Re: [GLOBAL-V6] Any consensus?

[Christian Schild <ipng@uni-muenster.de>]

Am Donnerstag, 14. Februar 2002 17:43 schrieb Takashi Arano:
> All,
>
> APNIC address policy SIG will be held just in a few weeks
> as Anne introduced.
>
> As the chair of SIG, I would really like to bring any consensus,
> in this mailing list, or anything more concrete if there is no consensus,
> to the meeting.... Otherwise, we would not make progress....
> Thank you for your constructive comments and your cooperation.

I think we have to decide between two different approaches now. Both have 
their pros and cons. 

The first one is to follow the way it is done in the 'IPv6 Address Allocation 
and Assignment Global Policy' draft. The advantage here is, that we actually
_have_ a very good draft, which is useful for all RIRs. A main concern 
is, that the criteria to get IPv6 address space as decribed in 5.2.1 is too 
strict. Only if one shows a 'need' for 776 /48 address blocks, one will get 
IPv6 address space, and that will hinder a lot of organziations to start 
IPv6 deployment. 
What we would have to do here, is to fine tune this criteria to meet everyones
desires. I believe this will be very difficult, because we will end up with 
too many exceptions from this rule. We have to define a good 'number' for 
research networks, small or large ISPs, DSL providers, global providers, 
large scale companies and maybe many more. 

The second way is to cut out that criteria entirely and give every LIR a
/32-prefix if they ask for it. This was concensus on the RIPE meeting and 
discussed here a lot. The advantage would be, that everyone who wants IPv6 
adress space could get it easy and it would fasten the deployment of IPv6.
The disadvantage is, that now it is far _too_ easy to obtain address space. 
Also LIR is not applicable to every region. While it is more difficult 
to become a LIR in the RIPE region due to heavy paperwork, in other regions 
one simply has to pay a fee to become a RIR-member. The idea on the RIPE
meeting was, that the heavy paperwork is hindering people to 'just buy' 
a /32-prefix. This may not work in other regions. 
Here we also would have to find a more specific restriction to cut out 
all the very small entities and 'end sites', who should not be able to get a 
/32. Takashi Arano already started doing so, proposing to replace LIR with 
'organizations who assign /48 to organizations other than itself'.

Sorry for this becoming kind of a longer summary, but it helped me to make 
up my mind (and maybe yours too).

So, im my opinion every enterprise who wants to offer any kind of IPv6 
services should be able to get IPv6 address space quick and easy, no matter 
how big or small it is and especially no matter how much customer it has. 

If we follow the first approach, the '776'-rule simply will fail. When 
we want to give IPv6 address space to everyone who has a need for it, 
we have to define an exception from this rule for every kind of organization 
and for every kind of service. I believe this is not possible. 

Considering the second approach, in fact it will be quick and easy for 
everyone to get IPv6 address space. 

I think that wiping out the disadvantage of approach two (by finding a 
better definition for 'LIR') is much easier than making rule exceptions 
for approach one. 

Regards,
     Christian 

-- 
JOIN - IP Version 6 in the WiN  Christian Schild
                 A DFN project  Westfaelische Wilhelms-Universitaet Muenster
Project Team email:             Zentrum fuer Informationsverarbeitung
          join@uni-muenster.de  Roentgenstrasse 9-13
http://www.join.uni-muenster.de D-48149 Muenster / Germany
email: schild@uni-muenster.de,phone: +49 251 83 31638, fax: +49 251 83 31653
-
- This list (global-v6) is handled by majordomo@lists.apnic.net