idnits 2.17.1 draft-ietf-bess-evpn-irb-mcast-03.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 : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (August 16, 2019) is 1708 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(See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-09) exists of draft-ietf-bess-evpn-df-election-framework-07 -- Possible downref: Non-RFC (?) normative reference: ref. 'EVPN-AR' -- Possible downref: Non-RFC (?) normative reference: ref. 'EVPN-BUM' == Outdated reference: A later version (-15) exists of draft-ietf-bess-evpn-inter-subnet-forwarding-05 == Outdated reference: A later version (-21) exists of draft-ietf-bess-evpn-igmp-mld-proxy-02 Summary: 0 errors (**), 0 flaws (~~), 4 warnings (==), 3 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 BESS W. Lin 3 Internet-Draft Z. Zhang 4 Intended status: Standards Track J. Drake 5 Expires: February 17, 2020 E. Rosen, Ed. 6 Juniper Networks, Inc. 7 J. Rabadan 8 Nokia 9 A. Sajassi 10 Cisco Systems 11 August 16, 2019 13 EVPN Optimized Inter-Subnet Multicast (OISM) Forwarding 14 draft-ietf-bess-evpn-irb-mcast-03 16 Abstract 18 Ethernet VPN (EVPN) provides a service that allows a single Local 19 Area Network (LAN), comprising a single IP subnet, to be divided into 20 multiple "segments". Each segment may be located at a different 21 site, and the segments are interconnected by an IP or MPLS backbone. 22 Intra-subnet traffic (either unicast or multicast) always appears to 23 the endusers to be bridged, even when it is actually carried over the 24 IP or MPLS backbone. When a single "tenant" owns multiple such LANs, 25 EVPN also allows IP unicast traffic to be routed between those LANs. 26 This document specifies new procedures that allow inter-subnet IP 27 multicast traffic to be routed among the LANs of a given tenant, 28 while still making intra-subnet IP multicast traffic appear to be 29 bridged. These procedures can provide optimal routing of the inter- 30 subnet multicast traffic, and do not require any such traffic to 31 leave a given router and then reenter that same router. These 32 procedures also accommodate IP multicast traffic that needs to travel 33 to or from systems that are outside the EVPN domain. 35 Status of This Memo 37 This Internet-Draft is submitted in full conformance with the 38 provisions of BCP 78 and BCP 79. 40 Internet-Drafts are working documents of the Internet Engineering 41 Task Force (IETF). Note that other groups may also distribute 42 working documents as Internet-Drafts. The list of current Internet- 43 Drafts is at https://datatracker.ietf.org/drafts/current/. 45 Internet-Drafts are draft documents valid for a maximum of six months 46 and may be updated, replaced, or obsoleted by other documents at any 47 time. It is inappropriate to use Internet-Drafts as reference 48 material or to cite them other than as "work in progress." 49 This Internet-Draft will expire on February 17, 2020. 51 Copyright Notice 53 Copyright (c) 2019 IETF Trust and the persons identified as the 54 document authors. All rights reserved. 56 This document is subject to BCP 78 and the IETF Trust's Legal 57 Provisions Relating to IETF Documents 58 (https://trustee.ietf.org/license-info) in effect on the date of 59 publication of this document. Please review these documents 60 carefully, as they describe your rights and restrictions with respect 61 to this document. Code Components extracted from this document must 62 include Simplified BSD License text as described in Section 4.e of 63 the Trust Legal Provisions and are provided without warranty as 64 described in the Simplified BSD License. 66 Table of Contents 68 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4 69 1.1. Background . . . . . . . . . . . . . . . . . . . . . . . 4 70 1.1.1. Segments, Broadcast Domains, and Tenants . . . . . . 4 71 1.1.2. Inter-BD (Inter-Subnet) IP Traffic . . . . . . . . . 5 72 1.1.3. EVPN and IP Multicast . . . . . . . . . . . . . . . . 6 73 1.1.4. BDs, MAC-VRFS, and EVPN Service Models . . . . . . . 7 74 1.2. Need for EVPN-aware Multicast Procedures . . . . . . . . 7 75 1.3. Additional Requirements That Must be Met by the Solution 8 76 1.4. Terminology . . . . . . . . . . . . . . . . . . . . . . . 10 77 1.5. Model of Operation: Overview . . . . . . . . . . . . . . 12 78 1.5.1. Control Plane . . . . . . . . . . . . . . . . . . . . 12 79 1.5.2. Data Plane . . . . . . . . . . . . . . . . . . . . . 14 80 2. Detailed Model of Operation . . . . . . . . . . . . . . . . . 17 81 2.1. Supplementary Broadcast Domain . . . . . . . . . . . . . 17 82 2.2. Detecting When a Route is About/For/From a Particular BD 18 83 2.3. Use of IRB Interfaces at Ingress PE . . . . . . . . . . . 21 84 2.4. Use of IRB Interfaces at an Egress PE . . . . . . . . . . 23 85 2.5. Announcing Interest in (S,G) . . . . . . . . . . . . . . 23 86 2.6. Tunneling Frames from Ingress PE to Egress PEs . . . . . 24 87 2.7. Advanced Scenarios . . . . . . . . . . . . . . . . . . . 25 88 3. EVPN-aware Multicast Solution Control Plane . . . . . . . . . 26 89 3.1. Supplementary Broadcast Domain (SBD) and Route Targets . 26 90 3.2. Advertising the Tunnels Used for IP Multicast . . . . . . 27 91 3.2.1. Constructing Routes for the SBD . . . . . . . . . . . 28 92 3.2.2. Ingress Replication . . . . . . . . . . . . . . . . . 28 93 3.2.3. Assisted Replication . . . . . . . . . . . . . . . . 29 94 3.2.4. BIER . . . . . . . . . . . . . . . . . . . . . . . . 30 95 3.2.5. Inclusive P2MP Tunnels . . . . . . . . . . . . . . . 31 96 3.2.5.1. Using the BUM Tunnels as IP Multicast Inclusive 97 Tunnels . . . . . . . . . . . . . . . . . . . . . 31 98 3.2.5.2. Using Wildcard S-PMSI A-D Routes to Advertise 99 Inclusive Tunnels Specific to IP Multicast . . . 32 100 3.2.6. Selective Tunnels . . . . . . . . . . . . . . . . . . 33 101 3.3. Advertising SMET Routes . . . . . . . . . . . . . . . . . 34 102 4. Constructing Multicast Forwarding State . . . . . . . . . . . 36 103 4.1. Layer 2 Multicast State . . . . . . . . . . . . . . . . . 36 104 4.1.1. Constructing the OIF List . . . . . . . . . . . . . . 37 105 4.1.2. Data Plane: Applying the OIF List to an (S,G) Frame . 38 106 4.1.2.1. Eligibility of an AC to Receive a Frame . . . . . 38 107 4.1.2.2. Applying the OIF List . . . . . . . . . . . . . . 39 108 4.2. Layer 3 Forwarding State . . . . . . . . . . . . . . . . 40 109 5. Interworking with non-OISM EVPN-PEs . . . . . . . . . . . . . 41 110 5.1. IPMG Designated Forwarder . . . . . . . . . . . . . . . . 43 111 5.2. Ingress Replication . . . . . . . . . . . . . . . . . . . 44 112 5.2.1. Ingress PE is non-OISM . . . . . . . . . . . . . . . 45 113 5.2.2. Ingress PE is OISM . . . . . . . . . . . . . . . . . 47 114 5.3. P2MP Tunnels . . . . . . . . . . . . . . . . . . . . . . 48 115 6. Traffic to/from Outside the EVPN Tenant Domain . . . . . . . 48 116 6.1. Layer 3 Interworking via EVPN OISM PEs . . . . . . . . . 49 117 6.1.1. General Principles . . . . . . . . . . . . . . . . . 49 118 6.1.2. Interworking with MVPN . . . . . . . . . . . . . . . 52 119 6.1.2.1. MVPN Sources with EVPN Receivers . . . . . . . . 53 120 6.1.2.1.1. Identifying MVPN Sources . . . . . . . . . . 53 121 6.1.2.1.2. Joining a Flow from an MVPN Source . . . . . 54 122 6.1.2.2. EVPN Sources with MVPN Receivers . . . . . . . . 56 123 6.1.2.2.1. General procedures . . . . . . . . . . . . . 56 124 6.1.2.2.2. Any-Source Multicast (ASM) Groups . . . . . . 58 125 6.1.2.2.3. Source on Multihomed Segment . . . . . . . . 58 126 6.1.2.3. Obtaining Optimal Routing of Traffic Between MVPN 127 and EVPN . . . . . . . . . . . . . . . . . . . . 59 128 6.1.2.4. Selecting the MEG SBD-DR . . . . . . . . . . . . 60 129 6.1.3. Interworking with 'Global Table Multicast' . . . . . 60 130 6.1.4. Interworking with PIM . . . . . . . . . . . . . . . . 60 131 6.1.4.1. Source Inside EVPN Domain . . . . . . . . . . . . 61 132 6.1.4.2. Source Outside EVPN Domain . . . . . . . . . . . 62 133 6.2. Interworking with PIM via an External PIM Router . . . . 63 134 7. Using an EVPN Tenant Domain as an Intermediate (Transit) 135 Network for Multicast traffic . . . . . . . . . . . . . . . . 64 136 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 67 137 9. Security Considerations . . . . . . . . . . . . . . . . . . . 67 138 10. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 67 139 11. References . . . . . . . . . . . . . . . . . . . . . . . . . 67 140 11.1. Normative References . . . . . . . . . . . . . . . . . . 67 141 11.2. Informative References . . . . . . . . . . . . . . . . . 69 142 Appendix A. Integrated Routing and Bridging . . . . . . . . . . 71 143 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 76 145 1. Introduction 147 1.1. Background 149 Ethernet VPN (EVPN) [RFC7432] provides a Layer 2 VPN (L2VPN) 150 solution, which allows IP backbone provider to offer ethernet service 151 to a set of customers, known as "tenants". 153 In this section (as well as in [EVPN-IRB]), we provide some essential 154 background information on EVPN. 156 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 157 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 158 "OPTIONAL" in this document are to be interpreted as described in BCP 159 14 [RFC2119] [RFC8174] when, and only when, they appear in all 160 capitals, as shown here. 162 1.1.1. Segments, Broadcast Domains, and Tenants 164 One of the key concepts of EVPN is the Broadcast Domain (BD). A BD 165 is essentially an emulated ethernet. Each BD belongs to a single 166 tenant. A BD typically consists of multiple ethernet "segments", and 167 each segment may be attached to a different EVPN Provider Edge 168 (EVPN-PE) router. EVPN-PE routers are often referred to as "Network 169 Virtualization Endpoints" or NVEs. However, this document will use 170 the term "EVPN-PE", or, when the context is clear, just "PE". 172 In this document, we use the term "segment" to mean the same as 173 "Ethernet Segment" or "ES" in [RFC7432]. 175 Attached to each segment are "Tenant Systems" (TSes). A TS may be 176 any type of system, physical or virtual, host or router, etc., that 177 can attach to an ethernet. 179 When two TSes are on the same segment, traffic between them does not 180 pass through an EVPN-PE. When two TSes are on different segments of 181 the same BD, traffic between them does pass through an EVPN-PE. 183 When two TSes, say TS1 and TS2 are on the same BD, then: 185 o If TS1 knows the MAC address of TS2, TS1 can send unicast ethernet 186 frames to TS2. TS2 will receive the frames unaltered. 188 o If TS1 broadcasts an ethernet frame, TS2 will receive the 189 unaltered frame. 191 o If TS1 multicasts an ethernet frame, TS2 will receive the 192 unaltered frame, as long as TS2 has been provisioned to receive 193 ethernet multicasts. 195 When we say that TS2 receives an unaltered frame from TS1, we mean 196 that the frame still contains TS1's MAC address, and that no 197 alteration of the frame's payload (and consequently, no alteration of 198 the payload's IP header) has been made. 200 EVPN allows a single segment to be attached to multiple PE routers. 201 This is known as "EVPN multi-homing". Suppose a given segment is 202 attached to both PE1 and PE2, and suppose PE1 receives a frame from 203 that segment. It may be necessary for PE1 to send the frame over the 204 backbone to PE2. EVPN has procedures to ensure that such a frame 205 cannot be sent by PE2 back to its originating segment. This is 206 particularly important for multicast, because a frame arriving at PE1 207 from a given segment will already have been seen by all the systems 208 on that segment that need to see it. If the frame were sent back to 209 the originating segment by PE2, receivers on that segment would 210 receive the packet twice. Even worse, the frame might be sent back 211 to PE1, which could cause an infinite loop. 213 1.1.2. Inter-BD (Inter-Subnet) IP Traffic 215 If a given tenant has multiple BDs, the tenant may wish to allow IP 216 communication among these BDs. Such a set of BDs is known as an 217 "EVPN Tenant Domain" or just a "Tenant Domain". 