This is all very well in theory. In practice, we obviously need to
limit ourselves to very short chains - maximum 2 would be prudent,
until we know whether it works at massive scale.
Decoupling the host and prefix is the *big* win for renumbering;
chaining up the ISP hierarchy is a marginal benefit.
Brian
Robert Elz wrote:
>
> Date: Fri, 19 Jan 2001 23:52:12 +0900
> From: [EMAIL PROTECTED]
> Message-ID: <[EMAIL PROTECTED]>
>
> | A6 chain will not stay in the same zone. it will go up from leaf site
> | to a small ISP, then to middle ISP, and then to large ISP. this is
> | the purpose of A6.
>
> Well, perhaps. This will depend upon how it is configured. The example
> you showed is most likely taking the thing towards absurdity (and yes,
> that is certainly possible).
>
> On the other hand, I suspect that many sites will not want to place their
> addresses into the hands of some other zone administrator, and will instead
> use
>
> foo.itojun.org. IN A6 64 ::9876:5432:9876:5432 segment0.itojun.org.
> segment0.itojun.org. IN A6 48 ::1234:0:0:0:0 ispX.itojun.org.
> ispX.itojun.org. IN A6 40 0:0:0012:: ispY.itojun.org.
> ispY.itojun.org. IN A6 32 0:0:2300:: ispZ.itojun.org.
> ispZ.itojun.org. IN A6 16 0:9876:: ispU.itojun.org.
> ispU.itojun.org. IN A6 0 1234::
>
> which gives almost all the same benefits (except you need to be
> involved when one of the prefixes changes), and gives you the
> added security (peace of mind I mean, not mathematical type security)
> that your addresses are all under your control.
>
> Then, having done that you will probably realise that you could
> just as easily make the third one be...
>
> ispX.itojun.org. IN A6 0 1234:9876:2312::
>
> (assuming I mentally did the correct transformation) and not bother
> with the long chain. It might mean a few more records to update if/when
> one of the ISP prefixes changes (not in this example, but in general
> it would), but not so many more that it becomes unreasonable.
>
> Then the chain is just 3 long, and all in the same zone.
>
> However, we do need more experimentation with A6 - and to get that
> we need people to actually use the thing, rather than just looking
> at the worst possible configuration of it and throwing up their hands
> and saying "too hard".
>
> kre
>
> --------------------------------------------------------------------
> IETF IPng Working Group Mailing List
> IPng Home Page: http://playground.sun.com/ipng
> FTP archive: ftp://playground.sun.com/pub/ipng
> Direct all administrative requests to [EMAIL PROTECTED]
> --------------------------------------------------------------------
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page: http://playground.sun.com/ipng
FTP archive: ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------
