idnits 2.17.1 draft-ietf-l2vpn-vpls-mcast-08.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an Introduction section. ** There are 2 instances of too long lines in the document, the longest one being 2 characters in excess of 72. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == Line 1379 has weird spacing: '...ertised by th...' == Line 1496 has weird spacing: '...pecting to se...' == Line 1501 has weird spacing: '...rt this situa...' == Line 1508 has weird spacing: '...rged by a par...' == Line 1509 has weird spacing: '...ponding to a ...' == (1 more instance...) == Using lowercase 'not' together with uppercase 'MUST', 'SHALL', 'SHOULD', or 'RECOMMENDED' is not an accepted usage according to RFC 2119. Please use uppercase 'NOT' together with RFC 2119 keywords (if that is what you mean). Found 'MUST not' in this paragraph: If the Selective tree is instantiated by a RSVP-TE P2MP LSP the PE at the root of the tree MUST establish the P2MP RSVP-TE LSP to the leaves. This LSP MAY have been established before the leaves receive the Selective tree binding, or MAY be established after the leaves receives the binding. A leaf MUST not switch to the Selective tree until it receives the binding and the RSVP-TE P2MP LSP is setup to the leaf. -- The document seems to contain a disclaimer for pre-RFC5378 work, and may have content which was first submitted before 10 November 2008. The disclaimer is necessary when there are original authors that you have been unable to contact, or if some do not wish to grant the BCP78 rights to the IETF Trust. If you are able to get all authors (current and original) to grant those rights, you can and should remove the disclaimer; otherwise, the disclaimer is needed and you can ignore this comment. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (October 25, 2010) is 4932 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-08) exists of draft-ietf-l2vpn-vpls-bgp-02 == Outdated reference: A later version (-09) exists of draft-ietf-l2vpn-vpls-ldp-03 == Outdated reference: A later version (-10) exists of draft-ietf-l3vpn-2547bis-mcast-08 == Outdated reference: A later version (-08) exists of draft-ietf-l3vpn-2547bis-mcast-bgp-06 -- No information found for draft-ietf-mpls-rsvp-te-no-php-obb-mapping - is the name correct? == Outdated reference: A later version (-07) exists of draft-ietf-l2vpn-vpls-mcast-reqts-05 -- Obsolete informational reference (is this intentional?): RFC 2385 (Obsoleted by RFC 5925) -- Obsolete informational reference (is this intentional?): RFC 4447 (Obsoleted by RFC 8077) Summary: 2 errors (**), 0 flaws (~~), 13 warnings (==), 5 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group R. Aggarwal (Editor) 3 Internet Draft Juniper Networks 4 Category: Standards Track 5 Expiration Date: April 2011 Y. Kamite 6 NTT Communications 8 L. Fang 9 Cisco Systems, Inc 11 October 25, 2010 13 Multicast in VPLS 15 draft-ietf-l2vpn-vpls-mcast-08.txt 17 Status of this Memo 19 This Internet-Draft is submitted to IETF in full conformance with the 20 provisions of BCP 78 and BCP 79. 22 Internet-Drafts are working documents of the Internet Engineering 23 Task Force (IETF), its areas, and its working groups. Note that other 24 groups may also distribute working documents as Internet-Drafts. 26 Internet-Drafts are draft documents valid for a maximum of six months 27 and may be updated, replaced, or obsoleted by other documents at any 28 time. It is inappropriate to use Internet-Drafts as reference 29 material or to cite them other than as "work in progress." 31 The list of current Internet-Drafts can be accessed at 32 http://www.ietf.org/ietf/1id-abstracts.txt. 34 The list of Internet-Draft Shadow Directories can be accessed at 35 http://www.ietf.org/shadow.html. 37 Copyright and License Notice 39 Copyright (c) 2010 IETF Trust and the persons identified as the 40 document authors. All rights reserved. 42 This document is subject to BCP 78 and the IETF Trust's Legal 43 Provisions Relating to IETF Documents 44 (http://trustee.ietf.org/license-info) in effect on the date of 45 publication of this document. Please review these documents 46 carefully, as they describe your rights and restrictions with respect 47 to this document. Code Components extracted from this document must 48 include Simplified BSD License text as described in Section 4.e of 49 the Trust Legal Provisions and are provided without warranty as 50 described in the Simplified BSD License. 52 This document may contain material from IETF Documents or IETF 53 Contributions published or made publicly available before November 54 10, 2008. The person(s) controlling the copyright in some of this 55 material may not have granted the IETF Trust the right to allow 56 modifications of such material outside the IETF Standards Process. 57 Without obtaining an adequate license from the person(s) controlling 58 the copyright in such materials, this document may not be modified 59 outside the IETF Standards Process, and derivative works of it may 60 not be created outside the IETF Standards Process, except to format 61 it for publication as an RFC or to translate it into languages other 62 than English. 64 Abstract 66 This document describes a solution for overcoming a subset of the 67 limitations of existing VPLS multicast solutions. It describes 68 procedures for VPLS multicast that utilize multicast trees in the 69 sevice provider (SP) network. One such multicast tree can be shared 70 between multiple VPLS instances. Procedures by which a single 71 multicast tree in the SP network can be used to carry traffic 72 belonging only to a specified set of one or more IP multicast streams 73 from one or more VPLSes are also described. 75 Table of Contents 77 1 Specification of requirements ......................... 4 78 2 Contributors .......................................... 4 79 3 Terminology ........................................... 5 80 4 Introduction .......................................... 5 81 5 Existing Limitations of VPLS Multicast ................ 6 82 6 Overview .............................................. 6 83 6.1 Inclusive and Selective Multicast Trees ............... 6 84 6.2 BGP-Based VPLS Membership Auto-Discovery .............. 8 85 6.3 IP Multicast Group Membership Discovery ............... 8 86 6.4 Advertising P-Multicast Tree to VPLS/C-Multicast Binding ..9 87 6.5 Aggregation ........................................... 9 88 6.6 Inter-AS VPLS Multicast ............................... 10 89 7 Intra-AS Inclusive P-Multicast Tree A-D/Binding ....... 11 90 7.1 Originating intra-AS VPLS auto-discovery routes ....... 12 91 7.2 Receiving intra-AS VPLS auto-discovery routes ......... 12 92 8 Demultiplexing P-Multicast Tree Traffic ............... 14 93 8.1 One P-Multicast Tree - One VPLS Mapping ............... 14 94 8.2 One P-Multicast Tree - Many VPLS Mapping .............. 14 95 9 Establishing P-Multicast Trees ........................ 15 96 9.1 Common Procedures ..................................... 15 97 9.2 RSVP-TE P2MP LSPs ..................................... 16 98 9.2.1 P2MP TE LSP - VPLS Mapping ............................ 16 99 9.3 Receiver Initiated MPLS Trees ......................... 17 100 9.3.1 P2MP LSP - VPLS Mapping ............................... 17 101 9.4 Encapsulation of Aggregate P-Multicast Trees .......... 17 102 10 Inter-AS Inclusive P-Multicast Tree A-D/Binding ....... 17 103 10.1 VSIs on the ASBRs ..................................... 18 104 10.1.1 Option (a) ............................................ 18 105 10.1.2 Option (e) ............................................ 18 106 10.2 Option (b) - Segmented Inter-AS Trees ................. 19 107 10.2.1 Segmented Inter-AS Trees VPLS Inter-AS A-D/Binding .... 19 108 10.2.2 Propagating BGP VPLS A-D routes to other ASes: Overview ...20 109 10.2.2.1 Propagating Intra-AS VPLS A-D routes in E-BGP ......... 21 110 10.2.2.2 Inter-AS A-D route received via E-BGP ................. 22 111 10.2.2.3 Leaf A-D Route received via E-BGP ..................... 24 112 10.2.2.4 Inter-AS A-D Route received via I-BGP ................. 24 113 10.3 Option (c) ............................................ 25 114 11 Optimizing Multicast Distribution via Selective Trees . 26 115 11.1 Protocol for Switching to Selective Trees ............. 28 116 11.2 Advertising C-(S, G) Binding to a Selective Tree ...... 29 117 11.3 Receiving S-PMSI A-D routes by PEs .................... 31 118 11.4 Inter-AS Selective Tree ............................... 32 119 11.4.1 VSIs on the ASBRs ..................................... 33 120 11.4.1.1 VPLS Inter-AS Selective Tree A-D Binding .............. 33 121 11.4.2 Inter-AS Segmented Selective Trees .................... 33 122 11.4.2.1 Handling S-PMSI A-D routes by ASBRs ................... 34 123 11.4.2.1.1 Merging Selective Tree into an Inclusive Tree ......... 35 124 11.4.3 Inter-AS Non-Segmented Selective trees ................ 36 125 12 BGP Extensions ........................................ 36 126 12.1 Inclusive Tree/Selective Tree Identifier .............. 36 127 12.2 MCAST-VPLS NLRI ....................................... 37 128 12.2.1 S-PMSI auto-discovery route ........................... 37 129 12.2.2 Leaf auto-discovery route ............................. 38 130 13 Aggregation Methodology ............................... 39 131 14 Data Forwarding ....................................... 39 132 14.1 MPLS Tree Encapsulation ............................... 39 133 14.1.1 Mapping multiple VPLS instances to a P2MP LSP ......... 39 134 14.1.2 Mapping one VPLS instance to a P2MP LSP ............... 40 135 15 VPLS Data Packet Treatment ............................ 41 136 16 Security Considerations ............................... 42 137 17 IANA Considerations ................................... 43 138 18 Acknowledgments ....................................... 43 139 19 Normative References .................................. 43 140 20 Informative References ................................ 44 141 21 Author's Address ...................................... 45 143 1. Specification of requirements 145 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 146 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 147 document are to be interpreted as described in [RFC2119]. 149 2. Contributors 151 Rahul Aggarwal 152 Yakov Rekhter 153 Juniper Networks 154 Yuji Kamite 155 NTT Communications 156 Luyuan Fang 157 AT&T 158 Chaitanya Kodeboniya 160 3. Terminology 162 This document uses terminology described in [RFC4761] and [RFC4762]. 164 4. Introduction 166 [RFC4761] and [RFC4762] describe a solution for VPLS multicast that 167 relies on ingress replication. This solution has certain limitations 168 for certain VPLS multicast traffic profiles. For example it may 169 result in highly non-optimal bandwidth utilization in the MPLS 170 network when large amount of multicast traffic is to be transported 172 This document describes procedures for overcoming the limitations of 173 existing VPLS multicast solutions. It describes procedures for VPLS 174 multicast that utilize multicast trees in the Sevice Provider (SP) 175 network. The procedures described in this document are applicable to 176 both [RFC4761] and [RFC4762]. 178 It provides mechanisms that allow a single multicast distribution 179 tree in the Service Provider (SP) network to carry all the multicast 180 traffic from one or VPLS sites connected to a given PE, irrespective 181 of whether these sites belong to the same or different VPLSes. Such a 182 tree is referred to as an "Inclusive tree" and more specifically as 183 an "Aggregate Inclusive tree" when the tree is used to carry 184 multicast traffic from more than one VPLS. 186 This document also provides procedures by which a single multicast 187 distribution tree in the SP network can be used to carry traffic 188 belonging only to a specified set of IP multicast streams, originated 189 in one or more VPLS sites connected to a given PE, irrespective of 190 whether these sites belong to the same or different VPLSes. Such a 191 tree is referred to as a "Selective tree" and more specifically as an 192 "Aggregate Selective tree" when the IP multicast streams belong to 193 different VPLSes. This allows multicast traffic, by default, to be 194 carried on an Inclusive tree, while traffic from some specific 195 multicast streams, e.g., high bandwidth streams, could be carried on 196 one of the "Selective trees". 198 5. Existing Limitations of VPLS Multicast 200 One of the limitations of existing VPLS multicast solutions described 201 in [RFC4761] and [RFC4762] is that they rely on ingress replication. 