219 If tenant systems TS1 and TS2 are not in the same BD, then they do 220 not receive unaltered ethernet frames from each other. In order for 221 TS1 to send traffic to TS2, TS1 encapsulates an IP datagram inside an 222 ethernet frame, and uses ethernet to send these frames to an IP 223 router. The router decapsulates the IP datagram, does the IP 224 processing, and re-encapsulates the datagram for ethernet. The MAC 225 source address field now has the MAC address of the router, not of 226 TS1. The TTL field of the IP datagram should be decremented by 227 exactly 1, even if the frame needs to be sent from one PE to another. 228 The structure of the provider's IP backbone is thus hidden from the 229 tenants. 231 EVPN accommodates the need for inter-BD communication within a Tenant 232 Domain by providing an integrated L2/L3 service for unicast IP 233 traffic. EVPN's Integrated Routing and Bridging (IRB) functionality 234 is specified in [EVPN-IRB]. Each BD in a Tenant Domain is assumed to 235 be a single IP subnet, and each IP subnet within a a given Tenant 236 Domain is assumed to be a single BD. EVPN's IRB functionality allows 237 IP traffic to travel from one BD to another, and ensures that proper 238 IP processing (e.g., TTL decrement) is done. 240 A brief overview of IRB, including the notion of an "IRB interface", 241 can be found in Appendix A. As explained there, an IRB interface is 242 a sort of virtual interface connecting an L3 routing instance to a 243 BD. A BD may have multiple attachment circuits (ACs) to a given PE, 244 where each AC connects to a different ethernet segment of the BD. 245 However, these ACs are not visible to the L3 routing function; from 246 the perspective of an L3 routing instance, a PE has just one 247 interface to each BD, viz., the IRB interface for that BD. 249 The "L3 routing instance" depicted in Appendix A is associated with a 250 single Tenant Domain, and may be thought of as an IP-VRF for that 251 Tenant Domain. 253 1.1.3. EVPN and IP Multicast 255 [EVPN-IRB] and [EVPN_IP_Prefix] cover inter-subnet (inter-BD) IP 256 unicast forwarding, but they do not cover inter-subnet IP multicast 257 forwarding. 259 [RFC7432] covers intra-subnet (intra-BD) ethernet multicast. The 260 intra-subnet ethernet multicast procedures of [RFC7432] are used for 261 ethernet Broadcast traffic, for ethernet unicast traffic whose MAC 262 Destination Address field contains an Unknown address, and for 263 ethernet traffic whose MAC Destination Address field contains an 264 ethernet Multicast MAC address. These three classes of traffic are 265 known collectively as "BUM traffic" (Broadcast/Unknown-Unicast/ 266 Multicast), and the procedures for handling BUM traffic are known as 267 "BUM procedures". 269 [IGMP-Proxy] extends the intra-subnet ethernet multicast procedures 270 by adding procedures that are specific to, and optimized for, the use 271 of IP multicast within a subnet. However,that document does not 272 cover inter-subnet IP multicast. 274 The purpose of this document is to specify procedures for EVPN that 275 provide optimized IP multicast functionality within an EVPN tenant 276 domain. This document also specifies procedures that allow IP 277 multicast packets to be sourced from or destined to systems outside 278 the Tenant Domain. We refer to the entire set of these procedures as 279 "OISM" (Optimized Inter-Subnet Multicast) procedures. 281 In order to support the OISM procedures specified in this document, 282 an EVPN-PE MUST also support [EVPN-IRB] and [IGMP-Proxy]. (However, 283 certain of the procedures in [IGMP-Proxy] are modified when OISM is 284 supported.) 286 1.1.4. BDs, MAC-VRFS, and EVPN Service Models 288 [RFC7432] defines the notion of "MAC-VRF". A MAC-VRF contains one or 289 more "Bridge Tables" (see section 3 of [RFC7432] for a discussion of 290 this terminology), each of which represents a single Broadcast 291 Domain. 293 In the IRB model (outlined in Appendix A) a L3 routing instance has 294 one IRB interface per BD, NOT one per MAC-VRF. This document does 295 not distinguish between a "Broadcast Domain" and a "Bridge Table", 296 and will use the terms interchangeably (or will use the acronym "BD" 297 to refer to either). The way the BDs are grouped into MAC-VRFs is 298 not relevant to the procedures specified in this document. 300 Section 6 of [RFC7432] also defines several different EVPN service 301 models: 303 o In the "vlan-based service", each MAC-VRF contains one "bridge 304 table", where the bridge table corresponds to a particular Virtual 305 LAN (VLAN). (See section 3 of [RFC7432] for a discussion of this 306 terminology.) Thus each VLAN is treated as a BD. 308 o In the "vlan bundle service", each MAC-VRF contains one bridge 309 table, where the bridge table corresponds to a set of VLANs. Thus 310 a set of VLANs are treated as constituting a single BD. 312 o In the "vlan-aware bundle service", each MAC-VRF may contain 313 multiple bridge tables, where each bridge table corresponds to one 314 BD. If a MAC-VRF contains several bridge tables, then it 315 corresponds to several BDs. 317 The procedures of this document are intended to work for all these 318 service models. 320 1.2. Need for EVPN-aware Multicast Procedures 322 Inter-subnet IP multicast among a set of BDs can be achieved, in a 323 non-optimal manner, without any specific EVPN procedures. For 324 instance, if a particular tenant has n BDs among which he wants to 325 send IP multicast traffic, he can simply attach a conventional 326 multicast router to all n BDs. Or more generally, as long as each BD 327 has at least one IP multicast router, and the IP multicast routers 328 communicate multicast control information with each other, 329 conventional IP multicast procedures will work normally, and no 330 special EVPN functionality is needed. 332 However, that technique does not provide optimal routing for 333 multicast. In conventional multicast routing, for a given multicast 334 flow, there is only one multicast router on each BD that is permitted 335 to send traffic of that flow to the BD. If that BD has receivers for 336 a given flow, but the source of the flow is not on that BD, then the 337 flow must pass through that multicast router. This leads to the 338 "hair-pinning" problem described (for unicast) in Appendix A. 340 For example, consider an (S,G) flow that is sourced by a TS S and 341 needs to be received by TSes R1 and R2. Suppose S is on a segment of 342 BD1, R1 is on a segment of BD2, but both are attached to PE1. 343 Suppose also that the tenant has a multicast router, attached to a 344 segment of BD1 and to a segment of BD2. However, the segments to 345 which that router is attached are both attached to PE2. Then the 346 flow from S to R would have to follow the path: 347 S-->PE1-->PE2-->Tenant Multicast Router-->PE2-->PE1-->R1. Obviously, 348 the path S-->PE1-->R would be preferred. 350 Now suppose that there is a second receiver, R2. R2 is attached to a 351 third BD, BD3. However, it is attached to a segment of BD3 that is 352 attached to PE1. And suppose also that the Tenant Multicast Router 353 is attached to a segment of BD3 that attaches to PE2. In this case, 354 the Tenant Multicast Router will make two copies of the packet, one 355 for BD2 and one for BD3. PE2 will send both copies back to PE1. Not 356 only is the routing sub-optimal, but PE2 sends multiple copies of the 357 same packet to PE1. This is a further sub-optimality. 359 This is only an example; many more examples of sub-optimal multicast 360 routing can easily be given. To eliminate sub-optimal routing and 361 extra copies, it is necessary to have a multicast solution that is 362 EVPN-aware, and that can use its knowledge of the internal structure 363 of a Tenant Domain to ensure that multicast traffic gets routed 364 optimally. The procedures of this document allow us to avoid all 365 such sub-optimalities when routing inter-subnet multicasts within a 366 Tenant Domain. 368 1.3. Additional Requirements That Must be Met by the Solution 370 In addition to providing optimal routing of multicast flows within a 371 Tenant Domain, the EVPN-aware multicast solution is intended to 372 satisfy the following requirements: 374 o The solution must integrate well with the procedures specified in 375 [IGMP-Proxy]. That is, an integrated set of procedures must 376 handle both intra-subnet multicast and inter-subnet multicast. 378 o With regard to intra-subnet multicast, the solution MUST maintain 379 the integrity of multicast ethernet service. This means: 381 * If a source and a receiver are on the same subnet, the MAC 382 source address (SA) of the multicast frame sent by the source 383 will not get rewritten. 385 * If a source and a receiver are on the same subnet, no IP 386 processing of the ethernet payload is done. The IP TTL is not 387 decremented, the header checksum is not changed, no 388 fragmentation is done, etc. 390 o On the other hand, if a source and a receiver are on different 391 subnets, the frame received by the receiver will not have the MAC 392 Source address of the source, as the frame will appear to have 393 come from a multicast router. Also, proper processing of the IP 394 header is done, e.g., TTL decrement by 1, header checksum 395 modification, possibly fragmentation, etc. 397 o If a Tenant Domain contains several BDs, it MUST be possible for a 398 multicast flow (even when the multicast group address is an "any 399 source multicast" (ASM) address), to have sources in one of those 400 BDs and receivers in one or more of the other BDs, without 401 requiring the presence of any system performing PIM Rendezvous 402 Point (RP) functions ([RFC7761]). Multicast throughout a Tenant 403 Domain must not require the tenant systems to be aware of any 404 underlying multicast infrastructure. 406 o Sometimes a MAC address used by one TS on a particular BD is also 407 used by another TS on a different BD. Inter-subnet routing of 408 multicast traffic MUST NOT make any assumptions about the 409 uniqueness of a MAC address across several BDs. 411 o If two EVPN-PEs attached to the same Tenant Domain both support 412 the OISM procedures, each may receive inter-subnet multicasts from 413 the other, even if the egress PE is not attached to any segment of 414 the BD from which the multicast packets are being sourced. It 415 MUST NOT be necessary to provision the egress PE with knowledge of 416 the ingress BD. 418 o There must be a procedure that that allows EVPN-PE routers 419 supporting OISM procedures to send/receive multicast traffic to/ 420 from EVPN-PE routers that support only [RFC7432], but that do not 421 support the OISM procedures or even the procedures of [EVPN-IRB]. 422 However, when interworking with such routers (which we call 423 "non-OISM PE routers"), optimal routing may not be achievable. 425 o It MUST be possible to support scenarios in which multicast flows 426 with sources inside a Tenant Domain have "external" receivers, 427 i.e., receivers that are outside the domain. It must also be 428 possible to support scenarios where multicast flows with external 429 sources (sources outside the Tenant Domain) have receivers inside 430 the domain. 432 This presupposes that unicast routes to multicast sources outside 433 the domain can be distributed to EVPN-PEs attached to the domain, 434 and that unicast routes to multicast sources within the domain can 435 be distributed outside the domain. 437 Of particular importance are the scenario in which the external 438 sources and/or receivers are reachable via L3VPN/MVPN, and the 439 scenario in which external sources and/or receivers are reachable 440 via IP/PIM. 442 The solution for external interworking MUST allow for deployment 443 scenarios in which EVPN does not need to export a host route for 444 every multicast source. 446 o The solution for external interworking must not presuppose that 447 the same tunneling technology is used within both the EVPN domain 448 and the external domain. For example, MVPN interworking must be 449 possible when MVPN is using MPLS P2MP tunneling, and EVPN is using 450 Ingress Replication or VXLAN tunneling. 452 o The solution must not be overly dependent on the details of a 453 small set of use cases, but must be adaptable to new use cases as 454 they arise. (That is, the solution must be robust.) 