[6bone] non-global address space for IXs (was: 2001:478:: as /48)

[Dan Reeder]

seeing as I can't resisit adding my AU$0.022 (ing GST), I thought I'd
contribute by raising this:

Wasn't one of the main features of a v6 internet supposed to be the
fantastic implementation of hierarchical design brought on by not only RIR
management, but co-ordinated and co-operative address aggregation iteself?

Whats to stop other people around the rest of the world from designing their
networks in this fashion? Personally i think we could then kiss the word
"hierarchy" goodbye if that happens. It'd be a veritable schemozzle.
I certainly don't agree with this attitude of "if you dont like it, filter
it."

Dan Reeder
----- Original Message ----- 
From: "John Fraizer" <tvo@EnterZone.Net>
To: "Gregg C Levine" <hansolofalcon@worldnet.att.net>
Cc: <6bone@ISI.EDU>
Sent: Sunday, September 07, 2003 11:22 AM
Subject: RE: [6bone] non-global address space for IXs (was: 2001:478:: as
/48)


>
>
> On Sat, 6 Sep 2003, Gregg C Levine wrote:
>
> > Hello (again) from Gregg C Levine
> > Would any of you mind doing me a favor, when this discussion reaches a
> > good resting point? Please reiterate the whole series of things that
> > led up to this discussion.
> >
> > Oh, and on the sub-subject of ISPs, I happen to know from an earlier
> > phone call with their support staff, that even though they provide DSL
> > service for consumer AT&T customers, they haven't plans at the time to
> > make available IPv6 services. I'm still working on that issue.
> > -------------------
> > Gregg C Levine hansolofalcon@worldnet.att.net
>
>
> Bill Manning made a general announcement (like he has for the two previous
> years) that folks *may* see chunks of 2001:478::/32 announced as /48's or
> possibly /64's since it is divided into /48's and /64's for use at IX's
> that go to EP.NET for address space management.  From there, the v6 police
> decided that since folks can (now) get address space from the RIRs, that
> every EP that gets address space from EP.NET should now renumber or that
> someone should announce 2001:478::/32.  Nevermind that no ONE entity
> currently has connectivity to every part of 2001:478::/32 or that this
> address space has been in use at EPs since BEFORE the RIRs had made
> address space available for use at IX's. (I don't care if they had policy
> for it - they weren't handing out address space for this purpose yet.)
>
> From there, Jeroen demonstrated that he didn't understand how exchange
> points REALLY work and the reason for having all participants in the
> exchange point share not only the same L2 fabric but also the same L3 LIS
> so that routes learned from the route-servers have next-hops that are
> "connected" routes and do not require any further configuration on the
> part of exchange point participants other than bringing up peering
> session(s) with the route-server(s).
>
> From there, Jeroen decided to make anti-American remarks, blah blah blah
> in reference to my lack of patience with people who want to bitch and moan
> without being in possession of even a fraction of the clue required to
> understand the technology involved and an obvious lack of experience
> operating in the environment that address space carved out of
> 2001:478::/32 is implemented in.
>
> It boils down to this: People took offense (for whatever reason) that a
> single /32 (2001:478::/32) that was not part of address space carved out
> by the RIRs for use at IX's was being used for addressing IX's.  Nevermind
> the LONG history (pre IPv6) of EP.NET providing address space and
> management of that address space for exchange points, nevermind the fact
> that nobody is FORCING them to accept those /48's, nevermind
> anything.  They simply wanted to get bent out of shape because someone
> besides the RIRs was making address space (from a SINGLE /32) available
> for use by folks who operate v6 exchange points.
>
> Typical 6bone pissing and moaning if you ask me.
>
> --
> John Fraizer
> EnterZone, Inc
> (13944+$|13944+_14813+$|13944+_17266+$)
> PGP Key = 6C5903C4
> Fingerprint = 2AA6 6614 1B5E EDD2 38AD C417 3E61 F975 6C59 03C4
>
> _______________________________________________
> 6bone mailing list
> 6bone@mailman.isi.edu
> http://mailman.isi.edu/mailman/listinfo/6bone
>