202 Thus the ingress PE replicates the multicast packet for each egress 203 PE and sends it to the egress PE using a unicast tunnel. 205 This may be an acceptable model when the bandwidth of the multicast 206 traffic is low or/and the number of replications performed on an 207 average on each outgoing interface for a particular customer VPLS 208 multicast packet is small. If this is not the case it is desirable to 209 utilize multicast trees in the SP network to transmit VPLS multicast 210 packets [MCAST-VPLS-REQ]. Note that unicast packets that are flooded 211 to each of the egress PEs, before the ingress PE learns the 212 destination MAC address of those unicast packets, MAY still use 213 ingress replication. 215 6. Overview 217 This document describes procedures for using multicast trees in the 218 SP network to transport VPLS multicast data packets. RSVP-TE P2MP 219 LSPs described in [RFC4875] are an example of such multicast trees. 220 The use of multicast trees in the SP network can be beneficial when 221 the bandwidth of the multicast traffic is high or when it is 222 desirable to optimize the number of copies of a multicast packet 223 transmitted on a given link. This comes at a cost of state in the SP 224 network to build multicast trees and overhead to maintain this state. 225 This document describes procedures for using multicast trees for VPLS 226 multicast when the provider tunneling technology is either P2MP RSVP- 227 PE or mLDP [MLDP]. The protocol architecture described herein is 228 considered to be flexible to support other P-tunneling technologies 229 as well. 231 This document uses the prefix 'C' to refer to the customer control or 232 data packets and 'P' to refer to the provider control or data 233 packets. An IP (multicast source, multicast group) tuple is 234 abbreviated to (S, G). 236 6.1. Inclusive and Selective Multicast Trees 238 Multicast trees used for VPLS can be of two types: 240 1. Inclusive trees. This option supports the use of a single 241 multicast distribution tree, referred to as an Inclusive P-Multicast 242 tree, in the SP network to carry all the multicast traffic from a 243 specified set of VPLS sites connected to a given PE. There is no 244 assumption made with respect to whether this traffic is IP 245 encapsulated or not. A particular P-Multicast tree can be set up to 246 carry the traffic originated by sites belonging to a single VPLS, or 247 to carry the traffic originated by sites belonging to different 248 VPLSes. The ability to carry the traffic of more than one VPLS on the 249 same tree is termed of the VPLSes that are using the tree. This 250 implies that a PE may receive multicast traffic for a multicast 251 stream even if it doesn't have any receivers that are interested in 252 receiving traffic for that stream. 254 An Inclusive P-Multicast tree as defined in this document is a P2MP 255 tree. A P2MP tree is used to carry traffic only for VPLS sites that 256 are connected to the PE that is the root of the tree. 258 2. Selective trees. A Selective P-Multicast tree is used by a PE 259 to send IP multicast traffic for one or IP more specific multicast 260 streams, originated within sites connected to the PE, that belong to 261 the same or different VPLSes, to a subset of the PEs that belong to 262 those VPLSes. Each of the PEs in the subset should be on the path to 263 a receiver of one or more multicast streams that are mapped onto the 264 tree. The ability to use the same tree for multicast streams that 265 belong to different VPLSes is termed a PE the ability to create 266 separate SP multicast trees for specific multicast streams, e.g. high 267 bandwidth multicast streams. This allows traffic for these multicast 268 streams to reach only those PE routers that have receivers in these 269 streams. This avoids flooding other PE routers in the VPLS. 271 A SP can use both Inclusive P-Multicast trees and Selective P- 272 Multicast trees or either of them for a given VPLS on a PE, based on 273 local configuration. Inclusive P-Multicast trees can be used for 274 both IP and non-IP data multicast traffic, while Selective P- 275 Multicast trees can be used only for IP multicast data traffic. 277 A variety of transport technologies may be used in the SP network. 278 For inclusive P-Multicast trees, these transport technologies include 279 point-to-multipoint LSPs created by RSVP-TE or mLDP. For selective P- 280 Multicast trees, only unicast PE-PE tunnels (using MPLS or IP/GRE 281 encapsulation) and P2MP LSPs are supported, and the supported P2MP 282 LSP signaling protocols are RSVP-TE, and mLDP. 284 This document also describes the data plane encapsulations for 285 supporting the various SP multicast transport options. 287 6.2. BGP-Based VPLS Membership Auto-Discovery 289 In order to establish Inclusive P-Multicast trees for one or more 290 VPLSes, when Aggregation is performed or when the tunneling 291 technology is P2MP RSVP-TE, the root of the tree must be able to 292 discover the other PEs that have membership in one or more of these 293 VPLSes. This document uses the BGP-based procedures described in 294 [RFC4761] and [L2VPN-SIG] for discovering the VPLS membership of all 295 PEs. 297 The leaves of the Inclusive P-Multicast trees must also be able to 298 auto-discover the identifier of the tree. This is described in 299 section 6.4. 301 6.3. IP Multicast Group Membership Discovery 303 The setup of a Selective P-Multicast tree for one or more IP 304 multicast (S, G)s, requires the ingress PE to learn the PEs that have 305 receivers in one or more of these (C-S, C-G)s, in the following 306 cases: 308 + When aggregation is used OR 310 + When the tunneling technology is P2MP RSVP-TE 312 + If ingress replication is used and the ingress PE wants to send 313 traffic for (C-S, C-G)s to only those PEs that are on the path to 314 receivers to the (C-S,C-G)s. 316 For discovering the IP multicast group membership, this document 317 describes procedures that allow an ingress PE to enable explicit 318 tracking. Thus an ingress PE can request the IP multicast membership 319 from egress PEs for one or more C-multicast streams. These procedures 320 are described in section "Optimizing Multicast Distribution via 321 Selective Trees". 323 These procedures are applicable when IGMP is used as the multicast 324 routing protocol between the VPLS CEs. They are also applicable when 325 PIM is used as the multicast routing protocol between the VPLS CEs 326 and PIM join suppression is disabled on all the CEs. However these 327 procedures do not apply when PIM is used as the multicast routing 328 protocol between the VPLS CEs and it not possible to disable PIM join 329 suppression on all the CEs. Procedures for this case are for further 330 study. 332 The leaves of the Selective P-Multicast trees must also be able to 333 discover the identifier of the tree. This is described in section 334 6.4. 336 6.4. Advertising P-Multicast Tree to VPLS/C-Multicast Binding 338 This document describes procedures based on BGP VPLS Auto-Discovery 339 (A-D) that are used by the root of an Aggregate P-Multicast tree to 340 advertise the Inclusive or Selective P-Multicast tree binding and the 341 de-multiplexing information to the leaves of the tree. This document 342 uses the PMSI Tunnel Attribute [BGP-MVPN] for this purpose. 344 Once a PE decides to bind a set of VPLSes or customer multicast 345 groups to an Inclusive P-Multicast tree or a Selective P-Multicast 346 tree, it needs to announce this binding to other PEs in the network. 347 This procedure is referred to as Inclusive P-Multicast tree or 348 Selective P-Multicast tree binding distribution and is performed 349 using BGP. 351 When an Aggregated Inclusive P-Multicast tree is used by an ingress 352 PE, this discovery implies that an ingress PE MUST announce the 353 binding of all VPLSes bound to the Inclusive P-Multicast tree to the 354 other PEs. The inner label assigned by the ingress PE for each VPLS 355 MUST be included, if more than one VPLS is bound to the same P- 356 Multicast tree. The Inclusive P-Multicast tree Identifier MUST be 357 included. 359 For a Selective P-Multicast tree this discovery implies announcing 360 all the specific entries bound to this P-Multicast tree 361 along with the Selective P-Multicast tree Identifier. The inner label 362 assigned for each MUST be included if s from 363 different VPLSes are bound to the same P-Multicast tree. The labels 364 MUST be distinct on a per VPLS basis and MAY be distinct on a per basis. The Selective P-Multicast tree Identifier MUST be 366 included. 368 6.5. Aggregation 370 As described above the ability to carry the traffic of more than one 371 VPLS on the same P-Multicast tree is termed 'Aggregation'. Both 372 Inclusive and Selective P-Multicast trees support aggregation. 374 Aggregation enables the SP to place a bound on the amount of 375 multicast tree forwarding and control plane state which the P routers 376 must have. Let us call the number of VPLSes aggregated onto a single 377 P-Multicast tree as the "Aggregation Factor". When Inclusive source 378 P-Multicast trees are used the number of trees that a PE is the root 379 of is proportional to: 381 + (Number of VPLSes on the PE / Aggregation Factor). 383 In this case the state maintained by a P router, is proportional to: 385 + ((Average number of VPLSes on a PE / Aggregation Factor) * number 386 of PEs) / (Average number of P-Multicast trees that transit a 387 given P router) 389 Thus the state does not grow linearly with the number of VPLSes. 391 Aggregation requires a mechanism for the egresses of the P-Multicast 392 tree to demultiplex the multicast traffic received over the P- 393 Multicast tree. This document describes how upstream-assigned labels 394 can be assigned and distributed by the root of aggregate P-Multicast 395 tree and then used by the egresses to perform this demultiplexing. 397 6.6. Inter-AS VPLS Multicast 399 This document supports three models of inter-AS VPLS service, option 400 (a), (b) and (c) which are very similar conceptually to option (a), 401 (b) and (c) specified in [RFC4364] for IP VPNs. The three options 402 described here are also similar to the three options described in in 403 [RFC4761], which in turn extends the concepts of [RFC4364] to inter- 404 AS VPLS. 406 For option (a) and option (b) support this document specifies a model 407 where Inter-AS VPLS service can be offered without requiring a single 408 P-Multicast tree to span multiple ASes. There are two variants of 409 this model and they are described in section 10. 411 For option (c) support this document specifies a model where Inter-AS 412 VPLS service is offered by requiring a single P-Multicast tree to 413 span multiple ASs. This is because in the case of option (c) the 414 ASBRs do not exchange BGP-VPLS NLRIs or A-D routes. 416 7. Intra-AS Inclusive P-Multicast Tree A-D/Binding 418 This section specifies procedures for the intra-AS auto-discovery (A- 419 D) of VPLS membership and the distribution of information used to 420 instantiate P-Multicast Tunnels. 422 VPLS auto-discovery/binding consists of two components: intra-AS and 423 inter-AS. The former provides VPLS auto-discovery/binding within a 424 single AS. The latter provides VPLS auto-discovery/binding across 425 multiple ASes. Inter-AS auto-discovery/binding is described in 426 section 10. 428 VPLS auto-discovery using BGP as described in [RFC4761, L2VPN-SIG] 429 enables a PE to learn the VPLS membership of other PEs. A PE that 430 belongs to a particular VPLS announces a BGP Network Layer 431 Reachability Information (NLRI) that identifies the Virtual Switch 432 Instance (VSI). This NLRI is constructed from the tuple. The 434 NLRI defined in [RFC4761] comprises the tuple and label 435 blocks for PW signaling. The VE-ID in this case is a two octet 436 number. The NLRI defined in [L2VPN-SIG] comprises only the where the VE-ID is a four octet number. 439 The procedures for constructing Inclusive intra-AS and inter-AS trees 440 as specified in this document require the BGP A-D NLRI to carry only 441 the . Hence these procedures can be used for both BGP-VPLS 442 and LDP-VPLS with BGP A-D. 444 It is to be noted that BGP A-D is an inherent feature of BGP-VPLS. 445 However it is not an inherent feature of LDP-VPLS. Infact there are 446 deployments and/or implementations of LDP-VPLS that require 447 configuration to enable a PE in a particular VPLS to determine other 448 PEs in the VPLS and exchange PW labels using FEC 128 [RFC4447]. The 449 use of BGP A-D for LDP-VPLS [L2VPN-SIG], to enable automatic setup of 450 PWs, requires FEC 129 [RFC4447]. However FEC 129 is not required in 451 order to use BGP A-D for the setup of P-Multicast trees for LDP-VPLS 452 as described in this document. An LDP-VPLS implementation that 453 supports P-Multicast trees described in this document, MUST support 454 the BGP A-D procedures to setup P-Multicast trees and it MAY support 455 FEC 129 to automate the signaling of PWs. 457 7.1. Originating intra-AS VPLS auto-discovery routes 459 To participate in the VPLS auto-discovery/binding a PE router that 460 has a given VSI of a given VPLS originates an BGP VPLS intra-AS auto- 461 discovery route and advertises this route in Multi-Protocol (MP) I- 462 BGP. The route is constructed as described in [RFC4761] and [L2VPN- 463 SIG]. 