456 1.4. Terminology 458 In this document we make frequent use of the following terminology: 460 o OISM: Optimized Inter-Subnet Multicast. EVPN-PEs that follow the 461 procedures of this document will be known as "OISM" PEs. EVPN-PEs 462 that do not follow the procedures of this document will be known 463 as "non-OISM" PEs. 465 o IP Multicast Packet: An IP packet whose IP Destination Address 466 field is a multicast address that is not a link-local address. 467 (Link-local addresses are IPv4 addresses in the 224/8 range and 468 IPv6 address in the FF02/16 range.) 470 o IP Multicast Frame: An ethernet frame whose payload is an IP 471 multicast packet (as defined above). 473 o (S,G) Multicast Packet: An IP multicast packet whose IP Source 474 Address field contains S and whose IP Destination Address field 475 contains G. 477 o (S,G) Multicast Frame: An IP multicast frame whose payload 478 contains S in its IP Source Address field and G in its IP 479 Destination Address field. 481 o Broadcast Domain (BD): an emulated ethernet, such that two systems 482 on the same BD will receive each other's link-local broadcasts. 484 Note that EVPN supports service models in which a single EVPN 485 Instance (EVI) contains only one BD, and service models in which a 486 single EVI contains multiple BDs. Both types of service model are 487 supported by this draft. In all models, a given BD belongs to 488 only one EVI. 490 o Designated Forwarder (DF). As defined in [RFC7432], an ethernet 491 segment may be multi-homed (attached to more than one PE). An 492 ethernet segment may also contain multiple BDs, of one or more 493 EVIs. For each such EVI, one of the PEs attached to the segment 494 becomes that EVI's DF for that segment. Since a BD may belong to 495 only one EVI, we can speak unambiguously of the BD's DF for a 496 given segment. 498 When the text makes it clear that we are speaking in the context 499 of a given BD, we will frequently use the term "a segment's DF" to 500 mean the given BD's DF for that segment. 502 o AC: Attachment Circuit. An AC connects the bridging function of 503 an EVPN-PE to an ethernet segment of a particular BD. ACs are not 504 visible at the router (L3) layer. 506 If a given ethernet segment, attached to a given PE, contains n 507 BDs, we will say that the PE has n ACs to that segment. 509 o L3 Gateway: An L3 Gateway is a PE that connects an EVPN tenant 510 domain to an external multicast domain by performing both the OISM 511 procedures and the Layer 3 multicast procedures of the external 512 domain. 514 o PEG (PIM/EVPN Gateway): A L3 Gateway that connects an EVPN Tenant 515 Domain to an external multicast domain whose Layer 3 multicast 516 procedures are those of PIM ([RFC7761]). 518 o MEG (MVPN/EVPN Gateway): A L3 Gateway that connects an EVPN Tenant 519 Domain to an external multicast domain whose Layer 3 multicast 520 procedures are those of MVPN ([RFC6513], [RFC6514]). 522 o IPMG (IP Multicast Gateway): A PE that is used for interworking 523 OISM EVPN-PEs with non-OISM EVPN-PEs. 525 o DR (Designated Router): A PE that has special responsibilities for 526 handling multicast on a given BD. 528 o FHR (First Hop Router): The FHR is a PIM router ([RFC7761]) with 529 special responsibilities. It is the first multicast router to see 530 (S,G) packets from source S, and if G is an "Any Source Multicast 531 (ASM)" group, the FHR is responsible for sending PIM Register 532 messages to the PIM Rendezvous Point for group G. 534 o LHR (Last Hop Router): The LHR is a PIM router ([RFC7761]) with 535 special responsibilities. Generally it is attached to a LAN, and 536 it determines whether there are any hosts on the LAN that need to 537 receive a given multicast flow. If so, it creates and sends the 538 PIM Join messages that are necessary to draw the flow. 540 o EC (Extended Community). A BGP Extended Communities attribute 541 ([RFC4360], [RFC7153]) is a BGP path attribute that consists of 542 one or more extended communities. 544 o RT (Route Target): A Route Target is a particular kind of BGP 545 Extended Community. A BGP Extended Community consists of a type 546 field, a sub-type field, and a value field. Certain type/sub-type 547 combinations indicate that a particular Extended Community is an 548 RT. RT1 and RT2 are considered to be the same RT if and only if 549 they have the same type, same sub-type, and same value fields. 551 o Use of the "C-" prefix. In many documents on VPN multicast, the 552 prefix "C-" appears before any address or wildcard that refers to 553 an address or addresses in a tenant's address space, rather than 554 to an address of addresses in the address space of the backbone 555 network. This document omits the "C-" prefix in many cases where 556 it is clear from the context that the reference is to the tenant's 557 address space. 559 This document also assumes familiarity with the terminology of 560 [RFC4364], [RFC6514], [RFC7432], [RFC7761], [IGMP-Proxy], 561 [EVPN_IP_Prefix] and [EVPN-BUM]. 563 1.5. Model of Operation: Overview 565 1.5.1. Control Plane 567 In this section, and in the remainder of this document, we assume the 568 reader is familiar with the procedures of IGMP/MLD (see [RFC2236] and 569 [RFC2710]), by which hosts announce their interest in receiving 570 particular multicast flows. 572 Consider a Tenant Domain consisting of a set of k BDs: BD1, ..., BDk. 573 To support the OISM procedures, each Tenant Domain must also be 574 associated with a "Supplementary Broadcast Domain" (SBD). An SBD is 575 treated in the control plane as a real BD, but it does not have any 576 ACs. The SBD has several uses; these will be described later in this 577 document (see Section 2.1 and Section 3). 579 Each PE that attaches to one or more of the BDs in a given tenant 580 domain will be provisioned to recognize that those BDs are part of 581 the same Tenant Domain. Note that a given PE does not need to be 582 configured with all the BDs of a given Tenant Domain. In general, a 583 PE will only be attached to a subset of the BDs in a given Tenant 584 Domain, and will be configured only with that subset of BDs. 585 However, each PE attached to a given Tenant Domain must be configured 586 with the SBD for that Tenant Domain. 588 Suppose a particular segment of a particular BD is attached to PE1. 589 [RFC7432] specifies that PE1 must originate an Inclusive Multicast 590 Ethernet Tag (IMET) route for that BD, and that the IMET route must 591 be propagated to all other PEs attached to the same BD. If the given 592 segment contains a host that has interest in receiving a particular 593 multicast flow, either an (S,G) flow or a (*,G) flow, PE1 will learn 594 of that interest by participating in the IGMP/MLD procedures, as 595 specified in [IGMP-Proxy]. In this case, we will say that: 597 o PE1 is interested in receiving the flow; 599 o The AC attaching the interested host to PE1 is also said to be 600 interested in the flow; 602 o The BD containing an AC that is interested in a particular flow is 603 also said to be interested in that flow. 605 Once PE1 determines that it has an AC that is interested in receiving 606 a particular flow or set of flows, it originates one or more 607 Selective Multicast Ethernet Tag (SMET) route to advertise that 608 interest. 610 Note that each IMET or SMET route is "for" a particular BD. The 611 notion of a route being "for" a particular BD is explained in 612 Section 2.2. 614 When OISM is being supported, the procedures of [IGMP-Proxy], are 615 modified as follows: 617 o The IMET route originated by a particular PE for a particular BD 618 is distributed to all other PEs attached to the Tenant Domain 619 containing that BD, even to those PEs that are not attached to 620 that particular BD. 622 o The SMET routes originated by a particular PE are originated on a 623 per-Tenant-Domain basis, rather than on a per-BD basis. That is, 624 the SMET routes are considered to be for the Tenant Domain's SBD, 625 rather than for any of its ordinary BDs. These SMET routes are 626 distributed to all the PEs attached to the Tenant Domain. 628 In this way, each PE attached to a given Tenant Domain learns, 629 from each other PE attached to the same Tenant Domain, the set of 630 flows that are of interest to each of those other PEs. 632 An OISM PE that is provisioned with several BDs in the same Tenant 633 Domain MUST originate an IMET route for each such BD. To indicate 634 its support of [IGMP-Proxy], it SHOULD attach the EVPN Multicast 635 Flags Extended Community to each such IMET route, but it MUST attach 636 the EC to at least one such IMET route. 638 Suppose PE1 is provisioned with both BD1 and BD2, and is provisioned 639 to consider them to be part of the same Tenant Domain. It is 640 possible that PE1 will receive from PE2 both an IMET route for BD1 641 and an IMET route for BD2. If either of these IMET routes has the 642 EVPN Multicast Flags Extended Community, PE1 MUST assume that PE2 is 643 supporting the procedures of [IGMP-Proxy] for ALL BDs in the Tenant 644 Domain. 646 If a PE supports OISM functionality, it indicates that by setting the 647 "OISM-supported" flag in the Multicast Flags Extended Community that 648 it attaches to some or all of its IMET routes. An OISM PE SHOULD 649 attach this EC with the OISM-supported flag set to all the IMET 650 routes it originates. However, if PE1 imports IMET routes from PE2, 651 and at least one of PE2's IMET routes indicates that PE2 is an OISM 652 PE, PE1 MUST assume that PE2 is following OISM procedures. 654 1.5.2. Data Plane 656 Suppose PE1 has an AC to a segment in BD1, and PE1 receives from that 657 AC an (S,G) multicast frame (as defined in Section 1.4). 659 There may be other ACs of PE1 on which TSes have indicated an 660 interest (via IGMP/MLD) in receiving (S,G) multicast packets. PE1 is 661 responsible for sending the received multicast packet out those ACs. 662 There are two cases to consider: 664 o Intra-Subnet Forwarding: In this case, an attachment AC with 665 interest in (S,G) is connected to a segment that is part of the 666 source BD, BD1. If the segment is not multi-homed, or if PE1 is 667 the Designated Forwarder (DF) (see [RFC7432]) for that segment, 668 PE1 sends the multicast frame on that AC without changing the MAC 669 SA. The IP header is not modified at all; in particular, the TTL 670 is not decremented. 672 o Inter-Subnet Forwarding: An AC with interest in (S,G) is connected 673 to a segment of BD2, where BD2 is different than BD1. If PE1 is 674 the DF for that segment (or if the segment is not multi-homed), 675 PE1 decapsulates the IP multicast packet, performs any necessary 676 IP processing (including TTL decrement), then re-encapsulates the 677 packet appropriately for BD2. PE1 then sends the packet on the 678 AC. Note that after re-encapsulation, the MAC SA will be PE1's 679 MAC address on BD2. The IP TTL will have been decremented by 1. 681 In addition, there may be other PEs that are interested in (S,G) 682 traffic. Suppose PE2 is such a PE. Then PE1 tunnels a copy of the 683 IP multicast frame (with its original MAC SA, and with no alteration 684 of the payload's IP header) to PE2. The tunnel encapsulation 685 contains information that PE2 can use to associate the frame with an 686 "apparent source BD". If the actual source BD of the frame is BD1, 687 then: 689 o If PE2 is attached to BD1, the tunnel encapsulation used to send 690 the frame to PE2 will cause PE2 to identify BD1 as the apparent 691 source BD. 693 o If PE2 is not attached to BD1, the tunnel encapsulation used to 694 send the frame to PE2 will cause PE2 to identify the SBD as the 695 apparent source BD. 697 Note that the tunnel encapsulation used for a particular BD will have 698 been advertised in an IMET route or S-PMSI route ([EVPN-BUM]) for 699 that BD. That route carries a PMSI Tunnel attribute, which specifies 700 how packets originating from that BD are encapsulated. This 701 information enables the PE receiving a tunneled packet to identify 702 the apparent source BD as stated above. See Section 3.2 for more 703 details. 705 When PE2 receives the tunneled frame, it will forward it on any of 706 its ACs that have interest in (S,G). 