RFC 6625

Wildcards in Multicast VPN Auto-Discovery Routes

📅 published May 2012, from ietf

RFC 6625: Wildcards in Multicast VPN Auto-Discovery Routes 🦋

Okay bestie, here's the tea on RFC 6625 💖 It's an IETF document on the Standards Track (so it's an actual Internet Standard, no cap ✨), dated May 2012, with ISSN 2070-1721.

It was written by four people:

  • E. Rosen, Ed. from Cisco Systems, Inc.
  • Y. Rekhter, Ed. from Juniper Networks, Inc.
  • W. Henderickx from Alcatel-Lucent
  • R. Qiu from Huawei

It updates RFC 6514 (Updates: 6514) 🎀

Abstract ✨

Okay so in Multicast VPNs (MVPNs), customer multicast flows ride through a service provider's network inside "tunnels" 🌈 The base specs already define BGP multicast VPN "auto-discovery routes" and explain how to use one to say "hey, this one customer flow is traveling through this specific tunnel." 💅

The problem is that they can't squish multiple customer flows into one route when those flows share a tunnel. They also can't say "this tunnel is the default for all customer flows," unless every provider edge router in the MVPN has joined that tunnel. Literally so limiting 😭

This document fixes all of that by letting "wildcard" elements show up in the customer flow identifiers. With wildcards, one auto-discovery route can cover multiple customer flows, or even all of them. Iconic 👑

Status of This Memo 🌸

This is an Internet Standards Track document.

It's a product of the Internet Engineering Task Force (IETF), and it represents the consensus of the IETF community. It went through public review and was approved for publication by the Internet Engineering Steering Group (IESG). More info on Internet Standards is in Section 2 of RFC 5741.

For the current status, any errata, and how to give feedback, check http://www.rfc-editor.org/info/rfc6625.

Copyright (c) 2012 IETF Trust and the persons identified as the document authors. All rights reserved.

This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) as they stand on the publication date. Read those carefully, bestie, because they explain your rights and restrictions for this doc. Any Code Components pulled out of this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions, and they come with no warranty, as the Simplified BSD License describes.

Table of Contents 🍓

   1. Introduction ....................................................3
      1.1. Terminology ................................................3
      1.2. Wildcards in S-PMSI A-D Routes .............................5
      1.3. Use Cases ..................................................5
   2. Encoding of Wildcards ...........................................7
   3. Finding the Matching S-PMSI A-D Route ...........................8
      3.1. Finding the Match for Data Transmission ....................8
      3.2. Finding the Match for Data Reception .......................9
           3.2.1. Finding the Match for (C-S,C-G) .....................9
           3.2.2. Finding the Wildcard Match for (C-*,C-G) ............9
   4. Procedures for S-PMSI A-D Routes with Wildcards ................10
      4.1. Procedures for All Kinds of Wildcards .....................10
      4.2. Procedures for (C-*,C-G) S-PMSI A-D Routes ................11
      4.3. Procedures for (C-S,C-*) S-PMSI A-D Routes ................12
      4.4. Procedures for (C-*,C-*) S-PMSI A-D Routes ................13
   5. Security Considerations ........................................15
   6. Acknowledgments ................................................15
   7. Normative References ...........................................15

1. Introduction ✨

Okay, same intro as the abstract, but with the details 💅 In MVPNs, customer multicast flows are carried in tunnels through a service provider's network. The base specs define BGP multicast VPN "auto-discovery routes" and explain how to use one to advertise that an individual customer multicast flow is riding in a particular tunnel. 🦋

Those specs can't put multiple customer flows in one route when they share one tunnel. They also can't advertise that a tunnel is the default for carrying all customer flows, unless every PE router of the MVPN has joined that tunnel. This document removes these restrictions by specifying the use of "wildcard" elements in the customer flow identifiers. With wildcards, a single auto-discovery route can refer to multiple customer flows, or even to all customer flows. Main character energy 👑

1.1. Terminology 🎀

This document uses terminology from [MVPN]. Specifically, it uses the prefixes "C-" and "P-" (as specified in Section 3.1 of [MVPN]) to tell apart addresses in the "customer address space" from addresses in the "provider address space." These are the terms and acronyms that matter most here:

