http://www.ietf.org/audio/ietf87/ietf87-charlottenburg1-20130730-1700-pm3.mp3

WG status
draft-ietf-mboned-v4v6-mcast-ps

WGLC yielded many comments, both for and against.  May need to further scope use cases before moving forward.   Stig promised to check if his comments are addressed in latest version.

udp-multicast-guidelines
This doc should help clear AMT IESG discusses, although
someone is saying a circuit breaker might be needed.

tarapore-mbone-multicast-cdni
Discusses different ways of doing multicast peerings.
Between 2 multicast enabled domains using native, GRE
or AMT. Also with just one multicast enabled domain and
using AMT for receivers in the other domain.
Requesting comments and adoption.
6 people read the draft, and 6 favored adoption.

Multicast traceroute for MVPNs
Extending mtracev2 for MVPNs. Allows for tracing and
providing accurate errors for multicast between PEs.
Attempts to work with all MVPN deployment scenarios.
Looking for a home, requesting adoption in mboned.

Toerless: Not tracing the P hops through the core?
Presenter:They are traced, as well as ASBRs etc.
Toerless:No multicast?
Presenter: It is sent on the provider tunnels data would use.
Toerless: Is it a single hop unicast packet?
Presenter: It is multicast
Ice: Can the request make it upstream but data not
   making it downstream? Ice asking if can be made more
   generic.
Presenter: Decided to make it very specific, but open for
   considerations.
Thomas (L3VPN cochair): says work should be done
in mboned.
Greg: Ask if could punt it back to them if we prefer.
Thomas: Says L3VPN can provide feedback but think work
should be done in mboned.

Multicast state damping
Thomas Morin presenting. A lot of membership dynamicity
downstream can result in high control plane load in the core.
Proposing delaying prune propagation. Means no traffic
disruption. But increases bandwidth usage.

Various discussion on how the rate limiting works, some
confusion. Not all commenters read the draft.

For BGP it is similar to BGP dampening with some small
changes.

Asking for feedback. Looking for a home. Presenter
suggests mboned might be better place than pim or l3vpn.

Greg: protocol changes or bcp...

Missing discussion here.

Dino: Suggesting PIM over TCP would help.
Thomas: Saying that it would not help.

Stig: Would like review in pim etc. to make sure
      doesn't affect pim.

Dino: Says it may just behave like a loss of prunes.

Dino: Does it need to be a formal change, can it be
      done just by vendor locally?

Thomas: Useful to document behavior.


On using LISP for multicast with unicast replication
in the core.

Missing some details here while notetaker in mic line.

Greg and Stig suggesting some ways of solving the problem
with new active sources in the source cloud that are
not directly connected to the LISP router.

Ice saying this is a bit similar to an mLDP scenario.

Thomas asking if could do something like MVPN with
head-end replication

LISP-RE Dino   Replication Engineering
Proposing replication points in the core so avoid
doing only head-end replication.
Discussion between Dino and Greg on how it works.

Jeffrey For MVPNs there is a bit similar case with
segmented provider tunnels.

draft-tsao-mboned-v4v6mcast-dynamic-conversion
Prefix configurations specifying rules for stateless
translation.

draft-ietf-6man-multicast-addr-arch-update-01
6man draft on IPv6 architecture update.