708 If PE2 determines from the tunnel encapsulation that the apparent 709 source BD is BD1, then 711 o For those ACs that connect PE2 to BD1, the intra-subnet forwarding 712 procedure described above is used, except that it is now PE2, not 713 PE1, carrying out that procedure. Unmodified EVPN procedures from 715 [RFC7432] are used to ensure that a packet originating from a 716 multi-homed segment is never sent back to that segment. 718 o For those ACs that do not connect to BD1, the inter-subnet 719 forwarding procedure described above is used, except that it is 720 now PE2, not PE1, carrying out that procedure. 722 If the tunnel encapsulation identifies the apparent source BD as the 723 SBD, PE2 applies the inter-subnet forwarding procedures described 724 above to all of its ACs that have interest in the flow. 726 These procedures ensure that an IP multicast frame travels from its 727 ingress PE to all egress PEs that are interested in receiving it. 728 While in transit, the frame retains its original MAC SA, and the 729 payload of the frame retains its original IP header. Note that in 730 all cases, when an IP multicast packet is sent from one BD to 731 another, these procedures cause its TTL to be decremented by 1. 733 So far we have assumed that an IP multicast packet arrives at its 734 ingress PE over an AC that belongs to one of the BDs in a given 735 Tenant Domain. However, it is possible for a packet to arrive at its 736 ingress PE in other ways. Since an EVPN-PE supporting IRB has an 737 IP-VRF, it is possible that the IP-VRF will have a "VRF interface" 738 that is not an IRB interface. For example, there might be a VRF 739 interface that is actually a physical link to an external ethernet 740 switch, or to a directly attached host, or to a router. When an 741 EVPN-PE, say PE1, receives a packet through such means, we will say 742 that the packet has an "external" source (i.e., a source "outside the 743 Tenant Domain"). There are also other scenarios in which a multicast 744 packet might have an external source, e.g., it might arrive over an 745 MVPN tunnel from an L3VPN PE. In such cases, we will still refer to 746 PE1 as the "ingress EVPN-PE". 748 When an EVPN-PE, say PE1, receives an externally sourced multicast 749 packet, and there are receivers for that packet inside the Tenant 750 Domain, it does the following: 752 o Suppose PE1 has an AC in BD1 that has interest in (S,G). Then PE1 753 encapsulates the packet for BD1, filling in the MAC SA field with 754 PE1's own MAC address on BD1. It sends the resulting frame on the 755 AC. 757 o Suppose some other EVPN-PE, say PE2, has interest in (S,G). PE1 758 encapsulates the packet for ethernet, filling in the MAC SA field 759 with PE1's own MAC address on the SBD. PE1 then tunnels the 760 packet to PE2. The tunnel encapsulation will identify the 761 apparent source BD as the SBD. Since the apparent source BD is 762 the SBD, PE2 will know to treat the frame as an inter-subnet 763 multicast. 765 When ingress replication is used to transmit IP multicast frames from 766 an ingress EVPN-PE to a set of egress PEs, then of course the ingress 767 PE has to send multiple copies of the frame. Each copy is the 768 original ethernet frame; decapsulation and IP processing take place 769 only at the egress PE. 771 If a Point-to-Multipoint (P2MP) tree or BIER ([EVPN-BIER]) is used to 772 transmit an IP multicast frame from an ingress PE to a set of egress 773 PEs, then the ingress PE only has to send one copy of the frame to 774 each of its next hops. Again, each egress PE receives the original 775 frame and does any necessary IP processing. 777 2. Detailed Model of Operation 779 The model described in Section 1.5.2 can be expressed more precisely 780 using the notion of "IRB interface" (see Appendix A). For a given 781 Tenant Domain: 783 o A given PE has one IRB for each BD to which it is attached. This 784 IRB interface connects L3 routing to that BD. When IP multicast 785 packets are sent or received on the IRB interfaces, the semantics 786 of the interface is modified from the semantics described in 787 Appendix A. See Section 2.3 for the details of the modification. 789 o Each PE also has an IRB interface that connects L3 routing to the 790 SBD. The semantics of this interface is different than the 791 semantics of the IRB interface to the real BDs. See Section 2.3. 793 In this section we assume that PIM is not enabled on the IRB 794 interfaces. In general, it is not necessary to enable PIM on the IRB 795 interfaces unless there are PIM routers on one of the Tenant Domain's 796 BDs, or unless there is some other scenario requiring a Tenant 797 Domain's L3 routing instance to become a PIM adjacency of some other 798 system. These cases will be discussed in Section 7. 800 2.1. Supplementary Broadcast Domain 802 Suppose a given Tenant Domain contains three BDs (BD1, BD2, BD3) and 803 two PEs (PE1, PE2). PE1 attaches to BD1 and BD2, while PE2 attaches 804 to BD2 and BD3. 806 To carry out the procedures described above, all the PEs attached to 807 the Tenant Domain must be provisioned with the SBD for that tenant 808 domain. A Route Target (RT) must be associated with the SBD, and 809 provisioned on each of those PEs. We will refer to that RT as the 810 "SBD-RT". 812 A Tenant Domain is also configured with an IP-VRF ([EVPN-IRB]), and 813 the IP-VRF is associated with an RT. This RT MAY be the same as the 814 SBD-RT. 816 Suppose an (S,G) multicast frame originating on BD1 has a receiver on 817 BD3. PE1 will transmit the packet to PE2 as a frame, and the 818 encapsulation will identify the frame's source BD as BD1. Since PE2 819 is not provisioned with BD1, it will treat the packet as if its 820 source BD were the SBD. That is, a packet can be transmitted from 821 BD1 to BD3 even though its ingress PE is not configured for BD3, and/ 822 or its egress PE is not configured for BD1. 824 EVPN supports service models in which a given EVPN Instance (EVI) can 825 contain only one BD. It also supports service models in which a 826 given EVI can contain multiple BDs. No matter which service model is 827 being used for a particular tenant, it is highly RECOMMENDED that an 828 EVI containing only the SBD be provisioned for that tenant. 830 If, for some reason, it is not feasible to provision an EVI that 831 contains only the SBD, it is possible to put the SBD in an EVI that 832 contains other BDs. However, in that case, the SBD-RT MUST be 833 different than the RT associated with any other BD. Otherwise the 834 procedures of this document (as detailed in Sections 2.2 and 3.1) 835 will not produce correct results. 837 2.2. Detecting When a Route is About/For/From a Particular BD 839 In this document, we frequently say that a particular multicast route 840 is "about" a particular BD, or is "from" a particular BD, or is "for" 841 a particular BD or is "related to" a particular BD or "is associated 842 with" a particular BD. These terms are used interchangeably. 843 Subsequent sections of this document explain when various routes must 844 be originated for particular BDs. In this section, we explain how 845 the PE originating a route marks the route to indicate which BD it is 846 about. We also explain how a PE receiving the route determines which 847 BD the route is about. 849 In EVPN, each BD is assigned a Route Target (RT). An RT is a BGP 850 extended community that can be attached to the BGP routes used by the 851 EVPN control plane. In some EVPN service models, each BD is assigned 852 a unique RT. In other service models, a set of BDs (all in the same 853 EVI) may be assigned the same RT. The RT that is assigned to the SBD 854 is called the "SBD-RT". 856 In those service models that allow a set of BDs to share a single RT, 857 each BD is assigned a non-zero Tag ID. The Tag ID appears in the 858 Network Layer Reachability Information (NLRI) of many of the BGP 859 routes that are used by the EVPN control plane. 861 A given route may be about the SBD, or about an "ordinary BD" (a BD 862 that is not the SBD). An RT that has been assigned to an ordinary BD 863 will be known as an "ordinary BD-RT". 865 When constructing an IMET, SMET, S-PMSI ([EVPN-BUM]), or Leaf 866 ([EVPN-BUM]) route that is about a given BD, the following rules 867 apply: 869 o If the route is about an ordinary BD, say BD1, then 871 * the route MUST carry the ordinary BD-RT associated with BD1, 872 and 874 * the route MUST NOT carry any RT that is associated with an 875 ordinary BD other than BD1. 877 o If the route is about the SBD, the route MUST carry the SBD-RT, 878 and MUST NOT carry any RT that is associated with any other BD. 880 o As detailed in subsequent sections, under certain circumstances a 881 route that is about BD1 may carry both the RT of BD1 and also the 882 SBD-RT. 884 The IMET route for the SBD MUST carry an Multicast Flags Extended 885 Community, in which an "OISM SBD" flag is set. 887 When receiving an IMET, SMET, S-PMSI or Leaf route, it is necessary 888 for the receiving PE to determine the BD to which the route belongs. 889 This is done by examining the RTs carried by the route, as well as 890 the Tag ID field of the route's NLRI. There are several cases to 891 consider. Some of these cases are error cases that arise when the 892 route has not been properly constructed. 894 When one of the error cases is detected, the route MUST be regarded 895 as a malformed route, and the "treat-as-withdraw" procedure of 896 [RFC7606] MUST be applied. Note though that these error cases are 897 only detectable by EVPN procedures at the receiving PE; BGP 898 procedures at intermediate nodes will generally not detect the 899 existence of such error cases, and in general SHOULD NOT attempt to 900 do so. 902 Case 1: The receiving PE recognizes more than one of the route's RTs 903 as being an SBD-RT (i.e., the route carries SBD-RTs of more 904 than one Tenant Domain). 906 This is an error case; the route has not been properly 907 constructed. 909 Case 2: The receiving PE recognizes one of the route's RTs as being 910 associated with an ordinary BD, and recognizes one of the 911 route's other RTs as being associated with a different 912 ordinary BD. 914 This is an error case; the route has not been properly 915 constructed. 917 Case 3: The receiving PE recognizes one of the route's RTs as being 918 associated with an ordinary BD in a particular Tenant 919 Domain, and recognizes another of the route's RTs as being 920 associated with the SBD of a different Tenant Domain. 922 This is an error case; the route has not been properly 923 constructed. 925 Case 4: The receiving PE does not recognize any of the route's RTs 926 as being associated with an ordinary BD in any of its tenant 927 domains, but does recognize one of the RTs as the SBD-RT of 928 one of its Tenant Domains. 930 In this case, receiving PE associates the route with the SBD 931 of that Tenant Domain. This association is made even if the 932 Tag ID field of the route's NLRI is not the Tag ID of the 933 SBD. 935 This is a normal use case where either (a) the route is for 936 a BD to which the receiving PE is not attached, or (b) the 937 route is for the SBD. In either case, the receiving PE 938 associates the route with the SBD. 940 Case 5: The receiving PE recognizes exactly one of the RTs as an 941 ordinary BD-RT that is associated with one of the PE's EVIs, 942 say EVI-1. The receiving PE also recognizes one of the RTs 943 as being the SBD-RT of the Tenant Domain containing EVI-1. 945 In this case, the route is associated with the BD in EVI-1 946 that is identified (in the context of EVI-1) by the Tag ID 947 field of the route's NLRI. (If EVI-1 contains only a single 948 BD, the Tag ID is likely to be zero.) 949 This is the case where the route is for a BD to which the 950 receiving PE is attached, but the route also carries the 951 SBD-RT. In this case, the receiving PE associates the route 952 with the ordinary BD, not with the SBD. 954 N.B.: According to the above rules, the mapping from BD to RT is a 955 many-to-one or one-to-one mapping. A route that an EVPN-PE 956 originates for a particular BD carries that BD's RT, and an EVPN-PE 957 that receives the route associates it with a BD as described above. 