> -----Original Message-----
> From: Marshall Eubanks [mailto:[EMAIL PROTECTED]]
> Sent: Monday, September 17, 2001 3:40 PM
> To: Thomas Narten
> Cc: Dave Thaler; Erik Nordmark; [EMAIL PROTECTED];
> [EMAIL PROTECTED]
> Subject: Re: uni-based-mcast and malloc-ipv6-guide
> 
> Thomas Narten wrote:
> 
> > > How so?  One can renew multicast addresses.  (Conceivably an
> > > implementation could support the ability to automatically renew
> > > multicast addresses as long as the unicast address stays valid.)
> >
> > Suppose I want to get a multicast address so I can advertise a
session
> > (e.g., on a web page). I do this two weeks in advance of the event.
I
> > want the address to last one day only. The key point is that I will
> > want to use the address staring two weeks from now.
> >
> > However, because unicast lifetimes are only a week, that's too far
> > into the future to actually get the address (yet), so the address
> > can't be obtained yet. This would seem like it might be a problem.
Is
> > it? (I don't know how this scenario fits with the way multicast is
> > actually used in IPv4, so this really is a question.)
> >
> > Thomas
> 
> In IPv4 I would say, use a GLOP address for this. (We'll rent you one
if
> you don't have any.)

GLOP has exactly the same issue for end sites that don't own
their own AS #.

When you renumber, chances are it's because you changed your provider,
in which case you can no longer use your old provider's GLOP space.

Uni-based mcast addresses don't require end sites to have their
own AS # (or any other non-aggregatable number).

-Dave
--------------------------------------------------------------------
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]
--------------------------------------------------------------------

Reply via email to