465 The route carries a single L2VPN NLRI with the RD set to the RD of 466 the VSI, and the VE-ID set to the VE-ID of the VSI. 468 If an Inclusive P-Multicast tree is used to instantiate the provider 469 tunnel for VPLS multicast on the PE, the advertising PE MUST 470 advertise the type and the identity of the P-Multicast tree in the 471 the PMSI Tunnel attribute [BGP-MVPN]. This attribute is described in 472 section 12.1. 474 A PE that uses an Inclusive P-Multicast tree to instantiate the 475 provider tunnel MAY aggregate two or more VPLSes present on the PE 476 onto the same tree. If the PE decides to perform aggregation after it 477 has already advertised the intra-AS VPLS auto-discovery routes for 478 these VPLSes, then aggregation requires the PE to re-advertise these 479 routes. The re-advertised routes MUST be the same as the original 480 ones, except for the PMSI Tunnel attribute. If the PE has not 481 previously advertised intra-AS auto-discovery routes for these 482 VPLSes, then the aggregation requires the PE to advertise (new) 483 intra-AS auto-discovery routes for these VPLSes. The P-Tunnel 484 attribute in the newly advertised/re-advertised routes MUST carry the 485 identity of the P-Multicast tree that aggregates the VPLSes, as well 486 as an MPLS upstream-assigned label [RFC5331]. Each re-advertised 487 route MUST have a distinct label. 489 Discovery of PE capabilities in terms of what tunnels types they 490 support is outside the scope of this document. Within a given AS PEs 491 participating in a VPLS are expected to advertise tunnel bindings 492 whose tunnel types are supported by all other PEs that are 493 participating in this VPLS and are part of the same AS. 495 7.2. Receiving intra-AS VPLS auto-discovery routes 497 When a PE receives a BGP Update message that carries an intra-AS A-D 498 route such that (a) the route was originated by some other PE within 499 the same AS as the local PE, (b) at least one of the Route Targets of 500 the route matches one of the import Route Targets configured for a 501 particular VSI on the local PE, (c) the BGP route selection 502 determines that this is the best route with respect to the NLRI 503 carried by the route, and (d) the route carries the PMSI Tunnel 504 attribute, the PE performs the following. 506 If the route carries the PMSI Tunnel attribute then: 508 + If the Tunnel Type in the PMSI Tunnel attribute is set to LDP 509 P2MP LSP, the PE SHOULD join the P-Multicast tree whose identity 510 is carried in the PMSI Tunnel Attribute. 512 + If the Tunnel Type in the PMSI Tunnel attribute is set to RSVP-TE 513 P2MP LSP, the receiving PE has to establish the appropriate state 514 to properly handle the traffic received over that LSP. The PE 515 that originated the route MUST establish an RSVP-TE P2MP LSP with 516 the local PE as a leaf. This LSP MAY have been established before 517 the local PE receives the route. 519 + If the PMSI Tunnel attribute does not carry a label, then all 520 packets that are received on the P-Multicast tree, as identified 521 by the PMSI Tunnel attribute, are forwarded using the VSI that 522 has at least one of its import Route Targets that matches one of 523 the Route Targets of the received auto-discovery route. 525 + If the PMSI Tunnel attribute has the Tunnel Type set to LDP P2MP 526 LSP or RSVP-TE P2MP LSP, and the attribute also carries an MPLS 527 label, then the egress PE MUST treat this as an upstream-assigned 528 label, and all packets that are received on the P-Multicast tree, 529 as identified by the PMSI Tunnel attribute, with that upstream 530 label are forwarded using the VSI that has at least one of its 531 import Route Target that matches one of the Route Targets of the 532 received intra-AS auto-discovery route. 534 If the local PE uses RSVP-TE P2MP LSP for sending (multicast) 535 traffic, originated by VPLS sites connected to the PE, to the sites 536 attached to other PEs then the local PE MUST use the Originating 537 Router's IP address information carried in the intra-AS A-D route to 538 add the PE, that originated the route, as a leaf node to the LSP. 539 This MUST be done irrespective of whether the received Intra-AS A-D 540 route carries the PMSI Tunnel attribute or not. 542 8. Demultiplexing P-Multicast Tree Traffic 544 Demultiplexing received VPLS traffic requires the receiving PE to 545 determine the VPLS instance the packet belongs to. The egress PE can 546 then perform a VPLS lookup to further forward the packet. It also 547 requires the egress PE to determine the identity of the ingress PE 548 for MAC learning, as described in section 15. 550 8.1. One P-Multicast Tree - One VPLS Mapping 552 When a P-Multicast tree is mapped to only one VPLS, determining the 553 tree on which the packet is received is sufficient to determine the 554 VPLS instance on which the packet is received. The tree is determined 555 based on the tree encapsulation. If MPLS encapsulation is used, eg: 556 RSVP-TE P2MP LSPs, the outer MPLS label is used to determine the 557 tree. Penultimate-hop-popping MUST be disabled on the MPLS LSP (RSVP- 558 TE P2MP LSP or LDP P2MP LSP). 560 8.2. One P-Multicast Tree - Many VPLS Mapping 562 As traffic belonging to multiple VPLSes can be carried over the same 563 tree, there is a need to identify the VPLS the packet belongs to. 564 This is done by using an inner label that determines to the VPLS for 565 which the packet is intended. The ingress PE uses this label as the 566 inner label while encapsulating a customer multicast data packet. 567 Each of the egress PEs must be able to associate this inner label 568 with the same VPLS and use it to demultimplex the traffic received 569 over the Aggregate Inclusive tree or the Aggregate Selective tree. 571 This document requires the use of upstream label assignment by the 572 ingress PE [RFC5331]. Hence the inner label is assigned by the 573 ingress PE. When the egress PE receives a packet over an Aggregate 574 tree, the outer encapsulation [in the case of MPLS P2MP LSPs, the 575 outer MPLS label] specifies the label space to perform the inner 576 label lookup. The same label space MUST be used by the egress PE for 577 all P-Multicast trees that have the same root [RFC5331]. 579 If the tree uses MPLS encapsulation, as in RSVP-TE P2MP LSPs, the 580 outer MPLS label and the incoming interface provides the label space 581 of the label beneath it. This assumes that penultimate-hop-popping is 582 disabled. The egress PE MUST NOT advertise IMPLICIT NULL or EXPLICIT 583 NULL for that tree. Once the label representing the tree is popped 584 off the MPLS label stack, the next label is the demultiplexing 585 information that allows the proper MVPN to be determined. 587 The ingress PE informs the egress PEs about the inner label as part 588 of the tree binding procedures described in section 12. 590 9. Establishing P-Multicast Trees 592 This document does not place any fundamental restrictions on the 593 multicast technology used to setup P-Multicast trees. However 594 specific procedures are specified only for RSVP-TE P2MP LSPs and LDP 595 P2MP LSPs. An implementation that supports this document MUST support 596 RSVP-TE P2MP LSPs and LDP P2MP LSPs. 598 The P-Multicast trees supported in this document are P2MP trees. A 599 P2MP tree is used to carry traffic originated in sites connected to 600 the PE which is the root of the tree, irrespective of whether these 601 sites belong to the same or different VPLSes. 603 9.1. Common Procedures 605 The following procedures apply to both RSVP-TE P2MP and LDP P2MP 606 LSPs. 608 Demultiplexing the C-multicast data packets at the egress PE requires 609 that the PE must be able to determine the P2MP LSP that the packets 610 are received on. This enables the egress PE to determine the VPLS 611 that the packet belongs to. To achieve this the LSP MUST be signaled 612 with penultimate-hop-popping (PHP) off as described in section 8. In 613 other words an egress PE MUST NOT advertise IMPLICIT NULL or EXPLICIT 614 NULL for a P2MP LSP that is carrying traffic for one or more VPLSes. 615 This is because the egress PE needs to rely on the MPLS label, that 616 it advertises to its upstream neighbor, to determine the P2MP LSP 617 that a C-multicast data packet is received on. 619 The egress PE also needs to identify the ingress PE to perform MAC 620 learning. When P2MP LSPs are used as P2MP trees, determining the 621 P2MP LSP that the packets are received on, is sufficient to determine 622 the ingress PE. This is because the ingress PE is the root of the 623 P2MP LSP. 625 The egress PE relies on receiving the PMSI Tunnel Attribute in BGP to 626 determine the VPLS instance to P2MP LSP mapping. 628 9.2. RSVP-TE P2MP LSPs 630 This section describes procedures that are specific to the usage of 631 RSVP-TE P2MP LSPs for instantiating a P-Multicast tree. Procedures in 632 [RFC4875] are used to signal the P2MP LSP. The LSP is signaled after 633 the root of the P2MP LSP discovers the leaves. The egress PEs are 634 discovered using the procedures described in section 7. Aggregation 635 as described in this document is supported. 637 9.2.1. P2MP TE LSP - VPLS Mapping 639 P2MP TE LSP to VPLS mapping is learned at the egress PEs using BGP 640 based advertisements of the P2MP TE LSP - VPLS mapping. They require 641 that the root of the tree include the P2MP TE LSP identifier as the 642 tunnel identifier in the BGP advertisements. This identifier contains 643 the following information elements: 644 - The type of the tunnel is set to RSVP-TE P2MP LSP 645 - RSVP-TE P2MP LSP's SESSION Object 647 This Tunnel Identifier is described in section 12.1. 649 Once the egress PE receives the P2MP TE LSP to VPLS mapping: 651 + If the egress PE already has RSVP-TE state for the P2MP TE LSP, 652 it MUST begin to assign a MPLS label from the non-reserved label 653 range, for the P2MP TE LSP and signal this to the previous hop of 654 the P2MP TE LSP. Further it MUST create forwarding state to 655 forward packets received on the P2MP LSP. 657 + If the egress PE does not have RSVP-TE state for the P2MP TE LSP, 658 it MUST retain this mapping. Subsequently when the egress PE 659 receives the RSVP-TE P2MP signaling message, it creates the RSVP- 660 TE P2MP LSP state. It MUST then assign a MPLS label from the 661 non-reserved label range, for the P2MP TE LSP, and signal this to 662 the previous hop of the P2MP TE LSP. 664 Note that if the signaling to set up an RSVP-TE P2MP LSPis 665 completed before a given egress PE learns, via a PMSI Tunnel 666 attribute, of the VPLS or set of VPLSes to which the LSP is 667 bound, the PE MUST discard any traffic received on that LSP until 668 the binding is received. In order for the egress PE to be able to 669 discard such traffic it needs to know that the LSP is associated 670 with one or more VPLSes and that the VPLS A-D route that binds 671 the LSP to a VPLS has not yet been received. This is provided by 672 extending [RFC4875] with [RSVP-OBB]. 674 9.3. Receiver Initiated MPLS Trees 676 Receiver initiated P2MP MPLS trees signaled using LDP [mLDP] can also 677 be used. Procedures in [MLDP] MUST be used to signal the P2MP LSP. 678 The LSP is signaled once the leaves receive the LDP FEC for the tree 679 from the root as described in section 7. An ingress PE is required to 680 discover the egress PEs when aggregation is used and this is achieved 681 using the procedures in section 7. 683 9.3.1. P2MP LSP - VPLS Mapping 685 P2MP LSP to VPLS mapping is learned at the egress PEs using BGP based 686 advertisements of the P2MP LSP - VPLS mapping. They require that the 687 root of the tree include the P2MP LSP identifier as the tunnel 688 identifier in the BGP advertisements. This identifier contains the 689 following information elements: 690 - The type of the tunnel is set to LDP P2MP LSP 691 - LDP P2MP FEC which includes an identifier generated by the 692 root. 694 Each egress PE SHOULD "join" the P2MP MPLS tree by sending LDP label 695 mapping messages for the LDP P2MP FEC, that was learned in the BGP 696 advertisement, using procedures described in [MLDP]. 