[magma] Re: [pim] IPv6 Inter-domain Multicasting and Address Assignments

["Marshall Eubanks" <[e-mail]>]

Hello;

In my opinion, this should be done on Mboned, and certainly not on
PIM. However, here is my take in this.


On 2/22/07, Tim Chown <[e-mail]> wrote:
> On Tue, Feb 20, 2007 at 03:41:33PM -0500, Gibbons, James wrote:
> >
> >    Hello,
> >
> >
> >    SI   International  represents  and  contracts  with  DITO  (DoD  IPv6
> >    Transition  Office)  on  a number of IPv6 issues.  In this case we are
> >    providing  DITO  with  a  white  paper  on IPv6 Inter-domain multicast
> >    address   assignments   and  issues/problems  (expected  or  realized)
> >    regarding IPv6 multicast implementations.
> >
> >    To begin, I have gone over a number of RFCs (including RFC 3306, 3956,
> >    3307,  2375,  4607,  ID draft-ietf-mboned-ipv6-multicast-issues-02.txt
> >    among  others)  but they are vague as to actual working implementation
> >    of  inter-domain  multicast  address  assignments and how they will or
> >    might work proceed.
>
> RFC3306 lets you create for yourself a globally unique multicast group
> address where your unicast prefix (which is unique by allocation) contributes
> towards the first 96 bits of the group address.
>
> The missing bit is how you manage the last 32 bits, but that could be
> application determined if for example you used a /64 per application.
> Or you can choose your own manually.
>
> >    In IPv4 GLOP was developed but not widely used, where an organizations

I disagree with this. Except for going to IANA - or making it up - this is the
only thing that is used in IPv4.

Also, you might be interested that we are working on instantiating eGLOP - see

http://www.apnic.net/docs/policy/proposals/prop-047-v001.html  -

similar proposals have been sent to RIPE and ARIN.

> >    ASN  is embedded in the 223.0.0.0 / 8 multicast range.  In IPv6 I have
> >    seen  V6UPBM  (RFC  3306)  and  "Embedded  RP"  (RFC  3956) as similar
> >    proposals for IPv6.
>
> Yep.
>
> >    However,  I  am  still  not  clear  of the "reality" of IPv6 multicast
> >    address  assignments  especially  with  regards to globally unique (by
> >    organization, domain, site, etc.) inter-domain multicast.
>
> We're using RFC3306 and Embedded-RP group addresses for international
> IPv6 multicast applications.

This has been used since at least early 2003; it was first presented
at the Summer 2003
IETF in Vienna - see

http://www.6net.org/events/workshop-2004/durand.pdf

>
> >    That is, how far have actual address assignments proceeded?
>
> For 3306 you don't need an authority to hand it out, it's just there for
> you to use implicitly by the format.
>
> >    What is just proposed versus being implemented or to be implemented?
> >
> >    Are there actual standards being followed?
>
> A growing base of routers support Embedded RP... we've run that between
> the UK and US academic networks; all router son path (mainly Cisco and
> Juniper) support it.   No RP required :)
>
> >    Are   there  known/expected  issues/problems  with  IPv6  inter-domain
> >    multicast and address assignments?
> >
> >    Maybe a quote from my actual assignment will further clarify what I am
> >    looking for:
> >
> >    "DITO  is  having  some concerns with IPv6 multicast address space and
> >    how  it  should  or  should  not be provisioned.  You should start the
> >    study  with  regards  to  how  IPv4 multicast addressing worked in the
> >    past.   If  you  ever  worked  with  it you would know there was never
> >    official  group  reservations  made  with  regards  to address blocks.
> >    However,  this  was  never a real issue since its popularity died.  So
> >    except  for  what  was deemed to be the well-known addresses there was
> >    never  any  reservations, unlike how unicast is reserved.  One thought
> >    might be the Army gets a block of addresses, Navy, and so on."
>
> A big win for v6 is that you dont need an ASN for GLOP like v4, you can
> just form multicast groups from your unicast allocated prefix(es).
>

Note that there has been a big discussion about assignment of IPv6
addresses - see

http://www.arin.net/policy/proposals/2005_1.html

> And with a lot of /64's in an IPv6 site, you can (as we have) assign an

Most assignments are /48s

> RFC3306 group range per application if you chose to do so.
>
> The other nice thing about v6 multicast is the explicit scoping bits, which
> make it easier to control where your traffic flows.
>

Yes. I expect more in the future here.

> --
> Tim

Regards
Marshall

>
> _______________________________________________
> pim mailing list
> [e-mail]
> https://www1.ietf.org/mailman/listinfo/pim
>

_______________________________________________
magma mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/magma