  • MVPN: Multicast Virtual Private Network, which is a VPN [L3VPN] offering multicast service.
  • VRF: VPN Routing and Forwarding table [L3VPN].
  • SP: Service Provider.
  • P-tunnel: A tunnel through the network of one or more SPs.
  • C-S: Multicast Source, which is a multicast source address in the address space of a customer network.
  • C-G: Multicast Group, which is a multicast group address (destination address) in the address space of a customer network.
  • C-multicast flow or C-flow: A customer multicast flow. Each C-flow is identified by the ordered pair (source address, group address), where both addresses are in the customer's address space. A C-flow's identifier is usually written as (C-S,C-G).
  • RP: A "Rendezvous Point," as defined in [PIM].
  • C-RP: A Rendezvous Point whose address is in the customer's address space.
  • Selective P-tunnel: A P-tunnel joined only by the Provider Edge (PE) routers that need to receive one or more of the C-flows traveling through that P-tunnel.
  • Inclusive P-tunnel: A P-tunnel joined by all the PE routers that attach to sites of a given MVPN.
  • S-PMSI A-D route: Selective Provider Multicast Service Interface Auto-Discovery route. These are carried in BGP Update messages and advertise the fact that particular C-flows are bound to (i.e., traveling through) particular P-tunnels. 🌈

Familiarity with multicast concepts and terminology [PIM] is also presupposed, so brush up on that first 🧚

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]. 🫶

1.2. Wildcards in S-PMSI A-D Routes 🌸

As specified in [MVPN] and [MVPN-BGP], an S-PMSI A-D route advertises that a particular C-flow is bound to a particular selective P-tunnel. 💖

The identifier of the specified C-flow (for example, (C-S,C-G)) is encoded into the Network Layer Reachability Information (NLRI) of the S-PMSI A-D route. The identifier of the specified P-tunnel is encoded into an attribute (the "PMSI Tunnel Attribute") of that route. So each S-PMSI A-D route specifies a single C-flow. To bind multiple C-flows to a single P-tunnel, you have to advertise one S-PMSI A-D route per C-flow, each specifying the same P-tunnel. Extra work, honestly 😭

This document defines OPTIONAL extensions to the procedures and encodings specified in [MVPN] and [MVPN-BGP]. These extensions let a single S-PMSI A-D route advertise that multiple C-multicast flows are bound to a single P-tunnel. 🔥

The extensions are built on allowing the NLRI of an S-PMSI A-D route to contain a "wildcard." In the NLRI encoding, a wildcard can replace the C-S, the C-G, or both. The notation "C-*" denotes a wildcard. The extensions allow three kinds of wildcards: (C-*,C-*), (C-S,C-*), and (C-*,C-G). ✨

Using wildcards, a PE may be able to reduce the number of S-PMSI A-D routes it originates, which improves the scalability of the control plane. There's no impact on data plane scalability, though, because the number of P-tunnels doesn't go down 🙈

The encoding and detailed procedures are specified in the subsequent sections of this document. Stay tuned, bestie 🦋

1.3. Use Cases 💅

okay so there are a bunch of situations where wildcards in the NLRI of an S-PMSI A-D route are lowkey SO useful 🦋