698 9.4. Encapsulation of Aggregate P-Multicast Trees 700 An Aggregate Inclusive P-Multicast tree or an Aggregate Selective P- 701 Multicast tree MUST use a MPLS encapsulation. The protocol type in 702 the data link header is as described in [RFC5332]. 704 10. Inter-AS Inclusive P-Multicast Tree A-D/Binding 706 This document supports three models of inter-AS VPLS service, option 707 (a), (b) and (c) which are very similar conceptually to option (a), 708 (b) and (c) specified in [RFC4364] for IP VPNs. The three options 709 described here are also similar to the three options described in 710 [RFC4761], which in turn extends the concepts of [RFC4364] to inter- 711 AS VPLS. An implementation MUST support all three of these models. 712 When there are multiple options for implementing one of these models, 713 this section specifies which option is mandatory. 715 For option (a) and option (b) support this section specifies a model 716 where inter-AS VPLS service can be offered without requiring a single 717 P-Multicast tree to span multiple ASes. This allows individual ASes 718 to potentially use different P-tunneling technologies. There are two 719 variants of this model. One that requires MAC lookup on the ASBRs and 720 another that does not require MAC lookup on the ASBRs and instead 721 builds segmented inter-AS trees. This applies to both Inclusive and 722 Selective trees. 724 For option (c) support this document specifies a model where Inter-AS 725 VPLS service is offered by requiring a single Inclusive P-Multicast 726 tree to span multiple ASs. This is referred to as a non-segmented P- 727 Multicast tree. This is because in the case of option (c) the ASBRs 728 do not exchange BGP-VPLS NLRIs or VPLS A-D routes. Selective inter-AS 729 trees for option (c) support may be segmented or non-segmented. 731 10.1. VSIs on the ASBRs 733 In this variant, the ASBRs MUST perform a MAC lookup, in addition to 734 any MPLS lookups, to determine the forwarding decision on a VPLS 735 packet. The P-Multicast trees are confined to an AS. An ASBR on 736 receiving a VPLS packet from another ASBR is required to perform a 737 MAC lookup to determine how to forward the packet. Thus an ASBR is 738 required to keep a VSI for the VPLS and MUST be configured with its 739 own VE ID for the VPLS. In this variant the BGP VPLS A-D routes 740 generated by PEs in an AS MUST NOT be propagated outside the AS. 742 10.1.1. Option (a) 744 When this variant is used with option (a) an ASBR in one AS treats an 745 adjoining ASBR in another AS as a CE and determines the VSI for 746 packets received from another ASBR based on the incoming ethernet 747 interface. In the case of option (a) the ASBRs do not exchange VPLS 748 A-D routes. 750 An implementation MUST support this variant for option (a). 752 10.1.2. Option (e) 754 The VSIs on the ASBRs variant can be used such that the interconnect 755 between the ASBRs is a PW and MPLS encapsulation is used between the 756 ASBRs. An ASBR in one AS treats an adjoining ASBR in another AS as a 757 CE and determines the VSI for packets received from another ASBR 758 based on the incoming MPLS encapsulation. The only VPLS A-D routes 759 that are propagated outside the AS are the ones originated by ASBRs. 760 This MPLS PW connects the VSIs on the ASBRs and MUST be signaled 761 using the procedures defined in [RFC4761] or [RFC4762]. 763 The P-Multicast trees for a VPLS are confined to each AS and the VPLS 764 auto-discovery/binding MUST follow the intra-AS procedures described 765 in section 8. An implementation MAY support option (e). 767 10.2. Option (b) - Segmented Inter-AS Trees 769 In this variant, an inter-AS P-Multicast tree, rooted at a particular 770 PE for a particular VPLS instance, consists of a number of 771 "segments", one per AS, which are stitched together at ASBRs. These 772 are known as "segmented inter-AS trees". Each segment of a segmented 773 inter-AS tree may use a different multicast transport technology. In 774 this variant, an ASBR is not required to keep a VSI for the VPLS and 775 is not required to perform a MAC lookup in order to forward the VPLS 776 packet. This implies that an ASBR is not required to be configured 777 with a VE ID for the VPLS. This variant is applicable to option (b). 778 An implementation MUST support this variant. 780 The construction of segmented Inter-AS trees requires the BGP-VPLS A- 781 D NLRI described in [RFC4761, RFC4762]. A BGP VPLS A-D route for a 782 tuple advertised outside the AS, to which the originating 783 PE belongs, will be referred to as an inter-AS VPLS auto-discovery 784 route (Though this route is originated by a PE as an intra-AS route 785 and is referred to as an inter-AS route outside the AS). 787 In addition to this, segmented inter-AS trees require support for the 788 PMSI Tunnel Attribute described in section 12.1. They also require 789 additional procedures in BGP to signal leaf A-D routes between ASBRs 790 as explained in subsequent sections. 792 10.2.1. Segmented Inter-AS Trees VPLS Inter-AS A-D/Binding 794 This section specifies the procedures for inter-AS VPLS A-D/binding 795 for segmented inter-AS trees. 797 An ASBR must be configured to support a particular VPLS as follows: 799 + An ASBR MUST be be configured with a set of (import) Route 800 Targets (RTs) that specifies the set of VPLSes supported by the 801 ASBR. These Route Targets control acceptance of BGP VPLS auto- 802 discovery routes by the ASBR. Note that instead of being 803 configured, the ASBR MAY obtain this set of (import) Route 804 Targets (RTs) by using Route Target Constrain [RFC4684]. 806 + The ASBR MUST be configured with the tunnel types for the intra- 807 AS segments of the VPLSes supported by the ASBR, as well as 808 (depending on the tunnel type) the information needed to create 809 the PMSI Tunnel attribute for these tunnel types. Note that 810 instead of being configured, the ASBR MAY derive the tunnel types 811 from the intra-AS auto-discovery routes received by the ASBR from 812 the PEs in its own AS. 814 If an ASBR is configured to support a particular VPLS, the ASBR MUST 815 participate in the intra-AS VPLS auto-discovery/binding procedures 816 for that VPLS within the ASBR's own AS, as defined in this document. 818 Moreover, in addition to the above the ASBR performs procedures 819 specified in the next section. 821 10.2.2. Propagating BGP VPLS A-D routes to other ASes: Overview 823 An auto-discovery route for a given VPLS, originated by an ASBR 824 within a given AS, is propagated via BGP to other ASes. The precise 825 rules for distributing and processing the inter-AS auto-discovery 826 routes are given in subsequent sections. 828 Suppose that an ASBR A receives and installs an auto-discovery route 829 for VPLS "X" and VE ID "V" that originated at a particular PE, PE1. 830 The BGP next hop of that received route becomes A's "upstream 831 neighbor" on a multicast distribution tree for (X, V) that is rooted 832 at PE1. When the auto-discovery routes have been distributed to all 833 the necessary ASes, they define a "reverse path" from any AS that 834 supports VPLS X and VE ID V back to PE1. For instance, if AS2 835 supports VPLS X, then there will be a reverse path for VPLS X and VE 836 ID V from AS2 to AS1. This path is a sequence of ASBRs, the first of 837 which is in AS2, and the last of which is in AS1. Each ASBR in the 838 sequence is the BGP next hop of the previous ASBR in the sequence on 839 the given auto-discovery route. 841 This reverse path information can be used to construct a 842 unidirectional multicast distribution tree for VPLS X and VE ID V, 843 containing all the ASes that support X, and having PE1 at the root. 844 We call such a tree an "inter-AS tree". Multicast data originating in 845 VPLS sites for VPLS X connected to PE1 will travel downstream along 846 the tree which is rooted at PE1. 848 The path along an inter-AS tree is a sequence of ASBRs. It is still 849 necessary to specify how the multicast data gets from a given ASBR to 850 the set of ASBRs which are immediately downstream of the given ASBR 851 along the tree. This is done by creating "segments": ASBRs in 852 adjacent ASes will be connected by inter-AS segments, ASBRs in the 853 same AS will be connected by "intra-AS segments". 855 For a given inter-AS tree and a given AS there MUST be only one ASBR 856 within that AS that accepts traffic flowing on that tree. Further for 857 a given inter-AS tree and a given AS there MUST be only one ASBR in 858 that AS that sends the traffic flowing on that tree to a particular 859 adjacent AS. The precise rules for accomplishing this are given in 860 subsequent sections. 862 An ASBR initiates creation of an intra-AS segment when the ASBR 863 receives an inter-AS auto-discovery route from an E-BGP neighbor. 864 Creation of the segment is completed as a result of distributing, via 865 I-BGP, this route within the ASBR's own AS. 867 For a given inter-AS tunnel each of its intra-AS segments could be 868 constructed by its own independent mechanism. Moreover, by using 869 upstream-assigned labels within a given AS multiple intra-AS segments 870 of different inter-AS tunnels of either the same or different VPLSes 871 may share the same P-Multicast tree. 873 If the P-Multicast tree instantiating a particular segment of an 874 inter-AS tunnel is created by a multicast control protocol that uses 875 receiver-initiated joins (e.g, mLDP), and this P-Multicast tree does 876 not aggregate multiple segments, then all the information needed to 877 create that segment will be present in the inter-AS auto-discovery 878 routes received by the ASBR from the neighboring ASBR. But if the P- 879 Multicast tree instantiating the segment is created by a protocol 880 that does not use receiver-initiated joins (e.g., RSVP-TE, ingress 881 unicast replication), or if this P-Multicast tree aggregates multiple 882 segments (irrespective of the multicast control protocol used to 883 create the tree), then the ASBR needs to learn the leaves of the 884 segment. These leaves are learned from A-D routes received from other 885 PEs in the AS, for the same VPLS (i.e. same VE-ID) as the one that 886 the segment belongs to. 888 The following sections specify procedures for propagation of inter-AS 889 auto-discovery routes across ASes in order to construct inter-AS 890 segmented trees. 892 10.2.2.1. Propagating Intra-AS VPLS A-D routes in E-BGP 894 For a given VPLS configured on an ASBR when the ASBR receives intra- 895 AS A-D routes originated PEs in its own AS, the ASBR MUST propagate 896 each of these route in E-BGP. This procedure MUST be performed for 897 each of the VPLSes configurd on the ASBR. Each of these routes is 898 constructed as follows: 900 + The route carries a single BGP VPLS A-D NLRI with the RD and VE 901 ID being the same as the NLRI in the received intra-AS A-D route. 903 + The Next Hop field of the MP_REACH_NLRI attribute is set to a 904 routable IP address of the ASBR. 906 + The route carries the PMSI Tunnel attribute with the Tunnel Type 907 set to Ingress Replication; the attribute carries no MPLS labels. 909 + The route MUST carry the export Route Target used by the VPLS. 911 10.2.2.2. Inter-AS A-D route received via E-BGP 913 When an ASBR receives from one of its E-BGP neighbors a BGP Update 914 message that carries an inter-AS auto-discovery route, if (a) at 915 least one of the Route Targets carried in the message matches one of 916 the import Route Targets configured on the ASBR, and (b) the ASBR 917 determines that the received route is the best route to the 918 destination carried in the NLRI of the route, the ASBR re-advertises 919 this inter-AS auto-discovery route to other PEs and ASBRs within its 920 own AS. The best route selection procedures MUST ensure that for the 921 same destination, all ASBRs in an AS pick the same route as the best 922 route. The best route selection procedures are specified in 923 [RFC4761] and clarified in [MULTI-HOMING]. The best route procedures 924 ensure that if multiple ASBRs, in an AS, receive the same inter-AS A- 925 D route from their E-BGP neighbors, only one of these ASBRs 926 propagates this route in I-BGP. This ASBR becomes the root of the 927 intra-AS segment of the inter-AS tree and ensures that this is the 928 only ASBR that accepts traffic into this AS from the inter-AS tree. 930 When re-advertising an inter-AS auto-discovery route the ASBR MUST 931 set the Next Hop field of the MP_REACH_NLRI attribute to a routable 932 IP address of the ASBR. 934 Depending on the type of a P-Multicast tree used to instantiate the 935 intra-AS segment of the inter-AS tunnel, the PMSI Tunnel attribute of 936 the re-advertised inter-AS auto-discovery route is constructed as 937 follows: 939 + If the ASBR uses ingress replication to instantiate the intra-AS 940 segment of the inter-AS tunnel, the re-advertised route MUST NOT 941 carry the PMSI Tunnel attribute. 