958 However, RTs are not used only to help identify the BD to which a 959 route belongs; they may also used by BGP to determine the path along 960 which the route is distributed, and to determine which PEs receive 961 the route. There may be cases where it is desirable to originate a 962 route about a particular BD, but have that route distributed to only 963 some of the EVPN-PEs attached to that BD. Or one might want the 964 route distributed to some intermediate set of systems, where it might 965 be modified or replaced before being propagated further. Such 966 situations are outside the scope of this document. 968 Additionally, there may be situations where it is desirable to 969 exchange routes among two or more different Tenant Domains ("EVPN 970 Extranet"). Such situations are outside the scope of this document. 972 2.3. Use of IRB Interfaces at Ingress PE 974 When an (S,G) multicast frame is received from an AC belonging to a 975 particular BD, say BD1: 977 1. The frame is sent unchanged to other EVPN-PEs that are interested 978 in (S,G) traffic. The encapsulation used to send the frame to 979 the other EVPN-PEs depends on the tunnel type being used for 980 multicast transmission. (For our purposes, we consider Ingress 981 Replication (IR), Assisted Replication (AR) and BIER to be 982 "tunnel types", even though IR, AR and BIER do not actually use 983 P2MP tunnels.) At the egress PE, the apparent source BD of the 984 frame can be inferred from the tunnel encapsulation. If the 985 egress PE is not attached to the actual source BD, it will infer 986 that the apparent source BD is the SBD. 988 Note that the the inter-PE transmission of a multicast frame 989 among EVPN-PEs of the same Tenant Domain does NOT involve the IRB 990 interfaces, as long as the multicast frame was received over an 991 AC attached to one of the Tenant Domain's BDs. 993 2. The frame is also sent up the IRB interface that attaches BD1 to 994 the Tenant Domain's L3 routing instance in this PE. That is, the 995 L3 routing instance, behaving as if it were a multicast router, 996 receives the IP multicast frames that arrive at the PE from its 997 local ACs. The L3 routing instance decapsulates the frame's 998 payload to extract the IP multicast packet, decrements the IP 999 TTL, adjusts the header checksum, and does any other necessary IP 1000 processing (e.g., fragmentation). 1002 3. The L3 routing instance keeps track of which BDs have local 1003 receivers for (S,G) traffic. (A "local receiver" is a TS, 1004 reachable via a local AC, that has expressed interest in (S,G) 1005 traffic.) If the L3 routing instance has an IRB interface to 1006 BD2, and it knows that BD2 has a LOCAL receiver interested in 1007 (S,G) traffic, it encapsulates the packet in an ethernet header 1008 for BD2, putting its own MAC address in the MAC SA field. Then 1009 it sends the packet down the IRB interface to BD2. 1011 If a packet is sent from the L3 routing instance to a particular BD 1012 via the IRB interface (step 3 in the above list), and if the BD in 1013 question is NOT the SBD, the packet is sent ONLY to LOCAL ACs of that 1014 BD. If the packet needs to go to other PEs, it has already been sent 1015 to them in step 1. Note that this is a change in the IRB interface 1016 semantics from what is described in [EVPN-IRB] and Figure 2. 1018 If a given locally attached segment is multi-homed, existing EVPN 1019 procedures ensure that a packet is not sent by a given PE to that 1020 segment unless the PE is the DF for that segment. Those procedures 1021 also ensure that a packet is never sent by a PE to its segment of 1022 origin. Thus EVPN segment multi-homing is fully supported; duplicate 1023 delivery to a segment or looping on a segment are thereby prevented, 1024 without the need for any new procedures to be defined in this 1025 document. 1027 What if an IP multicast packet is received from outside the tenant 1028 domain? For instance, perhaps PE1's IP-VRF for a particular tenant 1029 domain also has a physical interface leading to an external switch, 1030 host, or router, and PE1 receives an IP multicast packet or frame on 1031 that interface. Or perhaps the packet is from an L3VPN, or a 1032 different EVPN Tenant Domain. 1034 Such a packet is first processed by the L3 routing instance, which 1035 decrements TTL and does any other necessary IP processing. Then the 1036 packet is sent into the Tenant Domain by sending it down the IRB 1037 interface to the SBD of that Tenant Domain. This requires 1038 encapsulating the packet in an ethernet header. The MAC SA field 1039 will contain the PE's own MAC on the SBD. 1041 An IP multicast packet sent by the L3 routing instance down the IRB 1042 interface to the SBD is treated as if it had arrived from a local AC, 1043 and steps 1-3 are applied. Note that the semantics of sending a 1044 packet down the IRB interface to the SBD are thus slightly different 1045 than the semantics of sending a packet down other IRB interfaces. IP 1046 multicast packets sent down the SBD's IRB interface may be 1047 distributed to other PEs, but IP multicast packets sent down other 1048 IRB interfaces are distributed only to local ACs. 1050 If a PE sends a link-local multicast packet down the SBD IRB 1051 interface, that packet will be distributed (as an ethernet frame) to 1052 other PEs of the Tenant Domain, but will not appear on any of the 1053 actual BDs. 1055 2.4. Use of IRB Interfaces at an Egress PE 1057 Suppose an egress EVPN-PE receives an (S,G) multicast frame from the 1058 frame's ingress EVPN-PE. As described above, the packet will arrive 1059 as an ethernet frame over a tunnel from the ingress PE, and the 1060 tunnel encapsulation will identify the source BD of the ethernet 1061 frame. 1063 We define the notion of the frame's "apparent source BD" as follows. 1064 If the egress PE is attached to the actual source BD, the actual 1065 source BD is the apparent source BD. If the egress PE is not 1066 attached to the actual source BD, the SBD is the apparent source BD. 1068 The egress PE now takes the following steps: 1070 1. If the egress PE has ACs belonging to the apparent source BD of 1071 the frame, it sends the frame unchanged to any ACs of that BD 1072 that have interest in (S,G) packets. The MAC SA of the frame is 1073 not modified, and the IP header of the frame's payload is not 1074 modified in any way. 1076 2. The frame is also sent to the L3 routing instance by being sent 1077 up the IRB interface that attaches the L3 routing instance to the 1078 apparent source BD. Steps 2 and 3 of Section 2.3 are then 1079 applied. 1081 2.5. Announcing Interest in (S,G) 1083 [IGMP-Proxy] defines procedures used by an egress PE to announce its 1084 interest in a multicast flow or set of flows. If an egress PE 1085 determines it has LOCAL receivers in a particular BD, say BD1, that 1086 are interested in a particular set of flows, it originates one or 1087 more SMET routes for BD1. Each SMET route specifies a particular 1088 (S,G) or (*,G) flow. By originating an SMET route for BD1, a PE is 1089 announcing "I have receivers for (S,G) or (*,G) in BD1". Such an 1090 SMET route carries the Route Target (RT) for BD1, ensuring that it 1091 will be distributed to all PEs that are attached to BD1. 1093 The OISM procedures for originating SMET routes differ slightly from 1094 those in [IGMP-Proxy]. In most cases, the SMET routes are considered 1095 to be for the SBD, rather than for the BD containing local receivers. 1096 These SMET routes carry the SBD-RT, and do not carry any ordinary BD- 1097 RT. Details on the processing of SMET routes can be found in 1098 Section 3.3. 1100 Since the SMET routes carry the SBD-RT, every ingress PE attached to 1101 a particular Tenant Domain will learn of all other PEs (attached to 1102 the same Tenant Domain) that have interest in a particular set of 1103 flows. Note that a PE that receives a given SMET route does not 1104 necessarily have any BDs (other than the SBD) in common with the PE 1105 that originates that SMET route. 1107 If all the sources and receivers for a given (*,G) are in the Tenant 1108 Domain, inter-subnet "Any Source Multicast" traffic will be properly 1109 routed without requiring any Rendezvous Points, shared trees, or 1110 other complex aspects of multicast routing infrastructure. Suppose, 1111 for example, that: 1113 o PE1 has a local receiver, on BD1, for (*,G) 1115 o PE2 has a local source, on BD2, for (*,G). 1117 PE1 will originate an SMET(*,G) route for the SBD, and PE2 will 1118 receive that route, even if PE2 is not attached to BD1. PE2 will 1119 thus know to forward (S,G) traffic to PE1. PE1 does not need to do 1120 any "source discovery". (This does assume that source S does not 1121 send the same (S,G) datagram on two different BDs, and that the 1122 Tenant Domain does not contain two or more sources with the same IP 1123 address S. The use of multicast sources that have IP "anycast" 1124 addresses is outside the scope of this document.) 1126 If some PE attached to the Tenant Domain does not support [IGMP- 1127 Proxy], it will be assumed to be interested in all flows. Whether a 1128 particular remote PE supports [IGMP-Proxy] is determined by the 1129 presence of the Multicast Flags Extended Community in its IMET route; 1130 this is specified in [IGMP-Proxy]. 1132 2.6. Tunneling Frames from Ingress PE to Egress PEs 1134 [RFC7432] specifies the procedures for setting up and using "BUM 1135 tunnels". A BUM tunnel is a tunnel used to carry traffic on a 1136 particular BD if that traffic is (a) broadcast traffic, or (b) 1137 unicast traffic with an unknown MAC DA, or (c) ethernet multicast 1138 traffic. 1140 This document allows the BUM tunnels to be used as the default 1141 tunnels for transmitting IP multicast frames. It also allows a 1142 separate set of tunnels to be used, instead of the BUM tunnels, as 1143 the default tunnels for carrying IP multicast frames. Let's call 1144 these "IP Multicast Tunnels". 1146 When the tunneling is done via Ingress Replication or via BIER, this 1147 difference is of no significance. However, when P2MP tunnels are 1148 used, there is a significant advantage to having separate IP 1149 multicast tunnels. 1151 Other things being equal, it is desirable for an ingress PE to 1152 transmit a copy of a given (S,G) multicast frame on only one P2MP 1153 tunnel. All egress PEs interested in (S,G) packets then have to join 1154 that tunnel. If the source BD and PE for an (S,G) frame are BD1 an 1155 PE1 respectively, and if PE2 has receivers on BD2 for (S,G), then PE2 1156 must join the P2MP LSP on which PE1 transmits the (S,G) frame. PE2 1157 must join this P2MP LSP even if PE2 is not attached to the source BD 1158 (BD1). If PE1 were transmitting the multicast frame on its BD1 BUM 1159 tunnel, then PE2 would have to join the BD1 BUM tunnel, even though 1160 PE2 has no BD1 attachment circuits. This would cause PE2 to pull all 1161 the BUM traffic from BD1, most of which it would just have to 1162 discard. Thus we RECOMMEND that the default IP multicast tunnels be 1163 distinct from the BUM tunnels. 1165 Notwithstanding the above, link local IP multicast traffic MUST 1166 always be carried on the BUM tunnels, and ONLY on the BUM tunnels. 1167 Link local IP multicast traffic consists of IPv4 traffic with a 1168 destination address prefix of 224/8 and IPv6 traffic with a 1169 destination address prefix of FF02/16. In this document, the terms 1170 "IP multicast packet" and "IP multicast frame" are defined in 1171 Section 1.4 so as to exclude the link-local traffic. 1173 Note that it is also possible to use "selective tunnels" to carry 1174 particular multicast flows (see Section 3.2). When an (S,G) frame is 1175 transmitted on a selective tunnel, it is not transmitted on the BUM 1176 tunnel or on the default IP Multicast tunnel. 1178 2.7. Advanced Scenarios 1180 There are some deployment scenarios that require special procedures: 1182 1. Some multicast sources or receivers are attached to PEs that 1183 support [RFC7432], but do not support this document or 1184 [EVPN-IRB]. To interoperate with these "non-OISM PEs", it is 1185 necessary to have one or more gateway PEs that interface the 1186 tunnels discussed in this document with the BUM tunnels of the 1187 legacy PEs. This is discussed in Section 5. 1189 2. Sometimes multicast traffic originates from outside the EVPN 1190 domain, or needs to be sent outside the EVPN domain. This is 1191 discussed in Section 6. An important special case of this, 1192 integration with MVPN, is discussed in Section 6.1.2. 