  • Using a selective P-tunnel as the default tunnel. So [MVPN] and [MVPN-BGP] have procedures that let a PE say it'll use an inclusive P-tunnel as the P-tunnel it sends all C-flows on by "default" 🎀 But those docs have no way for a PE to say it'll use a selective P-tunnel as its "default" one 😭 With the extensions in this document, a PE can totally do that! It just advertises an S-PMSI A-D route whose NLRI contains (C-*,C-*) ✨
  • Binding multiple C-flows traveling along a customer's Protocol Independent Multicast - Sparse Mode (PIM-SM) shared tree to a single P-tunnel. A PE might be hooked up to an MVPN site that has a customer RP (C-RP) 👑 The C-RP can be the root of one or more shared trees, which multicast folks call (*,G) trees. If the PE advertises one S-PMSI A-D route whose NLRI contains the (C-*,C-G) wildcard, it can bind every C-flow on the customer's (*,G) tree to a single P-tunnel 🧚 This only works when C-G is a non-bidirectional ASM (Any Source Multicast) group, bestie, so keep that in mind 💜
  • Binding multiple C-flows with the same C-group address to a single P-tunnel, even if each such C-flow is traveling along a customer's PIM source tree. Picture an MVPN site with a bunch of multicast sources all sending to one common multicast group, each on a customer's PIM source trees 🌈 Or the PE could be connected to several sites, each with at least one source sending to that same group. Either way, one S-PMSI A-D route with an NLRI containing (C-*,C-G) lets the PE bind all of those C-flows to a single P-tunnel 🍓 This one also only applies when the C-group is a non-bidirectional ASM group 🙈
  • Binding multiple C-flows with the same C-group address to a single P-tunnel, when those C-flows are traveling along a customer's BIDIR-PIM shared tree. Same vibe as the last one, but this time it only applies when the C-group is a BIDIR-PIM group 🌸
  • Binding multiple C-flows from a given C-source to a given P-tunnel, irrespective of whether those C-flows all have the same C-group address. This comes in handy when the C-group addresses are SSM (Single Source Multicast) addresses 💖 Imagine one source sending out a bunch of "channels," each with its own C-group address. You can bind all those channels to one P-tunnel without advertising an S-PMSI A-D route for each one. Iconic efficiency, no cap ✨

Of course, a specific C-flow, (C-S,C-G), can always be given its own P-tunnel by advertising an S-PMSI A-D route whose NLRI contains (C-S,C-G) 🦋

Later on, in Section 4, we're gonna talk about an S-PMSI A-D route being "ignored" 👀 Quick heads up: when we say a route is "ignored," we are NOT saying its normal BGP processing gets skipped. What it means is the route doesn't get considered when picking which P-tunnel to use for receiving multicast data, and the MPLS label values it carries aren't used. We'll put "ignore" in quotes whenever we mean this 🫶

This document only gives procedures for "unidirectional" P-tunnels, aka point-to-multipoint ones. "Bidirectional" (multipoint-to-multipoint) P-tunnels are outside the scope of this doc, so we're not going there today 🙅‍♀️

2. Encoding of Wildcards 🎀

Per [MVPN-BGP] Section 4.3, the MCAST-VPN NLRI in an S-PMSI A-D route is encoded like this, and yes it's giving diagram energy 🧚

            +-----------------------------------+
            |      RD   (8 octets)              |
            +-----------------------------------+
            | Multicast Source Length (1 octet) |
            +-----------------------------------+
            |  Multicast Source (variable)      |
            +-----------------------------------+
            |  Multicast Group Length (1 octet) |
            +-----------------------------------+
            |  Multicast Group   (variable)     |
            +-----------------------------------+
            |   Originating Router's IP Addr    |
            +-----------------------------------+

So the "source length" and "group length" fields normally always have a non-zero value 💅 But this document says a "zero-length" source or group stands in for the matching wildcard, and that's the whole trick 🌈 Specifically:

  • A source wildcard is encoded as a zero-length source field. The "multicast source length" field holds 0x00, and the "multicast source" field is left out entirely 🍓
  • A group wildcard is encoded as a zero-length group field. The "multicast group length" field holds 0x00, and the "multicast group" field is left out 🌸

3. Finding the Matching S-PMSI A-D Route 🔍

okay this section is where we get the precise rules for figuring out which S-PMSI A-D route "matches" a given (C-S,C-G) or (C-*,C-G) 🧚 Section 4 leans on these matching rules, so pay attention, bestie 💜

All the matching rules assume you're in the context of a given VRF at a given PE 💖

Here's the tea: the rules a PE uses to find the matching route for a (C-S,C-G) C-flow it needs to transmit are slightly different from the rules it uses for a (C-S,C-G) C-flow it needs to receive 😭 Those are specified in Sections 3.1 and 3.2, respectively 🦋

The S-PMSI A-D route that matches a given (C-S,C-G) can change over time, because routes get withdrawn or new ones get originated and/or advertised 🌊 Especially, if (C-S,C-G) currently matches a route whose NLRI contains (C-*,C-*), then originating or receiving an S-PMSI A-D route whose NLRI contains (C-S,C-G) can make (C-S,C-G) match that newer route instead 👑 Also worth knowing: which S-PMSI A-D route matches a given (C-S,C-G) does NOT depend on the order the routes were originated or received in ✨

