Network Working Group                             Eric C. Rosen (Editor)
Internet Draft                                       Cisco Systems, Inc.
Expiration Date: December 2006
                                                 Rahul Aggarwal (Editor)
                                                        Juniper Networks

                                                               June 2006


                     Multicast in MPLS/BGP IP VPNs


                 draft-ietf-l3vpn-2547bis-mcast-02.txt

Status of this Memo

   By submitting this Internet-Draft, each author represents that any
   applicable patent or other IPR claims of which he or she is aware
   have been or will be disclosed, and any of which he or she becomes
   aware will be disclosed, in accordance with Section 6 of BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-
   Drafts.

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   The list of current Internet-Drafts can be accessed at
   http://www.ietf.org/ietf/1id-abstracts.txt.

   The list of Internet-Draft Shadow Directories can be accessed at
   http://www.ietf.org/shadow.html.

Abstract

   In order for IP multicast traffic within a BGP/MPLS IP VPN (Virtual
   Private Network) to travel from one VPN site to another, special
   protocols and procedures must be implemented by the VPN Service
   Provider.  These protocols and procedures are specified in this
   document.







Rosen & Raggarwa                                                [Page 1]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006




Table of Contents

    1          Specification of requirements  ......................   4
    2          Introduction  .......................................   4
    2.1        Optimality vs Scalability  ..........................   5
    2.1.1      Multicast Distribution Trees  .......................   7
    2.1.2      Ingress Replication through Unicast Tunnels  ........   8
    2.2        Overview  ...........................................   8
    2.2.1      Multicast Routing Adjacencies  ......................   8
    2.2.2      MVPN Definition  ....................................   8
    2.2.3      Auto-Discovery  .....................................   9
    2.2.4      PE-PE Multicast Routing Information  ................  10
    2.2.5      PE-PE Multicast Data Transmission  ..................  11
    2.2.6      Inter-AS MVPNs  .....................................  11
    2.2.7      Optional Deployment Models  .........................  12
    3          Concepts and Framework  .............................  12
    3.1        PE-CE Multicast Routing  ............................  12
    3.2        P-Multicast Service Interfaces (PMSIs)  .............  13
    3.2.1      Inclusive and Selective PMSIs  ......................  14
    3.2.2      Tunnels Instantiating PMSIs  ........................  15
    3.3        Use of PMSIs for Carrying Multicast Data  ...........  17
    3.3.1      MVPNs with Default MI-PMSIs  ........................  18
    3.3.2      When MI-PMSIs are Required  .........................  18
    3.3.3      MVPNs That Do Not Use MI-PMSIs  .....................  18
    4          BGP-Based Autodiscovery of MVPN Membership  .........  19
    5          PE-PE Transmission of C-Multicast Routing  ..........  21
    5.1        RPF Information for Unicast VPN-IP Routes  ..........  21
    5.2        PIM Peering  ........................................  23
    5.2.1      Full Per-MVPN PIM Peering Across a MI-PMSI  .........  23
    5.2.2      Lightweight PIM Peering Across a MI-PMSI  ...........  23
    5.2.3      Unicasting of PIM C-Join/Prune Messages  ............  24
    5.2.4      Details of Per-MVPN PIM Peering over MI-PMSI  .......  24
    5.2.4.1    PIM C-Instance Control Packets  .....................  25
    5.2.4.2    PIM C-instance RPF Determination  ...................  25
    5.3        Use of BGP for Carrying C-Multicast Routing  ........  26
    5.3.1      Sending BGP Updates  ................................  27
    5.3.2      Explicit Tracking  ..................................  29
    5.3.3      Withdrawing BGP Updates  ............................  29
    6          I-PMSI Instantiation  ...............................  29
    6.1        MVPN Membership and Egress PE Auto-Discovery  .......  30
    6.1.1      Auto-Discovery for Ingress Replication  .............  30
    6.1.2      Auto-Discovery for P-Multicast Trees  ...............  30
    6.2        C-Multicast Routing Information Exchange  ...........  31
    6.3        Aggregation  ........................................  31
    6.3.1      Aggregate Tree Leaf Discovery  ......................  31
    6.3.2      Aggregation Methodology  ............................  32



Rosen & Raggarwa                                                [Page 2]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


    6.3.3      Encapsulation of the Aggregate Tree  ................  33
    6.3.4      Demultiplexing C-multicast traffic  .................  33
    6.4        Mapping Received Packets to MVPNs  ..................  34
    6.4.1      Unicast Tunnels  ....................................  34
    6.4.2      Non-Aggregated P-Multicast Trees  ...................  35
    6.4.3      Aggregate P-Multicast Trees  ........................  36
    6.5        I-PMSI Instantiation Using Ingress Replication  .....  36
    6.6        Establishing P-Multicast Trees  .....................  37
    6.7        RSVP-TE P2MP LSPs  ..................................  38
    6.7.1      P2MP TE LSP Tunnel - MVPN Mapping  ..................  38
    6.7.2      Demultiplexing C-Multicast Data Packets  ............  39
    7          Optimizing Multicast Distribution via S-PMSIs  ......  39
    7.1        S-PMSI Instantiation Using Ingress Replication  .....  40
    7.2        Protocol for Switching to S-PMSIs  ..................  40
    7.2.1      A UDP-based Protocol for Switching to S-PMSIs  ......  40
    7.2.1.1    Binding a Stream to an S-PMSI  ......................  41
    7.2.1.2    Packet Formats and Constants  .......................  42
    7.2.2      A BGP-based Protocol for Switching to S-PMSIs  ......  42
    7.2.2.1    Advertising C-(S, G) Binding to a S-PMSI using BGP  .  42
    7.2.2.2    Explicit Tracking  ..................................  43
    7.2.2.3    Switching to S-PMSI  ................................  44
    7.3        Aggregation  ........................................  44
    7.4        Instantiating the S-PMSI with a PIM Tree  ...........  44
    7.5        Instantiating S-PMSIs using RSVP-TE P2MP Tunnels  ...  45
    8          Inter-AS Procedures  ................................  45
    8.1        Non-Segmented Inter-AS Tunnels  .....................  46
    8.1.1      Inter-AS MVPN Auto-Discovery  .......................  47
    8.1.2      Inter-AS MVPN Routing Information Exchange  .........  47
    8.1.3      Inter-AS I-PMSI  ....................................  47
    8.1.4      Inter-AS S-PMSI  ....................................  49
    8.2        Segmented Inter-AS Tunnels  .........................  49
    8.2.1      Inter-AS MVPN Auto-Discovery Tunnels  ...............  49
    8.2.1.1    Originating Inter-AS MVPN A-D Information  ..........  49
    8.2.1.2    Propagating Inter-AS MVPN A-D Information  ..........  50
    8.2.1.2.1  Inter-AS Auto-Discovery Route received via EBGP  ....  51
    8.2.1.2.2  Leaf Auto-Discovery Route received via EBGP  ........  52
    8.2.1.2.3  Inter-AS Auto-Discovery Route received via IBGP  ....  52
    8.2.2      Inter-AS MVPN Routing Information Exchange  .........  53
    8.2.3      Inter-AS I-PMSI  ....................................  54
    8.2.3.1    Support for Unicast VPN Inter-AS Methods  ...........  54
    8.2.4      Inter-AS S-PMSI  ....................................  55
    9          Duplicate Packet Detection and Single Forwarder PE  .  56
   10          Deployment Models  ..................................  59
   10.1        Co-locating C-RPs on a PE  ..........................  59
   10.1.1      Initial Configuration  ..............................  59
   10.1.2      Anycast RP Based on Propagating Active Sources  .....  60
   10.1.2.1    Receiver(s) Within a Site  ..........................  60
   10.1.2.2    Source Within a Site  ...............................  60



Rosen & Raggarwa                                                [Page 3]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   10.1.2.3    Receiver Switching from Shared to Source Tree  ......  61
   10.2        Using MSDP between a PE and a Local C-RP  ...........  61
   11          Encapsulations  .....................................  62
   11.1        Encapsulations for Single PMSI per Tunnel  ..........  62
   11.1.1      Encapsulation in GRE  ...............................  62
   11.1.2      Encapsulation in IP  ................................  64
   11.1.3      Encapsulation in MPLS  ..............................  64
   11.2        Encapsulations for Multiple PMSIs per Tunnel  .......  65
   11.2.1      Encapsulation in GRE  ...............................  65
   11.2.2      Encapsulation in IP  ................................  65
   11.3        Encapsulations for Unicasting PIM Control Messages  .  66
   11.4        General Considerations for IP and GRE Encaps  .......  66
   11.4.1      MTU  ................................................  66
   11.4.2      TTL  ................................................  66
   11.4.3      Differentiated Services  ............................  67
   11.4.4      Avoiding Conflict with Internet Multicast  ..........  67
   12          Security Considerations  ............................  67
   13          IANA Considerations  ................................  67
   14          Other Authors  ......................................  67
   15          Other Contributors  .................................  67
   16          Authors' Addresses  .................................  68
   17          Normative References  ...............................  69
   18          Informative References  .............................  70
   19          Full Copyright Statement  ...........................  71
   20          Intellectual Property  ..............................  71






1. Specification of requirements

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in [RFC2119].


2. Introduction

   [RFC4364] specifies the set of procedures which a Service Provider
   (SP) must implement in order to provide a particular kind of VPN
   service ("BGP/MPLS IP VPN") for its customers.  The service described
   therein allows IP unicast packets to travel from one customer site to
   another, but it does not provide a way for IP multicast traffic to
   travel from one customer site to another.

   This document extends the service defined in  [RFC4364] so that it



Rosen & Raggarwa                                                [Page 4]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   also includes the capability of handling iP multicast traffic.  This
   requires a number of different protocols to work together.  The
   document provides a framework describing how the various protocols
   fit together, and also provides detailed specification of some of the
   protocols.   The detailed specification of some of the other
   protocols is found in pre-existing documents or in companion
   documents.


2.1. Optimality vs Scalability

   In a "BGP/MPLS IP VPN" [RFC4364], unicast routing of VPN packets is
   achieved without the need to keep any per-VPN state in the core of
   the SP's network (the "P routers").  Routing information from a
   particular VPN is maintained only by the Provider Edge routers (the
   "PE routers", or "PEs") that attach directly to sites of that VPN.
   Customer data travels through the P routers in tunnels from one PE to
   another (usually MPLS Label Switched Paths, LSPs), so to support the
   VPN service the P routers only need to have routes to the PE routers.
   The PE-to-PE routing is optimal, but the amount of associated state
   in the P routers depends only on the number of PEs, not on the number
   of VPNs.

   However, in order to provide optimal multicast routing for a
   particular multicast flow, the P routers through which that flow
   travels have to hold state which is specific to that flow.
   Scalability would be poor if the amount of state in the P routers
   were proportional to the number of multicast flows in the VPNs.
   Therefore, when supporting multicast service for a BGP/MPLS IP VPN,
   the optimality of the multicast routing must be traded off against
   the scalability of the P routers.   We explain this below in more
   detail.

   If a particular VPN is transmitting "native" multicast traffic over
   the backbone,  we refer to it as an "MVPN".  By "native" multicast
   traffic, we mean packets that a CE sends to a PE, such that the IP
   destination address of the packets is a multicast group address, or
   the packets are multicast control packets addressed to the PE router
   itself, or the packets are IP multicast data packets encapsulated in
   MPLS.

   We say that the backbone multicast routing for a particular multicast
   group in a particular VPN is "optimal" if and only if all of the
   following conditions hold:







Rosen & Raggarwa                                                [Page 5]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


     - When a PE router receives a multicast data packet of that group
       from a CE router, it transmits the packet in such a way that the
       packet is received by every other PE router which is on the path
       to a receiver of that group;

     - The packet is not received by any other PEs;

     - While in the backbone, no more than one copy of the packet ever
       traverses any link.

     - While in the backbone, if bandwidth usage is to be optimized, the
       packet traverses minimum cost trees rather than shortest path
       trees.


   Optimal routing for a particular multicast group requires that the
   backbone maintain one or more source-trees which are specific to that
   flow.  Each such tree requires that state be maintained in all the P
   routers that are in the tree.

   This would potentially require an unbounded amount of state in the P
   routers, since the SP has no control of the number of multicast
   groups in the VPNs that it supports. Nor does the SP have any control
   over the number of transmitters in each group, nor of the
   distribution of the receivers.

   The procedures defined in this document allow an SP to provide
   multicast VPN service without requiring the amount of state
   maintained by the P routers to be proportional to the number of
   multicast data flows in the VPNs.  The amount of state is traded off
   against the optimality of the multicast routing.  Enough flexibility
   is provided so that a given SP can make his own tradeoffs between
   scalability and optimality.  An SP can even allow some multicast
   groups in some VPNs to receive optimal routing, while others do not.
   Of course, the cost of this flexibility is an increase in the number
   of options provided by the protocols.

   The basic technique for providing scalability is to aggregate a
   number of customer multicast flows onto a single multicast
   distribution tree through the P routers.  A number of aggregation
   methods are supported.

   The procedures defined in this document also accommodate the SP that
   does not want to build multicast distribution trees in his backbone
   at all; the ingress PE can replicate each multicast data packet and
   then unicast each replica through a tunnel to each egress PE that
   needs to receive the data.




Rosen & Raggarwa                                                [Page 6]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


2.1.1. Multicast Distribution Trees

   This document supports the use of a single multicast distribution
   tree in the backbone to carry all the multicast traffic from a
   specified set of one or more MVPNs.  Such a tree is referred to as an
   "Inclusive Tree". An Inclusive Tree which carries the traffic of more
   than one MVPN is an "Aggregate Inclusive Tree".  An Inclusive Tree
   contains, as its members, all the PEs that attach to any of the MVPNs
   using the tree.

   With this option, even if each tree supports only one MVPN, the upper
   bound on the amount of state maintained by the P routers is
   proportional to the number of VPNs supported, rather than to the
   number of multicast flows in those VPNs.  If the trees are
   unidirectional, it would be more accurate to say that the state is
   proportional to the product of the number of VPNs and the average
   number of PEs per VPN.  The amount of state maintained by the P
   routers can be further reduced by aggregating more MVPNs onto a
   single tree.  If each such tree supports a set of MVPNs, (call it an
   "MVPN aggregation set"), the state maintained by the P routers is
   proportional to the product of the number of MVPN aggregation sets
   and the average number of PEs per MVPN. Thus the state does not grow
   linearly with the number of MVPNs.

   However, as data from many multicast groups is aggregated together
   onto a single "Inclusive Tree", it is likely that some PEs will
   receive multicast data for which they have no need, i.e., some degree
   of optimality has been sacrificed.

   This document also provides procedures which enable a single
   multicast distribution tree in the backbone to be used to carry
   traffic belonging only to a specified set of one or more multicast
   groups, from one or more MVPNs. Such a tree is referred to as a
   "Selective Tree" and more specifically as an "Aggregate Selective
   Tree" when the multicast groups belong to different MVPNs.  By
   default, traffic from most multicast groups could be carried by an
   Inclusive Tree, while traffic from, e.g., high bandwidth groups could
   be carried in one of the "Selective Trees".  When setting up the
   Selective Trees, one should include only those PEs which need to
   receive multicast data from one or more of the groups assigned to the
   tree.  This provides more optimal routing than can be obtained by
   using only Inclusive Trees, though it requires additional state in
   the P routers.








Rosen & Raggarwa                                                [Page 7]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


2.1.2. Ingress Replication through Unicast Tunnels

   This document also provides procedures for carry MVPN data traffic
   through unicast tunnels from the ingress PE to each of the egress
   PEs. The ingress PE replicates the multicast data packet received
   from a CE and sends it to each of the egress PEs using the unicast
   tunnels.  This requires no multicast routing state in the P routers
   at all, but it puts the entire replication load on the ingress PE
   router, and makes no attempt to optimize the multicast routing.


2.2. Overview

2.2.1. Multicast Routing Adjacencies

   In BGP MPLS IP VPNs [RFC4364], each CE ("Customer Edge") router is a
   unicast routing adjacency of a PE router, but CE routers at different
   sites do not become unicast routing adjacencies of each other. This
   important characteristic is retained for multicast routing -- a CE
   router becomes a multicast routing adjacency of a PE router, but CE
   routers at different sites do not become multicast routing
   adjacencies of each other.

   The multicast routing protocol on the PE-CE link is presumed to be
   PIM.  The Sparse Mode, Dense Mode, Single Source Mode, and
   Bidirectional Modes are supported. A CE router exchanges "ordinary"
   PIM control messages with the PE router to which it is attached.

   The PEs attaching to a particular MVPN then have to exchange the
   multicast routing information with each other.  Two basic methods for
   doing this are defined: (1) PE-PE PIM, and (2) BGP.  In the former
   case, the PEs need to be multicast routing adjacencies of each other.
   In the latter case, they do not.  For example, each PE may be a BGP
   adjacency of a Route Reflector (RR), and not of any other PEs.

   To support the "Carrier's Carrier" model of [RFC4364], mLDP or BGP
   can be used on the PE-CE interface. This will be described in
   subsequent versions of this document.


2.2.2. MVPN Definition

   An MVPN is defined by two sets of sites, Sender Sites set and
   Receiver Sites set, with the following properties:







Rosen & Raggarwa                                                [Page 8]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


     -  Hosts within the Sender Sites set could originate multicast
       traffic for receivers in the Receiver Sites set.

     -  Receivers not in the Receiver Sites set should not be able to
       receive this traffic.

     -  Hosts within the Receiver Sites set could receive multicast
       traffic originated by any host in the Sender Sites set.

     -  Hosts within the Receiver Sites set should not be able to
       receive multicast traffic originated by any host that is not in
       the Sender Sites set.