1194 3. In some scenarios, one or more of the tenant systems is a PIM 1195 router, and the Tenant Domain is used for as a transit network 1196 that is part of a larger multicast domain. This is discussed in 1197 Section 7. 1199 3. EVPN-aware Multicast Solution Control Plane 1201 3.1. Supplementary Broadcast Domain (SBD) and Route Targets 1203 As discussed in Section 2.1, every Tenant Domain is associated with a 1204 single Supplementary Broadcast Domain (SBD). Recall that a Tenant 1205 Domain is defined to be a set of BDs that can freely send and receive 1206 IP multicast traffic to/from each other. If an EVPN-PE has one or 1207 more ACs in a BD of a particular Tenant Domain, and if the EVPN-PE 1208 supports the procedures of this document, that EVPN-PE MUST be 1209 provisioned with the SBD of that Tenant Domain. 1211 At each EVPN-PE attached to a given Tenant Domain, there is an IRB 1212 interface leading from the L3 routing instance of that Tenant Domain 1213 to the SBD. However, the SBD has no ACs. 1215 Each SBD is provisioned with a Route Target (RT). All the EVPN-PEs 1216 supporting a given SBD are provisioned with that RT as an import RT. 1217 That RT MUST NOT be the same as the RT associated with any other BD. 1219 We will use the term "SBD-RT" to denote the RT has has been assigned 1220 to the SBD. Routes carrying this RT will be propagated to all 1221 EVPN-PEs in the same Tenant Domain as the originator. 1223 Section 2.2 specifies the rules by which an EVPN-PE that receives a 1224 route determines whether a received route "belongs to" a particular 1225 ordinary BD or SBD. 1227 Section 2.2 also specifies additional rules that must be following 1228 when constructing routes that belong to a particular BD, including 1229 the SBD. 1231 The SBD SHOULD be in an EVPN Instance (EVI) of its own. Even if the 1232 SBD is not in an EVI of its own, the SBD-RT MUST be different than 1233 the RT associated with any other BD. This restriction is necessary 1234 in order for the rules of Sections 2.2 and 3.1 to work correctly. 1236 Note that an SBD, just like any other BD, is associated on each 1237 EVPN-PE with a MAC-VRF. Per [RFC7432], each MAC-VRF is associated 1238 with a Route Distinguisher (RD). When constructing a route that is 1239 "about" an SBD, an EVPN-PE will place the RD of the associated 1240 MAC-VRF in the "Route Distinguisher" field of the NLRI. (If the 1241 Tenant Domain has several MAC-VRFs on a given PE, the EVPN-PE has a 1242 choice of which RD to use.) 1244 If Assisted Replication (AR, see [EVPN-AR]) is used, each 1245 AR-REPLICATOR for a given Tenant Domain must be provisioned with the 1246 SBD of that Tenant Domain, even if the AR-REPLICATOR does not have 1247 any L3 routing instance. 1249 3.2. Advertising the Tunnels Used for IP Multicast 1251 The procedures used for advertising the tunnels that carry IP 1252 multicast traffic depend upon the type of tunnel being used. If the 1253 tunnel type is neither Ingress Replication, Assisted Replication, nor 1254 BIER, there are procedures for advertising both "inclusive tunnels" 1255 and "selective tunnels". 1257 When IR, AR or BIER are used to transmit IP multicast packets across 1258 the core, there are no P2MP tunnels. Once an ingress EVPN-PE 1259 determines the set of egress EVPN-PEs for a given flow, the IMET 1260 routes contain all the information needed to transport packets of 1261 that flow to the egress PEs. 1263 If AR is used, the ingress EVPN-PE is also an AR-LEAF and the IMET 1264 route coming from the selected AR-REPLICATOR contains the information 1265 needed. The AR-REPLICATOR will behave as an ingress EVPN-PE when 1266 sending a flow to the egress EVPN-PEs. 1268 If the tunneling technique requires P2MP tunnels to be set up (e.g., 1269 RSVP-TE P2MP, mLDP, PIM), some of the tunnels may be selective 1270 tunnels and some may be inclusive tunnels. 1272 Selective P2MP tunnels are always advertised by the ingress PE using 1273 S-PMSI A-D routes ([EVPN-BUM]). 1275 For inclusive tunnels, there is a choice between using a BD's 1276 ordinary "BUM tunnel" [RFC7432] as the default inclusive tunnel for 1277 carrying IP multicast traffic, or using a separate IP multicast 1278 tunnel as the default inclusive tunnel for carrying IP multicast. In 1279 the former case, the inclusive tunnel is advertised in an IMET route. 1280 In the latter case, the inclusive tunnel is advertised in a (C-*,C-*) 1281 S-PMSI A-D route ([EVPN-BUM]). Details may be found in subsequent 1282 sections. 1284 3.2.1. Constructing Routes for the SBD 1286 There are situations in which an EVPN-PE needs to originate IMET, 1287 SMET, and/or SPMSI routes for the SBD. Throughout this document, we 1288 will refer to such routes respectively as "SBD-IMET routes", 1289 "SBD-SMET routes", and "SBD-SPMSI routes". Subsequent sections 1290 detail the conditions under which these routes need to be originated. 1292 When an EVPN-PE needs to originate an SBD-IMET, SBD-SMET, or 1293 SBD-SPMSI route, it constructs the route as follows: 1295 o the RD field of the route's NLRI is set to the RD of the MAC-VRF 1296 that is associated with the SBD; 1298 o the SBD-RT is attached to the route; 1300 o the "Tag ID" field of the route's NLRI is set to the Tag ID that 1301 has been assigned to the SBD. This is most likely 0 if a 1302 VLAN-based or VLAN-bundle service is being used, but non-zero if a 1303 VLAN-aware bundle service is being used. 1305 3.2.2. Ingress Replication 1307 When Ingress Replication (IR) is used to transport IP multicast 1308 frames of a given Tenant Domain, each EVPN-PE attached to that Tenant 1309 Domain MUST originate an SBD-IMET route (see Section 3.2.1). 1311 The SBD-IMET route MUST carry a PMSI Tunnel attribute (PTA), and the 1312 MPLS label field of the PTA MUST specify a downstream-assigned MPLS 1313 label that maps uniquely (in the context of the originating EVPN-PE) 1314 to the SBD. 1316 Following the procedures of [RFC7432], an EVPN-PE MUST also originate 1317 an IMET route for each BD to which it is attached. Each of these 1318 IMET routes carries a PTA specifying a downstream-assigned label that 1319 maps uniquely, in the context of the originating EVPN-PE, to the BD 1320 in question. These IMET routes need not carry the SBD-RT. 1322 When an ingress EVPN-PE needs to use IR to send an IP multicast frame 1323 from a particular source BD to an egress EVPN-PE, the ingress PE 1324 determines whether the egress PE has originated an IMET route for 1325 that BD. If so, that IMET route contains the MPLS label that the 1326 egress PE has assigned to the source BD. The ingress PE uses that 1327 label when transmitting the packet to the egress PE. Otherwise, the 1328 ingress PE uses the label that the egress PE has assigned to the SBD 1329 (in the SBD-IMET route originated by the egress). 1331 Note that the set of IMET routes originated by a given egress PE, and 1332 installed by a given ingress PE, may change over time. If the egress 1333 PE withdraws its IMET route for the source BD, the ingress PE MUST 1334 stop using the label carried in that IMET route, and instead MUST use 1335 the label carried in the SBD-IMET route from that egress PE. 1336 Implementors must also take into account that an IMET route from a 1337 particular PE for a particular BD may arrive after that PE's SBD-IMET 1338 route. 1340 3.2.3. Assisted Replication 1342 When Assisted Replication is used to transport IP multicast frames of 1343 a given Tenant Domain, each EVPN-PE (including the AR-REPLICATOR) 1344 attached to the Tenant Domain MUST originate an SBD-IMET route (see 1345 Section 3.2.1). 1347 An AR-REPLICATOR attached to a given Tenant Domain is considered to 1348 be an EVPN-PE of that Tenant Domain. It is attached to all the BDs 1349 in the Tenant Domain, but it has no IRB interfaces. 1351 As with Ingress Replication, the SBD-IMET route carries a PTA where 1352 the MPLS label field specifies the downstream-assigned MPLS label 1353 that identifies the SBD. However, the AR-REPLICATOR and AR-LEAF 1354 EVPN-PEs will set the PTA's flags differently, as per [EVPN-AR]. 1356 In addition, each EVPN-PE originates an IMET route for each BD to 1357 which it is attached. As in the case of Ingress Replication, these 1358 routes carry the downstream-assigned MPLS labels that identify the 1359 BDs and do not carry the SBD-RT. 1361 When an ingress EVPN-PE, acting as AR-LEAF, needs to send an IP 1362 multicast frame from a particular source BD to an egress EVPN-PE, the 1363 ingress PE determines whether there is any AR-REPLICATOR that 1364 originated an IMET route for that BD. After the AR-REPLICATOR 1365 selection (if there are more than one), the AR-LEAF uses the label 1366 contained in the IMET route of the AR-REPLICATOR when transmitting 1367 packets to it. The AR-REPLICATOR receives the packet and, based on 1368 the procedures specified in [EVPN-AR], transmits the packets to the 1369 egress EVPN-PEs using the labels contained in the IMET routes 1370 received from the egress PEs. 1372 If an ingress AR-LEAF for a given BD has not received any IMET route 1373 for that BD from an AR-REPLICATOR, the ingress AR-LEAF follows the 1374 procedures in Section 3.2.2. 1376 3.2.4. BIER 1378 When BIER is used to transport multicast packets of a given Tenant 1379 Domain, and a given EVPN-PE attached to that Tenant Domain is a 1380 possible ingress EVPN-PE for traffic originating outside that Tenant 1381 Domain, the given EVPN-PE MUST originate an SBD-IMET route, (see 1382 Section 3.2.1). 1384 In addition, IMET routes that are originated for other BDs in the 1385 Tenant Domain MUST carry the SBD-RT. 1387 Each IMET route (including but not limited to the SBD-IMET route) 1388 MUST carry a PMSI Tunnel attribute (PTA). The MPLS label field of 1389 the PTA MUST specify an upstream-assigned MPLS label that maps 1390 uniquely (in the context of the originating EVPN-PE) to the BD for 1391 which the route is originated. 1393 Suppose an ingress EVPN-PE, PE1, needs to use BIER to tunnel an IP 1394 multicast frame to a set of egress EVPN-PEs. And suppose the frame's 1395 source BD is BD1. The frame is encapsulated as follows: 1397 o A four-octet MPLS label stack entry ([RFC3032]) is prepended to 1398 the frame. The Label field is set to the upstream-assigned label 1399 that PE1 has assigned to BD1. 1401 o The resulting MPLS packet is then encapsulated in a BIER 1402 encapsulation ([RFC8296], [EVPN-BIER]). The BIER BitString is set 1403 to identify the egress EVPN-PEs. The BIER "proto" field is set to 1404 the value for "MPLS packet with upstream-assigned label at top of 1405 stack". 1407 Note: It is possible that the packet being tunneled from PE1 1408 originated outside the Tenant Domain. In this case, the actual 1409 source BD (BD1) is considered to be the SBD, and the 1410 upstream-assigned label it carries will be the label that PE1 1411 assigned to the SBD, and advertised in its SBD-IMET route. 1413 Suppose an egress PE, PE2, receives such a BIER packet. The BFIR-id 1414 field of the BIER header allows PE2 to determine that the ingress PE 1415 is PE1. There are then two cases to consider: 1417 1. PE2 has received and installed an IMET route for BD1 from PE1. 1419 In this case, the BIER packet will be carrying the 1420 upstream-assigned label that is specified in the PTA of that IMET 1421 route. This enables PE2 to determine the "apparent source BD" 1422 (as defined in Section 2.4). 1424 2. PE2 has not received and installed an IMET route for BD1 from 1425 PE1. 1427 In this case, PE2 will not recognize the upstream-assigned label 1428 carried in the BIER packet. PE2 MUST discard the packet. 1430 Further details on the use of BIER to support EVPN can be found in 1431 [EVPN-BIER]. 1433 3.2.5. Inclusive P2MP Tunnels 1435 3.2.5.1. Using the BUM Tunnels as IP Multicast Inclusive Tunnels 1437 The procedures in this section apply only when 1439 (a) it is desired to use the BUM tunnels to carry IP multicast 1440 traffic across the backbone, and 1442 (b) the BUM tunnels are P2MP tunnels (i.e., neither IR, AR, nor BIER 1443 are being used to transport the BUM traffic). 