3.1. Finding the Match for Data Transmission 📡

Let's say we've got a given PE, and we'll call it PE1 🎀 At any given time, for a given VRF at PE1, there's a (possibly empty) set of S-PMSI A-D routes that PE1 has originated and advertised but not withdrawn 🌸 We call these routes "currently originated" by PE1. Now suppose PE1 needs to transmit a particular C-flow (C-S,C-G) to one or more other PEs. Here's the algorithm it uses to find the S-PMSI A-D route that the C-flow "matches" 💅

  • If there's an S-PMSI A-D route currently originated by PE1 whose NLRI contains (C-S,C-G), then the (C-S,C-G) C-flow matches that route 🦋
  • Otherwise, if there's an S-PMSI A-D route currently originated by PE1 whose NLRI contains (C-S,C-*), AND if C-G is an SSM group address, then the (C-S,C-G) C-flow matches that route 💖
  • Otherwise, if there's an S-PMSI A-D route currently originated by PE1 whose NLRI contains (C-*,C-G), AND if C-G is an ASM group address, then the (C-S,C-G) C-flow matches that route 🌈
  • Otherwise, if there's an S-PMSI A-D route currently originated by PE1 whose NLRI contains (C-*,C-*), then the (C-S,C-G) C-flow matches that route 🧚

3.2. Finding the Match for Data Reception 📥

We say an S-PMSI A-D route is "installed" (in a given VRF) if the BGP decision process has selected it as the preferred route for its NLRI 👑

An S-PMSI A-D route is considered "originated by a given PE" if that PE's IP address is contained in the "Originating Router's IP Address" field in the MCAST-VPN NLRI of the route 🫶

3.2.1. Finding the Match for (C-S,C-G) 🎯

Suppose a PE router (we'll call it PE1) needs to receive (C-S,C-G), and PE1 has picked another PE router (we'll call it PE2) as the "upstream PE" [MVPN] for that flow 💜

  • If there's an installed S-PMSI A-D route originated by PE2 whose NLRI contains (C-S,C-G), then (C-S,C-G) matches that route 🌸
  • Otherwise, if there's an installed S-PMSI A-D route originated by PE2 whose NLRI contains (C-S,C-*), AND if C-G is an SSM multicast group address, then (C-S,C-G) matches that route 🍓
  • Otherwise, if there's an installed S-PMSI A-D route originated by PE2 whose NLRI contains (C-*,C-G), AND if C-G is an ASM multicast group address, then (C-S,C-G) matches that route 🦋
  • Otherwise, if there's an installed S-PMSI A-D route originated by PE2 whose NLRI contains (C-*,C-*), then (C-S,C-G) matches that route ✨

3.2.2. Finding the Wildcard Match for (C-*,C-G) 🌈

Suppose a PE router (we'll call it PE1) needs to receive (C-*,C-G) traffic 💖 Heads up: even if (C-*,C-G) matches a non-wildcard S-PMSI A-D route (like Section 12.3 of [MVPN-BGP] describes), it might also match one or more wildcard S-PMSI A-D routes, as spelled out below 🧚

If on PE1 there's an installed S-PMSI A-D route originated by PE2 whose NLRI contains (C-*,C-G), then (C-*,C-G) matches this route if any one of these conditions holds 👑

  • PE1 determines that PE2 is the "upstream" PE [MVPN] for the C-RP of C-G, or
  • PE1 has installed one or more Source Active A-D routes for C-G originated by PE2, and for at least one of these routes, PE1 does not have a corresponding (C-S,C-G) state, or
  • C-G is a BIDIR-PIM group, or
  • Source Active A-D routes are not being used.

If (C-*,C-G) does not match a (C-*,C-G) S-PMSI A-D route from PE2, but PE1 has an installed (C-*,C-*) S-PMSI A-D route from PE2, then (C-*,C-G) matches the (C-*,C-*) route if any one of these conditions holds 🌸

  • PE1 determines that PE2 is the "upstream" PE [MVPN] for the C-RP of C-G, or
  • PE1 has installed one or more Source Active A-D routes for C-G originated by PE2, and for at least one of these routes, PE1 does not have a corresponding (C-S,C-G) state, or
  • C-G is a BIDIR-PIM group, or
  • Source Active A-D routes are not being used.