943 + If the ASBR uses a P-Multicast tree to instantiate the intra-AS 944 segment of the inter-AS tunnel, the PMSI Tunnel attribute MUST 945 contain the identity of the tree that is used to instantiate the 946 segment (note that the ASBR could create the identity of the tree 947 prior to the actual instantiation of the segment). If in order to 948 instantiate the segment the ASBR needs to know the leaves of the 949 tree, then the ASBR obtains this information from the auto- 950 discovery routes received from other PEs/ASBRs in ASBR's own AS. 952 + An ASBR that uses a P-Multicast tree to instantiate the intra-AS 953 segment of the inter-AS tunnel MAY aggregate two or more VPLSes 954 present on the ASBR onto the same tree. If the ASBR already 955 advertises inter-AS auto-discovery routes for these VPLSes, then 956 aggregation requires the ASBR to re-advertise these routes. The 957 re-advertised routes MUST be the same as the original ones, 958 except for the PMSI Tunnel attribute. If the ASBR has not 959 previously advertised inter-AS auto-discovery routes for these 960 VPLSes, then the aggregation requires the ASBR to advertise (new) 961 inter-AS auto-discovery routes for these VPLSes. The PMSI Tunnel 962 attribute in the newly advertised/re-advertised routes MUST carry 963 the identity of the P-Multicast tree that aggregates the VPLSes, 964 as well as an MPLS upstream-assigned label [RFC5331]. Each re- 965 advertised route MUST have a distinct label. 967 In addition the ASBR MUST send to the E-BGP neighbor, from whom it 968 receives the inter-AS auto-discovery route, a BGP Update message that 969 carries a "leaf auto-discovery route". The exact encoding of this 970 route is described in section 12. This route contains the following 971 information elements: 973 + The route carries a single NLRI with the Route Key field set to 974 the tuple of the BGP VPLS A-D NLRI of the inter-AS 975 auto-discovery route received from that neighbor. The NLRI also 976 carries the IP address of the ASBR (this MUST be a routable IP 977 address). 979 + The leaf auto-discovery route MUST include the PMSI Tunnel 980 attribute with the Tunnel Type set to Ingress Replication, and 981 the Tunnel Identifier set to a routable address of the 982 advertising router. The PMSI Tunnel attribute MUST carry a 983 downstream assigned MPLS label that is used to demultiplex the 984 VPLS traffic received over a unicast tunnel by the advertising 985 router. 987 + The Next Hop field of the MP_REACH_NLRI attribute of the route 988 SHOULD be set to the same IP address as the one carried in the 989 Originating Router's IP Address field of the route. 991 + To constrain the distribution scope of this route the route MUST 992 carry the NO_ADVERTISE BGP community ([RFC1997]). 994 + The ASBR constructs an IP-based Route Target extended community 995 by placing the IP address carried in the next hop of the received 996 Inter-AS VPLS A-D route in the Global Administrator field of the 997 community, with the Local Administrator field of this community 998 set to 0, and sets the Extended Communities attribute of the Leaf 999 A-D route to that community. Note that this Route Target is the 1000 same as the ASBR Import RT of the EBGP neighbor from which the 1001 ASBR received the inter-AS VPLS A-D route. 1003 10.2.2.3. Leaf A-D Route received via E-BGP 1005 When an ASBR receives via E-BGP a leaf auto-discovery route, the ASBR 1006 accepts the route only if if (a) at least one of the Route Targets 1007 carried in the message matches one of the import Route Targets 1008 configured on the ASBR, and (b) the ASBR determines that the received 1009 route is the best route to the destination carried in the NLRI of the 1010 route. 1012 If the ASBR accepts the leaf auto-discovery route, the ASBR finds an 1013 auto-discovery route whose BGP-VPLS A-D NLRI has the same value as 1014 the field of the the leaf auto-discovery route. 1016 The MPLS label carried in the PMSI Tunnel attribute of the leaf auto- 1017 discovery route is used to stitch a one hop ASBR-ASBR LSP to the tail 1018 of the intra-AS tunnel segment associated with the found auto- 1019 discovery route. 1021 10.2.2.4. Inter-AS A-D Route received via I-BGP 1023 In the context of this section we use the term "PE/ASBR router" to 1024 denote either a PE or an ASBR router. 1026 Note that a given inter-AS auto-discovery route is advertised within 1027 a given AS by only one ASBR as described above. 1029 When a PE/ASBR router receives from one of its I-BGP neighbors a BGP 1030 Update message that carries an inter-AS auto-discovery route, if (a) 1031 at least one of the Route Targets carried in the message matches one 1032 of the import Route Targets configured on the PE/ASBR, and (b) the 1033 PE/ASBR determines that the received route is the best route to the 1034 destination carried in the NLRI of the route, the PE/ASBR performs 1035 the following operations. The best route determination is based as 1036 described in [RFC4761] and clarified in [MULTI-HOMING]. 1038 If the router is an ASBR then the ASBR propagates the route to its E- 1039 BGP neighbors. When propagating the route to the E-BGP neighbors the 1040 ASBR MUST set the Next Hop field of the MP_REACH_NLRI attribute to a 1041 routable IP address of the ASBR. 1043 If the received inter-AS auto-discovery route carries the PMSI Tunnel 1044 attribute with the Tunnel Type set to LDP P2MP LSP, the PE/ASBR 1045 SHOULD join the P-Multicast tree whose identity is carried in the 1046 PMSI Tunnel Attribute. 1048 If the received inter-AS auto-discovery route carries the PMSI Tunnel 1049 attribute with the Tunnel Identifier set to RSVP-TE P2MP LSP, then 1050 the ASBR that originated the route MUST establish an RSVP-TE P2MP LSP 1051 with the local PE/ASBRas a leaf. This LSP MAY have been established 1052 before the local PE/ASBR receives the route, or MAY be established 1053 after the local PE receives the route. 1055 If the received inter-AS auto-discovery route carries the PMSI Tunnel 1056 attribute with the Tunnel Type set to LDP P2MP LSP, or RSVP-TE P2MP 1057 LSP, but the attribute does not carry a label, then the P-Multicast 1058 tree, as identified by the PMSI Tunnel Attribute, is an intra-AS LSP 1059 segment that is part of the inter-AS Tunnel for the 1060 advertised by the inter-AS auto-discovery route and rooted at the PE 1061 that originated the auto-discovery route. If the PMSI Tunnel 1062 attribute carries a (upstream-assigned) label, then a combination of 1063 this tree and the label identifies the intra-AS segment. If the 1064 received router is an ASBR, this intra-AS segment may further be 1065 stitched to ASBR-ASBR inter-AS segment of the inter-AS tunnel. If the 1066 PE/ASBR has local receivers in the VPLS, packets received over the 1067 intra-AS segment must be forwarded to the local receivers using the 1068 local VSI. 1070 10.3. Option (c) 1072 In this method, there is a multi-hop E-BGP peering between the PEs 1073 (or a Route Reflector) in one AS and the PEs (or Route Reflector) in 1074 another AS. The PEs exchange BGP-VPLS NLRI or BGP-VPLS A-D NLRI, 1075 along with PMSI Tunnel Attribute, as in the intra-AS case described 1076 in section 8. An implementation MUST support this method. 1078 The PEs in different ASs use a non-segmented inter-AS P2MP tunnel for 1079 VPLS multicast. A non-segmented inter-AS tunnel is a single tunnel 1080 which spans AS boundaries. The tunnel technology cannot change from 1081 one point in the tunnel to the next, so all ASes through which the 1082 tunnel passes must support that technology. In essence, AS boundaries 1083 are of no significance to a non-segmented inter-AS P2MP tunnel. 1085 This method requires no VPLS A-D routes in the control or VPLS MAC 1086 address learning in the data plane on the ASBRs. The ASBRs only need 1087 to participate in the non-segmented P2MP tunnel setup in the control 1088 plane, and do MPLS label forwarding in the data plane. 1090 The setup of non-segmented inter-AS P2MP tunnels MAY require the P- 1091 routers in one AS to have IP reachability to the loopback addresses 1092 of the PE routers in another AS, depending on the tunneling 1093 technology chosen. If this is the case, reachability to the loopback 1094 addresses of PE routers in one AS MUST be present in the IGP in 1095 another AS. 1097 The data forwarding in this model is the same as in the intra-AS case 1098 described in section 8. 1100 11. Optimizing Multicast Distribution via Selective Trees 1102 Whenever a particular multicast stream is being sent on an Inclusive 1103 P-Multicast tree, it is likely that the data of that stream is being 1104 sent to PEs that do not require it as the sites connected to these 1105 PEs may have no receivers for the stream. If a particular stream has 1106 a significant amount of traffic, it may be beneficial to move it to a 1107 Selective P-Multicast tree which has at its leaves only those PEs, 1108 connected to sites that have receivers for the multicast stream (or 1109 at least includes fewer PEs that are attached to sites with no 1110 receivers compared to an Inclusive tree). 1112 A PE connected to the multicast source of a particular multicast 1113 stream may be performing explicit tracking i.e. it may know the PEs 1114 that have receivers in the multicast stream. Section 11.2.1 describes 1115 procedures that enable explicit tracking. If this is the case 1116 Selective P-Multicast trees can also be triggered on other criteria. 1117 For instance there could be a "pseudo wasted bandwidth" criteria: 1118 switching to a Selective tree would be done if the bandwidth 1119 multiplied by the number of uninterested PEs (PE that are receiving 1120 the stream but have no receivers) is above a specified threshold. The 1121 motivation is that (a) the total bandwidth wasted by many sparsely 1122 subscribed low-bandwidth groups may be large, and (b) there's no 1123 point to moving a high-bandwidth group to a Selective tree if all the 1124 PEs have receivers for it. 1126 Switching a (C-S, C-G) stream to a Selective P-Multicast tree may 1127 require the root of the tree to determine the egress PEs that need to 1128 receive the (C-S, C-G) traffic. This is true in the following cases: 1130 + If the tunnel is a P2MP tree, such as a RSVP-TE P2MP Tunnel, the 1131 PE needs to know the leaves of the tree before it can instantiate 1132 the Selective tree. 1134 + If a PE decides to send traffic for multicast streams, belonging 1135 to different VPLSes, using one P-Multicast Selective tree, such a 1136 tree is termed an Aggregate tree with a selective mapping. The 1137 setting up of such an Aggregate tree requires the ingress PE to 1138 know all the other PEs that have receivers for multicast groups 1139 that are mapped onto the tree. 1141 + If ingress replication is used and the ingress PE wants to send 1142 traffic for (C-S, C-G)s to only those PEs that are on the path to 1143 receivers to the (C-S,C-G)s. 1145 For discovering the IP multicast group membership, for the above 1146 cases, this document describes procedures that allow an ingress PE to 1147 enable explicit tracking. Thus an ingress PE can request the IP 1148 multicast membership from egress PEs for one or more C-multicast 1149 streams. These procedures are described in section 11.2.1. 1151 These procedures are applicable when IGMP is used as the multicast 1152 routing protocol between the VPLS CEs. They are also applicable when 1153 PIM is used as the multicast routing protocol between the VPLS CEs 1154 and PIM join suppression is disabled on all the CEs. However these 1155 procedures do not apply when PIM is used as the multicast routing 1156 protocol between the VPLS CEs and it not possible to disable PIM join 1157 suppression on all the CEs. Procedures that allow the setup of 1158 Selective trees for this case are for further study. 1160 The root of the Selective P-Multicast tree MAY decide to do explicit 1161 tracking of the IP multicast stream only after it has determined to 1162 move the stream to a Selective tree, or it MAY have been doing 1163 explicit tracking all along. This document also describes explicit 1164 tracking for a wild-card source and/or group in section 11.2.1, which 1165 facilitates a Selective P-Multicast tree only mode in which IP 1166 multicast streams are always carried on a Selective P-Multicast tree. 1167 In the description on Selective P-Multicast trees the notation C-S, 1168 is intended to representeither a specific source address or a 1169 wildcard. Similarly C-G is intended to represent either a specific 1170 group address or a wildcard. 