1445 In this case, an IP multicast frame (whether inter-subnet or 1446 intra-subnet) will be carried across the backbone in the BUM tunnel 1447 belonging to its source BD. Each EVPN-PE attached to a given Tenant 1448 Domain needs to join the BUM tunnels for every BD in the Tenant 1449 Domain, even those BDs to which the EVPN-PE is not locally attached. 1450 This ensures that an IP multicast packet from any source BD can reach 1451 all PEs attached to the Tenant Domain. 1453 Note that this will cause all the BUM traffic from a given BD in a 1454 Tenant Domain to be sent to all PEs that attach to that Tenant 1455 Domain, even the PEs that don't attach to the given BD. To avoid 1456 this, it is RECOMMENDED that the BUM tunnels not be used as IP 1457 Multicast inclusive tunnels, and that the procedures of 1458 Section 3.2.5.2 be used instead. 1460 If a PE is a possible ingress EVPN-PE for traffic originating outside 1461 the Tenant Domain, the PE MUST originate an SBD-IMET route (see 1462 Section 3.2.1). This route MUST carry a PTA specifying the P2MP 1463 tunnel used for transmitting IP multicast packets that originate 1464 outside the tenant domain. All EVPN-PEs of the Tenant Domain MUST 1465 join the tunnel specified in the PTA of an SBD-IMET route: 1467 o If the tunnel is an RSVP-TE P2MP tunnel, the originator of the 1468 route MUST use RSVP-TE P2MP procedures to add each PE of the 1469 Tenant Domain to the tunnel, even PEs that have not originated an 1470 SBD-IMET route. 1472 o If the tunnel is an mLDP or PIM tunnel, each PE importing the 1473 SBD-IMET route MUST add itself to the tunnel, using mLDP or PIM 1474 procedures, respectively. 1476 Whether or not a PE originates an SBD-IMET route, it will of course 1477 originate an IMET route for each BD to which it is attached. Each of 1478 these IMET routes MUST carry the SBD-RT, as well as the RT for the BD 1479 to which it belongs. 1481 If a received IMET route is not the SBD-IMET route, it will also be 1482 carrying the RT for its source BD. The route's NLRI will carry the 1483 Tag ID for the source BD. From the RT and the Tag ID, any PE 1484 receiving the route can determine the route's source BD. 1486 If the MPLS label field of the PTA contains zero, the specified P2MP 1487 tunnel is used only to carry frames of a single source BD. 1489 If the MPLS label field of the PTA does not contain zero, it MUST 1490 contain an upstream-assigned MPLS label that maps uniquely (in the 1491 context of the originating EVPN-PE) to the source BD (or, in the case 1492 of an SBD-IMET route, to the SBD). The tunnel may then be used to 1493 carry frames of multiple source BDs. The apparent source BD of a 1494 particular packet is inferred from the label carried by the packet. 1496 IP multicast traffic originating outside the Tenant Domain is 1497 transmitted with the label corresponding to the SBD, as specified in 1498 the ingress EVPN-PE's SBD-IMET route. 1500 3.2.5.2. Using Wildcard S-PMSI A-D Routes to Advertise Inclusive 1501 Tunnels Specific to IP Multicast 1503 The procedures of this section apply when (and only when) it is 1504 desired to transmit IP multicast traffic on an inclusive tunnel, but 1505 not on the same tunnel used to transmit BUM traffic. 1507 However, these procedures do NOT apply when the tunnel type is 1508 Ingress Replication or BIER, EXCEPT in the case where it is necessary 1509 to interwork between non-OISM PEs and OISM PEs, as specified in 1510 Section 5. 1512 Each EVPN-PE attached to the given Tenant Domain MUST originate an 1513 SBD-SPMSI A-D route. The NLRI of that route MUST contain (C-*,C-*) 1514 (see [RFC6625]). Additional rules for constructing that route are 1515 given in Section 3.2.1. 1517 In addition, an EVPN-PE MUST originate an S-PMSI A-D route containing 1518 (C-*,C-*) in its NLRI for each of the other BDs, in the given Tenant 1519 Domain, to which it is attached. All such routes MUST carry the 1520 SBD-RT. This ensures that those routes are imported by all EVPN-PEs 1521 attached to the Tenant Domain. 1523 A PE receiving these routes follows the procedures of Section 2.2 to 1524 determine which BD the route is for. 1526 If the MPLS label field of the PTA contains zero, the specified 1527 tunnel is used only to carry frames of a single source BD. 1529 If the MPLS label field of the PTA does not contain zero, it MUST 1530 specify an upstream-assigned MPLS label that maps uniquely (in the 1531 context of the originating EVPN-PE) to the source BD. The tunnel may 1532 be used to carry frames of multiple source BDs, and the apparent 1533 source BD for a particular packet is inferred from the label carried 1534 by the packet. 1536 The EVPN-PE advertising these S-PMSI A-D route routes is specifying 1537 the default tunnel that it will use (as ingress PE) for transmitting 1538 IP multicast packets. The upstream-assigned label allows an egress 1539 PE to determine the apparent source BD of a given packet. 1541 3.2.6. Selective Tunnels 1543 An ingress EVPN-PE for a given multicast flow or set of flows can 1544 always assign the flow to a particular P2MP tunnel by originating an 1545 S-PMSI A-D route whose NLRI identifies the flow or set of flows. The 1546 NLRI of the route could be (C-*,C-G), or (C-S,C-G). The S-PMSI A-D 1547 route MUST carry the SBD-RT, so that it is imported by all EVPN-PEs 1548 attached to the Tenant Domain. 1550 An S-PMSI A-D route is "for" a particular source BD. It MUST carry 1551 the RT associated with that BD, and it MUST have the Tag ID for that 1552 BD in its NLRI. 1554 When an EVPN-PE imports an S-PMSI A-D route, it applies the rules of 1555 Section 2.2 to associate the route with a particular BD. 1557 Each such route MUST contain a PTA, as specified in Section 3.2.5.2. 1559 An egress EVPN-PE interested in the specified flow or flows MUST join 1560 the specified tunnel. Procedures for joining the specified tunnel 1561 are specific to the tunnel type. (Note that if the tunnel type is 1562 RSVP-TE P2MP LSP, the Leaf Information Required (LIR) flag of the PTA 1563 SHOULD NOT be set. An ingress OISM PE knows which OISM EVPN PEs are 1564 interested in any given flow, and hence can add them to the RSVP-TE 1565 P2MP tunnel that carries such flows.) 1566 If the PTA does not specify a non-zero MPLS label, the apparent 1567 source BD of any packets that arrive on that tunnel is considered to 1568 be the BD associated with the route that carries the PTA. If the PTA 1569 does specify a non-zero MPLS label, the apparent source BD of any 1570 packets that arrive on that tunnel carrying the specified label is 1571 considered to be the BD associated with the route that carries the 1572 PTA. 1574 It should be noted that when either IR or BIER is used, there is no 1575 need for an ingress PE to use S-PMSI A-D routes to assign specific 1576 flows to selective tunnels. The procedures of Section 3.3, along 1577 with the procedures of Section 3.2.2, Section 3.2.3, or 1578 Section 3.2.4, provide the functionality of selective tunnels without 1579 the need to use S-PMSI A-D routes. 1581 3.3. Advertising SMET Routes 1583 [IGMP-Proxy] allows an egress EVPN-PE to express its interest in a 1584 particular multicast flow or set of flows by originating an SMET 1585 route. The NLRI of the SMET route identifies the flow or set of 1586 flows as (C-*,C-*) or (C-*,C-G) or (C-S,C-G). 1588 Each SMET route belongs to a particular BD. The Tag ID for the BD 1589 appears in the NLRI of the route, and the route carries the RT 1590 associated that that BD. From this pair, other EVPN-PEs 1591 can identify the BD to which a received SMET route belongs. 1592 (Remember though that the route may be carrying multiple RTs.) 1594 There are three cases to consider: 1596 o Case 1: It is known that no BD of a Tenant Domain contains a 1597 multicast router. 1599 In this case, an egress PE advertises its interest in a flow or 1600 set of flows by originating an SMET route that belongs to the SBD. 1601 We refer to this as an SBD-SMET route. The SBD-SMET route carries 1602 the SBD-RT, and has the Tag ID for the SBD in its NLRI. SMET 1603 routes for the individual BDs are not needed, because there is no 1604 need for a PE that receives an SMET route to send a corresponding 1605 IGMP Join message out any of its ACs. 1607 o Case 2: It is known that more than one BD of a Tenant Domain may 1608 contain a multicast router. 1610 This is very like Case 1. An egress PE advertises its interest in 1611 a flow or set of flows by originating an SBD-SMET route. The 1612 SBD-SMET route carries the SBD-RT, and has the Tag ID for the SBD 1613 in its NLRI. 1615 In this case, it is important to be sure that SMET routes for the 1616 individual BDs are not originated. Suppose, for example, that PE1 1617 had local receivers for a given flow on both BD1 and BD2, and that 1618 it originated SMET routes for both those BDs. Then PEs receiving 1619 those SMET routes might send IGMP Joins on both those BDs. This 1620 could cause externally sourced multicast traffic to enter the 1621 Tenant Domain at both BDs, which could result in duplication of 1622 data. 1624 N.B.: If it is possible that more than one BD contains a tenant 1625 multicast router, then in order to receive multicast data 1626 originating from outside EVPN, the PEs MUST follow the procedures 1627 of Section 6. 1629 o Case 3: It is known that only a single BD of a Tenant Domain 1630 contains a multicast router. 1632 Suppose that an egress PE is attached to a BD on which there might 1633 be a tenant multicast router. (The tenant router is not 1634 necessarily on a segment that is attached to that PE.) And 1635 suppose that the PE has one or more ACs attached to that BD which 1636 are interested in a given multicast flow. In this case, IN 1637 ADDITION to the SMET route for the SBD, the egress PE MAY 1638 originate an SMET route for that BD. This will enable the ingress 1639 PE(s) to send IGMP/MLD messages on ACs for the BD, as specified in 1640 [IGMP-Proxy]. As long as that is the only BD on which there is a 1641 tenant multicast router, there is no possibility of duplication of 1642 data. 1644 This document does not specify procedures for dynamically determining 1645 which of the three cases applies to a given deployment; the PEs of a 1646 given Tenant Domain MUST be provisioned to know which case applies. 1648 As detailed in [IGMP-Proxy], an SMET route carries a Multicast Flags 1649 EC containing flags indicating whether it is to result in the 1650 propagation of IGMP v1, v2, or v3 messages on the ACs of the BD to 1651 which the SMET route belongs. These flags SHOULD be set to zero in 1652 an SBD-SMET route. 1654 Note that a PE only needs to originate the set of SBD-SMET routes 1655 that are needed to pull in all the traffic in which it is interested. 1656 Suppose PE1 has ACs attached to BD1 that are interested in (C-*,C-G) 1657 traffic, and ACs attached to BD2 that are interested in (C-S,C-G) 1658 traffic. A single SBD-SMET route specifying (C-*,C-G) will pull in 1659 all the necessary flows. 1661 As another example, suppose the ACs attached to BD1 are interested in 1662 (C-*,C-G) but not in (C-S,C-G), while the ACs attached to BD2 are 1663 interested in (C-S,C-G). A single SBD-SMET route specifying 1664 (C-*,C-G) will pull in all the necessary flows. 1666 In other words, to determine the set of SBD-SMET routes that have to 1667 be sent for a given C-G, the PE has to merge the IGMP/MLD state for 1668 all the BDs (of the given Tenant Domain) to which it is attached. 1670 Per [IGMP-Proxy], importing an SMET route for a particular BD will 1671 cause IGMP/MLD state to be instantiated for the IRB interface to that 1672 BD. This applies as well when the BD is the SBD. 1674 However, traffic that originates in one of the actual BDs of a 1675 particular Tenant Domain MUST NOT be sent down the IRB interface that 1676 connects the L3 routing instance of that Tenant Domain to the SBD. 1677 That would cause duplicate delivery of traffic, since such traffic 1678 will have already been distributed throughout the Tenant Domain. 1679 Therefore, when setting up the IGMP/MLD state based on SBD-SMET 1680 routes, care must be taken to ensure that the IRB interface to the 1681 SBD is not added to the Outgoing Interface (OIF) list if the traffic 1682 originates within the Tenant Domain. 