4. Procedures for S-PMSI A-D Routes with Wildcards ✨

4.1. Procedures for All Kinds of Wildcards

okay bestie, this section sets the rules for three wildcard flavours in the NLRI of an S-PMSI A-D route 💖 :

  • (C-*,C-G): Source wildcard, group specified.
  • (C-S,C-*): Source specified, group wildcard.
  • (C-*,C-*): Source wildcard, group wildcard.

Every other wildcard is outside the scope of this doc 🙈

Originating S-PMSI A-D routes with a particular wildcard kind is OPTIONAL 🎀. BUT if your PE can originate a kind, it MUST be able to interpret and correctly process S-PMSI A-D routes with that kind, and it SHOULD be able to handle all three kinds 👑

For a given MVPN, a PE MUST NOT originate S-PMSI A-D routes with a particular kind of wildcard unless it is known a priori that every PE attached to that MVPN can interpret and correctly process that kind 🫶

Whether to originate or withdraw wildcard routes is decided locally by the originating PE, so that's its call alone 🦋

As [MVPN-BGP] says, an S-PMSI A-D route rides in the NLRI field of an MP_REACH_NLRI attribute (see [BGP-MP]) ✨ Each S-PMSI A-D route has one address family, IPv4 or IPv6, which lives in the Address Family Identifier (AFI) field of that MP_REACH_NLRI attribute. A wildcard only ever refers to multicast flows of that same address family, so no crossing the streams 💅

These procedures only apply when the PMSI Tunnel Attribute of an S-PMSI A-D route specifies a "unidirectional" P-tunnel. Bidirectional P-tunnels (like Multipoint-to-Multipoint Label Switched Paths or BIDIR-PIM trees) are outside the scope of this doc 🌈

Throughout the next sections, an S-PMSI A-D route whose NLRI contains (C-*,C-G), (C-S,C-*), or (C-*,C-*) gets nicknamed a "(C-*,C-G) route", a "(C-S,C-*) route", or a "(C-*,C-*)" route, respectively 🍓

4.2. Procedures for (C-*,C-G) S-PMSI A-D Routes

This doc only covers (C-*,C-G) S-PMSI A-D routes when C-G is an ASM group address. Using them where C-G is an SSM group address is outside the scope of this doc 🙈 So if a PE gets a (C-*,C-G) S-PMSI A-D route and can tell that C-G is an SSM group address, the PE SHOULD "ignore" that route 🚫

By default, the Route Targets carried by a (C-*,C-G) S-PMSI A-D route originated by a given VRF are the same as the Route Targets carried in the (unicast) VPN-IP routes that came from that VRF 💜 An implementation MUST allow those Route Targets to be set by configuration. If there's no configured set, the route MUST carry the default set of Route Targets 🎀

If a PE needs to transmit packets of a (C-S,C-G) C-flow, and (C-S,C-G) matches a (C-*,C-G) S-PMSI A-D route under the rules of Section 3.1, then the PE MUST use the P-tunnel advertised in that route for sending that C-flow 🔥 (Quick fun fact: a given (C-S,C-G) can't match both a (C-*,C-G) wildcard and a (C-S,C-*) wildcard, so there's no drama there 💀)

If PIM is the PE-PE control protocol, and the PE has (C-*,C-G) and/or (C-S,C-G) state that matches an S-PMSI A-D route (following the procedures of Section 3.2), then the PE MUST join the P-tunnel in the PMSI Tunnel Attribute of that route 🦋

If BGP is the PE-PE control protocol, then:

  • If a PE has currently originated a C-multicast Shared Tree Join for (C-*,C-G), and (C-*,C-G) matches a (C-*,C-G) S-PMSI A-D route, then the PE applies the procedures of Section 12.3 ("Receiving S-PMSI A-D Routes by PEs") of [MVPN-BGP] to that S-PMSI A-D route 🌸
  • Otherwise (the PE does not have a currently originated C-multicast Shared Tree Join for (C-*,C-G)), if there are one or more values of C-S for which the PE has a currently originated Source Tree Join C-multicast route for (C-S,C-G), the PE MUST join the tunnels advertised by the S-PMSI A-D routes that match each such (C-S,C-G) (according to Section 3.2) ✨
  • Otherwise, the PE "ignores" the route 🙈

