Updates on H&S framework draft

Hi all,

Carlos and I finally got an update on the H&S draft out
yesterday.
http://www.ietf.org/internet-drafts/draft-i
etf-softwire-hs-framework- 
l2tpv2-03.txt

Now we ask to anyone on the list to take 10 minutes of their
time and  
read this version carefully.
Any remarks and comments are welcome ! From these feedbacks
we hope  
to release a new version before the cut-off for Prague.

Enjoy your read !
-- 
Bruno STEVANT - ENST Bretagne



_______________________________________________
Softwires mailing list
Softwiresietf.org
http
s://www1.ietf.org/mailman/listinfo/softwires

For the Mesh framework, it makes sense to have multicast  
considerations included in the Framework document.
For the H&S framework, deploying multicast is mostly to
give to the  
involved components of a Softwire a role in the multicast
topology.  
No special encap., signaling, etc.
But there is still some scenarios to explore : ASM/SSM,
emmitter and  
receiver between SI ... This will probably be part of a
separate  
document, which will be guidelines for deployment.


Le 19 mars 07 à 18:07, Greg Shepherd a écrit :

> While I agree with the mcast portion of the draft -
mcast protocols  
> run over the tunnels - there's no mention of the
protocol end- 
> points; ie - is the hub expected to be part of the
mcast topology,  
> replicating accordingly?
>
> Thanks,
> Greg
>
> On 2/16/07, Bruno STEVANT <bruno.stevantenst-bretagne.fr> wrote:  
> Hi all,
>
> Carlos and I finally got an update on the H&S draft
out yesterday.
> http://www.ietf.org/internet-drafts/draft-i
etf-softwire-hs-framework-
> l2tpv2-03.txt
>
> Now we ask to anyone on the list to take 10 minutes of
their time and
> read this version carefully.
> Any remarks and comments are welcome ! From these
feedbacks we hope
> to release a new version before the cut-off for
Prague.
>
> Enjoy your read !
> --
> Bruno STEVANT - ENST Bretagne
>
>
>
> _______________________________________________
> Softwires mailing list
> Softwiresietf.org
> http
s://www1.ietf.org/mailman/listinfo/softwires
>

-- 
Bruno STEVANT - ENST Bretagne



_______________________________________________
Softwires mailing list
Softwiresietf.org
http
s://www1.ietf.org/mailman/listinfo/softwires

Le 19 mars 07 à 18:57, Greg Shepherd a écrit :
>
> On 3/19/07, Bruno STEVANT <bruno.stevantenst-bretagne.fr> wrote:  
> For the Mesh framework, it makes sense to have
multicast
> considerations included in the Framework document.
> For the H&S framework, deploying multicast is
mostly to give to the
> involved components of a Softwire a role in the
multicast topology.
> No special encap., signaling, etc.
>
> I was addressing the question as to whether the hub is
part of the  
> L3 infrastructure (and therefore running PIM) or is
just an L2 dumb  
> hub (transiting/broadcasting PIM messages to the other
spokes) -  
> neither appear as deployment-only issues to me. Either
way it will  
> need some mcast smarts and should be addressed in the
draft.
>

The hub (or Softwires Concentrator as we call it) has one
main  
connection to the IPv6 network and several tunnels.
In Phase 0, using L2TPv2, bridging between tunnel and main
connection  
(Ethernet) is not possible (or am I wrong somewhere ?)
So SC is part of L3 infrastructure as a router, and indeed
should  
include features to enable multicast.

Softwires Problem Statement mentions Multicast as one of the
 
requirement as :
"Existing multicast solutions can be used over the
softwire."

I think we addressed that requirement sufficiently with :
"Multicast protocols simply run over L2TPv2 Softwires
transparently  
together with other regular IP traffic."

Now what about the SC role in the multicast infrastructure
?
We shall also ask the same question about the SI.
Should they be proxy ? Multicast router ?

I think that answering such question is already talking
about  
deployment. And therefore this should be addressed in a
separate  
document.
-- 
Bruno STEVANT - ENST Bretagne



_______________________________________________
Softwires mailing list
Softwiresietf.org
http
s://www1.ietf.org/mailman/listinfo/softwires