1172 The PE at the root of the tree MUST signal the leaves of the tree 1173 that the (C-S, C-G) stream is now bound to the to the Selective Tree. 1175 Note that the PE could create the identity of the P-Multicast tree 1176 prior to the actual instantiation of the tunnel. 1178 If the Selective tree is instantiated by a RSVP-TE P2MP LSP the PE at 1179 the root of the tree MUST establish the P2MP RSVP-TE LSP to the 1180 leaves. This LSP MAY have been established before the leaves receive 1181 the Selective tree binding, or MAY be established after the leaves 1182 receives the binding. A leaf MUST not switch to the Selective tree 1183 until it receives the binding and the RSVP-TE P2MP LSP is setup to 1184 the leaf. 1186 11.1. Protocol for Switching to Selective Trees 1188 Selective trees provide a PE the ability to create separate P- 1189 Multicast trees for certain streams. The source PE, that 1190 originates the Selective tree, and the egress PEs, MUST use the 1191 Selective tree for the streams that are mapped to it. This 1192 may require the source and egress PEs to switch to the Selective tree 1193 from an Inclusive tree if they were already using an Inclusive tree 1194 for the streams mapped to the Selective tree. 1196 Once a source PE decides to setup an Selective tree, it MUST announce 1197 the mapping of the streams (which may be in different 1198 VPLSes) that are mapped to the tree to the other PEs using BGP. After 1199 the egress PEs receive the announcement they setup their forwarding 1200 path to receive traffic on the Selective tree if they have one or 1201 more receivers interested in the streams mapped to the 1202 tree. Setting up the forwarding path requires setting up the 1203 demultiplexing forwarding entries based on the top MPLS label (if 1204 there is no inner label) or the inner label (if present) as described 1205 in section 9. The egress PEs MAY perform this switch to the Selective 1206 tree once the advertisement from the ingress PE is received or wait 1207 for a preconfigured timer to do so, after receiving the 1208 advertisement, when the P2MP LSP protocol is mLDP. When the P2MP LSP 1209 protocol is P2MP RSVP-TE an egress PE MUST perform this switch to the 1210 Selective tree only after the advertisement from the ingress PE is 1211 received and the RSVP-TE P2MP LSP has been setup to the egress PE. 1212 This switch MAY be done after waiting for a preconfigured timer after 1213 these two steps have been accomplished. 1215 A source PE MUST use the following approach to decide when to start 1216 transmitting data on the Selective tree, if it was already using an 1217 Inclusive tree. A certain pre-configured delay after advertising the 1218 streams mapped to an Selective tree, the source PE begins 1219 to send traffic on the Selective tree. At this point it stops to send 1220 traffic for the streams, that are mapped on the Selective 1221 tree, on the Inclusive tree. This traffic is instead transmitted on 1222 the Selective tree. 1224 11.2. Advertising C-(S, G) Binding to a Selective Tree 1226 The ingress PE informs all the PEs that are on the path to receivers 1227 of the (C-S, C-G) of the binding of the Selective tree to the (C-S, 1228 C-G), using BGP. The BGP announcement is done by sending update for 1229 the MCAST-VPLS address family using what is referred to as the S-PMSI 1230 A-D route. The format of the NLRI is described in section 12.1. The 1231 NLRI MUST be constructed as follows: 1233 + The RD MUST be set to the RD configured locally for the VPLS. 1234 This is required to uniquely identify the as the 1235 addresses could overlap between different VPLSes. This MUST be 1236 the same RD value used in the VPLS auto-discovery process. 1238 + The Multicast Source field MUST contain the source address 1239 associated with the C-multicast stream, and the Multicast Source 1240 Length field is set appropriately to reflect this. If the source 1241 address is a wildcard the source address is set to 0. 1243 + The Multicast Group field MUST contain the group address 1244 associated with the C-multicast stream, and the Multicast Group 1245 Length field is set appropriately to reflect this. If the group 1246 address is a wildcard the group address is set to 0. 1248 + The Originating Router's IP Address field MUST be set to the IP 1249 address that the (local) PE places in the BGP next-hop of the 1250 BGP-VPLS A-D routes. Note that the tuple uniquely identifies a given VPLS. 1253 The PE constructs the rest of the Selective A-D route as follows. 1255 Depending on the type of a P-Multicast tree used for the P-tunnel, 1256 the PMSI tunnel attribute of the S-PMSI A-D route is constructed as 1257 follows: 1259 + The PMSI tunnel attribute MUST contain the identity of the P- 1260 Multicast tree (note that the PE could create the identity of the 1261 tree prior to the actual instantiation of the tree). 1263 + If in order to establish the P-Multicast tree the PE needs to 1264 know the leaves of the tree within its own AS, then the PE 1265 obtains this information from the Leaf A-D routes received from 1266 other PEs/ASBRs within its own AS (as other PEs/ASBRs originate 1267 Leaf A-D routes in response to receiving the S-PMSI A-D route) by 1268 setting the Leaf Information Required flag in the PMSI Tunnel 1269 attribute to 1. This enables explicit tracking for the multicast 1270 stream(s) advertised by the S-PMSI A-D route. 1272 + If a PE originates S-PMSI A-D routes with the Leaf Information 1273 Required flag in the PMSI Tunnel attribute set to 1, then the PE 1274 MUST be (auto)configured with an import Route Target, which 1275 controls acceptance of Leaf A-D routes by the PE. (Procedures for 1276 originating Leaf A-D routes by the PEs that receive the S-PMSI A- 1277 D route are described in section "Receiving S-PMSI A-D routes by 1278 PEs.) 1280 This Route Target is IP address specific. The Global 1281 Administrator field of this Route Target MUST be set to the IP 1282 address carried in the Next Hop of all the S-PMSI A-D routes 1283 advertised by this PE (if the PE uses different Next Hops, then 1284 the PE MUST be (auto)configured with multiple import RTs, one per 1285 each such Next Hop). The Local Administrator field of this Route 1286 Target MUST be set to 0. 1288 If the PE supports Route Target Constrain [RFC4684], the PE 1289 SHOULD advertise this import Route Target within its own AS using 1290 Route Target Constrains. To constrain distribution of the Route 1291 Target Constrain routes to the AS of the advertising PE these 1292 routes SHOULD carry the NO_EXPORT Community ([RFC1997]). 1294 + A PE MAY aggregate two or more S-PMSIs originated by the PE onto 1295 the same P-Multicast tree. If the PE already advertises S-PMSI A- 1296 D routes for these S-PMSIs, then aggregation requires the PE to 1297 re-advertise these routes. The re-advertised routes MUST be the 1298 same as the original ones, except for the PMSI tunnel attribute. 1299 If the PE has not previously advertised S-PMSI A-D routes for 1300 these S-PMSIs, then the aggregation requires the PE to advertise 1301 (new) S-PMSI A-D routes for these S-PMSIs. The PMSI Tunnel 1302 attribute in the newly advertised/re-advertised routes MUST carry 1303 the identity of the P-Multicast tree that aggregates the S-PMSIs. 1304 If at least some of the S-PMSIs aggregated onto the same P- 1305 Multicast tree belong to different MVPNs, then all these routes 1306 MUST carry an MPLS upstream assigned label [RFC5331]. If all 1307 these aggregated S-PMSIs belong to the same MVPN, then the routes 1308 MAY carry an MPLS upstream assigned label [RFC5331]. The labels 1309 MUST be distinct on a per MVPN basis, and MAY be distinct on a 1310 per route basis. 1312 The Next Hop field of the MP_REACH_NLRI attribute of the route SHOULD 1313 be set to the same IP address as the one carried in the Originating 1314 Router's IP Address field. 1316 By default the set of Route Targets carried by the route MUST be the 1317 same as the Route Targets carried in the BGP-VPLS A-D route 1318 originated from the VSI. The default could be modified via 1319 configuration. 1321 11.3. Receiving S-PMSI A-D routes by PEs 1323 Consider a PE that receives an S-PMSI A-D route. If one or more of 1324 the VSIs on the PE have their import Route Targets that contain one 1325 or more of the Route Targets carried by the received S-PMSI A-D 1326 route, then for each such VSI the PE performs the following. 1328 Procedures for receiving an S-PMSI A-D route by a PE (both within and 1329 outside of the AS of the PE that originates the route) are the same 1330 as specified in Section "Inter-AS A-D route received via IBGP" except 1331 that (a) instead of Inter-AS I-PMSI A-D routes the procedures apply 1332 to S-PMSI A-D routes, and (b) the rules for determining whether the 1333 received S-PMSI A-D route is the best route to the destination 1334 carried in the NLRI of the route, are the same as BGP path selection 1335 rules and may be modified by policy, and (c) a PE performs procedures 1336 specified in that section only if in addition to the criteria 1337 specified in that section the following is true: 1339 + If as a result of IGMP or PIM snooping on the PE-CE interfaces, 1340 the PE has snooped state for at least one multicast join that 1341 matches the multicast source and group advertised in the S-PMSI 1342 A-D route. Further if the oif (outgoing interfaces) for this 1343 state contains one or more interfaces to the locally attached 1344 CEs. 1346 The snooped state is said to "match" the S-PMSI A-D route if any of 1347 the following is true: 1349 + The S-PMSI A-D route carries (C-S, C-G) and the snooped state is 1350 for (C-S, C-G). OR 1352 + The S-PMSI A-D route carries (C-*, C-G) and (a) the snooped 1353 state is for (C-*, C-G) OR (b) the snooped state is for at least 1354 one multicast join with the multicast group address equal to C-G 1355 and there doesn't exist another S-PMSI A-D route that carries (C- 1356 S, C-G) where C-S is the source address of the snooped state. 1358 + The S-PMSI A-D route carries (C-S, C-*) and (a) the snooped 1359 state is for at least one multicast join with the multicast 1360 source address equal to C-S, and (b) there doesn't exist another 1361 S-PMSI A-D route that carries (C-S, C-G) where C-G is the group 1362 address of the snooped state. 1364 + The S-PMSI A-D route carries (C-*, C-*) 1366 Note if the above conditions are true, and if the received S-PMSI A-D 1367 route has a PMSI Tunnel attribute with the Leaf Information Required 1368 flag set to 1, then the PE originates a Leaf A-D route. The Route Key 1369 of the Leaf A-D route is set to the MCAST-VPLS NLRI of the S-PMSI A-D 1370 route. The rest of the Leaf A-D route is constructed using the same 1371 procedures as specified in section "Originating Leaf A-D route into 1372 IBGP", except that instead of originating Leaf A-D routes in response 1373 to receiving Inter-AS A-D routes the procedures apply to originating 1374 Leaf A-D routes in response to receiving S-PMSI A-D routes. 1376 In addition to the procedures specified in Section "Inter-AS A-D 1377 route received via IBGP" the PE MUST set up its forwarding path to 1378 receive traffic, for each multicast stream in the matching snooped 1379 state, from the tunnel advertised by the S-PMSI A-D route (the PE 1380 MUST switch to the Selective tree). 1382 11.4. Inter-AS Selective Tree 1384 Inter-AS Selective trees support all three models of inter-AS VPLS 1385 service, option (a), (b) and (c), that are supported by Inter-AS 1386 Inclusive trees. They are constructed in a manner that is very 1387 similar to Inter-AS Inclusive trees. 1389 For option (a) and option (b) support inter-AS Selective trees are 1390 constructed without requiring a single P-Multicast tree to span 1391 multiple ASes. This allows individual ASes to potentially use 1392 different P-tunneling technologies.There are two variants of this 1393 model. One that requires MAC and IP multicast lookup on the ASBRs and 1394 another that does not require MAC/IP multicast lookup on the ASBRs 1395 and instead builds segmented inter-AS Selective trees. 1397 Segmented Inter-AS Selective trees can also be used with option (c) 1398 unlike Segmented Inter-AS Inclusive trees. This is because the S-PMSI 1399 A-D routes can be exchanged via ASBRs (even though BGP VPLS A-D 1400 routes are not exchanged via ASBRs). 