4.3. Procedures for (C-S,C-*) S-PMSI A-D Routes

This doc covers (C-S,C-*) S-PMSI A-D routes for only the C-multicast flows where C-G is an SSM group address 💖 Using them for other C-multicast flows is outside the scope of this doc. Specifically, if a PE gets a (C-S,C-*) S-PMSI A-D route and can tell that C-G is not an SSM group address, the PE SHOULD "ignore" that route 🚫

By default, the Route Targets carried by a (C-S,C-*) S-PMSI A-D route originated by a given VRF are the intersection of two sets: the Route Targets in the Intra-AS I-PMSI A-D route from that VRF, and the Route Targets in the unicast VPN-IP route to C-S from that VRF 🧚 An implementation MUST allow those Route Targets to be set by configuration. With no configured set, the route MUST carry the default set of Route Targets 👑

If a PE needs to transmit packets of a (C-S,C-G) C-flow, and (C-S,C-G) matches a (C-S,C-*) S-PMSI A-D route under the rules of Section 3.1, then the PE MUST use the P-tunnel advertised in that route for sending that C-flow 💅 (Again, a given (C-S,C-G) can't match both a (C-*,C-G) wildcard and a (C-S,C-*) wildcard 🙃)

If PIM is the PE-PE control protocol for distributing C-multicast routing, and a PE needs to receive a (C-S,C-G) flow, and (C-S,C-G) matches the (C-S,C-*) S-PMSI A-D route (following Section 3.2), then the PE MUST join the P-tunnel in the PMSI Tunnel Attribute of that route 🌈

If BGP is the PE-PE control protocol for distributing C-multicast routing, and there's some (C-S,C-G) where (a) the PE has a currently originated (C-S,C-G) Source Tree Join C-multicast route, AND (b) that (C-S,C-G) matches (following Section 3.2) a (C-S,C-*) S-PMSI A-D route, then PE1 applies the procedures of Section 12.3 ("Receiving S-PMSI A-D Routes by PEs") of [MVPN-BGP] to the matching S-PMSI A-D route 🍓

4.4. Procedures for (C-,C-) S-PMSI A-D Routes

(C-*,C-*) S-PMSI A-D routes get used when, for a given MVPN, a PE has a policy not to use an I-PMSI for carrying multicast data traffic from the MVPN's site(s) connected to that PE 🌸 When the (C-*,C-*) wildcard is paired with BGP C-multicast routing, you get the "S-PMSI only" model, where no I-PMSIs are used at all for that MVPN 🧁

A (C-*,C-*) S-PMSI A-D route is originated for a given MVPN by a given PE only if that PE has been provisioned with the policy to do so 🔒

When it's been provisioned that way, the PE MAY originate the (C-*,C-*) S-PMSI A-D route as soon as it's enabled to support the given MVPN 🚀 Or, the PE MAY hold off on originating the route until one of these conditions holds:

  • The PE-PE protocol for distributing C-multicast routing is PIM, and for the given MVPN, the PE has some (C-S,C-G) or (C-*,C-G) state whose upstream interface is one of the VRF interfaces for that MVPN 🦋
  • The PE-PE protocol for distributing C-multicast routing is BGP, and the PE has received and installed either of the following:
    • a Source Tree Join C-multicast route, where the C-S in the route's NLRI is reachable via one of that MVPN's VRF interfaces, or
    • a Shared Tree Join C-multicast route, where the C-RP carried in that route is reachable via one of that MVPN's VRF interfaces 🌷

By default, the Route Targets carried by a (C-*,C-*) S-PMSI A-D route originated from a given VRF are the same as the Route Targets carried in the VPN-IP unicast routes from that VRF 💖 An implementation MUST allow those Route Targets to be set by configuration. With no configured set, the route MUST carry the default set of Route Targets, as described above 🎀

If a PE needs to transmit packets of a (C-S,C-G) C-flow, and (C-S,C-G) matches a (C-*,C-*) S-PMSI A-D route under the rules of Section 3.1, then the PE MUST use the P-tunnel advertised in that route for sending that C-flow 🔥 (Note that a given (C-S,C-G) can't match both a (C-*,C-*) wildcard and any other wildcard 💅)

