>
> Seriously, the main reason behind IPv6 in threedegrees is that is was
> the easiest way to do NAT traversal with Teredo. This has even been said
> semi-publicly both by Tony Hain and Christian Huitema.
>
I'm not convinced that is entirely true... I'm sure they could have tunneled
it all across port 80 (Micr0$0ft is getting very good at hiding the entire
IP stack inside HTTP(s), and it's an increasingly disturbing trend).


My point wasn't so much that 3degree was or was not worth-while, as, it
was the only non-cellular use of IPv6 I had noticed outside of the research
community.  Personally, I think 3degree (or any other application form
Micr0$0ft that requires V6) is a good reason to resist deploying V6.
I would be happy if Micr0$0ft wer to disappear tomorrow.

Owen