   A site could be both in the Sender Sites set and Receiver Sites set,
   which implies that hosts within such a site could both originate and
   receive multicast traffic. An extreme case is when the Sender Sites
   set is the same as the Receiver Sites set, in which case all sites
   could originate and receive multicast traffic from each other.

   Sites within a given MVPN may be either within the same, or in
   different organizations, which implies than an MVPN can be either an
   Intranet or na Extranet.

   A given site may be in more than one MVPN, which implies that MVPNs
   may overlap.

   Not all sites of a given MVPN have to be connected to the same
   service provider, which implies that an MVPN can span multiple
   service providers.

   Another way to look at MVPN is to say that an MVPN is defined by a
   set of administrative policies. Such policies determine both Sender
   Sites set and Receiver Site set. Such policies are established by
   MVPN customers, but implemented/realized by MVPN Service Providers
   using the existing BGP/MPLS VPN mechanisms, such as Route Targets,
    with extensions, as necessary.


2.2.3. Auto-Discovery

   In order for the PE routers attaching to a given MVPN to exchange
   MVPN control information with each other, each one needs to discover
   all the other PEs that attach to the same MVPN.  (Strictly speaking,
   a PE in the receiver sites set need only discover the other PEs in
   the sender sites set and a PE in the sender sites set need only
   discover the other PEs in the receiver sites set.) This is referred
   to as "MVPN Auto-Discovery".



Rosen & Raggarwa                                                [Page 9]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   This document discusses two ways of providing MVPN autodiscovery:

     - BGP can be used for discovering and maintaining MVPN membership.
       The PE routers advertise their MVPN membership to other PE
       routers using BGP. A PE is considered to be a "member" of a
       particular MVPN if it contains a VRF (Virtual Routing and
       Forwarding table, see [RFC4364]) which is configured to contain
       the multicast routing information of that MVPN.  This auto-
       discovery option does not make any assumptions about the methods
       used for transmitting MVPN multicast data packets through the
       backbone.

     - If it is known that the multicast data packets of a particular
       MVPN are to be transmitted (at least, by default) through a non-
       aggregated Inclusive Tree which is to be set up by PIM-SM or
       PIM-Bidir, and if the PEs attaching to that MVPN are configured
       with the group address corresponding to that tree, then the PEs
       can auto-discover each other simply by joining the tree and then
       multicasting PIM Hellos over the tree.


2.2.4. PE-PE Multicast Routing Information

   The BGP/MPLS IP VPN [RFC4364] specification requires a PE to maintain
   at most one BGP peering with every other PE in the network. This
   peering is used to exchange VPN routing information. The use of Route
   Reflectors further reduces the number of BGP adjacencies maintained
   by a PE to exchange VPN routing information with other PEs. This
   document describes various options for exchanging MVPN control
   information between PE routers based on the use of PIM or BGP. These
   options have different overheads with respect to the number of
   routing adjacencies that a PE router needs to maintain to exchange
   MVPN control information with other PE routers. Some of these options
   allow the retention of the unicast BGP/MPLS VPN model letting a PE
   maintain at most one routing adjacency with other PE routers to
   exchange MVPN control information.

   The solution in [RFC4364] uses BGP to exchange VPN routing
   information between PE routers. This document describes various
   solutions for exchanging MVPN control information. One option is the
   use of BGP, providing reliable transport. Another option is the use
   of the currently existing, "soft state" PIM standard [PIM-SM].









Rosen & Raggarwa                                               [Page 10]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


2.2.5. PE-PE Multicast Data Transmission

   Like [RFC4364], this document decouples the procedures for exchanging
   routing information from the procedures for transmitting data
   traffic. Hence a variety of transport technologies may be used in the
   backbone. For inclusive trees, these transport technologies include
   unicast PE-PE tunnels (using MPLS or IP/GRE encapsulation), multicast
   distribution trees created by PIM-SSM, PIM-SM, or PIM-Bidir (using
   IP/GRE encapsulation), point-to-multipoint LSPs created by RSVP-TE or
   mLDP, and multipoint-to-multipoint LSPs created by mLDP.  (However,
   techniques for aggregating the traffic of multiple MVPNs onto a
   single multipoint-to-multipoint LSP or onto a single bidirectional
   multicast distribution tree are for further study.) For selective
   trees, only unicast PE-PE tunnels (using MPLS or IP/GRE
   encapsulation) and unidirectional single-source trees are supported,
   and the supported tree creation protocols are PIM-SSM (using IP/GRE
   encapsulation), RSVP-TE, and mLDP.

   In order to aggregate traffic from multiple MVPNs onto a single
   multicast distribution tree, it is necessary to have a mechanism to
   enable the egresses of the tree to demultiplex the multicast traffic
   received over the tree and to associate each received packet with a
   particular MVPN.  This document specifies a mechanism whereby
   upstream label assignment [MPLS-UPSTREAM-LABEL] is used by the root
   of the tree to assign a label to each flow.  This label is used by
   the receivers to perform the demultiplexing. This document also
   describes procedures based on BGP that are used by the root of an
   Aggregate Tree to advertise the Inclusive and/or Selective binding
   and the demultiplexing information to the leaves of the tree.

   This document also describes the data plane encapsulations for
   supporting the various SP multicast transport options.

   This document assumes that when SP multicast trees are used, traffic
   for a particular multicast group is transmitted by a particular PE on
   only one SP multicast tree. The use of multiple SP multicast trees
   for transmitting traffic belonging to a particular multicast group is
   for further study.


2.2.6. Inter-AS MVPNs

   [RFC4364] describes different options for supporting Inter-AS
   BGP/MPLS unicast VPNs. This document describes how Inter-AS MVPNs can
   be supported for each of the unicast BGP/MPLS VPN Inter-AS options.
   This document also specifies a model where Inter-AS MVPN service can
   be offered without requiring a single SP multicast tree to span
   multiple ASes. In this model, an inter-AS multicast tree consists of



Rosen & Raggarwa                                               [Page 11]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   a number of "segments", one per AS, which are stitched together at AS
   boundary points. These are known as "segmented inter-AS trees".  Each
   segment of a segmented inter-AS tree may use a different multicast
   transport technology.

   It is also possible to support Inter-AS MVPNs with non-segmented
   source trees that extend across AS boundaries.


2.2.7. Optional Deployment Models

   The document also discusses an optional MVPN deployment model in
   which PEs take on all or part of the role of a PIM RP (Rendezvous
   Point).  The necessary prtoocol extensions to support this are
   defined.


3. Concepts and Framework

3.1. PE-CE Multicast Routing

   Support of multicast in BGP/MPLS IP VPNs is modeled closely after
   support of unicast in BGP/MPLS IP VPNs. That is, a multicast routing
   protocol will be run on the PE-CE interfaces, such that PE and CE are
   multicast routing adjacencies on that interface.  CEs at different
   sites do not become multicast routing adjacencies of each other.

   If a PE attaches to n VPNs for which multicast support is provided
   (i.e., to n "MVPNs"), the PE will run n independent instances of a
   multicast routing protocol.  We will refer to these multicast routing
   instances as "VPN-specific multicast routing instances", or more
   briefly as "multicast C-instances". The notion of a "VRF" ("Virtual
   Routing and Forwarding Table"), defined in [RFC4364], is extended to
   include multicast routing entries as well as unicast routing entries.
   Each multicast routing entry is thus associated with a particular
   VRF.

   Whether a particular VRF belongs to an MVPN  or not is determined by
   configuration.

   In this document, we will not attempt to provide support for every
   possible multicast routing protocol that could possibly run on the
   PE-CE link.  Rather, we consider multicast C-instances only for the
   following multicast routing protocols:







Rosen & Raggarwa                                               [Page 12]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


     - PIM Sparse Mode (PIM-SM)

     - PIM Single Source Mode (PIM-SSM)

     - PIM Bidirectional Mode (PIM-Bidir)

     - PIM Dense Mode (PIM-DM)

   In order to support the "Carrier's Carrier" model of [RFC4364], mLDP
   or BGP will also be supported on the PE-CE interface; however, this
   is not described in this revision.

   As the document only supports PIM-based C-instances, we will
   generally use the term "PIM C-instances" to refer to the multicast
   C-instances.

   A PE router may also be running a "provider-wide" instance of PIM, (a
   "PIM P-instance"), in which it has a PIM adjacency with, e.g., each
   of its IGP neighbors (i.e., with P routers), but NOT with any CE
   routers, and not with other PE routers (unless another PE router
   happens to be an IGP adjacency).  In this case, P routers would also
   run the P-instance of PIM, but NOT a C-instance.  If there is a PIM
   P-instance, it may or may not have a role to play in support of VPN
   multicast; this is discussed in later sections.  However, in no case
   will the PIM P-instance contain VPN-specific multicast routing
   information.

   In order to help clarify when we are speaking of the PIM P-instance
   and when we are speaking of a PIM C-instance, we will also apply the
   prefixes "P-" and "C-" respectively to control messages, addresses,
   etc.  Thus a P-Join would be a PIM Join which is processed by the PIM
   P-instance, and a C-Join would be a PIM Join which is processed by a
   C-instance.  A P-group address would be a group address in the SP's
   address space, and a C-group address would be a group address in a
   VPN's address space.


3.2. P-Multicast Service Interfaces (PMSIs)

   Multicast data packets received by a PE over a PE-CE interface must
   be forwarded to one or more of the other PEs in the same MVPN for
   delivery to one or more other CEs.

   We define the notion of a "P-Multicast Service Interface" (PMSI).  If
   a particular MVPN is supported by a particular set of PE routers,
   then there will be a PMSI connecting those PE routers.  A PMSI is a
   conceptual "overlay" on the P network with the following property: a
   PE in a given MVPN can give a packet to the PMSI, and the packet will



Rosen & Raggarwa                                               [Page 13]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   be delivered to some or all of the other PEs in the MVPN, such that
   any PE receiving such a packet will be able to tell which MVPN the
   packet belongs to.

   As we discuss below, a PMSI may be instantiated by a number of
   different transport mechanisms, depending on the particular
   requirements of the MVPN and of the SP.  We will refer to these
   transport mechanisms as "tunnels".

   For each MVPN, there are one or more PMSIs that are used for
   transmitting the MVPN's multicast data from one PE to others.  We
   will use the term "PMSI" such that a single PMSI belongs to a single
   MVPN.  However, the transport mechanism which is used to instantiate
   a PMSI may allow a single "tunnel" to carry the data of multiple
   PMSIs.

   In this document we make a clear distinction between the multicast
   service (the PMSI) and its instantiation.  This allows us to separate
   the discussion of different services from the discussion of different
   instantiations of each service.  The term "tunnel" is used to refer
   only to the transport mechanism that instantiates a service.

   [This is a significant change from previous drafts on the topic of
   MVPN, which have used the term "Multicast Tunnel" to refer both to
   the multicast service (what we call here the PMSI) and to its
   instantiation.]


3.2.1. Inclusive and Selective PMSIs

   We will distinguish between three different kinds of PMSI:

     - "Multidirectional Inclusive" PMSI (MI-PMSI)

       A Multidirectional Inclusive PMSI is one which enables ANY PE
       attaching to a particular MVPN to transmit a message such that it
       will be received by EVERY other PE attaching to that MVPN.

       There is at most one MI-PMSI per MVPN.  (Though the tunnel which
       instantiates an MI-PMSI may actually carry the data of more than
       one PMSI.)

       An MI-PMSI can be thought of as an overlay broadcast network
       connecting the set of PEs supporting a particular MVPN.

       [The "Default MDTs" of rosen-08 provide the transport service of
       MI-PMSIs, in this terminology.]




Rosen & Raggarwa                                               [Page 14]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


     - "Unidirectional Inclusive" PMSI (UI-PMSI)

       A Unidirectional Inclusive PMSI is one which enables a particular
       PE, attached to a particular MVPN, to transmit a message such
       that it will be received by all the other PEs attaching to that
       MVPN.  There is at most one UI-PMSI per PE per MVPN, though the
       "tunnel" which instantiates a UI-PMSI may in fact carry the data
       of more than one PMSI.

     - "Selective" PMSI (S-PMSI).

       A Selective PMSI is one which provides a mechanism wherein a
       particular PE in an MVPN can multicast messages so that they will
       be received by a subset of the other PEs of that MVPN.  There may
       be an arbitrary number of S-PMSIs per PE per MVPN.  Again, the
       "tunnel" which instantiates a given S-PMSI may carry data from
       multiple S-PMSIs.

       [The "Data MDTs" of earlier drafts provide the transport service
       of "Selective PMSIs" in the terminology of this draft.]

   We will see in later sections the role played by these different
   kinds of PMSI.  We will use the term "I-PMSI" when we are not
   distinguishing between "MI-PMSIs" and "UI-PMSIs".


3.2.2. Tunnels Instantiating PMSIs

   A number of different tunnel setup techniques can be used to create
   the tunnels that instantiate the PMSIs.  Among these are:

     - PIM

       A PMSI can be instantiated as (a set of) Multicast Distribution
       Trees created by the PIM P-instance ("P-trees").

       PIM-SSM, PIM-Bidir, or PIM-SM can be used to create P-trees.
       (PIM-DM  is not supported for this purpose.)

       A single MI-PMSI can be instantiated by a single shared P-tree,
       or by a number of source P-trees (one for each PE of the MI-
       PMSI).  P-trees may be shared by multiple MVPNs (i.e., a given
       P-tree may be the instantiation of multiple PMSIs), as long as
       the enacapsulation provides some means of demultiplexing the data
       traffic by MVPN.

       Selective PMSIs are most instantiated by source P-trees, and are
       most naturally created by PIM-SSM, since by definition only one



Rosen & Raggarwa                                               [Page 15]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


       PE is the source of the multicast data on a Selective PMSI.

       [The "Default MDTs" of [rosen-08] are MI-PMSIs instantiated as
       PIM trees.  The "data MDTs" of [rosen-08] are S-PMSIs
       instantiated as PIM trees.]

     - MLDP

       A PMSI may be instantiated as one or more mLDP Point-to-
       Multipoint (P2MP) LSPs, or as an mLDP Multipoint-to-Point(MP2MP)
       LSP.  A Selective PMSI or a Unidirectional Inclusive PMSI would
       be instantiated as a single mLDP P2MP LSP, whereas a
       Multidirectional Inclusive PMSI could be instantiated either as a
       set of such LSPs (one for each PE in the MVPN) or as a single
       M2PMP LSP.

       MLDP P2MP LSPs can be shared across multiple MVPNs.

     - RSVP-TE

       A PMSI may be instantiated as one or more RSVP-TE Point-to-
       Multipoint (P2MP) LSPs.  A Selective PMSI or a Unidirectional
       Inclusive PMSI would be instantiated as a single RSVP-TE P2MP
       LSP, whereas a Multidirectional Inclusive PMSI would be
       instantiated as a set of such LSPs, one for each PE in the MVPN.
       RSVP-TE P2MP LSPs can be shared across multiple MVPNs.

     - A Mesh of Unicast Tunnels.

       If a PMSI is implemented as a mesh of unicast tunnels, a PE
       wishing to transmit a packet through the PMSI would replicate the
       packet, and send a copy to each of the other PEs.

       An MI-PMSI for a given MVPN can be instantiated as a full mesh of
       unicast tunnels among that MVPN's PEs.  A UI-PMSI or an S-PMSI
       can be instantiated as a partial mesh.


     - Unicast Tunnels to the Root of a P-Tree.

       Any type of PMSI can be instantiated through a method in which
       there is a single P-tree (created, for example, via PIM-SSM or
       via RSVP-TE), and a PE transmits a packet to the PMSI by sending
       it in a unicast tunnel to the root of that P-tree.  All PEs in
       the given MVPN would need to be leaves of the tree.

       When this instantiation method is used, the transmitter of the
       multicast data may receive its own data back.  Methods for



Rosen & Raggarwa                                               [Page 16]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


       avoiding this are for further study.

   It can be seen that each method of implementing PMSIs has its own
   area of applicability.  This specification therefore allows for the
   use of any of these methods.  At first glance, this may seem like an
   overabundance of options.  However, the history of multicast
   development and deployment should make it clear that there is no one
   option which is always acceptable.  The use of segmented inter-AS
   trees does allow each SP to select the option which it finds most
   applicable in its own environment, without causing any other SP to
   choose that same option.

   Specifying the conditions under which a particular tree building
   method is applicable is outside the scope of this document.


3.3. Use of PMSIs for Carrying Multicast Data

   Each PE supporting a particular MVPN must have a way of discovering:

     - The set of other PEs in its AS that are attached to sites of that
       MVPN, and the set of other ASes that have PEs attached to sites
       of that MVPN.  However, if segmented inter-AS trees are not used
       (see section 8.2), then each PE needs to know the entire set of
       PEs attached to sites of that MVPN.

     - If segmented inter-AS trees are to be used, the set of border
       routers in its AS that support inter-AS connectivity for that
       MVPN

     - If the MVPN is configured to use a default MI-PMSI, the
       information needed to set up and to use the tunnels instantiating
       the default MI-PMSI,

     - For each other PE, whether the PE supports Aggregate Trees for
       the MVPN, and if so, the demultiplexing information which must be
       provided so that the other PE can determine whether a packet
       which it received on an aggregate tree belongs to this MVPN.

   In some cases this information is provided by means of the BGP-based
   auto-discovery procedures detailed in section 4.  In other cases,
   this information is provided after discovery is complete, by means of
   procedurs defined in section 6.1.2.  In either case, the information
   which is provided must be sufficient to enable the PMSI to be bound
   to the identified tunnel, to enable the tunnel to be created if it
   does not already exist, and to enable the different PMSIs which may
   travel on the same tunnel to be properly demultiplexed.