If PIM is the PE-PE control protocol for distributing C-multicast routing, and a given PE, say PE1, needs to receive a (C-S,C-G) flow, and (C-S,C-G) matches the (C-*,C-*) S-PMSI A-D route (following Section 3.2), then PE1 MUST join the P-tunnel in the PMSI Tunnel Attribute of that route ✨

If BGP is the PE-PE control protocol for distributing C-multicast routing, then the PE applies the procedures of Section 12.3 ("Receiving S-PMSI A-D Routes by PEs") of [MVPN-BGP] to the matching S-PMSI A-D route if, and only if, one of these conditions holds:

  • The PE has a currently originated C-multicast Source Tree Join route for (C-S,C-G) that matches (following Section 3.2) the (C-*,C-*) S-PMSI A-D route, or
  • The PE has a currently originated C-multicast Shared Tree Join route for (C-*,C-G) that matches (following Section 3.2) the (C-*,C-*) S-PMSI A-D route. 🌈

5. Security Considerations ✨

There are no additional security considerations beyond those of [MVPN] and [MVPN-BGP]. 🔒 Just the ones you already know from those docs, bestie, so go check them out 🫶

6. Acknowledgments 💜

The authors wish to thank Arjen Boers, Dongling Duan, Apoorva Karan, Thomas Morin, Keyur Patel, Karthik Subramanian, and Kurt Windisch for many helpful discussions. Such a squad 👑🫶

7. Normative References 📚

  • [BGP-MP] Bates, T., Chandra, R., Katz, D., and Y. Rekhter, "Multiprotocol Extensions for BGP-4", RFC 4760, January 2007.
  • [L3VPN] Rosen, E. and Y. Rekhter, "BGP/MPLS IP Virtual Private Networks (VPNs)", RFC 4364, February 2006.
  • [MVPN] Rosen, E., Ed., and R. Aggarwal, Ed., "Multicast in MPLS/BGP IP VPNs", RFC 6513, February 2012.
  • [MVPN-BGP] Aggarwal, R., Rosen, E., Morin, T., and Y. Rekhter, "BGP Encodings and Procedures for Multicast in MPLS/BGP IP VPNs", RFC 6514, February 2012.
  • [PIM] Fenner, B., Handley, M., Holbrook, H., and I. Kouvelas, "Protocol Independent Multicast - Sparse Mode (PIM-SM): Protocol Specification (Revised)", RFC 4601, August 2006.
  • [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997.

Authors' Addresses 💖

Okay bestie, this is the last stop! 🧚 Here are the brilliant humans who built this whole thing, and how to reach each of them. Main character energy, fr fr 💅

Rahul Aggarwal 🦋 is at Arktan:

Rahul Aggarwal
Arktan

EMail: raggarwa_1@yahoo.com

Yiqun Cai 🌸 is at Microsoft, which is giving big tech vibes:

Yiqun Cai
Microsoft
1065 La Avenida
Mountain View, CA  94043

EMail: yiqunc@microsoft.com

Wim Henderickx 🎀 is at Alcatel-Lucent:

Wim Henderickx
Alcatel-Lucent

EMail: wim.henderickx@alcatel-lucent.be

Praveen Muley 🫶 is also at Alcatel-Lucent, iykyk:

Praveen Muley
Alcatel-Lucent

EMail: Praveen.Muley@alcatel-lucent.com

Ray (Lei) Qiu 👑 is reachable at this address:

Ray (Lei) Qiu
2330 Central Expressway
Santa Clara, CA  95050
USA

EMail: rayq@huawei.com

Yakov Rekhter (editor) 🔥 is at Juniper Networks, the one who kept everything organized:

Yakov Rekhter (editor)
Juniper Networks
1194 North Mathilda Ave.
Sunnyvale, CA  94089

EMail: yakov@juniper.net

Eric C. Rosen (editor) 💜 is at Cisco Systems, Inc., and co-ran the editing, no cap:

Eric C. Rosen (editor)
Cisco Systems, Inc.
1414 Massachusetts Avenue
Boxborough, MA  01719

EMail: erosen@cisco.com

IJsbrand Wijnands 🌈 is also at Cisco Systems, Inc., over in Belgium:

IJsbrand Wijnands
Cisco Systems, Inc.
De kleetlaan 6a Diegem 1831
Belgium

EMail: ice@cisco.com
📦 where this page comes from