1402 In the case of Option (c) an Inter-AS Selective tree may also be a 1403 non-segmented P-Multicast tree that spans multiple ASs. 1405 11.4.1. VSIs on the ASBRs 1407 The requirements on ASBRs in this model include the requirements 1408 presented in section 10. The source ASBR (that receives traffic from 1409 another AS) may independently decide whether it wishes to use 1410 Selective trees or not. If it uses Selective trees the source ASBR 1411 MUST perform a MAC lookup to determine the Selective tree to forward 1412 the VPLS packet on. 1414 11.4.1.1. VPLS Inter-AS Selective Tree A-D Binding 1416 The mechanisms for propagating S-PMSI A-D routes are the same as the 1417 intra-AS case described in section 12.2. The BGP Selective tree A-D 1418 routes generated by PEs in an AS MUST NOT be propagated outside the 1419 AS. 1421 11.4.2. Inter-AS Segmented Selective Trees 1423 Inter-AS Segmented Selective trees MUST be used when option (b) is 1424 used to provide the inter-AS VPLS service. They MAY be used when 1425 option (c) is used to provide the inter-AS VPLS service. 1427 A Segmented inter-AS Selective Tunnel is constructed similar to an 1428 inter-AS Segmented Inclusive Tunnel. Namely, such a tunnel is 1429 constructed as a concatenation of tunnel segments. There are two 1430 types of tunnel segments: an intra-AS tunnel segment (a segment that 1431 spans ASBRs within the same AS), and inter-AS tunnel segment (a 1432 segment that spans adjacent ASBRs in adjacent ASes). ASes that are 1433 spanned by a tunnel are not required to use the same tunneling 1434 mechanism to construct the tunnel - each AS may pick up a tunneling 1435 mechanism to construct the intra-AS tunnel segment of the tunnel, in 1436 its AS. 1438 The PE that decides to set up a Selective tree, advertises the 1439 Selective tree to multicast stream binding using a S-PMSI A-D route 1440 as per procedures in section 11.2, to the routers in its own AS. 1442 A S-PMSI A-D route advertised outside the AS, to which the 1443 originating PE belongs, will be referred to as an inter-AS Selective 1444 Tree A-D route (Although this route is originated by a PE as an 1445 intra-AS route it is referred to as an inter-AS route outside the 1446 AS). 1448 11.4.2.1. Handling S-PMSI A-D routes by ASBRs 1450 Procedures for handling an S-PMSI A-D route by ASBRs (both within and 1451 outside of the AS of the PE that originates the route) are the same 1452 as specified in Section "Propagating VPLS BGP A-D routes to other 1453 ASes", except that instead of Inter-AS BGP-VPLS A-D routes and the 1454 BGP-VPLS A-D NLRI these procedures apply to S-PMSI A-D routes and the 1455 S-PMSI A-D NLRI. 1457 In addition to these procedures an ASBR advertises a Leaf A-D route 1458 in response to a S-PMSI A-D route only if: 1460 + The S-PMSI A-D route was received via EBGP from another ASBR and 1461 the ASBR merges the S-PMSI A-D route into an Inter-AS BGP VPLS A- 1462 D route as described in the next section. OR 1464 + The ASBR receives a Leaf A-D route from a downstream PE or ASBR 1465 in response to the S-PMSI A-D route, received from an upstream PE 1466 or ASBR, that the ASBR propagated inter-AS to downstream ASBRs 1467 and PEs. 1469 + The ASBR has snooped state from local CEs that matches the NLRI 1470 carried in the S-PMSI A-D route as per the following rules: 1472 i) The NLRI encodes (C-S, C-G) which is the same as the snooped 1473 (C-S, C-G) ii) The NLRI encodes (*, C-G) and there is snooped 1474 state for at least one (C-S, C-G) and there is no other matching 1475 SPMSI A-D route for (C-S, C-G) OR there is snooped state for (*, 1476 C-G) iii) The NLRI encodes (*, *) and there is snooped state for 1477 at least one (C-S, C-G) or (*, C-G) and there is no other 1478 matching SPMSI A-D route for that (C-S, C-G) or (*, C-G) 1479 respecively. 1481 The C-multicast data traffic is sent on the Selective tree by the 1482 originating PE. When it reaches an ASBR that is on the Inter-AS 1483 segmented tree, it is delivered to local receivers, if any. It is 1484 then forwarded on any inter-AS or intra-AS segments that exist on the 1485 Inter-AS Selective Segmented tree. If the Inter-AS Segmented 1486 Selective Tree is merged onto an Inclusive tree, as described in the 1487 next section, the data traffic is forwarded onto the Inclusive tree. 1489 11.4.2.1.1. Merging Selective Tree into an Inclusive Tree 1491 Consider the situation where: 1493 + An ASBR is receiving (or expecting to receive) inter-AS (C-S, C- 1494 G) data from upstream via a Selective tree. 1496 + The ASBR is sending (or expecting to send) the inter-AS (C-S, 1497 C-G) data downstream via an Inclusive tree. 1499 This situation may arise if the upstream providers have a policy of 1500 using Selective trees but the downstream providers have a policy of 1501 using Inclusive trees. To support this situation, an ASBR MAY, 1502 under certain conditions, merge one or more upstream Selective trees 1503 into a downstream Inclusive tree. Note that this can be the case only 1504 for option (b) and not for option (c) as for option (c) the ASBRs do 1505 not have Inclusive tree state. 1507 A Selective tree (corresponding to a particular S-PMSI A-D route) MAY 1508 be merged by a particular ASBR into an Inclusive tree 1509 (corresponding to a particular Inter-AS BGP VPLS A-D route) if and 1510 only if the following conditions all hold: 1512 + The S-PMSI A-D route and the Inter-AS BGP VPLS A-D route 1513 originate in the same AS. The Inter-AS BGP VPLS A-D route carries 1514 the originating AS in the AS_PATH attribute of the route. The S- 1515 PMSI A-D route carries the originating AS in the AS_PATH 1516 attribute of the route. 1518 + The S-PMSI A-D route and the Inter-AS BGP VPLS A-D route have 1519 exactly the same set of RTs. 1521 An ASBR performs merging by stitching the tail end of the P-tunnel, 1522 as specified in the the PMSI Tunnel Attribute of the S-PMSI A-D route 1523 received by the ASBR, to the to the head of the P-tunnel, as 1524 specified in the PMSI Tunnel Attribute of the Inter-AS BGP VPLS A-D 1525 route re-advertised by the ASBR. 1527 An ASBR that merges an S-PMSI A-D route into an Inter-AS BGP VPLS A-D 1528 route MUST NOT re-advertise the S-PMSI A-D route. 1530 11.4.3. Inter-AS Non-Segmented Selective trees 1532 Inter-AS Non-segmented Selective trees MAY be used in the case of 1533 option (c). 1535 In this method, there is a multi-hop E-BGP peering between the PEs 1536 (or a Route Reflector) in one AS and the PEs (or Route Reflector) in 1537 another AS. The PEs exchange BGP Selective tree A-D routes, along 1538 with PMSI Tunnel Attribute, as in the intra-AS case described in 1539 section 10.3. 1541 The PEs in different ASs use a non-segmented Selective inter-AS P2MP 1542 tunnel for VPLS multicast. 1544 This method requires no VPLS information (in either the control or 1545 the data plane) on the ASBRs. The ASBRs only need to participate in 1546 the non-segmented P2MP tunnel setup in the control plane, and do MPLS 1547 label forwarding in the data plane. 1549 The data forwarding in this model is the same as in the intra-AS case 1550 described in section 9. 1552 12. BGP Extensions 1554 This section describes the encoding of the BGP extensions required by 1555 this document. 1557 12.1. Inclusive Tree/Selective Tree Identifier 1559 Inclusive P-Multicast tree and Selective P-Multicast tree 1560 advertisements carry the P-Multicast tree identifier. 1562 This document reuses the BGP attribute, called PMSI Tunnel Attribute 1563 that is defined in [MVPN]. 1565 Only the following Tunnel Types MUST be used when PMSI Tunnel 1566 attribute is carried in VPLS A-D or VPLS S-PMSI A-D routes: 1568 + 0 - No tunnel information present 1569 + 1 - RSVP-TE P2MP LSP 1570 + 2 - LDP P2MP LSP 1571 + 6 - Ingress Replication 1573 12.2. MCAST-VPLS NLRI 1575 This document defines a new BGP NLRI, called the MCAST-VPLS NLRI. 1577 Following is the format of the MCAST-VPLS NLRI: 1579 +-----------------------------------+ 1580 | Route Type (1 octet) | 1581 +-----------------------------------+ 1582 | Length (1 octet) | 1583 +-----------------------------------+ 1584 | Route Type specific (variable) | 1585 +-----------------------------------+ 1587 The Route Type field defines encoding of the rest of MCAST-VPLS NLRI 1588 (Route Type specific MCAST-VPLS NLRI). 1590 The Length field indicates the length in octets of the Route Type 1591 specific field of MCAST-VPLS NLRI. 1593 This document defines the following Route Types for auto-discovery 1594 routes: 1595 + 3 - Selective Tree auto-discovery route; 1596 + 4 - Leaf auto-discovery route. 1598 The MCAST-VPLS NLRI is carried in BGP using BGP Multiprotocol 1599 Extensions [RFC4760] with an AFI of 25 (L2VPN AFI), and an SAFI of 1600 MCAST-VPLS [To be assigned by IANA]. The NLRI field in the 1601 MP_REACH_NLRI/MP_UNREACH_NLRI attribute contains the MCAST-VPLS NLRI 1602 (encoded as specified above). 1604 In order for two BGP speakers to exchange labeled MCAST-VPLS NLRI, 1605 they must use BGP Capabilities Advertisement to ensure that they both 1606 are capable of properly processing such NLRI. This is done as 1607 specified in [RFC4760], by using capability code 1 (multiprotocol 1608 BGP) with an AFI of 25 and an SAFI of MCAST-VPLS. 1610 The following describes the format of the Route Type specific MCAST- 1611 VPLS NLRI for various Route Types defined in this document. 1613 12.2.1. S-PMSI auto-discovery route 1615 An S-PMSI A-D route type specific MCAST-VPLS NLRI consists of the 1616 following: 1618 +-----------------------------------+ 1619 | RD (8 octets) | 1620 +-----------------------------------+ 1621 | Multicast Source Length (1 octet) | 1622 +-----------------------------------+ 1623 | Multicast Source (Variable) | 1624 +-----------------------------------+ 1625 | Multicast Group Length (1 octet) | 1626 +-----------------------------------+ 1627 | Multicast Group (Variable) | 1628 +-----------------------------------+ 1629 | Originating Router's IP Addr | 1630 +-----------------------------------+ 1632 The RD is encoded as described in [RFC4364]. 1634 The Multicast Source field contains the C-S address i.e the address 1635 of the multicast source. This may be 0 to indicate a wildcard. If the 1636 Multicast Source field contains an IPv4 address, then the value of 1637 the Multicast Source Length field is 32. If the Multicast Source 1638 field contains an IPv6 address, then the value of the Multicast 1639 Source Length field is 128. 1641 The Multicast Group field contains the C-G address i.e. the address 1642 of the multicast group. This may be 0 to indicate a wildcard. If the 1643 Multicast Group field contains an IPv4 address, then the value of the 1644 Multicast Group Length field is 32. If the Multicast Group field 1645 contains an IPv6 address, then the value of the Multicast Group 1646 Length field is 128. 1648 Usage of Selective Tree auto-discovery routes is described in Section 1649 12. 1651 12.2.2. Leaf auto-discovery route 1653 A leaf auto-discovery route type specific MCAST-VPLS NLRI consists of 1654 the following: 1656 +-----------------------------------+ 1657 | Route Key (variable) | 1658 +-----------------------------------+ 1659 | Originating Router's IP Addr | 1660 +-----------------------------------+ 1662 Usage of Leaf auto-discovery routes is described in sections "Inter- 1663 AS Inclusive P-Multicast tree A-D/Binding" and "Optimizing Multicast 1664 Distribution via Selective trees". 1666 13. Aggregation Methodology 1668 In general the herustics used to decide which VPLS instances or entries to aggregate is implementation dependent. It is also 1670 conceivable that offline tools can be used for this purpose. This 1671 section discusses some tradeoffs with respect to aggregation. 1673 The "congruency" of aggregation is defined by the amount of overlap 1674 in the leaves of the client trees that are aggregated on a SP tree. 1675 For Aggregate Inclusive trees the congruency depends on the overlap 1676 in the membership of the VPLSes that are aggregated on the Aggregate 1677 Inclusive tree. If there is complete overlap aggregation is perfectly 1678 congruent. As the overlap between the VPLSes that are aggregated 1679 reduces, the congruency reduces. 1681 If aggregation is done such that it is not perfectly congruent a PE 1682 may receive traffic for VPLSes to which it doesn't belong. As the 1683 amount of multicast traffic in these unwanted VPLSes increases 1684 aggregation becomes less optimal with respect to delivered traffic. 1685 Hence there is a tradeoff between reducing state and delivering 1686 unwanted traffic. 