Re: I-D ACTION:draft-ietf-v6ops-addr-select-req-00.txt

["tom.petch" <cfinss@dial.pipex.com>]

I think that 'own' is always the wrong word for address prefixes.  They are
assigned, they are allocated but ownership, if it exists, should vest in IANA,
else the system breaks down.

RFC2050 is the relevant reference.

Tom Petch


----- Original Message -----
From: "Ruri Hiromi" <hiromi@inetcore.com>
To: "Brian E Carpenter" <brc@zurich.ibm.com>
Cc: "IPv6 Operations" <v6ops@ops.ietf.org>
Sent: Thursday, January 11, 2007 4:21 AM
Subject: Re: I-D ACTION:draft-ietf-v6ops-addr-select-req-00.txt


> Brian,
>
> excuse my late responding to this, but may I clarify your comment?
>
> Now we can't think of how to set preference for ISPs because the
> RFC3484(or its update) solve this, but does it suggests we should
> consider additional section about network topologies with how to
> operate multiple policies?
>
> Or does it mean simply adding definition of xSP?  If yes, I will add
> like this.
> (Addition into Section 2 at Terminology.)
> xSP: a service provider who is an owner of the address prefix(es). It
> inputs the address prefix to the Policy Broker.
>
> Regards,
>
> On 2006/11/25, at 1:13, Brian E Carpenter wrote:
>
> > It seems to me that in addition to mentioning an xSP as the
> > source of address selection, a corporate or campus network
> > operator should also be mentioned*. I'm especially thinking
> > of a corporate network with many points of connection to
> > several ISPs; that can make the policy quite complex
> > and location dependent.
> >
> > I don't think this will change the actual technical
> > requirements much, but it needs to be considered.
> >
> >     Brian
> >
> > *The draft doesn't define xSP. If it supposed to
> > include corporate or campus operators, this needs
> > to be written.
> >
> >
>
> -------------------------------
> Ruri Hiromi
> hiromi@inetcore.com
>
>
>
>