Rosen & Raggarwa                                               [Page 17]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


3.3.1. MVPNs with Default MI-PMSIs

   If an MVPN uses an MI-PMSI, then the MI-PMSI for that MVPN will be
   created as soon as the necessary information has been obtained.
   Creating a PMSI means creating the tunnel which carries it (unless
   that tunnel already exists), as well as binding the PMSI to the
   tunnel. The MI-PMSI for that MVPN is then used as the default method
   of transmitting multicast data packets for that MVPN.  In effect, all
   the multicast streams for the MVPN are, by default, aggregated onto
   the MI-MVPN.

   If a particular multicast stream from a particular source PE has
   certain characteristics, it can be desirable to migrate it from the
   MI-PMSI to an S-PMSI.  Procedures for migrating a stream from an MI-
   PMSI to an S-PMSI are discussed in section 7.


3.3.2. When MI-PMSIs are Required

   MI-PMSIs are required under the following conditions:

     - The MVPN is using PIM-DM, or some other protocol (such as BSR)
       which relies upon flooding.  Only with an MI-PMSI can the C-data
       (or C-control-packets) received from any CE be flooded to all
       PEs.

     - If the procedure for carrying C-multicast routes from PE to PE
       involves the multicasting of P-PIM control messages among the PEs
       (see sections 5.2.1, 5.2.2, and 5.2.4).


3.3.3. MVPNs That Do Not Use MI-PMSIs

   If a particular MVPN does not use a default MI-PMSI, then its
   multicast data may be sent by default on a UI-PMSI.

   It is also possible to send all the multicast data on an S-PMSI,
   omitting any usage of I-PMSIs.  This prevents PEs from receiving data
   which they don't need, at the cost of requiring additional tunnels.
   However, in order to cost-effectively instantiate S-PMSIs with
   Aggregate P-trees, it is necessary for the transmitting PE to know
   which PEs need to receive which multicast streams. This is known as
   "explicit tracking", and the procedures to enable explicit tracking
   may themselves impose a cost.  This is further discussed in section
   7.2.2.2.






Rosen & Raggarwa                                               [Page 18]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


4. BGP-Based Autodiscovery of MVPN Membership

   BGP-based autodiscovery is done by means of a new address family, the
   MCAST-VPN address family. (This address family also has other uses,
   as will be seen later.)  Any PE which attaches to an MVPN must issue
   a BGP update message containing an NLRI in this address family, along
   with a specific set of attributes.  In this document, we specify the
   information which must be contained in these BGP updates in order to
   provide auto-discovery.  The encoding details, along with the
   complete set of detailed procedures, are specified in a separate
   document [MVPN-BGP].

   This section specifies the intra-AS BGP-based autodiscovery
   procedures.  When segmented inter-AS trees are used, additional
   procedures are needed, as specified in section 8.  Further detail may
   be found in [MVPN-BGP].  (When segmented inter-AS trees are not used,
   the inter-AS procedures are almost identical to the intra-AS
   procedures.)

   BGP-based autodiscovery uses a particular kind of MCAST-VPN route
   known as an "auto-discovery routes", or "A-D route".

   An "intra-AS A-D route" is a particular kind of A-D route that is
   never distributed outside its AS of origin.  Intra-AS A-D routes are
   originated by the PEs that are (directly) connected to the site(s) of
   that MVPN.

   For the purpose of auto-discovery, each PE attached to a site in a
   given MVPN must originate an intra-AS auto-discovery route.  The NLRI
   of that route must the following information:

     - The route type (i.e., intra-AS A-D route)

     - IP address of the originating PE

     - An RD configured locally for the MVPN.  This is an RD which can
       be prepended to that IP address to form a globally unique VPN-IP
       address of the PE.

   The A-D route must also carry the following attributes:

     - One or more Route Target attributes.  If any other PE has one of
       these Route Targets configured for import into a VRF, it treats
       the advertising PE as a member in the MVPN to which the VRF
       belongs. This allows each PE to discover the PEs that belong to a
       given MVPN.  More specifically it allows a PE in the receiver
       sites set to discover the PEs in the sender sites set of the MVPN
       and the PEs in the sender sites set of the MVPN to discover the



Rosen & Raggarwa                                               [Page 19]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


       PEs in the receiver sites set of the MVPN. The PEs in the
       receiver sites set would be configured to import the Route
       Targets advertised in the BGP Auto-Discovery routes by PEs in the
       sender sites set. The PEs in the sender sites set would be
       configured to import the Route Targets advertised in the BGP
       Auto-Discovery routes by PEs in the receiver sites set.

     * PMSI tunnel attribute.  This attribute is present if and only if
       a default MI-PMSI is to be used for the MVPN.  It contains the
       following information:

           whether the MI-PMSI is instantiated by

             + A PIM-Bidir tree,

             + a set of PIM-SSM trees,

             + a set of PIM-SM trees

             + a set of RSVP-TE point-to-multipoint LSPs

             + a set of mLDP point-to-multipoint LSPs

             + an mLDP multipoint-to-multipoint LSP

             + a set of unicast tunnels

             + a set of unicast tunnels to the root of a shared tree (in
               this case the root must be identified)

         * If the PE wishes to setup a default tunnel to instantiate the
           I-PMSI, a unique identifier for the tunnel used to
           instantiate the I-PMSI.

           All the PEs attaching to a given MVPN (within a given AS)
           must have been configured with the same PMSI tunnel attribute
           for that MVPN.  They are also expected to know the
           encapsulation to use.

           Note that a default tunnel can be identified at discovery
           time only if the tunnel already exists (e.g., it was
           constructed by means of configuration), or if it can be
           constructed without each PE knowing the the identities of all
           the others (e.g., it is constructed by a receiver-initiated
           join technique such as PIM or mLDP).

           In other cases, a default tunnel cannot be identified until
           the PE has discovered one or more of the other PEs.   This



Rosen & Raggarwa                                               [Page 20]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


           will be the case, for example, if the tunnel is an RSVP-TE
           P2MP LSP, which must be set up from the head end.  In these
           cases, a PE will first send an A-D route without a tunnel
           identifier, and then will send another one with a tunnel
           identifier after discovering one or more of the other PEs.

         * Whether the tunnel used to instantiate the I-PMSI for this
           MVPN is aggregating I-PMSIs from multiple MVPNs.  This will
           affect the encapsulation used.  If aggregation is to be used,
           a demultiplexor value to be carried by packets for this
           particular MVPN must also be specified. The demultiplexing
           mechanism and signaling procedures are described in section
           6.
       Further details of the use of this information are provided in
       subsequent sections.


5. PE-PE Transmission of C-Multicast Routing

   As a PE attached to a given MVPN receives C-Join/Prune messages from
   its CEs in that MVPN, it must convey the information contained in
   those messages to other PEs that are attached to the same MVPN.

   There are several different methods for doing this. As these methods
   are not interoperable, the method to be used for a particular MVPN
   must either be configured, or discovered as part of the BGP-based
   auto-discovery process.


5.1. RPF Information for Unicast VPN-IP Routes

   When a PE receives a C-Join/Prune message from a CE, the message
   identifies a particular multicast flow as belong either to a source
   tree (S,G) or to a shared tree (*,G).  We use the term C-source to
   refer to S, in the case of a source tree, or to the Rendezvous Point
   (RP) for G, in the case of (*,G).  The PE needs to find the "upstream
   multicast hop" for the (S,G) or (*,G) flow, and it does this by
   looking up the C-source in the unicast VRF associated with the PE-CE
   interfaces over which the C-Join/Prune was received.  To facilitate
   this, all unicast VPN-IP routes from an MVPN will carry RPF
   information, which identifies the PE that originated the route, as
   well as identifying the Autonomous System containing that PE.  This
   information is consulted when a PE does an "RPF lookup" of the C-
   source as part of processing the C-Join/Prune messages.  This RPF
   information contains the following:






Rosen & Raggarwa                                               [Page 21]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


     - Source AS Extended Community

       To support MVPN a PE that originates a (unicast) route to VPN-
       IPv4 addresses MUST include in the BGP Update message that
       carries this route the Source AS extended community, except if it
       is known a priori that none of these addresses will act as
       multicast sources and/or RP, in which case the (unicast) route
       need not carry the Source AS extended community.  The Global
       Administrator field of this community MUST be set to the
       autonomous system number of the PE. The Local Administrator field
       of this community SHOULD be set to 0. This community is described
       further in [MVPN-BGP].

     - Route Import Extended Community

       To support MVPN in addition to the import/export Route Target(s)
       used by the unicast routing, each VRF on a PE MUST have an import
       Route Target that is unique to this VRF, except if it is known a
       priori that none of the (local) MVPN sites associated with the
       VRF contain multicast source(s) and/or RP, in which case the VRF
       need not have this import Route Target. This Route Target MUST be
       IP address specific, and is constructed as follows:

     + The Global Administrator field of the Route Target MUST be set to
       an IP address of the PE. This address MUST be a routable IP
       address.  This address MAY be common for all the VRFs on the PE
       (e.,g., this address may be PE's loopback address).

     + The Local Administrator field of the Route Target associated with
       a given VRF contains a 2 octets long number that uniquely
       identifies that VRF within the PE that contains the VRF
       (procedures for assigning such numbers are purely local to the
       PE, and outside the scope of this document).

   A PE that originates a (unicast) route to VPN-IPv4 addresses MUST
   include in the BGP Updates message that carries this route the Route
   Import extended community that has the value of this Route Target,
   except if it is known a priori that none of these addresses will act
   as multicast sources and/or RP, in which case the (unicast) route
   need not carry the Route Import extended community.

   The Route Import Extended Community is described further in [MVPN-
   BGP].








Rosen & Raggarwa                                               [Page 22]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


5.2. PIM Peering

5.2.1. Full Per-MVPN PIM Peering Across a MI-PMSI

   If the set of PEs attached to a given MVPN are connected via a MI-
   PMSI, the PEs can form "normal" PIM adjacencies with each other.
   Since the MI-PMSI functions as a broadcast network, the standard PIM
   procedures for forming and maintaining adjacencies over a LAN can be
   applied.

   As a result, the C-Join/Prune messages which a PE receives from a CE
   can be multicast to all the other PEs of the MVPN.  PIM "join
   suppression" can be enabled and the PEs can send Asserts as needed.

   [This is the procedure specified in [rosen-08].]


5.2.2. Lightweight PIM Peering Across a MI-PMSI

   The procedure of the previous section has the following
   disadvantages:

     - Periodic Hello messages must be sent by all PEs.

       Standard PIM procedures require that each PE in a particular MVPN
       periodically multicast a Hello to all the other PEs in that MVPN.
       If the number of MVPNs becomes very large, sending and receiving
       these Hellos can become a substantial overhead for the PE
       routers.

     - Periodic retransmission of C-Join/Prune messages.

       PIM is a "soft-state" protocol, in which reliability is assured
       through frequent retransmissions (refresh) of control messages.
       This too can begin to impose a large overhead on the PE routers
       as the number of MVPNs grows.

   The first of these disadvantages is easily remedied.  The reason for
   the periodic PIM Hellos is to ensure that each PIM speaker on a LAN
   knows who all the other PIM speakers on the LAN are.  However, in the
   context of MVPN, PEs in a given MVPN can learn the identities of all
   the other PEs in the MVPN by means of the BGP-based auto-discovery
   procedure of section 4.  In that case, the periodic Hellos would
   serve no function, and could simply be eliminated.  (Of course, this
   does imply a change to the standard PIM procedures.)

   When Hellos are suppressed, we may speak of "lightweight PIM
   peering".



Rosen & Raggarwa                                               [Page 23]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   The periodic refresh of the C-Join/Prunes is not as simple to
   eliminate.  The L3VPN WG has asked the PIM WG to specify "refresh
   reduction" procedures for PIM, so as to eliminate the need for the
   periodic refreshes.  If and when such procedures have been specified,
   it will be very useful to incorporate them, so as to make the
   lightweight PIM peering procedures even more lightweight.


5.2.3. Unicasting of PIM C-Join/Prune Messages

   PIM does not require that the C-Join/Prune messages which a PE
   receives from a CE to be multicast to all the other PEs; it allows
   them to be unicast to a single PE, the one which is upstream on the
   path to the root of the multicast tree mentioned in the Join/Prune
   message. Note that when the C-Join/Prune messages are unicast, there
   is no such thing as "join suppression".  Therefore PIM Refresh
   Reduction may be considered to be a pre-requisite for the procedure
   of unicasting the C-Join/Prune messages.

   When the C-Join/Prunes are unicast, they are not transmitted on a
   PMSI at all.  Note that the procedure of unicasting the C-Join/Prunes
   is different than the procedure of transmitting the C-Join/Prunes on
   an MI-PMSI which is instantiated as a mesh of unicast tunnels.

   If there are multiple PEs that can be used to reach a given C-source,
   procedures described in section 9 MUST be used to ensue that, at
   least within a single AS, all PEs choose the same PE to reach the C-
   source.


5.2.4. Details of Per-MVPN PIM Peering over MI-PMSI

   In this section, we assume that inter-AS MVPNs will be supported by
   means of non-segmented inter-AS trees.  Support for segmented inter-
   AS trees with PIM peering is for further study.

   When an MVPN uses an MI-PMSI, the C-instances of that MVPN can treat
   the MI-PMSI as a LAN interface, and form either full PIM adjacencies
   or lightweight PIM adjacencies with each other over that "LAN
   interface".

   To form a full PIM adjacency, the PEs execute the PIM LAN procedures,
   including the generation and processing of PIM Hello, Join/Prune,
   Assert, DF election and other PIM control packets.  These are
   executed independently for each C-instance.  PIM "join suppression"
   SHOULD be enabled.

   If it is known that all C-instances of a particular MVPN can support



Rosen & Raggarwa                                               [Page 24]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   lightweight adjacencies, then lightweight adjacencies MUST be used.
   If it is not known that all such C-instances support lightweight
   instances, then full adjacencies MUST be used.  Whether all the C-
   instances support lightweight adjacencies is known by virtue of the
   BGP-based auto-discovery procedures (combined with configuration).
   This knowledge might change over time, so the PEs must be able to
   switch in real time between the use of full adjacencies and
   lightweight adjacencies.

   The difference between a lightweight adjacency and a full adjacency
   is that no PIM Hellos are sent or received on a lightweight
   adjacency.  The function which Hellos usually provide in PIM can be
   provided in MVPN by the BGP-based auto-discovery procedures, so the
   Hellos become superfluous.

   Whether or not Hellos are sent, if PIM Refresh Reduction procedures
   are available, and all the PEs supporting the  MVPN are known to
   support these procedures, then the refresh reduction procedures MUST
   be used.


5.2.4.1. PIM C-Instance Control Packets

   All PIM C-Instance control packets of a particular MVPN are addressed
   to the ALL-PIM-ROUTERS (224.0.0.13) IP destination address, and
   transmitted over the MI-PMSI of that MVPN.  While in transit in the
   P-network, the packets are encapsulated as required for the
   particular kind of tunnel that is being used to instantiate the MI-
   PMSI.  Thus the C-instance control packets are not processed by the P
   routers, and MVPN-specific PIM routes can be extended from site to
   site without appearing in the P routers.


5.2.4.2. PIM C-instance RPF Determination

   Although the MI-PMSI is treated by PIM as a LAN interface, unicast
   routing is NOT run over it, and there are no unicast routing
   adjacencies over it.  It is therefore necessary to specify special
   procedures for determining when the MI-PMSI is to be regarded as the
   "RPF Interface" for a particular C-address.

   When a PE needs to determine the RPF interface of a particular C-
   address, it looks up the C-address in the VRF. If the route matching
   it (call this the "RPF route") is not a VPN-IP route learned from
   MP-BGP as described in [RFC4364], or if that route's outgoing
   interface is one of the interfaces associated with the VRF, then
   ordinary PIM procedures for determining the RPF interface apply.




Rosen & Raggarwa                                               [Page 25]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   However, if the RPF route is a VPN-IP route whose outgoing interface
   is not one of the interfaces associated with the VRF, then PIM will
   consider the outgoing interface to be the MI-PMSI associated with the
   VPN-specific PIM instance.

   Once PIM has determined that the RPF interface for a particular C-
   address is the MI-PMSI, it is necessary for PIM to determine the RPF
   neighbor for that C-address.  This will be one of the other PEs that
   is a PIM adjacency over the MI-PMSI.

   When a PE distributes a given VPN-IP route via BGP, the PE must
   determine whether that route might possibly be regarded, by another
   PE, as an RPF route. (If a given VRF is part of an MVPN, it may be
   simplest to regard every route exported from that VRF to be a
   potential RPF route.)  If the given VPN-IP route is a potential RPF
   route, then when the VPN-IP route is distributed by BGP, it SHOULD be
   accompanied by "RPF information".

   The RPF information contains an IP address which the PE will use as
   its Source IP address in any PIM control messages which it transmits
   to other PEs in the same MVPN.

   When a PE has determined that the RPF interface for a particular C-
   address is the MI-PMSI, it must look up the RPF information that was
   distributed along with the VPN-IP address corresponding to that C-
   address.  The IP address in this RPF information will be considered
   to be the IP address of the RPF adjacency for the C-address.

   If the RPF information is not present, but the "BGP Next Hop" for the
   C-address is one of the PEs that is a PIM adjacency over the MI-PMSI,
   then that PE should be treated as the RPF adjacency for that C-
   address.  However, if the MVPN spans multiple Autonomous Systems, the
   BGP Next Hop might not be a PIM adjacency, and if that is the case
   the RPF check will not succeed unless the RPF information is used.


5.3. Use of BGP for Carrying C-Multicast Routing

   It is possible to use BGP to carry C-multicast routing information
   from PE to PE, dispensing entirely with the transmission of C-
   Join/Prune messages from PE to PE. This section describes the
   procedures for carrying intra-AS multicast routing information.
   Inter-AS procedures are described in section 8.