1688 An implementation should provide knobs to control the congruency of 1689 aggregation. This will allow a SP to deploy aggregation depending on 1690 the VPLS membership and traffic profiles in its network. If 1691 different PEs or shared roots' are setting up Aggregate Inclusive 1692 trees this will also allow a SP to engineer the maximum amount of 1693 unwanted VPLSes that a particular PE may receive traffic for. 1695 The state/bandwidth optimality trade-off can be further improved by 1696 having a versatile many-to-many association between client trees and 1697 provider trees. Thus a VPLS can be mapped to multiple Aggregate 1698 trees. The mechanisms for achieving this are for further study. Also 1699 it may be possible to use both ingress replication and an Aggregate 1700 tree for a particular VPLS. Mechanisms for achieving this are also 1701 for further study. 1703 14. Data Forwarding 1705 14.1. MPLS Tree Encapsulation 1707 14.1.1. Mapping multiple VPLS instances to a P2MP LSP 1709 The following diagram shows the progression of the VPLS IP multicast 1710 packet as it enters and leaves the SP network when MPLS trees are 1711 being used for multiple VPLS instances. RSVP-TE P2MP LSPs are 1712 examples of such trees. 1714 Packets received Packets in transit Packets forwarded 1715 at ingress PE in the service by egress PEs 1716 provider network 1718 +---------------+ 1719 |MPLS Tree Label| 1720 +---------------+ 1721 | VPLS Label | 1722 ++=============++ ++=============++ ++=============++ 1723 ||C-Ether Hdr || || C-Ether Hdr || || C-Ether Hdr || 1724 ++=============++ >>>>> ++=============++ >>>>> ++=============++ 1725 || C-IP Header || || C-IP Header || || C-IP Header || 1726 ++=============++ >>>>> ++=============++ >>>>> ++=============++ 1727 || C-Payload || || C-Payload || || C-Payload || 1728 ++=============++ ++=============++ ++=============++ 1730 The receiver PE does a lookup on the outer MPLS tree label and 1731 determines the MPLS forwarding table in which to lookup the inner 1732 MPLS label. This table is specific to the tree label space. The inner 1733 label is unique within the context of the root of the tree (as it is 1734 assigned by the root of the tree, without any coordination with any 1735 other nodes). Thus it is not unique across multiple roots. So, to 1736 unambiguously identify a particular VPLS one has to know the label, 1737 and the context within which that label is unique. The context is 1738 provided by the outer MPLS label [RFC5331]. 1740 The outer MPLS label is stripped. The lookup of the resulting MPLS 1741 label determines the VSI in which the receiver PE needs to do the C- 1742 multicast data packet lookup. It then strips the inner MPLS label and 1743 sends the packet to the VSI for multicast data forwarding. 1745 14.1.2. Mapping one VPLS instance to a P2MP LSP 1747 The following diagram shows the progression of the VPLS IP multicast 1748 packet as it enters and leaves the SP network when a given MPLS tree 1749 is being used for a single VPLS instance. RSVP-TE P2MP LSPs are 1750 examples of such trees. 1752 Packets received Packets in transit Packets forwarded 1753 at ingress PE in the service by egress PEs 1754 provider network 1755 +---------------+ 1756 |MPLS Tree Label| 1757 ++=============++ ++=============++ ++=============++ 1758 ||C-Ether Hdr || || C-Ether Hdr || || C-Ether Hdr || 1759 ++=============++ >>>>> ++=============++ >>>>> ++=============++ 1760 || C-IP Header || || C-IP Header || || C-IP Header || 1761 ++=============++ >>>>> ++=============++ >>>>> ++=============++ 1762 || C-Payload || || C-Payload || || C-Payload || 1763 ++=============++ ++=============++ ++=============++ 1765 The receiver PE does a lookup on the outer MPLS tree label and 1766 determines the VSI in which the receiver PE needs to do the C- 1767 multicast data packet lookup. It then strips the inner MPLS label and 1768 sends the packet to the VSI for multicast data forwarding. 1770 15. VPLS Data Packet Treatment 1772 If the destination MAC address of a VPLS packet received by a PE from 1773 a VPLS site is a multicast adddress, a P-Multicast tree SHOULD be 1774 used to transport the packet, if possible. If the packet is an IP 1775 multicast packet and a Selective tree exists for that multicast 1776 stream, the Selective tree SHOULD be used. Else if an Inclusive tree 1777 exists for the VPLS, it SHOULD be used. 1779 If the destination MAC address of a VPLS packet is a broadcast 1780 address, it is flooded. If Inclusive tree is already established, PE 1781 SHOULD flood over it. If Inclusive tree cannot be used for some 1782 reason, PE MUST flood over multiple PWs, based on [RFC4761] or 1783 [RFC4762]. 1785 If the destination MAC address of a packet is a unicast address and 1786 it has not been learned, the packet MUST be sent to all PEs in the 1787 VPLS. Inclusive P-Multicast trees or a Selective P-Multicast tree 1788 bound to (C-*, C-*) SHOULD be used for sending unknown unicast MAC 1789 packets to all PEs. When this is the case the receiving PEs MUST 1790 support the ability to perform MAC address learning for packets 1791 received on a multicast tree. In order to perform such learning, the 1792 receiver PE MUST be able to determine the sender PE when a VPLS 1793 packet is received on a P-Multicast tree. This further implies that 1794 the MPLS P-Multicast tree technology MUST allow the egress PE to 1795 determine the sender PE from the received MPLS packet. 1797 When a receiver PE receives a VPLS packet with a source MAC address, 1798 that has not yet been learned, on a P-Multicast tree, the receiver PE 1799 determines the PW to the sender PE. The receiver PE then creates 1800 forwarding state in the VPLS instance with a destination MAC address 1801 being the same as the source MAC address being learned, and the PW 1802 being the PW to the sender PE. 1804 It should be noted that when a sender PE that is sending packets 1805 destined to an unknown unicast MAC address over a P-Multicast tree 1806 learns the PW to use for forwarding packets destined to this unicast 1807 MAC address, it might immediately switch to transport such packets 1808 over this particular PW. Since the packets were initially being 1809 forwarded using a P-Multicast tree, this could lead to packet 1810 reordering. This constraint should be taken into consideration if 1811 unknown unicast frames are forwarded using a P-Multicast tree, 1812 instead of multiple PWs based on [RFC4761] or [RFC4762]. 1814 An implementation MUST support the ability to transport unknown 1815 unicast traffic over Inclusive P-Multicast trees. Further an 1816 implementation MUST support the ability to perform MAC address 1817 learning for packets received on a P-Multicast tree. 1819 16. Security Considerations 1821 Security considerations discussed in [RFC4761] and [RFC4762] apply to 1822 this document. This section describes additional considerations. 1824 As mentioned in [RFC4761], there are two aspects to achieving data 1825 privacy in a VPLS: securing the control plane and protecting the 1826 forwarding path. Compromise of the control plane could result in a PE 1827 sending multicast data belonging to some VPLS to another VPLS, or 1828 blackholing VPLS multicast data, or even sending it to an 1829 eavesdropper; none of which are acceptable from a data privacy point 1830 of view. The mechanisms in this document use BGP for the control 1831 plane. Hence techniques such as in [RFC2385] help authenticate BGP 1832 messages, making it harder to spoof updates (which can be used to 1833 divert VPLS traffic to the wrong VPLS) or withdraws (denial-of- 1834 service attacks). In the multi-AS methods (b) and (c) described in 1835 Section 11, this also means protecting the inter-AS BGP sessions, 1836 between the ASBRs, the PEs, or the Route Reflectors. 1838 Note that [RFC2385] will not help in keeping MPLS labels, associated 1839 with P2MP LSPs or the upstream MPLS labels used for aggregation, 1840 private -- knowing the labels, one can eavesdrop on VPLS traffic. 1841 However, this requires access to the data path within a Service 1842 Provider network. 1844 One of the requirements for protecting the data plane is that the 1845 MPLS labels are accepted only from valid interfaces. This applies 1846 both to MPLS labels associated with P2MP LSPs and also applies to the 1847 upstream assigned MPLS labels. For a PE, valid interfaces comprise 1848 links from P routers. For an ASBR, a valid interface is a link from 1849 another ASBR in an AS that is part of a given VPLS. It is especially 1850 important in the case of multi-AS VPLSes that one accept VPLS packets 1851 only from valid interfaces. 1853 17. IANA Considerations 1855 This document defines a new NLRI, called MCAST-VPLS, to be carried in 1856 BGP using multiprotocol extensions. It requires assignment of a new 1857 SAFI. This is to be assigned by IANA. 1859 This document defines a BGP optional transitive attribute, called 1860 PMSI Attribute. This is the same attribute as the one defined in 1861 [BGP-MVPN] and the code point for this attribute has already been 1862 assigned by IANA as 22 [BGP-IANA]. Hence no further action is 1863 required from IANA regarding this attribute. 1865 18. Acknowledgments 1867 Many thanks to Thomas Morin for his support of this work. We would 1868 also like to thank authors of [BGP-MVPN] and [MVPN] as the details of 1869 the inter-AS segmented tree procedures in this document have 1870 benefited from those in [BGP-MVPN] and [MVPN]. We would also like to 1871 thank Wim Henderickx for his comments. 1873 19. Normative References 1875 [RFC2119] "Key words for use in RFCs to Indicate Requirement 1876 Levels.", Bradner, March 1997 1878 [RFC4761] K. Kompella, Y. Rekther, "Virtual Private LAN Service", 1879 draft-ietf-l2vpn-vpls-bgp-02.txt 1881 [RFC4762] M. Lasserre, V. Kompella, "Virtual Private LAN Services 1882 over MPLS", draft-ietf-l2vpn-vpls-ldp-03.txt 1884 [RFC4760] T. Bates, et. al., "Multiprotocol Extensions for BGP-4", 1885 January 2007 1887 [RFC5331] R. Aggarwal, Y. Rekhter, E. Rosen, "MPLS Upstream Label 1888 Assignment and Context Specific Label Space", RFC 5331, August 2008 1890 20. Informative References 1892 [L2VPN-SIG] E. Rosen et. al., "Provisioning, Autodiscovery, and 1893 Signaling in L2VPNs", draft-ietf-l2vpn-signaling-08.txt 1895 [RFC5332] T. Eckert, E. Rosen, R. Aggarwal, Y. Rekhter, "MPLS 1896 Multicast Encapsulations", RFC 5332, August 2008 1898 [MVPN] E. Rosen, R. Aggarwal, "Multicast in 2547 VPNs", draft-ietf- 1899 l3vpn-2547bis-mcast-08.txt" 1901 [BGP-MVPN] R. Aggarwal, E. Rosen, Y. Rekhter, T. Morin, C. 1902 Kodeboniya. "BGP Encodings for Multicast in 2547 VPNs", draft-ietf- 1903 l3vpn-2547bis-mcast-bgp-06.txt 1905 [RFC4875] R. Aggarwal et. al, "Extensions to RSVP-TE for Point to 1906 Multipoint TE LSPs", draft-ietf-mpls-rsvp-te-p2mp-07.txt 1908 [RSVP-OBB] Z. Ali, G. Swallow, R. Aggarwal, "Non PHP behavior and 1909 out-of-band mapping for RSVP-TE LSPs", draft-ietf-mpls-rsvp-te-no- 1910 php-obb-mapping, work in progress. 1912 [MLDP] I. Minei et. al, "Label Distribution Protocol Extensions for 1913 Point-to-Multipoint and Multipoint-to-Multipoint Label Switched 1914 Paths", draft-ietf-mpls-ldp-p2mp, work in progress. 1916 [RFC4364] "BGP MPLS VPNs", E. Rosen, Y.Rekhter, February 2006 1918 [MCAST-VPLS-REQ] Y. kamite, et. al., "Requirements for Multicast 1919 Support in Virtual Private LAN Services", draft-ietf-l2vpn-vpls- 1920 mcast-reqts-05.txt 1922 [RFC1997] R. Chandra, et. al., "BGP Communities Attribute", August 1923 1996 1925 [BGP-IANA] http://www.iana.org/assignments/bgp-parameters 1927 [RFC4684] P. Marques et. al., "Constrained Route Distribution for 1928 Border Gateway Protocol/MultiProtocol Label Switching (BGP/MPLS) 1929 Internet Protocol (IP) Virtual Private Networks (VPNs)", RFC 4684, 1930 November 2006 1932 [RFC2385] Heffernan, A., "Protection of BGP Sessions via the TCP MD5 1933 Signature Option", RFC 2385, August 1998. 1935 [RFC4447] L. Martini et. al., "Pseudowire Setup and Maintenance Using 1936 the Label Distribution Protocol (LDP)", RFC 4447 April 2006 1938 [MULTI-HOMING] K. Kompella et. al., "Multi-homing in BGP-based 1939 Virtual Private LAN Service", draft-kompella-l2vpn-vpls- 1940 multihoming-02.txt 1942 21. Author's Address 1944 Rahul Aggarwal 1945 Juniper Networks 1946 1194 North Mathilda Ave. 1947 Sunnyvale, CA 94089 1948 USA 1949 Phone: +1-408-936-2720 1950 Email: rahul@juniper.net 1952 Yuji Kamite 1953 NTT Communications Corporation 1954 Tokyo Opera City Tower 1955 3-20-2 Nishi Shinjuku, Shinjuku-ku, 1956 Tokyo 163-1421, 1957 Japan 1959 Email: y.kamite@ntt.com 1961 Luyuan Fang 1962 Cisco Systems 1963 300 Beaver Brook Road 1964 BOXBOROUGH, MA 01719 1965 USA 1967 Email: lufang@cisco.com 1969 Yakov Rekhter 1970 Juniper Networks 1971 1194 North Mathilda Ave. 1972 Sunnyvale, CA 94089 1973 USA 1975 Email: yakov@juniper.net 1977 Chaitanya Kodeboniya