[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Draft on Globally Unique IPv6 Local Unicast Addresses



Michel Py wrote:
> 
> Brian,
> 
> >> Michel Py wrote:
> >> It has nothing to do with the prefix itself but with the scope.
> 
> > Brian E Carpenter wrote:
> > Of course. And why is the scope issue different, just because
> > we have removed the ambiguity?
> 
> I read this the same way Jim does:
> 
> > Jim Bound wrote:
> > What is so elegant about this proposal is that it completely
> > eliminates the need for an implemented scoping architercture
> > within products today
> 
> There is "no scoping" written all over Bob's draft. Without getting into
> the "is this a good thing or not" stuff so far I have not seen any
> proposal that could mitigate the risk of removing ambiguity without
> creating a swamp if there is no scoping to back it up. In other words,
> if scoping goes the ambiguity will remain.

Well, I am too literal minded to have read that in the proposal. I think
it changes the scoping debate, but doesn't abolish it. As Tony Hain has
pointed out, scope is a fact of life, whether we architect it or not.

> 
> >>>> 2. We can't enforce operators to filter the prefix. I dropped
> >>>> filters on my BGP feeds and from one of my peers (a well-known
> >>>> player) I get all kinds of crud including /64 routes, /48 routes,
> >>>> /41 routes, name it I get it.
> 
> >>> I think there will be Darwinian effects here, as we saw during the
> >>> first years of CIDR. But yes, if someone pays enough, they will get
> >>> their /64 announced. That will happen whatever we write in RFCs.
> 
> >> My point exactly. We can't rely on this alone.
> 
> > Now explain why it's a problem, if the customers pay
> > enough to cover the costs.
> 
> I though we were on agreement that routing table bloat was not a good
> idea?

Indeed. But if a few Megacorps are willing to pay the ISPs to carry
specific routes, that won't bloat the tables. If a lot of medium size
companies tried the same, it would be bad.

    Brian

> 
> Michel.

-- 
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Brian E Carpenter 
Distinguished Engineer, Internet Standards & Technology, IBM 
On assignment at the IBM Zurich Laboratory, Switzerland
--------------------------------------------------------------------
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 majordomo@sunroof.eng.sun.com
--------------------------------------------------------------------