Rosen & Raggarwa                                               [Page 26]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


5.3.1. Sending BGP Updates

   The MCAST-VPN address family is used for this purpose.  MCAST-VPN
   routes used for the purpose of carrying C-multicast routing
   information are distinguished from those used for the purpose of
   carrying auto-discovery information by means of a "route type" field
   which is encoded into the NLRI.  The following information is
   required in BGP to advertise the MVPN routing information.  The NLRI
   contains:

     - The type of C-multicast route.

       There are two types:

         * source tree join

         * shared tree join

     - The RD configured, for the MVPN, on the PE that is advertising
       the information.  This is required to uniquely identify the <C-
       Source, C-Group> as the addresses could overlap between different
       MVPNs.

     - The C-Source address. (Omitted if the route type is "shared tree
       join")

     - The C-Group address.

     - The RD from the VPN-IP route to the C-source.

       That is, the route to the C-source is looked up in the local
       unicast VRF associated with the CE-PE interface over which the
       C-multicast control packet arrived.   The corresponding VPN-IP
       route is then examined, and the RD from that route is placed into
       the  C-multicast route.

       Note that this RD is NOT necessarily one which is configured on
       the local PE.  Rather it is one which is configured on the remote
       PE that is on the path to the C-source.

   The following attribute must also be included:

     - The upstream multicast hop.

       If a PE receives a C-Join (*, G) from a CE, the C-source is
       considered to be the C-RP for the particular C-G.  When the C-
       multicast route represents a "shared tree join", it is presumed
       that the root of the tree (e.g., the RP) is determined by some



Rosen & Raggarwa                                               [Page 27]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


       means outside the scope of this specification.

       When the PE processes a C-PIM Join/Prune message, the route to
       the C-source is looked up in the local unicast VRF associated
       with the CE-PE interface over which the C-multicast control
       packet arrived.  The corresponding VPN-IP route is then examined.
       If the AS specified therein is the local AS, or if no AS is
       specified therein, then the PE specified therein becomes the
       upstream multicast hop.  If the AS specified therein is a remote
       AS, the BGP next hop on the route to the  MVPN Auto-Discovery
       route advertised by the remote AS, becomes the upstream multicast
       hop.

       N.B.: It is possible that here is more than one unicast VPN-IP
       route to the C-source.  In this case, the route that was
       installed in the VRF is not necessarily the route that must be
       chosen by the PE.  In order to choose the proper route, the
       procedures followed in section 9 MUST be followed.

   The upstream multicast hop is identified in an Extended Communities
   attribute to facilitate the optional use of filters which can prevent
   the distribution of the update to BGP speakers other than the
   upstream multicast hop.

   When a PE distributes this information via BGP, it must include a
   Route Import Extended Communities attribute learned from the RPF
   information.

   Note that for these procedures to work the VPN-IP route MUST contain
   the RPF information.


   Note that there is no C-multicast route corresponding to the PIM
   function of pruning a source off the shared tree when a PE switches
   from a <C-*, C-G> tree to a <C-S, C-G> tree.  Section 9 of this
   document specifies a mandatory procedure that ensures that if any PE
   joins a <C-S, C-G> source tree, all other PEs that have joined or
   will join the <C-*, C-G> shared tree will also join the <C-S, C-G>
   source tree.  This eliminates the need for a C-multicast route that
   prunes C-S off the <C-*, C-G> shared tree when switching from <C-*,
   C-G> to <C-S, C-G> tree.










Rosen & Raggarwa                                               [Page 28]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


5.3.2. Explicit Tracking

   Note that the upstream multicast hop is NOT part of the NLRI in the
   C-multicast BGP routes.  This means that if several PEs join the same
   C-tree, the BGP routes they distribute to do so are regarded by BGP
   as comparable routes, and only one will be installed.  If a route
   reflector is being used, this further means that the PE which is used
   to reach the C-source will know only that one or more of the other
   PEs have joined the tree, but it won't know which one.  That is, this
   BGP update mechanism does not provide "explicit tracking".  Explicit
   tracking is not provided by default because it increases the amount
   of state needed and thus decreases scalability.  Also, as
   constructing the C-PIM messages to send "upstream" for a given tree
   does not depend on knowing all the PEs that are downstream on that
   tree, there is no reason for the C-multicast route type updates to
   provide explicit tracking.

   There are some cases in which explicit tracking is necessary in order
   for the PEs to set up certain kinds of P-trees.  There are other
   cases in which explicit tracking is desirable in order to determine
   how to optimally aggregate multicast flows onto a given aggregate
   tree.  As these functions have to do with the setting up of
   infrastructure in the P-network, rather than with the dissemination
   of C-multicast routing information, any explicit tracking that is
   necessary is handled by sending the "source active" A-D routes, that
   are described in sections 9 and 10.  Detailed procedures for turning
   on explicit tracking can be found in [MVPN-BGP].


5.3.3. Withdrawing BGP Updates

   A PE removes itself from a C-multicast tree (shared or source) by
   withdrawing the corresponding BGP update.

   If a PE has pruned a C-source from a shared C-multicast tree, and it
   needs to "unprune" that source from that tree, it does so by
   withdrawing the route that pruned the source from the tree.


6. I-PMSI Instantiation

   This section describes how tunnels in the SP network can be used to
   instantiate an I-PMSI for an MVPN on a PE.   When C-multicast data is
   delivered on an I-PMSI, the data will go to all PEs that are on the
   path to receivers for that C-group, but may also go to PEs that are
   not on the path to receivers for that C-group.

   The tunnels which instantiate I-PMSIs can be either PE-PE unicast



Rosen & Raggarwa                                               [Page 29]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   tunnels or P-multicast trees. When PE-PE unicast tunnels are used the
   PMSI is said to be instantiated using ingress replication.

   [Editor's Note: MD trees described in [ROSEN-8, MVPN-BASE] are an
   example of P-multicast trees. Also Aggregate Trees described in
   [RAGGARWA-MCAST] are an example of P-multicast trees.]


6.1. MVPN Membership and Egress PE Auto-Discovery

   As described in section 4 a PE discovers the MVPN membership
   information of other PEs using BGP auto-discovery mechanisms or using
   a mechanism that instantiates a MI-PMSI interface. When a PE supports
   only a UI-PMSI service for an MVPN, it MUST rely on the BGP auto-
   discovery mechanisms for discovering this information. This
   information also results in a PE in the sender sites set discovering
   the leaves of the P-multicast tree, which are the egress PEs that
   have sites in the receiver sites set in one or more MVPNs mapped onto
   the tree.


6.1.1. Auto-Discovery for Ingress Replication

   In order for a PE to use Unicast Tunnels to send a C-multicast data
   packet for a particular MVPN to a set of remote PEs, the remote PEs
   must be able to correctly decapsulate such packets and to assign each
   one to the proper MVPN. This requires that the encapsulation used for
   sending packets through the tunnel have demultiplexing information
   which the receiver can associate with a particular MVPN.

   If ingress replication is being used for an MVPN, the PEs announce
   this as part of the BGP based MVPN membership auto-discovery process,
   described in section 4.  The PMSI tunnel attribute specifies ingress
   replication.  The demultiplexor value is a downstream-assigned MPLS
   label (i.e., assigned by the PE that originated the A-D route, to be
   used by other PEs when they send multicast packets on a unicast
   tunnel to that PE).

   Other demultiplexing procedures for unicast are under consideration.


6.1.2. Auto-Discovery for P-Multicast Trees

   A PE announces the P-multicast technology it supports for a specified
   MVPN, as part of the BGP MVPN membership discovery. This allows other
   PEs to determine the P-multicast technology they can use for building
   P-multicast trees to instantiate an I-PMSI. If a PE has a default
   tree instantiation of an I-PMSI, it also announces the tree



Rosen & Raggarwa                                               [Page 30]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   identifier as part of the auto-discovery, as well as announcing its
   aggregation capability.

   The announcement of a tree identifier at discovery time is only
   possible if the tree already exists (e.g., a preconfigured "traffic
   engineered" tunnel), or if the tree can be constructed dynamically
   without any PE having to know in advance all the other PEs on the
   tree (e.g., the tree is created by receiver-initiated joins).


6.2. C-Multicast Routing Information Exchange

   When a PE doesn't support the use of a MI-PMSI for a given MVPN, it
   MUST either unicast MVPN routing information using PIM or else use
   BGP for exchanging the MVPN routing information.


6.3. Aggregation

   A P-multicast tree can be used to instantiate a PMSI service for only
   one MVPN or for more than one MVPN. When a P-multicast tree is shared
   across multiple MVPNs it is termed an Aggregate Tree [RAGGARWA-
   MCAST]. The procedures described in this document allow a single SP
   multicast tree to be shared across multiple MVPNs. The procedures
   that are specific to aggregation are optional and are explicitly
   pointed out. Unless otherwise specified a P-multicast tree technology
   supports aggregation.

   Aggregate Trees allow a single P-multicast tree to be used across
   multiple MVPNs and hence state in the SP core grows per-set-of-MVPNs
   and not per MVPN.  Depending on the congruency of the aggregated
   MVPNs, this may result in trading off optimality of multicast
   routing.

   An Aggregate Tree can be used by a PE to provide an UI-PMSI or MI-
   PMSI service for more than one MVPN. When this is the case the
   Aggregate Tree is said to have an inclusive mapping.


6.3.1. Aggregate Tree Leaf Discovery

   BGP MVPN membership discovery allows a PE to determine the different
   Aggregate Trees that it should create and the MVPNs that should be
   mapped onto each such tree. The leaves of an Aggregate Tree are
   determined by the PEs, supporting aggregation, that belong to all the
   MVPNs that are mapped onto the tree.

   If an Aggregate Tree is used to instantiate one or more S-PMSIs, then



Rosen & Raggarwa                                               [Page 31]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   it may be desirable for the PE at the root of the tree to know which
   PEs (in its MVPN) are receivers on that tree.  This enables the PE to
   decide when to aggregate two S-PMSIs, based on congruence (as
   discussed in the next seciton).  Thus explicit tracking may be
   required.  Since the procedures for disseminating C-multicast routes
   do not provide explicit tracking, a type of A-D route known as a
   "Leaf A-D Route" is used.  The PE which wants to assign a particular
   C-multicast flow to a particular Aggregate Tree can send an A-D route
   which elicits Leaf A-D routes from the PEs that need to receive that
   C-multicast flow.  This provides the explicit tracking information
   needed to support the aggregation methodology discussed in the next
   section.


6.3.2. Aggregation Methodology

   This document does not specify the mandatory implementatino of any
   particular set of rules for determining whether or not the PMSIs of
   two particular MVPNs are to be instantiated by the same Aggregate
   Tree.  This determination can be made by implementation-specific
   heuristics, by configuration, or even perhaps by the use of offline
   tools.

   It is the intention of this document that the control procedures will
   always result in all the PEs of an MVPN to agree on the PMSIs which
   are to be used and on the tunnels used to instantiate those PMSIs.

   This section discusses potential methodologies with respect to
   aggregation.

   The "congruency" of aggregation is defined by the amount of overlap
   in the leaves of the customer trees that are aggregated on a SP tree.
   For Aggregate Trees with an inclusive mapping the congruency depends
   on the overlap in the membership of the MVPNs that are aggregated on
   the tree. If there is complete overlap i.e. all MVPNs have exactly
   the same sites, aggregation is perfectly congruent. As the overlap
   between the MVPNs that are aggregated reduces, i.e. the number of
   sites that are common across all the MVPNs reduces, the congruency
   reduces.

   If aggregation is done such that it is not perfectly congruent a PE
   may receive traffic for MVPNs to which it doesn't belong. As the
   amount of multicast traffic in these unwanted MVPNs increases
   aggregation becomes less optimal with respect to delivered traffic.
   Hence there is a tradeoff between reducing state and delivering
   unwanted traffic.

   An implementation should provide knobs to control the congruency of



Rosen & Raggarwa                                               [Page 32]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   aggregation. These knobs are implementation dependent. Configuring
   the percentage of sites that MVPNs must have in common to be
   aggregated, is an example of such a knob. This will allow a SP to
   deploy aggregation depending on the MVPN membership and traffic
   profiles in its network.  If different PEs or servers are setting up
   Aggregate Trees this will also allow a service provider to engineer
   the maximum amount of unwanted MVPNs hat a particular PE may receive
   traffic for.


6.3.3. Encapsulation of the Aggregate Tree

   An Aggregate Tree may use an IP/GRE encapsulation or an MPLS
   encapsulation.  The protocol type in the IP/GRE header in the former
   case and the protocol type in the data link header in the latter need
   further explanation. This will be specified in a separate document.


6.3.4. Demultiplexing C-multicast traffic

   When multiple MVPNs are aggregated onto one P-Multicast tree,
   determining the tree over which the packet is received is not
   sufficient to determine the MVPN to which the packet belongs.  The
   packet must also carry some demultiplexing information to allow the
   egress PEs to determine the MVPN to which the packet belongs.  Since
   the packet has been multicast through the P network, any given
   demultiplexing value must have the same meaning to all the egress
   PEs.  The demultiplexing value is a MPLS label that corresponds to
   the multicast VRF to which the packet belongs. This label is placed
   by the ingress PE immediately beneath the P-Multicast tree header.
   Each of the egress PEs must be able to associate this MPLS label with
   the same MVPN.  If downstream label assignment were used this would
   require all the egress PEs in the MVPN to agree on a common label for
   the MVPN. Instead the MPLS label is upstream assigned [MPLS-
   UPSTREAM-LABEL]. The label bindings are advertised via BGP updates
   originated the ingress PEs.

   This procedure requires each egress PE to support a separate label
   space for every other PE. The egress PEs create a forwarding entry
   for the upstream assigned MPLS label, allocated by the ingress PE, in
   this label space. Hence when the egress PE receives a packet over an
   Aggregate Tree, it first determines the tree that the packet was
   received over. The tree identifier determines the label space in
   which the upstream assigned MPLS label lookup has to be performed.
   The same label space may be used for all P-multicast trees rooted at
   the same ingress PE, or an implementation may decide to use a
   separate label space for every P-multicast tree.




Rosen & Raggarwa                                               [Page 33]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   The encapsulation format is either MPLS or MPLS-in-something (e.g.
   MPLS-in-GRE [MPLS-IP]). When MPLS is used, this label will appear
   immediately below the label that identifies the P-multicast tree.
   When MPLS-in-GRE is used, this label will be the top MPLS label that
   appears when the GRE header is stripped off.

   When IP encapsulation is used for the P-multicast Tree, whatever
   information that particular encapsulation format uses for identifying
   a particular tunnel is used to determine the label space in which the
   MPLS label is looked up.

   If the P-multicast tree uses MPLS encapsulation, the P-multicast tree
   is itself identified by an MPLS label.  The egress PE MUST NOT
   advertise IMPLICIT NULL or EXPLICIT NULL for that tree.  Once the
   label representing the tree is popped off the MPLS label stack, the
   next label is the demultiplexing information that allows the proper
   MVPN to be determined.

   This specification requires that, to support this sort of
   aggregation, there be at least one upstream-assigned label per MVPN.
   It does not require that there be only one.  For example, an ingress
   PE could assign a unique label to each C-(S,G).  (This could be done
   using the same technique this is used to assign a particular C-(S,G)
   to an S-PMSI, see section 7.3.)


6.4. Mapping Received Packets to MVPNs

   When an egress PE receives a C-multicast data packet over a P-
   multicast tree, it needs to forward the packet to the CEs that have
   receivers in the packet's C-multicast group. It also needs to
   determine the RPF interface for the C-multicast data packet. In order
   to do this the egress PE needs to determine the tunnel that the
   packet was received on. The PE can then determine the MVPN that the
   packet belongs to and if needed do any further lookups that are
   needed to forward the packet.


6.4.1. Unicast Tunnels

   When ingress replication is used, the MVPN to which the received C-
   multicast data packet belongs can be determined by the MPLS label
   that was allocated by the egress. This label is distributed by the
   egress.  This also determines the RPF interface for the C-multicast
   data packet.






Rosen & Raggarwa                                               [Page 34]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


6.4.2. Non-Aggregated P-Multicast Trees

   If a P-multicast tree is associated with only one MVPN, determining
   the P-multicast tree on which a packet was received is sufficient to
   determine the packet's MVPN. All that the egress PE needs to know is
   the MVPN the P-multicast tree is associated with.

   There are different ways in which the egress PE can learn this
   association:

      a) Configuration. The P-multicast tree that a particular MVPN
         belongs to is configured on each PE.

         [Editor's Note: PIM-SM Default MD trees in [ROSEN-8] and
         [MVPN-BASE] are examples of configuring the P-multicast tree
         and MVPN association]

      b) BGP based advertisement of the P-multicast tree - MPVN mapping
         after the root of the tree discovers the leaves of the tree.
         The root of the tree sets up the tree after discovering each of
         the PEs that belong to the MVPN.  It then advertises the P-
         multicast tree - MVPN mapping to each of the leaves.  This
         mechanism can be used with both source initiated trees [e.g.
         RSVP-TE P2MP LSPs] and receiver initiated trees [e.g. PIM
         trees].

         [Editor's Note: Aggregate tree advertisements in [RAGGARWA-
         MCAST] are examples of this.]

      c) BGP based advertisment of the P-multicast tree - MVPN mapping
         as part of the MVPN membership discovery. The root of the tree
         advertises, to each of the other PEs that belong to the MVPN,
         the P-multicast tree that the MVPN is associated with. This
         implies that the root doesn't need to know the leaves of the
         tree beforehand. This is possible only for receiver initiated
         trees e.g. PIM based trees.

         [Editor's Note: PIM-SSM discovery in [ROSEN-8] is an example of
         the above]

   Both of the above require the BGP based advertisement to contain the
   P-multicast tree identifier. This identifier is encoded as a BGP
   attribute and contains the following elements:

     - Tunnel Type.






Rosen & Raggarwa                                               [Page 35]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


     - Tunnel identifier. The semantics of the identifier is determined
       by the tunnel type.



6.4.3. Aggregate P-Multicast Trees

   Once a PE sets up an Aggregate Tree it needs to announce the C-
   multicast groups being mapped to this tree to other PEs in the
   network. This procedure is referred to as Aggregate Tree discovery.
   For an Aggregate Tree with an inclusive mapping this discovery
   implies announcing:

     - The mapping of all MVPNs mapped to the Tree.

     - For each MVPN mapped onto the tree the inner label allocated for
       it by the ingress PE. The use of this label is explained in the
       demultiplexing procedures of section 6.3.4.

     - The P-multicast tree Identifier

   The egress PE creates a logical interface corresponding to the tree
   identifier. This interface is the RPF interface for all the <C-
   Source, C-Group> entries mapped to that tree.

   When PIM is used to setup P-multicast trees, the egress PE also Joins
   the P-Group Address corresponding to the tree. This results in setup
   of the PIM P-multicast tree.


6.5. I-PMSI Instantiation Using Ingress Replication

   As described in section 3 a PMSI can be instantiated using Unicast
   Tunnels between the PEs that are participating in the MVPN. In this
   mechanism the ingress PE replicates a C-multicast data packet
   belonging to a particular MVPN and sends a copy to all or a subset of
   the PEs that belong to the MVPN. A copy of the packet is tunnelled to
   a remote PE over an Unicast Tunnel to the remote PE. IP/GRE Tunnels
   or MPLS LSPs are examples of unicast tunnels that may be used. Note
   that the same Unicast Tunnel can be used to transport packets
   belonging to different MVPNs.

   Ingress replication can be used to instantiate a UI-PMSI. The PE sets
   up unicast tunnels to each of the remote PEs that support ingress
   replication. For a given MVPN all C-multicast data packets are sent
   to each of the remote PEs in the MVPN that support ingress
   replication. Hence a remote PE may receive C-multicast data packets
   for a group even if it doesn't have any receivers in that group.



Rosen & Raggarwa                                               [Page 36]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   Ingress replication can also be used to instantiate a MI-PMSI. In
   this case each PE has a mesh of unicast tunnels to every other PE in
   that MVPN.

   However when ingress replication is used it is recommended that only
   S-PMSIs be used. Instantiation of S-PMSIs with ingress replication is
   described in section 7.2.  Note that this requires the use of
   explicit tracking, i.e., a PE must know which of the other PEs have
   receivers for each C-multicast tree.


6.6. Establishing P-Multicast Trees

   It is believed that the architecture outlined in this document places
   no limitations on the protocols used to instantiate P-multicast
   trees. However, the only protocols being explicitly considered are
   PIM-SM, PIM-SSM, PIM-Bidir, RSVP-TE, and mLDP.

   A P-multicast tree can be either a source tree or a shared tree. A
   source tree is used to carry traffic only for the multicast VRFs that
   exist locally on the root of the tree i.e. for which the root has
   local CEs. The root is a PE router. Source P-multicast trees can be
   instantiated using PIM-SM, PIM-SSM, RSVP-TE P2MP LSPs, and mLDP P2MP
   LSPs.

   A shared tree on the other hand can be used to carry traffic
   belonging to VRFs that exist on other PEs as well. The root of a
   shared tree is not necessarily one of the PEs in the MVPN. All PEs
   that use the shared tree will send MVPN data packets to the root of
   the shared tree; if PIM is being used as the control protocol, PIM
   control packets also get sent to the root of the shared tree.  This
   may require an unicast tunnel between each of these PEs and the root.
   The root will then send them on the shared tree and all the PEs that
   are leaves of the shared tree will receive the packets. For example a
   RP based PIM-SM tree would be a shared tree. Shared trees can be
   instantiated using PIM-SM, PIM-SSM, PIM-Bidir, RSVP-TE P2MP LSPs,
   mLDP P2MP LSPs, and mLDP MP2MP LSPs.. Aggregation support for
   bidirectional P-trees (i.e., PIM-Bidir trees or mLDP MP2MP trees) is
   for further study. Shared trees require all the PEs to discover the
   root of the shared tree for a MVPN. To achieve this the root of a
   shared tree advertises as part of the BGP based MVPN membership
   discovery:

     - The capability to setup a shared tree for a specified MVPN.







Rosen & Raggarwa                                               [Page 37]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


     - A downstream assigned label that is to be used by each PE to
       encapsulate a MVPN data packet, when they send this packet to the
       root of the shared tree.

     - A downstream assigned label that is to be used by each PE to
       encapsulate a MVPN control packet, when they send this packet to
       the root of the shared tree.


   Both a source tree and a shared tree can be used to instantiate an
   I-PMSI.  If a source tree is used to instantiate an UI-PMSI for a
   MVPN, all the other PEs that belong to the MVPN, must be leaves of
   the source tree. If a shared tree is used to instantiate a UI-PMSI
   for a MVPN, all the PEs that are members of the MVPN must be leaves
   of the shared tree.


6.7. RSVP-TE P2MP LSPs

   This section describes procedures that are specific to the usage of
   RSVP-TE P2MP LSPs for instantiating a UI-PMSI. The RSVP-TE P2MP LSP
   can be either a source tree or a shared tree. Procedures in [RSVP-
   P2MP] are used to signal the LSP. The LSP is signalled after the root
   of the LSP discovers the leaves. The egress PEs are discovered using
   the MVPN membership procedures described in section 4. RSVP-TE P2MP
   LSPs can optionally support aggregation.


6.7.1. P2MP TE LSP Tunnel - MVPN Mapping

   P2MP TE LSP Tunnel to MVPN mapping can be learned at the egress PEs
   using either option (a) or option (b) described in section 6.4.2.
   Option (b) i.e. BGP based advertisements of the P2MP TE LSP Tunnel -
   MPVN mapping require that the root of the tree include the P2MP TE
   LSP Tunnel identifier as the tunnel identifier in the BGP
   advertisements. This identifier contains the following information
   elements:
     - The type of the tunnel is set to RSVP-TE P2MP Tunnel

     - RSVP-TE P2MP Tunnel's SESSION Object

     - Optionally RSVP-TE P2MP LSP's SENDER_TEMPLATE Object. This object
       is included when it is desired to identify a particular P2MP TE
       LSP.







Rosen & Raggarwa                                               [Page 38]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


6.7.2. Demultiplexing C-Multicast Data Packets

   Demultiplexing the C-multicast data packets at the egress PE follow
   procedures described in section 6.3.4. The RSVP-TE P2MP LSP Tunnel
   must be signaled with penultimate-hop-popping (PHP) off. Signalling
   the P2MP TE LSP Tunnel with PHP off requires an extension to RSVP-TE
   which will be described later.


7. Optimizing Multicast Distribution via S-PMSIs

   Whenever a particular multicast stream is being sent on an I-PMSI, it
   is likely that the data of that stream is being sent to PEs that do
   not require it.  If a particular stream has a significant amount of
   traffic,  it may be beneficial to move it to an S-PMSI which includes
   only those PEs that are transmitters and/or receivers (or at least
   includes fewer PEs that are neither).

   If explicit tracking is being done, S-PMSI creation can also be
   triggered on other criteria.  For instance there could be a "pseudo
   wasted bandwidth" criteria: switching to an S-PMSI would be done if
   the bandwidth multiplied by the number of uninterested PEs (PE that
   are receiving the stream but have no receivers) is above a specified
   threshold. The motivation is that (a) the total bandwidth wasted by
   many sparsely subscribed low-bandwidth groups may be large, and (b)
   there's no point to moving a high-bandwidth group to an S-PMSI if all
   the PEs have receivers for it.

   Switching to a S-PMSI requires the root of the S-PMSI to:

   a) Discover the egress PEs that will receive traffic using the S-
   PMSI, if this is required to setup the S-PMSI. There are two cases in
   which the PE needs to know the egress PEs:

     -  If the tunnel is a source initiated tree, such as a RSVP-TE P2MP
       Tunnel, the PE needs to know the leaves of the tree before it can
       instantiate the S-PMSI.

     -  If a PE instantiates multiple S-PMSIs, belonging to different
       MVPNs, using one P-multicast tree, such a tree is termed an
       Aggregate Tree with a selective mapping. The setting up of such
       an Aggregate Tree requires the ingress PE to know all the other
       PEs that have receivers for multicast groups that are mapped onto
       the tree.

   Discovering the egress PEs that will receive traffic using the S-PMSI
   requires doing explicit tracking for the particular (C-S, C-G) if the
   root hasn't already been doing it.



Rosen & Raggarwa                                               [Page 39]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   b) Setup the S-PMSI and

   c) If required, signal the binding of the multicast stream to the S-
   PMSI to the leaves of the tunnel used to instantiate the S-PMSI.

   Step (c) is required only when the tunnel is a P-multicast tree.  We
   specify two methods for achieving this.


7.1. S-PMSI Instantiation Using Ingress Replication

   As described in section 6.1.1, ingress replication can be used to
   instantiate a UI-PMSI. However this can result in a PE receiving
   packets for a multicast group for which it doesn't have any
   receivers. This can be avoided if the ingress PE tracks the remote
   PEs which have receivers in a particular C-multicast group.  In order
   to do this it needs to receive C-Joins from each of the remote PEs.
   It then replicates the C-multicast data packet and sends it to only
   those egress PEs which are on the path to a receiver of that C-group.
   It is possible that each PE that is using ingress replication
   instantiates only S-PMSIs. It is also possible that some PEs
   instantiate UI-PMSIs while others instantiate only S-PMSIs. In both
   these cases the PE MUST either unicast MVPN routing information using
   PIM or use BGP for exchanging the MVPN routing information. This is
   because there may be no MI-PMSI available for it to exchange MVPN
   routing information.

   Note that the use of ingress replication doesn't require any extra
   procedures for signaling the binding of the S-PMSI from the ingress
   PE to the egress PEs.  The procedures described for I-PMSIs are
   sufficient.


7.2. Protocol for Switching to S-PMSIs

   We describe two protocols for switching to S-PMSIs.  These protocols
   can be used when the tunnel that instantiates the S-PMSI is a P-
   multicast tree.


7.2.1. A UDP-based Protocol for Switching to S-PMSIs

   This procedure can be used for any MVPN which has an MI-PMSI.
   Traffic from all multicast streams in a given MPVN is sent, by
   default, on the MI-PMSI.  Consider a single multicast stream within a
   given MVPN, and consider a PE which is attached to a source of
   multicast traffic for that stream.  The PE can be configured to move
   the stream from the MI-PMSI to an S-PMSI if certain configurable



Rosen & Raggarwa                                               [Page 40]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   conditions are met.  To do this, it needs to inform all the PEs which
   attach to receivers for stream.  These PEs need to start listening
   for traffic on the S-PMSI, and the transmitting PE may start sending
   traffic on the S-PMSI when it is reasonably certain that all
   receiving PEs are listening on the S-PMSI.


7.2.1.1. Binding a Stream to an S-PMSI

   When a PE which attaches to a transmitter for a particular multicast
   stream notices that the conditions for moving the stream to an S-PMSI
   are met, it begins to periodically send an "S-PMSI Join Message" on
   the MI-PMSI.  The S-PMSI Join is a UDP-encapsulated message whose
   destination address is ALL-PIM-ROUTERS (224.0.0.13), and whose
   destination port is 3232.

   The S-PMSI Join Message contains the following information:

     - An identifier for the particular multicast stream which is to be
       bound to the S-PMSI.   This can be represented as an (S,G) pair.

     - An identifier for the particular S-PMSI to which the stream is to
       be bound.  This identifier is a structured field which includes
       the following information:

         * The type of tunnel used to instantiate the S-PMSI

         * An identifier for the tunnel.  The form of the identifier
           will depend upon the tunnel type.  The combination of tunnel
           identifier and tunnel type should contain enough information
           to enable all the PEs to "join" the tunnel and receive
           messages from it.

         * Any demultiplexing information needed by the tunnel
           encapsulation protocol to identify the particular S-PMSI.
           This allows a single tunnel to aggregate multiple S-PMSIs.
           If a particular tunnel is not aggregating multiple S-PMSIs,
           then no demultiplexing information is needed.

   A PE router which is not connected to a receiver will still receive
   the S-PMSI Joins, and MAY cache the information contained therein.
   Then if the PE later finds that it is attached to a receiver, it can
   immediately start listening to the S-PMSI.

   Upon receiving the S-PMSI Join, PE routers connected to receivers for
   the specified stream will take whatever action is necessary to start
   receiving multicast data packets on the S-PMSI.  The precise action
   taken will depend upon the tunnel type.



Rosen & Raggarwa                                               [Page 41]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   After a configurable delay, the PE router which is sending the S-PMSI
   Joins will start transmitting the stream's data packets on the S-
   PMSI.

   When the pre-configured conditions are no longer met for a particular
   stream, e.g. the traffic stops, the PE router connected to the source
   stops announcing S-PMSI Joins for that stream.  Any PE that does not
   receive, over a configurable interval, an S-PMSI Join for a
   particular stream will stop listening to the S-PMSI.


7.2.1.2. Packet Formats and Constants

   To be included.


7.2.2. A BGP-based Protocol for Switching to S-PMSIs

   This procedure can be used for a MVPN that is using either a UI-PMSI
   or a MI-PMSI. Consider a single multicast stream for a C-(S, G)
   within a given MVPN, and consider a PE which is attached to a source
   of multicast traffic for that stream. The PE can be configured to
   move the stream from the MI-PMSI or UI-PMSI to an S-PMSI if certain
   configurable conditions are met. Once a PE decides to move the C-(S,
   G) for a given MVPN to a S-PMSI, it needs to instantiate the S-PMSI
   using a tunnel and announce to all the egress PEs, that are on the
   path to receivers of the C-(S, G), of the binding of the S-PMSI to
   the C-(S, G). The announcement is done using BGP.  Depending on the
   tunneling technology used, this announcement may be done before or
   after setting up the tunnel. The source and egress PEs have to switch
   to using the S-PMSI for the C-(S, G).


7.2.2.1. Advertising C-(S, G) Binding to a S-PMSI using BGP

   The ingress PE informs all the PEs that are on the path to receivers
   of the C-(S, G) of the binding of the S-PMSI to the C-(S, G). The BGP
   announcement is done by sending update for the MCAST-VPN address
   family.  An A-D route is used, containing the following information:

      a) IP address of the originating PE

      b) The RD configured locally for the MVPN. This is required to
         uniquely identify the <C-Source, C-Group> as the addresses
         could overlap between different MVPNs.  This is the same RD
         value used in the auto-discovery process.





Rosen & Raggarwa                                               [Page 42]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


      c) The C-Source address. This address can be a prefix in order to
         allow a range of C-Source addresses to be mapped to an
         Aggregate Tree.

      d) The C-Group address. This address can be a range in order to
         allow a range of C-Group addresses to be mapped to an Aggregate
         Tree.

      e) If the S-PMSI is instantiated using an Aggregate Tree with a
         selective mapping, an inner label allocated by the Aggregate
         Tree root for the <C-Source, C-Group>. This allows a single
         tunnel to aggregate multiple S-PMSIs. This is the upstream
         label whose usage is described in section 6.3.4.

   When a PE distributes this information via BGP, it must include the
   following:

      1. An identifier for the particular S-PMSI to which the stream is
         to be bound.  This identifier is a structured field which
         includes the following information:

           * The type of tunnel used to instantiate the S-PMSI

           * An identifier for the tunnel.  The form of the identifier
             will depend upon the tunnel type.  The combination of
             tunnel identifier and tunnel type should contain enough
             information to enable all the PEs to "join" the tunnel and
             receive messages from it.

      2. Route Target Extended Communities attribute. This is used as
         described in section 4.


7.2.2.2. Explicit Tracking

   If the PE wants to enable explicit tracking for the specified flow,
   it also indicates this in the A-D route it uses to bind the flow to a
   particular S-PMSI.  Then any PE which receives the A-D route will
   respond with a "Leaf A-D Route" in which it identifies itself as a
   receiver of the specified flow.  The Leaf A-D route will be withdrawn
   when the PE is no longer a receiver for the flow.

   If the PE needs to enable explicit tracking for a flow before binding
   the flow to an S-PMSI, it can do so by sending an A-D route
   identifying the flow but not specifying an S-PMSI.  This will elicit
   the Leaf A-D Routes.  This is useful when the PE needs to know the
   receivers before selecting an S-PMSI.




Rosen & Raggarwa                                               [Page 43]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


7.2.2.3. Switching to S-PMSI

   After the egress PEs receive the announcement they setup their
   forwarding path to receive traffic on the S-PMSI if they have one or
   more receivers interested in the <C-S, C-G> bound to the S-PMSI. This
   involves changing the RPF interface for the relevant <C-S, C-G>
   entries to the interface that is used to instantiate the S-PMSI. If
   an Aggregate Tree is used to instantiate a S-PMSI this also implies
   setting up the demultiplexing forwarding entries based on the inner
   label as described in section 6.3.4.  The egress PEs may perform the
   switch to the S-PMSI once the advertisement from the ingress PE is
   received or wait for a preconfigured timer to do so.

   A source PE may use one of two approaches to decide when to start
   transmitting data on the S-PMSI. In the first approach once the
   source PE instantiates the S-PMSI, it starts sending multicast
   packets for <C-S, C-G> entries mapped to the S-PMSI on both that as
   well as on the I-PMSI, which is currently used to send traffic for
   the <C-S, C-G>. After some preconfigured timer the PE stops sending
   multicast packets for <C-S, C-G> on the I-PMSI. In the second
   approach after a certain pre-configured delay after advertising the
   <C-S, C-G> entry bound to a S-PMSI,  the source PE begins to send
   traffic on the S-PMSI. At this point it stops to send traffic for the
   <C-S, C-G> on the I-PMSI. This traffic is instead transmitted on the
   S-PMSI.


7.3. Aggregation

   S-PMSIs can be aggregated on a P-multicast tree. The S-PMSI to C-(S,
   G) binding advertisement supports aggregation. Furthermore the
   aggregation procedures of section 6.3 apply. It is also possible to
   aggregate both S-PMSIs and I-PMSIs on the same P-multicast tree.


7.4. Instantiating the S-PMSI with a PIM Tree

   The procedures of section 7.3 tell a PE when it must start listening
   and stop listening to a particular S-PMSI.  Those procedures also
   specify the method for instantiating the S-PMSI.  In this section, we
   provide the procedures to be used when the S-PMSI is instantiated as
   a PIM tree.  The PIM tree is created by the PIM P-instance.

   If a single PIM tree is being used to aggregate multiple S-PMSIs,
   then the PIM tree to which a given stream is bound may have already
   been joined by a given receiving PE.  If the tree does not already
   exist, then the appropriate PIM procedures to create it must be
   executed in the P-instance.



Rosen & Raggarwa                                               [Page 44]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   If the S-PMSI for a particular multicast stream is instantiated as a
   PIM-SM or PIM-Bidir tree, the S-PMSI identifier will specify the RP
   and the group P-address, and the PE routers which have receivers for
   that stream must build a shared tree toward the RP.

   If the S-PMSI is instantiated as a PIM-SSM tree, the PE routers build
   a source tree toward the PE router that is advertising the S-PMSI
   Join.  The IP address root of the tree is the same as the source IP
   address which appears in the S-PMSI Join.  In this case, the tunnel
   identifier in the S-PMSI Join will only need to specify a group P-
   address.

   The above procedures assume that each PE router has a set of group
   P-addresses that it can use for setting up the PIM-trees.  Each PE
   must be configured with this set of P-addresses.  If PIM-SSM is used
   to set up the tunnels, then the PEs may be with overlapping sets of
   group P-addresses.  If PIM-SSM is not used, then each PE must be
   configured with a unique set of group P-addresses (i.e., having no
   overlap with the set configured at any other PE router).  The
   management of this set of addresses is thus greatly simplified when
   PIM-SSM is used, so the use of PIM-SSM is strongly recommended
   whenever PIM trees are used to instantiate S-PMSIs.

   If it is known that all the PEs which need to receive data traffic on
   a given S-PMSI can support aggregation of multiple  S-PMSIs on a
   single PIM tree, then the transmitting PE, may, at its discretion,
   decide to bind the S-PMSI to a PIM  tree which is already bound to
   one or more other S-PMSIs, from the same or from different MVPNs.  In
   this case, appropriate demultiplexing information must be signaled.


7.5. Instantiating S-PMSIs using RSVP-TE P2MP Tunnels

   RSVP-TE P2MP Tunnels can be used for instantiating S-PMSIs.
   Procedures described in the context of I-PMSIs in section 6.7 apply.


8. Inter-AS Procedures

   If an MVPN has sites in more than one AS, it requires one or more
   PMSIs to be instantiated by inter-AS tunnels.  This document
   describes two different types of inter-AS tunnel:

      1. "Segmented Inter-AS tunnels"

         A segmented inter-AS tunnel consists of a number of independent
         segments which are stitched together at the ASBRs.  There are
         two types of segment, inter-AS segments and intra-AS segments.



Rosen & Raggarwa                                               [Page 45]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


         The segmented inter-AS tunnel consists of alternating intra-AS
         and inter-AS segments.

         Inter-AS segments connect adjacent ASBRs of different ASes;
         these "one-hop" segments are instantiated as unicast tunnels.

         Intra-AS segments connect ASBRs and PEs which are in the same
         AS.  An intra-AS segment may be of whatever technology is
         desired by the SP that administers the that AS.  Different
         intra-AS segments may be of different technologies.

         Note that an intra-AS segment of an inter-AS tunnel is distinct
         from any intra-AS tunnel in the AS.

         A segmented inter-AS tunnel can be thought of as a tree which
         is rooted at a particular AS, and which has as its leaves the
         other ASes which need to receive multicast data from the root
         AS.

      2. "Non-segmented Inter-AS tunnels"

         A non-segmented inter-AS tunnel is a single tunnel which spans
         AS boundaries.  The tunnel technology cannot change from one
         point in the tunnel to the next, so all ASes through which the
         tunnel passes must support that technology.  In essence, AS
         boundaries are of no significance to a non-segmented inter-AS
         tunnel.

         [Editor's Note: This is the model in [ROSEN-8] and [MVPN-
         BASE].]

   Section 10 of [RFC4364] describes three different options for
   supporting unicast Inter-AS BGP/MPLS IP VPNs, known as options A, B,
   and C.  We describe below how both segmented and non-segmented
   inter-AS trees can be supported when option B or option C is used.
   (Option A does not pass any routing information through an ASBR at
   all, so no special inter-AS procedures are needed.)


8.1. Non-Segmented Inter-AS Tunnels

   In this model, the previously described discovery and tunnel setup
   mechanisms are used, even though the PEs belonging to a given MVPN
   may be in different ASes.  The ASBRs play no special role, but
   function merely as P routers.






Rosen & Raggarwa                                               [Page 46]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


8.1.1. Inter-AS MVPN Auto-Discovery

   The previously described BGP-based auto-discovery mechanisms work "as
   is" when an MVPN contains PEs that are in different Autonomous
   Systems.


8.1.2. Inter-AS MVPN Routing Information Exchange

   MVPN routing information exchange can be done by PIM peering (either
   lightweight or full) across an MI-PMSI, or by unicasting PIM
   messages.  The method of using BGP to send MVPN routing information
   can also be used.

   If any form of PIM peering is used, a PE that sends C-PIM Join/Prune
   messages for a particular C-(S,G) must be able to identify the PE
   which is its PIM adjacency on the path to S.  The identity of the PIM
   adjacency is determined from the RPF information associated with the
   VPN-IP route to S.

   If no RPF information is present, then the identity of the PIM
   adjacency is taken from the BGP Next Hop attribute of the VPN-IP
   route to S.  Note that this will not give the correct result if
   option b of section 10 of [RFC4364] is used.  To avoid this
   possibility of error, the RPF information SHOULD always be present if
   MVPN routing information is to be distributed by PIM.

   If BGP (rather than PIM) is used to distribute the MVPN routing
   information, and if option b of section 10 of [RFC4364] is in use,
   then the MVPN routes will be installed in the ASBRs along the path
   from each multicast source in the MVPN to each multicast receiver in
   the MVPN.  If option b is not in use, the MVPN routes are not
   installed in the ASBRs.  The handling of MVPN routes in either case
   is thus exactly analogous to the handling of unicast VPN-IP routes in
   the corresponding case.



8.1.3. Inter-AS I-PMSI

   The procedures described earlier in this document can be used to
   instantiate an I-PMSI with inter-AS tunnels. Specific tunneling
   techniques require some explanation:

      1. If ingress replication is used, the inter-AS PE-PE tunnels will
         use the inter-AS tunneling procedures for the tunneling
         technology used.




Rosen & Raggarwa                                               [Page 47]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


      2. Inter-AS PIM-SM or PIM-SSM based trees rely on a PE joining a
         (P-S, P-G) tuple where P-S is the address of a PE in another
         AS. This (P-S, P-G) tuple is learned using the MVPN membership
         and BGP MVPN-tunnel binding procedures described earlier.
         However, if the source of the tree is in a different AS than a
         particular P router, it is possible that the P router will not
         have a route to the source.  For example, the remote AS may be
         using BGP to distribute a route to the source, but a particular
         P router may be part of a "BGP-free core", in which the P
         routers are not aware of BGP-distributed routes.

         In such a case it is necessary for a PE to to tell PIM to
         construct the tree through a particular BGP speaker, the "BGP
         next hop" for the tree source.  This can be accomplished with a
         PIM extension, in which the P-PIM Join/Prune messages carry a
         new "proxy" field which contains the address of that BGP next
         hop.  As the P-multicast tree is constructed, it is built
         towards the proxy (the BGP next hop) rather than towards P-S,
         so the P routers will not need to have a route to P-S.

         Support for inter-AS trees using PIM-Bidir are for further
         study.

         When the BGP-based discovery procedures for MVPN are in place,
         one can distinguish two different inter-AS routes to a
         particular P-S:

           - BGP will install a unicast route to P-S along a particular
             path, using the IP AFI/SAFI ;

           - A PE's MVPN auto-discovery information is advertised by
             sending a BGP update whose  NLRI  is in a special address
             family (AFI/SAFI) used for this purpose.  The  NLRI of the
             address family contains the  IP address of the PE, as well
             as an RD.  If the NLRI contains the IP address of P-S, this
             in effect creates a second route to P-S.  This route might
             follow a different path than the route in the unicast IP
             family.

         When building a PIM tree towards P-S, it may be desirable to
         build it along the route on which the MVPN auto-discovery
         AFI/SAFI is installed, rather than along the route on which the
         IP AFI/SAFI is installed.  This enables the inter-AS portion of
         the tree to follow a path which is specifically chosen for
         multicast (i.e., it allows the inter-AS multicast topology to
         be "non-congruent" to the inter-AS unicast topology).

         In order for P routers to send P-Join/Prune messages along this



Rosen & Raggarwa                                               [Page 48]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


         path, they need to make use of the "proxy" field extension
         discussed above.  The PIM message must also contain the full
         NLRI in the MVPN auto-discovery family, so that the BGP
         speakers can look up that NLRI to find the BGP next hop.

      3. Procedures in [RSVP-P2MP] are used for inter-AS RSVP-TE P2MP
         Tunnels.


8.1.4. Inter-AS S-PMSI

   The leaves of the tunnel are discovered using the MVPN routing
   information.  Procedures for setting up the tunnel are similar to the
   ones described in section 8.2.3 for an inter-AS I-PMSI.


8.2. Segmented Inter-AS Tunnels

8.2.1. Inter-AS MVPN Auto-Discovery Tunnels

   The BGP based MVPN membership discovery procedures of section 4 are
   used to auto-discover the intra-AS MVPN membership. This section
   describes the additional procedures for inter-AS MVPN membership
   discovery. It also describes the procedures for contructing segmented
   inter-AS tunnels.

   In this case, for a given MVPN in an AS, the objective is to form a
   spanning tree of MVPN membership, rooted at the AS. The nodes of this
   tree are ASes.  The leaves of this tree are only those ASes that have
   at least one PE with a member in the MVPN. The inter-AS tunnel used
   to instantiate an inter-AS PMSI must traverse this spanning tree. A
   given AS needs to announce to another AS only the fact that it has
   membership in a given MVPN. It doesn't need to announce the
   membership of each PE in the AS to other ASes.

   This section defines an inter-AS auto-discovery route as a route that
   carries information about an AS that has one or more PEs (directly)
   connected to the site(s) of that MVPN. Further it defines an inter-AS
   leaf auto-discovery route (leaf auto-discovery route) as a route used
   to inform the root of an intra-AS segment, of an inter-AS tunnel, of
   a leaf of that intra-AS segment.


8.2.1.1. Originating Inter-AS MVPN A-D Information

   A PE in a given AS advertises its MVPN membership to all its IBGP
   peers.  This IBGP peer may be a route reflector which in turn
   advertises this information to only its IBGP peers. In this manner



Rosen & Raggarwa                                               [Page 49]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   all the PEs and ASBRs in the AS learn this membership information.

   An Autonomous System Border Router (ASBR) may be configured to
   support a particular MVPN. If an ASBR is configured to support a
   particular MVPN, the ASBR MUST participate in the intra-AS MVPN
   auto-discovery/binding procedures for that MVPN within the AS that
   the ASBR belongs to, as defined in this document.

   Each ASBR then advertises the "AS MVPN membership" to its neighbor
   ASBRs using EBGP. This inter-AS auto-discovery route must not be
   advertised to the PEs/ASBRs in the same AS as this ASBR. The
   advertisement carries the following information elements:

      a. A Route Distinguisher for the MVPN. For a givem MVPN each ASBR
         in the AS must use the same RD when advertising this
         information to other ASBRs. To accomplish this all the ASBRs
         within that AS, that are configured to support the MVPN, MUST
         be configured with the same RD for that MVPN. This RD MUST be
         of Type 0, MUST embed the autonomous system number of the AS.

      b. The announcing ASBR's local address as the next-hop for the
         above information elements.

      c. By default the BGP Update message MUST carry export Route
         Targets used by the unicast routing of that VPN. The default
         could be modified via configuration by having a set of Route
         Targets used for the inter-AS auto-discovery routes being
         distinct from the ones used by the unicast routing of that VPN.



8.2.1.2. Propagating Inter-AS MVPN A-D Information

   As an inter-AS auto-discovery route originated by an ASBR within a
   given AS is propagated via BGP to other ASes, this results in
   creation of a data plane tunnel that spans multiple ASes. This tunnel
   is used to carry (multicast) traffic from the MVPN sites connected to
   the PEs of the AS to the MVPN sites connected to the PEs that are in
   the other ASes. Such tunnel consists of multiple intra-AS segments
   (one per AS) stitched at ASBRs' boundaries by single hop <ASBR-ASBR>
   LSP segments.

   An ASBR originates creation of an intra-AS segment when the ASBR
   receives an inter-AS auto-discovery route from an EBGP neighbor.
   Creation of the segment is completed as a result of distributing via
   IBGP this route within the ASBR's own AS.

   For a given inter-AS tunnel each of its intra-AS segments could be



Rosen & Raggarwa                                               [Page 50]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   constructed by its own independent mechanism. Moreover, by using
   upstream labels within a given AS multiple intra-AS segments of
   different inter-AS tunnels of either the same or different MVPNs may
   share the same P-Multicast Tree.

   Since (aggregated) inter-AS auto-discovery routes have granularity of
   <AS, MVPN>, an MVPN that is present in N ASes would have total of N
   inter-AS tunnels. Thus for a given MVPN the number of inter-AS
   tunnels is independent of the number of PEs that have this MVPN.

   The following sections specify procedures for propagation of
   (aggregated) inter-AS auto-discovery routes across ASes.


8.2.1.2.1. Inter-AS Auto-Discovery Route received via EBGP

   When an ASBR receives from one of its EBGP neighbors a BGP Update
   message that carries the inter-AS auto-discovery route if (a) at
   least one of the Route Targets carried in the message matches one of
   the import Route Targets configured on the ASBR, and (b) the ASBR
   determines that the received route is the best route to the
   destination carried in the NLRI of the route, the ASBR:


      a) Re-advertises this inter-AS auto-discovery route within its own
         AS.

         If the ASBR uses ingress replication to instantiate the intra-
         AS segment of the inter-AS tunnel, the re-advertised route
         SHOULD carry a Tunnel attribute with the Tunnel Identifier set
         to Ingress Replication, but no MPLS labels.

         If a P-Multicast Tree is used to instantiate the intra-AS
         segment of the inter-AS tunnel, and in order to advertise the
         P-Multicast tree identifier the ASBR doesn't need to know the
         leaves of the tree beforehand, then the advertising ASBR SHOULD
         advertise the P-Multicast tree identifier in the Tunnel
         Identifier of the Tunnel attribute. This, in effect, creates a
         binding between the inter-AS auto-discovery route and the P-
         Multicast Tree.

         If a P-Multicast Tree is used to instantiate the intra-AS
         segment of the inter-AS tunnel, and in order to advertise the
         P-Multicast tree identifier the advertising ASBR needs to know
         the leaves of the tree beforehand, the ASBR first discovers the
         leaves using the Auto-Discovery procedures, as specified
         further down. It then advertises the binding of the tree to the
         inter-AS auto-discovery route using the the original auto-



Rosen & Raggarwa                                               [Page 51]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


         discovery route with the addition of carrying in the route the
         Tunnel attribute that contains the type and the identity of the
         tree (encoded in the Tunnel Identifier of the attribute).

      b) Re-advertises the received inter-AS auto-discovery route to its
         EBGP peers, other than the EBGP neighbor from which the best
         inter-AS auto-discovery route was received.

      c) Advertises to its neighbor ASBR, from which it received the
         best inter-AS autodiscovery route to the destination carried in
         the NRLI of the route, a leaf auto-discovery route that carries
         an ASBR-ASBR tunnel binding with the tunnel identifier set to
         ingress replication. This binding as described in section 6 can
         be used by the neighbor ASBR to send traffic to this ASBR.



8.2.1.2.2. Leaf Auto-Discovery Route received via EBGP

   When an ASBR receives via EBGP a leaf auto-discovery route, the ASBR
   finds an inter-AS auto-discovery route that has the same RD as the
   leaf auto-discovery route. The MPLS label carried in the leaf auto-
   discovery route is used to stitch a one hop ASBR-ASBR LSP to the tail
   of the intra-AS tunnel segment associated with the inter-AS auto-
   discovery route.


8.2.1.2.3. Inter-AS Auto-Discovery Route received via IBGP

   If a given inter-AS auto-discovery route is advertised within an AS
   by multiple ASBRs of that AS, the BGP best route selection performed
   by other PE/ASBR routers within the AS does not require all these
   PE/ASBR routers to select the route advertised by the same ASBR - to
   the contrary different PE/ASBR routers may select routes advertised
   by different ASBRs.

   Further when a PE/ASBR receives from one of its IBGP neighbors a BGP
   Update message that carries a AS MVPN membership tree , if (a) the
   route was originated outside of the router's own AS, (b) at least one
   of the Route Targets carried in the message matches one of the import
   Route Targets configured on the PE/ASBR, and (c) the PE/ASBR
   determines that the received route is the best route to the
   destination carried in the NLRI of the route, if the router is an
   ASBR then the ASBR propagates the route to its EBGP neighbors. In
   addition the PE/ASBR performs the following.

   If the received inter-AS auto-discovery route carries the Tunnel
   attribute with the Tunnel Identifier set to LDP P2MP LSP, or PIM-SSM



Rosen & Raggarwa                                               [Page 52]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   tree, or PIM-SM tree, the PE/ASBR SHOULD join the P-Multicast tree
   whose identity is carried in the Tunnel Identifier.

   If the received source auto-discovery route carries the Tunnel
   attribute with the Tunnel Identifier set to RSVP-TE P2MP LSP, then
   the ASBR that originated the route MUST signal the local PE/ASBR as
   one of leaf LSRs of the RSVP-TE P2MP LSP. This signaling MAY have
   been completed before the local PE/ASBR receives the BGP Update
   message.

   If the NLRI of the route does not carry a label, then this tree is an
   intra-AS LSP segment that is part of the inter-AS Tunnel for the MVPN
   advertised by the inter-AS auto-discovery route. If the NLRI carries
   a (upstream) label, then a combination of this tree and the label
   identifies the intra-AS segment.

   If this is an ASBR, this intra-AS segment may further be stitched to
   ASBR-ASBR inter-AS segment of the inter-AS tunnel. If the PE/ASBR has
   local receivers in the MVPN, packets received over the intra-AS
   segment must be forwarded to the local receivers using the local VRF.

   If the received inter-AS auto-discovery route either does not carry
   the Tunnel attribute, or carries the Tunnel attribute with the Tunnel
   Identifier set to ingress replication, then the PE/ASBR originates a
   new auto-discovery route to allow the ASBR from which the auto-
   discovery route was received, to learn of this ASBR as a leaf of the
   intra-AS tree.

   Thus the AS MVPN membership information propagates across multiple
   ASes along a spanning tree. BGP AS-Path based loop prevention
   mechanism prevents loops from forming as this information propagates.


8.2.2. Inter-AS MVPN Routing Information Exchange

   All of the MVPN routing information exchange methods specified in
   section 5 can be supported across ASes.

   The objective in this case is to propagate the MVPN routing
   information to the remote PE that originates the unicast route to C-
   S/C-RP, in the reverse direction of the AS MVPN membership
   information announced by the remote PE's origin AS. This information
   is processed by each ASBR along this reverse path.

   To achieve this the PE that is generating the MVPN routing
   advertisement, first determines the source AS of the unicast route to
   C-S/C-RP. It then determines from the received AS MVPN membership
   information, for the source AS, the ASBR that is the next-hop for the



Rosen & Raggarwa                                               [Page 53]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   best path of the source AS MVPN membership. The BGP MVPN routing
   update is sent to this ASBR and the ASBR then further propagates the
   BGP advertisement. BGP filtering mechanisms ensure that the BGP MVPN
   routing information updates flow only to the upstream router on the
   reverse path of the inter-AS MVPN membership tree. Details of this
   filtering mechanism and the relevant encoding will be specified in a
   separate document.


8.2.3. Inter-AS I-PMSI

   All PEs in a given AS, use the same inter-AS heterogenous tunnel,
   rooted at the AS, to instantiate an I-PMSI for an inter-AS MVPN
   service. As explained earlier the intra-AS tunnel segments that
   comprise this tunnel can be built using different tunneling
   technologies. To instantiate an MI-PMSI service for a MVPN there must
   be an inter-AS tunnel rooted at each AS that has at least one PE that
   is a member of the MVPN.

   A C-multicast data packet is sent using an intra-AS tunnel segment by
   the PE that first receives this packet from the MVPN customer site.
   An ASBR forwards this packet to any locally connected MVPN receivers
   for the multicast stream. If this ASBR has received a tunnel binding
   for the AS MVPN membership that it advertised to a neighboring ASBR,
   it also forwards this packet to the neighboring ASBR. In this case
   the packet is encapsulated in the downstream MPLS label received from
   the neighboring ASBR. The neighboring ASBR delivers this packet to
   any locally connected MVPN receivers for that multicast stream. It
   also transports this packet on an intra-AS tunnel segment, for the
   inter-AS MVPN tunnel, and the other PEs and ASBRs in the AS then
   receive this packet.  The other ASBRs then repeat the procedure
   followed by the ASBR in the origin AS and the packet traverses the
   overlay inter-AS tunnel along a spanning tree.


8.2.3.1. Support for Unicast VPN Inter-AS Methods

   The above procedures for setting up an inter-AS I-PMSI can be
   supported for each of the unicast VPN inter-AS models described in
   [RFC4364]. These procedures do not depend on the method used to
   exchange unicast VPN routes. For Option B and Option C they do
   require MPLS encapsulation between the ASBRs.









Rosen & Raggarwa                                               [Page 54]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


8.2.4. Inter-AS S-PMSI

   An inter-AS tunnel for an S-PMSI is constructed similar to an inter-
   AS tunnel for an I-PMSI. Namely, such a tunnel is constructed as a
   concatenation of tunnel segments. There are two types of tunnel
   segments: an intra-AS tunnel segment (a segment that spans ASBRs
   within the same AS), and inter-AS tunnel segment (a segment that
   spans adjacent ASBRs in adjacent ASes). ASes that are spanned by a
   tunnel are not required to use the same tunneling mechanism to
   construct the tunnel - each AS may pick up a tunneling mechanism to
   construct the intra-AS tunnel segment of the tunnel on its

   The PE that decides to set up a S-PMSI, advertises the S-PMSI tunnel
   binding using procedures in section 7.3.2 to the routers in its own
   AS. The <C-S, C-G> membership for which the S-PMSI is instantiated,
   is propagated along an inter-AS spanning tree. This spanning tree
   traverses the same ASBRs as the AS MVPN membership spanning tree. In
   addition to the information elements described in section 7.3.2
   (Origin AS, RD, next-hop) the C-S and C-G is also advertised.

   An ASBR that receives the AS <C-S, C-G> information from its upstream
   ASBR using EBGP sends back a tunnel binding for AS <C-S, C-G>
   information if a) at least one of the Route Targets carried in the
   message matches one of the import Route Targets configured on the
   ASBR, and (b) the ASBR determines that the received route is the best
   route to the destination carried in the NLRI of the route. If the
   ASBR instantiates a S-PMSI for the AS <C-S, C-G> it sends back a
   downstream label that is used to forward the packet along its intra-
   AS S-PMSI for the <C-S, C-G>. However the ASBR may decide to use an
   AS MVPN membership I-PMSI instead, in which case it sends back the
   same label that it advertised for the AS MVPN membership I-PMSI. If
   the downstream ASBR instantiates a S-PMSI, it further propagates the
   <C-S, C-G> membership to its downstream ASes, else it does not.

   An AS can instantiate an intra-AS S-PMSI for the inter-AS S-PMSI
   tunnel only if the upstream AS instantiates a S-PMSI. The procedures
   allow each AS to determine whether it wishes to setup a S-PMSI or not
   and the AS is not forced to setup a S-PMSI just because the upstream
   AS decides to do so.

   The leaves of an intra-AS S-PMSI tunnel will be the PEs that have
   local receivers that are interested in <C-S, C-G> and the ASBRs that
   have received MVPN routing information for <C-S, C-G>. Note that an
   AS can determine these ASBRs as the MVPN routing information is
   propagated and processed by each ASBR on the AS MVPN membership
   spanning tree.

   The C-multicast data traffic is sent on the S-PMSI by the originating



Rosen & Raggarwa                                               [Page 55]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   PE.  When it reaches an ASBR that is on the spanning tree, it is
   delivered to local receivers, if any, and is also forwarded to the
   neighbor ASBR after being encapsulated in the label advertised by the
   neighbor. The neighbor ASBR either transports this packet on the S-
   PMSI for the multicast stream or an I-PMSI, delivering it to the
   ASBRs in its own AS. These ASBRs in turn repeat the procedures of the
   origin AS ASBRs and the multicast packet traverses the spanning tree.


9. Duplicate Packet Detection and Single Forwarder PE

   An egress PE may receive duplicate multicast data packets, from more
   than one ingress PE, for a MVPN when a a site that contains C-S or
   C-RP is multihomed to more than one PE. An egress PE may also receive
   duplicate data packets for a MVPN, from two different ingress PEs,
   when the CE-PE routing protocol is PIM-SM and a router or a CE in a
   site switches from the C-RP tree to C-S tree.

   For a given <C-S, C-G> a PE, say PE1, expects to receive C-data
   packets from the upstream PE, say PE2, which PE1 identified as the
   upstream multicast hop in the C-Multicast Routing Update that PE1
   sent in order to join <C-S, C-G>. If PE1 can determine that a data
   packet for <C-S, C-G> was received from the expected upstream PE,
   PE2, PE1 will accept the packet.  Otherwise, PE1 will drop the
   packet.  (But see section 10 for an exception case where PE1 will
   accept a packet even if it is from an unexpected upstream PE.) This
   determination can be performed only if the PMSI on which the packets
   are being received and the tunneling technology used to instantiate
   the PMSI allows the PE to determine the source PE that sent the
   packet. However this determination may not always be possible.

   Therefore, procedures are needed to ensure that packets are received
   at a PE only from a single upstream PE.  This is called single
   forwarder PE selection.

   Single forwarder PE selection is achieved by the following set of
   procedures:


      a. If there is more than one PE within the same AS through which
         C-S or C-RP of a given MVPN could be reached, and in the case
         of C-S not every such PE advertises an S-PMSI for <C-S, C-G>,
         all PEs that have this MVPN MUST send the MVPN routing
         information update for <C-S, C-G> or <C-*, C-G> to the same
         upstream PE.  This is achieved using the following procedure:

         If the next-hop interface on a PE's route to C-S/C-RP, that is
         installed in the VRF, is a VRF interface than the PE should use



Rosen & Raggarwa                                               [Page 56]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


         that route to reach C-S/C-RP. Else : from all the VPN-IP routes
         that could be imported into the VRF and have exactly the same
         IP prefix as the route in the VRF, the PE uses the one with the
         highest next-hop address to determine the upstream multicast
         hop to identify in the C-multicast route.

      b. The above procedure ensures that if C-S or C-RP is multi-homed
         to PEs within a single AS, a PE will not receive duplicate
         traffic as long as all the PEs in that AS are on either the C-S
         or C-RP tree.

         However the PE may receive duplicate traffic if C-S or C-RP is
         multi-homed to different ASes. In this case the PE can detect
         duplicate traffic as such duplicate traffic will arrive on a
         different tunnel - if the PE was expecting the traffic on an
         inter-AS tunnel, duplicate traffic will arrive on an intra-AS
         tunnel [this is not an intra-AS tunnel segment, of an inter-AS
         tunnel] and vice-versa.

         To achieve the above the PE has to keep track of which (inter-
         AS) auto-discovery route the PE uses for sending MVPN multicast
         routing information towards C-S/C-RP. Then the PE should
         receive (multicast) traffic originated by C-S/C-RP only from
         the (inter-AS) tunnel that was carried in the best source
         auto-discovery route for the MVPN and was originated by the AS
         that contains C-S/C-RP (where "the best" is determined by the
         PE). All other multicast traffic originated by C-S/C-RP, but
         received on any other tunnel should be discarded as duplicated.

         The PE may also receive duplicate traffic during a <C-*, C-G>
         to <C-S, C-G> switch. The issue and the solution are described
         next.

      c. If the tunneling technology in use for a particular MVPN does
         not allow the egress PEs to identify the ingress PE, then
         having all the PEs select the same PE to be the upstream
         multicast hop is not sufficient to prevent packet duplication.
         The reason is that a single tunnel may be carrying traffic on
         both the (C-*, C-G) tree and the (C-S, C-G) tree.  If some of
         the egress PEs have joined the source tree, but others expect
         to receive (S,G) packets from the shared tree, then two copies
         of data packet will travel on the tunnel, and the egress PEs
         will have no way to determine that only one copy should be
         accepted.

         To avoid this, it is necessary to ensure that once any PE joins
         the (C-S, C-G) tree, any other PE that has joined the (C-*, C-
         G) tree also switches to the (C-S, C-G) tree  (selecting, of



Rosen & Raggarwa                                               [Page 57]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


         course, the same upstream multicast hop, as specified above).

         Whenever a PE creates an <C-S,C-G> state as a result of
         receiving a C-multicast route for <C-S, C-G> from some other
         PE, and the C-G group is a Sparse Mode group, the PE that
         creates the state MUST originate an auto-discovery route as
         specified below. The route is being advertised using the same
         procedures as the MVPN auto-discovery/binding (both intra-AS
         and inter-AS) specified in this document with the following
         modifications:


            1. The Multicast Source field MUST be set to C-S.  The
               Multicast Source Length field is set appropriately to
               reflect this.

            2. The Multicast Group field MUST be set to C-G.  The
               Multicast Group Length field is set appropriately to
               reflect this.

         The route goes to all the PEs of the MVPN. When a PE receives
         this route, it checks whether there are any receivers in the
         MVPN sites attached to the PE for the group carried in the
         route. If yes, then it generates a C-multicast route indicating
         Join for <C-S, C-G>.  This forces all the PEs (in all ASes) to
         switch to the C-S tree for <C-S, C-G> from the C-RP tree.

         This is the same type of A-D route used to report active
         sources in the scenarios described in section 10.

         Note that when a PE thus joins the <C-S, C-G> tree, it may need
         to send a PIM (S,G,RPT-bit) prune to one of its CE PIM
         neighbors, as determined by ordinary PIM procedures..

         Whenever the PE deletes the <C-S, C-G> state that was
         previousely created as a result of receiving a C-multicast
         route for <C-S, C-G> from some other PE, the PE that deletes
         the state also withdraws the auto-discovery route that was
         advertised when the state was created.

         N.B.: SINCE ALL PES WITH RECEIVERS FOR GROUP C-G WILL JOIN THE
         C-S SOURCE TREE IF ANY OF THEM DO, IT IS NEVER NECESSARY TO
         DISTRIBUTE A BGP C-MULTICAST ROUTE FOR THE PURPOSE OF PRUNING
         SOURCES FROM THE SHARED TREE.


   In summary when the CE-PE routing protocol for all PEs that belong to
   a MVPN is not PIM-SM, selection of a consistant upstream PE to reach



Rosen & Raggarwa                                               [Page 58]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   C-S is sufficient to eliminate duplicates when C-S is multi-homed to
   a single AS. When C-S is multi-homed to multiple ASes, duplicate
   packet detection can be performed as the receiver PE can always
   determine whether packets arrived on the wrong tunnel. When the CE-PE
   routing protocol is PIM-SM, additional procedures as described above
   are required to force all PEs within all ASes to switch to the C-S
   tree from the C-RP tree when any PE switches to the C-S tree.


10. Deployment Models

   This section describes some optional deployment models and specific
   procedures for those deployment models.


10.1. Co-locating C-RPs on a PE

   [MVPN-REQ] describes C-RP engineering as an issue when PIM-SM (or
   bidir-PIM) is used in ASM mode on the VPN customer site. To quote
   from [MVPN-REQ]:

   "In some cases this engineering problem is not trivial: for instance,
   if sources and receivers are located in VPN sites that are different
   than that of the RP, then traffic may flow twice through the SP
   network and the CE-PE link of the RP (from source to RP, and then
   from RP to receivers) ; this is obviously not ideal.  A multicast VPN
   solution SHOULD propose a way to help on solving this RP engineering
   issue."

   One of the C-RP deployment models is for the customer to outsource
   the RP to the provider. In this case the provider may co-locate the
   RP on the PE that is connected to the customer site [MVPN-REQ]. This
   model is introduced in [RP-MVPN]. This section describes how
   anycast-RP can be used for achieving this by advertising active
   sources. This is described below.


10.1.1. Initial Configuration

   For a particular MVPN, at least one or more PEs that have sitse in
   that MVPN, act as an RP for the sites of that MVPN connected to these
   PEs.  Within each MVPN all these RPs use the same (anycast) address.
   All these RPs use the Anycast RP technique.








Rosen & Raggarwa                                               [Page 59]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


10.1.2. Anycast RP Based on Propagating Active Sources

   This mechanism is based on propagating active sources between RPs.

   [Editor's Note: This is derived from the model in [RP-MVPN].]


10.1.2.1. Receiver(s) Within a Site

   The PE which receives C-Join for (*,G) or (S,G) does not send the
   information that it has receiver(s) for G until it receives
   information about active sources for G from an upstream PE.

   On receiving this (described in the next section), the downstream PE
   will respond with Join for C-(S,G). Sending this information could be
   done using any of the procedures described in section 5. If BGP is
   used, the ingress address is set to the upstream PE's address which
   has triggered the source active information. Only the upstream PE
   will process this information. If unicast PIM is used then a unicast
   PIM message will have to be sent to the PE upstream PE that has
   triggered the source active information. If a MI-PMSI is used than
   further clarification is needed on the upstream neighbor address of
   the PIM message and will be provided in a future revision.


10.1.2.2. Source Within a Site

   When a PE receives PIM-Register from a site that belongs to a given
   VPN, PE follows the normal PIM anycast RP procedures. It then
   advertises the source and group of the multicast data packet caried
   in PIM-Register message to other PEs in BGP using the following
   information elements:

     - Active source address

     - Active group address

     - Route target of the MVPN.

   This advertisement goes to all the PEs that belong to that MVPN. When
   a PE receives this advertisement, it checks whether there are any
   receivers in the sites attached to the PE for the group carried in
   the source active advertisement. If yes, then it generates an
   advertisement for C-(S,G) as specified in the previous section.

   Note that the mechanism described in section 7.3.2. can be leveraged
   to advertise a S-PMSI binding along with the source active messages.




Rosen & Raggarwa                                               [Page 60]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


10.1.2.3. Receiver Switching from Shared to Source Tree

   No additional procedures are required when multicast receivers in
   customer's site shift from shared tree to source tree.


10.2. Using MSDP between a PE and a Local C-RP

   Section 10.1 describes the case where each PE is a C-RP.  This
   enables the PEs to know the active multicast sources for each MVPN,
   and they can then use BGP to distribute this information to each
   other.  As a result, the PEs do not have to join any shared C-trees,
   and this results in a simplification of the PE operation.

   In another deployment scenario, the PEs are not themselves C-RPs, but
   use MSDP to talk to the C-RPs.  In particular, a PE which attaches to
   a site that contains a C-RP becomes an MSDP peer of that C-RP.  That
   PE then uses BGP to distribute the information about the active
   sources to the other PEs.  When the PE determines, by MSDP, that a
   particular source is no longer active, then it withdraws the
   corresponding BGP update.  Then the PEs do not have to join any
   shared C-trees, but they do not have to be C-RPs either.

   MSDP provides the capability for a Source Active message to carry an
   encapsulated data packet.  This capability can be used to allow an
   MSDP speaker to receive the first (or first several) packet(s) of an
   (S,G) flow, even though the MSDP speaker hasn't yet joined the (S,G)
   tree.  (Presumably it will join that tree as a result of receiving
   the SA message which carries the encapsulated data packet.)  If this
   capability is not used, the first several data packets of an (S,G)
   stream may be lost.

   A PE which is talking MSDP to an RP may receive such an encapsulated
   data packet from the RP.  The data packet should be decapsulated and
   transmitted to the other PEs in the MVPN.  If the packet belongs to a
   particular (S,G) flow, and if the PE is a transmitter for some S-PMSI
   to which (S,G) has already been bound, the decapsulated data packet
   should be transmitted on that S-PMSI.  Otherwise, if an I-PMSI exists
   for that MVPN, the decapsulated data packet should be transmitted on
   it.  (If a default MI-PMSI exists, this would typically be used.)  If
   neither of these conditions hold, the decapsulated data packet is not
   transmitted to the other PEs in the MVPN.  The decision as to whether
   and how to transmit the decapsulated data packet does not effect the
   processing of the SA control message itself.

   Suppose that PE1 transmits a multicast data packet on a PMSI, where
   that data packet is part of an (S,G) flow, and PE2 receives that
   packet form that PMSI.  According to section 9, PE1 is not the PE



Rosen & Raggarwa                                               [Page 61]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   that PE2 expects to be transmitting (S,G) packets, then PE2 must
   discard the packet.  If an MSDP-encapsulated data packet is
   transmitted on a PMSI as specified above, this rule from section 9
   would likely result in the packet's getting discarded.  Therefore, if
   MSDP-encapsulated data packets being decapsulated and transmitted on
   a PMSI, we need to modify the rules of section 9 as follows:

      1. If the receiving PE, PE1, has already joined the (S,G) tree,
         and has chosen PE2 as the upstream PE for the (S,G) tree, but
         this packet does not come from PE2, PE1 must discard the
         packet.

      2. If the receiving PE, PE1, has not already joined the (S,G)
         tree, but is a PIM adjacency to a CE which is downstream on the
         (*,G) tree, the packet should be forwarded to the CE.


11. Encapsulations

   The BGP-based auto-discovery procedures will ensure that the PEs in a
   single MVPN only use tunnels that they can all support, and for a
   given kind of tunnel, that they only use encapsulations that they can
   all support.


11.1. Encapsulations for Single PMSI per Tunnel

11.1.1. Encapsulation in GRE

   GRE encapsulation can be used for any PMSI that is instantiated by a
   mesh of unicast tunnels, as well as for any PMSI that is instantiated
   by one or more PIM tunnels of any sort.



















Rosen & Raggarwa                                               [Page 62]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006



   Packets received        Packets in transit      Packets forwarded
   at ingress PE           in the service          by egress PEs
                           provider network

                           +---------------+
                           |  P-IP Header  |
                           +---------------+
                           |      GRE      |
   ++=============++       ++=============++       ++=============++
   || C-IP Header ||       || C-IP Header ||       || C-IP Header ||
   ++=============++ >>>>> ++=============++ >>>>> ++=============++
   || C-Payload   ||       || C-Payload   ||       || C-Payload   ||
   ++=============++       ++=============++       ++=============++


   The IP Protocol Number field in the P-IP Header must be set to 47.
   The Protocol Type field of the GRE Header must be set to 0x800.

   When an encapsulated packet is transmitted by a particular PE, the
   source IP address in the P-IP header must be the same address as is
   advertised by that PE in the RPF information.

   If the PMSI is instantiated by a PIM tree, the destination IP address
   in the P-IP header is the group P-address associated with that tree.
   The GRE key field value is omitted.

   If the PMSI is instantiated by unicast tunnels, the destination IP
   address is the address of the destination PE, and the optional GRE
   Key field is used to identify a particular MVPN.  In this case, each
   PE would have to advertise a key field value for each MVPN; each PE
   would assign the key field value that it expects to receive.

   [RFC2784] specifies an optional GRE checksum, and [RFC2890] specifies
   an optional GRE sequence number fields.

   The GRE sequence number field is not needed because the transport
   layer services for the original application will be provided by the
   C-IP Header.

   The use of GRE checksum field must follow [RFC2784].

   To facilitate high speed implementation, this document recommends
   that the ingress PE routers encapsulate VPN packets without setting
   the checksum, or sequence fields.






Rosen & Raggarwa                                               [Page 63]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


11.1.2. Encapsulation in IP

   IP-in-IP [RFC1853] is also a viable option.  When it is used, the
   IPv4 Protocol Number field is set to 4. The following diagram shows
   the progression of the packet as it enters and leaves the service
   provider network.


   Packets received        Packets in transit      Packets forwarded
   at ingress PE           in the service          by egress PEs
                           provider network

                           +---------------+
                           |  P-IP Header  |
   ++=============++       ++=============++       ++=============++
   || C-IP Header ||       || C-IP Header ||       || C-IP Header ||
   ++=============++ >>>>> ++=============++ >>>>> ++=============++
   || C-Payload   ||       || C-Payload   ||       || C-Payload   ||
   ++=============++       ++=============++       ++=============++



11.1.3. Encapsulation in MPLS

   If the PMSI is instantiated as a P2MP MPLS LSP, MPLS encapsulation is
   used. Penultimate-hop-popping must be disabled for the P2MP MPLS LSP.
   If the PMSI is instantiated as an RSVP-TE P2MP LSP, additional MPLS
   encapsulation procedures are used, as specified in [RSVP-P2MP].

   If other methods of assigning MPLS labels to multicast distribution
   trees are in use, these multicast distribution trees may be used as
   appropriate to instantiate PMSIs, and any additional MPLS
   encapsulation procedures may be used.


















Rosen & Raggarwa                                               [Page 64]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006



   Packets received        Packets in transit      Packets forwarded
   at ingress PE           in the service          by egress PEs
                           provider network

                           +---------------+
                           | P-MPLS Header |
   ++=============++       ++=============++       ++=============++
   || C-IP Header ||       || C-IP Header ||       || C-IP Header ||
   ++=============++ >>>>> ++=============++ >>>>> ++=============++
   || C-Payload   ||       || C-Payload   ||       || C-Payload   ||
   ++=============++       ++=============++       ++=============++



11.2. Encapsulations for Multiple PMSIs per Tunnel

   The encapsulations for transmitting multicast data messages when
   there are multiple PMSIs per tunnel are based on the encapsulation
   for a single PMSI per tunnel, but with an MPLS label used for
   demultiplexing.

   The label is upstream-assigned and distributed via BGP as specified
   in section 4.  The label must enable the receiver to select the
   proper VRF, and may enable the receiver to select a particular
   multicast routing entry within that VRF.


11.2.1. Encapsulation in GRE

   Rather than the IP-in-GRE encapsulation discussed in section 11.1.1,
   we use the MPLS-in-GRE encapsulation.  This is specified in [draft-
   mpls-in-ip-or-gre].  The GRE protocol type MUST be set to 0x8847.
   [The reason for using the unicast rather than the multicast value is
   specified in [MPLS-MCAST-ENCAPS].


11.2.2. Encapsulation in IP

   Rather than the IP-in-IP encapsulation discussed in section 12.1.2,
   we use the MPLS-in-IP encapsulation.  This is specified in [draft-
   mpls-in-ip-or-gre].  The IP protocol number MUST be set to the value
   identifying the payload as an MPLS unicast packet. [There is no "MPLS
   multicast packet" protocol number.]







Rosen & Raggarwa                                               [Page 65]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


11.3. Encapsulations for Unicasting PIM Control Messages

   When PIM control messages are unicast, rather than being sent on an
   MI-PMSI, the the receiving PE needs to determine the particular MVPN
   whose multicast routing information is being carried in the PIM
   message.  One method is to use a downstream-assigned MPLS label which
   the receiving PE has allocated for this specific purpose.  The label
   would be distributed via BGP.  This can be used with an MPLS, MPLS-
   in-GRE, or MPLS-in-IP encapsulation.

   A possible alternative to modify the PIM messages themselves so that
   they carry information which can be used to identify a particular
   MVPN, such as an RT.

   This area is still under consideration.



11.4. General Considerations for IP and GRE Encaps

   These apply also to the MPLS-in-IP and MPLS-in-GRE encapsulations.


11.4.1. MTU

   Path MTU discovery cannot be relied upon to ensure that the
   transmitter sends packets which are small enough to reach all the
   destinations.  This requires that:

      1. The ingress PE router (one that does the encapsulation) must
         not set the DF bit in the outer header, and

      2. If the "DF" bit is cleared in the IP header of the C-Packet,
         fragment the C-Packet before encapsulation if appropriate.
         This is very important in practice due to the fact that the
         performance of reassembly function is significantly lower than
         that of decapsulating and forwarding packets on today's router
         implementations.


11.4.2. TTL

   The ingress PE should not copy the TTL field from the payload IP
   header received from a CE router to the delivery IP or MPLS header.
   The setting of the TTL of the delivery header is determined by the
   local policy of the ingress PE router.





Rosen & Raggarwa                                               [Page 66]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


11.4.3. Differentiated Services

   By default, the setting of the DS field in the delivery IP header
   should follow the guidelines outlined in [RFC2983].  Setting the EXP
   field in the delivery MPLS header should follow the guidelines in
   [REF]. An SP may also choose to deploy any of the additional
   mechanisms the PE routers support.


11.4.4. Avoiding Conflict with Internet Multicast

   If the SP is providing Internet multicast, distinct from its VPN
   multicast services, and using PIM based P-multicast trees, it must
   ensure that the group P-addresses which it used in support of MPVN
   services are distinct from any of the group addresses of the Internet
   multicasts it supports.  This is best done by using administratively
   scoped addresses [ADMIN-ADDR].

   The group C-addresses need not be distinct from either the group P-
   addresses or the Internet multicast addresses.


12. Security Considerations

   To be supplied.


13. IANA Considerations

   To be supplied.


14. Other Authors

   Sarveshwar Bandi, Yiqun Cai, Thomas Morin, Yakov Rekhter, IJsbrands
   Wijnands, Seisho Yasukawa


15. Other Contributors

   Significant contributions were made Arjen Boers, Toerless Eckert,
   Adrian Farrel, Luyuan Fang, Dino Farinacci, Lenny Guiliano, Shankar
   Karuna, Anil Lohiya, Tom Pusateri, Ted Qian, Robert Raszuk, Tony
   Speakman, Dan Tappan.







Rosen & Raggarwa                                               [Page 67]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


16. Authors' Addresses


      Rahul Aggarwal (Editor)
      Juniper Networks
      1194 North Mathilda Ave.
      Sunnyvale, CA 94089
      Email: rahul@juniper.net



      Sarveshwar Bandi
      Motorola
      Vanenburg IT park, Madhapur,
      Hyderabad, India
      Email: sarvesh@motorola.com



      Yiqun Cai
      Cisco Systems, Inc.
      170 Tasman Drive
      San Jose, CA, 95134
      E-mail: ycai@cisco.com



      Thomas Morin
      France Telecom R & D
      2, avenue Pierre-Marzin
      22307 Lannion Cedex
      France
      Email: thomas.morin@francetelecom.com



      Yakov Rekhter
      Juniper Networks
      1194 North Mathilda Ave.
      Sunnyvale, CA 94089
      Email: yakov@juniper.net










Rosen & Raggarwa                                               [Page 68]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006



      Eric C. Rosen (Editor)
      Cisco Systems, Inc.
      1414 Massachusetts Avenue
      Boxborough, MA, 01719
      E-mail: erosen@cisco.com



      IJsbrand Wijnands
      Cisco Systems, Inc.
      170 Tasman Drive
      San Jose, CA, 95134
      E-mail: ice@cisco.com



      Seisho Yasukawa
      NTT Corporation
      9-11, Midori-Cho 3-Chome
      Musashino-Shi, Tokyo 180-8585,
      Japan
      Phone: +81 422 59 4769
      Email: yasukawa.seisho@lab.ntt.co.jp



17. Normative References

   [MVPN-REQ] T. Morin, Ed., "Requirements for Multicast in L3
   Provider-Provisioned VPNs", draft-ietf-l3vpn-ppvpn-mcast-reqts-
   08.txt, May 2006

   [RFC4364] "BGP/MPLS IP VPNs", Rosen, Rekhter, et. al., February 2006

   [RFC2119] "Key words for use in RFCs to Indicate Requirement
   Levels.", Bradner, March 1997

   [PIM-SM]  "Protocol Independent Multicast - Sparse Mode (PIM-SM)",
   Fenner, Handley, Holbrook, Kouvelas, March 2006, draft-ietf-pim- sm-
   v2-new-12.txt

   [RSVP-P2MP] R. Aggarwal, et. al., "Extensions to RSVP-TE for Point to
   Multipoint TE LSPs", draft-ietf-mpls-rsvp-te-p2mp-05.txt, May 2006

   [MPLS-IP] T. Worster, Y. Rekhter, E. Rosen, "Encapsulating MPLS in IP
   or Generic Routing Encapsulation (GRE)", RFC 4023, March 2005




Rosen & Raggarwa                                               [Page 69]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


   [MPLS-MCAST-ENCAPS] T. Eckert, E. Rosen, R. Aggarwal, Y. Rekhter,
   "MPLS Multicast Encapsulations", draft-ietf-mpls-multicast-encaps-
   00.txt, February 2006

   [MPLS-UPSTREAM-LABEL] R. Aggarwal, Y. Rekhter, E. Rosen, "MPLS
   Upstream Label Assignment and Context Specific Label Space", draft-
   ietf-mpls-upstream-label-01.txt, February 2006

   [MVPN-BGP], R. Aggarwal, E. Rosen,  T. Morin, Y. Rekhter,  C.
   Kodeboniya, "BGP Encodings for Multicast in MPLS/BGP IP VPNs",
   draft-raggarwa-l3vpn-2547bis-mcast-bgp-02.txt, June 2006


18. Informative References

   [ROSEN-8] E. Rosen, Y. Cai, I. Wijnands, "Multicast in MPLS/BGP IP
   VPNs", draft-rosen-vpn-mcast-08.txt

   [MVPN-BASE] R. Aggarwal, A. Lohiya, T. Pusateri, Y. Rekhter, "Base
   Specification for Multicast in MPLS/BGP VPNs", draft-raggarwa-l3vpn-
   2547-mvpn-00.txt

   [RAGGARWA-MCAST] R. Aggarwal, et. al., "Multicast in BGP/MPLS VPNs
   and VPLS", draft-raggarwa-l3vpn-mvpn-vpls-mcast-01.txt".

   [RP-MVPN] S. Yasukawa, et. al., "BGP/MPLS IP Multicast VPNs", draft-
   yasukawa-l3vpn-p2mp-mcast-00.txt

   [RFC2784] D. Farinacci, et. al., "Generic Routing Encapsulation",
   March 2000

   [RFC2890] G. Dommety, "Key and Sequence Number Extensions to GRE",
   September 2000

   [RFC1853] W. Simpson, "IP in IP Tunneling", October 1995

   [RFC2983] D. Black, "Differentiated Services and Tunnels", October
   2000













Rosen & Raggarwa                                               [Page 70]


Internet Draft   draft-ietf-l3vpn-2547bis-mcast-02.txt         June 2006


19. Full Copyright Statement

   Copyright (C) The Internet Society (2006).

   This document is subject to the rights, licenses and restrictions
   contained in BCP 78, and except as set forth therein, the authors
   retain all their rights.

   This document and the information contained herein are provided on an
   "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
   OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY AND THE INTERNET
   ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED,
   INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE
   INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
   WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.


20. Intellectual Property

   The IETF takes no position regarding the validity or scope of any
   Intellectual Property Rights or other rights that might be claimed to
   pertain to the implementation or use of the technology described in
   this document or the extent to which any license under such rights
   might or might not be available; nor does it represent that it has
   made any independent effort to identify any such rights.  Information
   on the procedures with respect to rights in RFC documents can be
   found in BCP 78 and BCP 79.

   Copies of IPR disclosures made to the IETF Secretariat and any
   assurances of licenses to be made available, or the result of an
   attempt made to obtain a general license or permission for the use of
   such proprietary rights by implementers or users of this
   specification can be obtained from the IETF on-line IPR repository at
   http://www.ietf.org/ipr.

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights that may cover technology that may be required to implement
   this standard.  Please address the information to the IETF at ietf-
   ipr@ietf.org.











Rosen & Raggarwa                                               [Page 71]