1684 There are some multicast scenarios that make use of "anycast 1685 sources". For example, two different sources may share the same 1686 anycast IP address, say S1, and each may transmit an (S1,G) multicast 1687 flow. In such a scenario, the two (S1,G) flows are typically 1688 identical. Ordinary PIM procedures will cause only one the flows to 1689 be delivered to each receiver that has expressed interest in either 1690 (*,G) or (S1,G). However, the OISM procedures described in this 1691 document will result in both of the (S1,G) flows being distributed in 1692 the Tenant Domain, and duplicate delivery will result. Therefore, if 1693 there are receivers for (*,G) in a given Tenant Domain, there MUST 1694 NOT be anycast sources for G within that Tenant Domain. (This 1695 restriction can be lifted by defining additional procedures; however 1696 that is outside the scope of this document.) 1698 4. Constructing Multicast Forwarding State 1700 4.1. Layer 2 Multicast State 1702 An EVPN-PE maintains "layer 2 multicast state" for each BD to which 1703 it is attached. 1705 Let PE1 be an EVPN-PE, and BD1 be a BD to which it is attached. At 1706 PE1, BD1's layer 2 multicast state for a given (C-S,C-G) or (C-*,C-G) 1707 governs the disposition of an IP multicast packet that is received by 1708 BD1's layer 2 multicast function on an EVPN-PE. 1710 An IP multicast (S,G) packet is considered to have been received by 1711 BD1's layer 2 multicast function in PE1 in the following cases: 1713 o The packet is the payload of an ethernet frame received by PE1 1714 from an AC that attaches to BD1. 1716 o The packet is the payload of an ethernet frame whose apparent 1717 source BD is BD1, and which is received by the PE1 over a tunnel 1718 from another EVPN-PE. 1720 o The packet is received from BD1's IRB interface (i.e., has been 1721 transmitted by PE1's L3 routing instance down BD1's IRB 1722 interface). 1724 According to the procedures of this document, all transmission of IP 1725 multicast packets from one EVPN-PE to another is done at layer 2. 1726 That is, the packets are transmitted as ethernet frames, according to 1727 the layer 2 multicast state. 1729 Each layer 2 multicast state (S,G) or (*,G) contains a set "output 1730 interfaces" (OIF list). The disposition of an (S,G) multicast frame 1731 received by BD1's layer 2 multicast function is determined as 1732 follows: 1734 o The OIF list is taken from BD1's layer 2 (S,G) state, or if there 1735 is no such (S,G) state, then from BD1's (*,G) state. (If neither 1736 state exists, the OIF list is considered to be null.) 1738 o The rules of Section 4.1.2 are applied to the OIF list. This will 1739 generally result in the frame being transmitted to some, but not 1740 all, elements of the OIF list. 1742 Note that there is no RPF check at layer 2. 1744 4.1.1. Constructing the OIF List 1746 In this document, we have extended the procedures of [IGMP-Proxy] so 1747 that IMET and SMET routes for a particular BD are distributed not 1748 just to PEs that attach to that BD, but to PEs that attach to any BD 1749 in the Tenant Domain. In this way, each PE attached to a given 1750 Tenant Domain learns, from each other PE attached to the same Tenant 1751 Domain, the set of flows that are of interest to each of those other 1752 PEs. (If some PE attached to the Tenant Domain does not support 1753 [IGMP-Proxy], it will be assumed to be interested in all flows. 1754 Whether a particular remote PE supports [IGMP-Proxy] is determined by 1755 the presence of an Extended Community in its IMET route; this is 1756 specified in [IGMP-Proxy].) If a set of remote PEs are interested in 1757 a particular flow, the tunnels used to reach those PEs are added to 1758 the OIF list of the multicast states corresponding to that flow. 1760 An EVPN-PE may run IGMP/MLD procedures on each of its ACs, in order 1761 to determine the set of flows of interest to each AC. (An AC is said 1762 to be interested in a given flow if it connects to a segment that has 1763 tenant systems interested in that flow.) If IGMP/MLD procedures are 1764 not being run on a given AC, that AC is considered to be interested 1765 in all flows. For each BD, the set of ACs interested in a given flow 1766 is determined, and the ACs of that set are added to the OIF list of 1767 that BD's multicast state for that flow. 1769 The OIF list for each multicast state must also contain the IRB 1770 interface for the BD to which the state belongs. 1772 Implementors should note that the OIF list of a multicast state will 1773 change from time to time as ACs and/or remote PEs either become 1774 interested in, or lose interest in, particular multicast flows. 1776 4.1.2. Data Plane: Applying the OIF List to an (S,G) Frame 1778 When an (S,G) multicast frame is received by the layer 2 multicast 1779 function of a given EVPN-PE, say PE1, its disposition depends (a) the 1780 way it was received, (b) upon the OIF list of the corresponding 1781 multicast state (see Section 4.1.1), (c) upon the "eligibility" of an 1782 AC to receive a given frame (see Section 4.1.2.1 and (d) upon its 1783 apparent source BD (see Section 3.2 for information about determining 1784 the apparent source BD of a frame received over a tunnel from another 1785 PE). 1787 4.1.2.1. Eligibility of an AC to Receive a Frame 1789 A given (S,G) multicast frame is eligible to be transmitted by a 1790 given PE, say PE1, on a given AC, say AC1, only if one of the 1791 following conditions holds: 1793 1. ESI labels are being used, PE1 is the DF for the segment to which 1794 AC1 is connected, and the frame did not originate from that same 1795 segment (as determined by the ESI label), or 1797 2. The ingress PE for the frame is a remote PE, say PE2, local bias 1798 is being used, and PE2 is not connected to the same segment as 1799 AC1. 1801 4.1.2.2. Applying the OIF List 1803 Assume a given (S,G) multicast frame has been received by a given PE, 1804 say PE1. PE1 determines the apparent source BD of the frame, finds 1805 the layer 2 (S,G) state for that BD (or the (*,G) state if there is 1806 no (S,G) state), and takes the OIF list from that state. (Note that 1807 if PE1 is not attached to the actual source BD, the apparent source 1808 BD will be the SBD.) 1810 Suppose PE1 has determined the frame's apparent source BD to be BD1 1811 (which may or may not be the SBD.) There are the following cases to 1812 consider: 1814 1. The frame was received by PE1 from a local AC, say AC1, that 1815 attaches to BD1. 1817 a. The frame MUST be sent out all local ACs of BD1 that appear 1818 in the OIF list, except for AC1 itself. 1820 b. The frame MUST also be delivered to any other EVPN-PEs that 1821 have interest in it. This is achieved as follows: 1823 i. If (a) AR is being used, and (b) PE1 is an AR-LEAF, and 1824 (c) the OIF list is non-null, PE1 MUST send the frame 1825 to the AR-REPLICATOR. 1827 ii. Otherwise the frame MUST be sent on all tunnels in the 1828 OIF list. 1830 c. The frame MUST be sent to the local L3 routing instance by 1831 being sent up the IRB interface of BD1. It MUST NOT be sent 1832 up any other IRB interfaces. 1834 2. The frame was received by PE1 over a tunnel from another PE. 1835 (See Section 3.2 for the rules to determine the apparent source 1836 BD of a packet received from another PE. Note that if PE1 is not 1837 attached to the source BD, it will regard the SBD as the apparent 1838 source BD.) 1840 a. The frame MUST be sent out all local ACs in the OIF list that 1841 connect to BD1 and that are eligible (per Section 4.1.2.1) to 1842 receive the frame. 1844 b. The frame MUST be sent up the IRB interface of the apparent 1845 source BD. (Note that this may be the SBD.) The frame MUST 1846 NOT be sent up any other IRB interfaces. 1848 c. If PE1 is not an AR-REPLICATOR, it MUST NOT send the frame to 1849 any other EVPN-PEs. However, if PE1 is an AR-REPLICATOR, it 1850 MUST send the frame to all tunnels in the OIF list, except 1851 for the tunnel over which the frame was received. 1853 3. The frame was received by PE1 from the BD1 IRB interface (i.e., 1854 the frame has been transmitted by PE1's L3 routing instance down 1855 the BD1 IRB interface), and BD1 is NOT the SBD. 1857 a. The frame MUST be sent out all local ACs in the OIF list that 1858 are eligible (per Section 4.1.2.1 to receive the frame. 1860 b. The frame MUST NOT be sent to any other EVPN-PEs. 1862 c. The frame MUST NOT be sent up any IRB interfaces. 1864 4. The frame was received from the SBD IRB interface (i.e., has been 1865 transmitted by PE1's L3 routing instance down the SBD IRB 1866 interface). 1868 a. The frame MUST be sent on all tunnels in the OIF list. This 1869 causes the frame to be delivered to any other EVPN-PEs that 1870 have interest in it. 1872 b. The frame MUST NOT be sent on any local ACs. 1874 c. The frame MUST NOT be sent up any IRB interfaces. 1876 4.2. Layer 3 Forwarding State 1878 If an EVPN-PE is performing IGMP/MLD procedures on the ACs of a given 1879 BD, it processes those messages at layer 2 to help form the layer 2 1880 multicast state. If also sends those messages up that BD's IRB 1881 interface to the L3 routing instance of a particular tenant domain. 1882 This causes layer 2 (C-S,C-G) or (C-*,C-G) L3 state to be created/ 1883 updated. 1885 A layer 3 multicast state has both an Input Interface (IIF) and an 1886 OIF list. 1888 To set the IIF of an (C-S,C-G) state, the EVPN-PE must determine the 1889 source BD of C-S. This is done by looking up S in the local 1890 MAC-VRF(s) of the given Tenant Domain. 1892 If the source BD is present on the PE, the IIF is set to the IRB 1893 interface that attaches to that BD. Otherwise the IIF is set to the 1894 SBD IRB interface. 1896 For (C-*,C-G) states, traffic can arrive from any BD, so the IIF 1897 needs to be set to a wildcard value meaning "any IRB interface". 1899 The OIF list of these states includes one or more of the IRB 1900 interfaces of the Tenant Domain. In general, maintenance of the OIF 1901 list does not require any EVPN-specific procedures. However, there 1902 is one EVPN-specific rule: 1904 If the IIF is one of the IRB interfaces (or the wild card meaning 1905 "any IRB interface"), then the SBD IRB interface MUST NOT be added 1906 to the OIF list. Traffic originating from within a particular 1907 EVPN Tenant Domain must not be sent down the SBD IRB interface, as 1908 such traffic has already been distributed to all EVPN-PEs attached 1909 to that Tenant Domain. 1911 Please also see Section 6.1.1, which states a modification of this 1912 rule for the case where OISM is interworking with external Layer 3 1913 multicast routing. 1915 5. Interworking with non-OISM EVPN-PEs 1917 It is possible that a given Tenant Domain will be attached to both 1918 OISM PEs and non-OISM PEs. Inter-subnet IP multicast should be 1919 possible and fully functional even if not all PEs attaching to a 1920 Tenant Domain can be upgraded to support OISM functionality. 1922 Note that the non-OISM PEs are not required to have IRB support, or 1923 support for [IGMP-Proxy]. It is however advantageous for the 1924 non-OISM PEs to support [IGMP-Proxy]. 1926 In this section, we will use the following terminology: 1928 o PE-S: the ingress PE for an (S,G) flow. 1930 o PE-R: an egress PE for an (S,G) flow. 1932 o BD-S: the source BD for an (S,G) flow. PE-S must have one or more 1933 ACs attached BD-S, at least one of which attaches to host S. 1935 o BD-R: a BD that contains a host interested in the flow. The host 1936 is attached to PE-R via an AC that belongs to BD-R. 1938 To allow OISM PEs to interwork with non-OISM PEs, a given Tenant 1939 Domain needs to contain one or more "IP Multicast Gateways" (IPMGs). 1940 An IPMG is an OISM PE with special responsibilities regarding the 1941 interworking between OISM and non-OISM PEs. 1943 If a PE is functioning as an IPMG, it MUST signal this fact by 1944 setting the "IPMG" flag in the Multicast Flags EC that it attaches to 1945 its IMET routes. An IPMG SHOULD attach this EC with the IPMG flag 1946 set to all IMET routes it originates. However, if PE1 imports any 1947 IMET route from PE2 that has the EC present with the "IPMG" flag set, 1948 then the PE1 will assume that PE2 is an IPMG. 1950 An IPMG Designated Forwarder (IPMG-DF) selection procedure is used to 1951 ensure that, at any given time, there is exactly one active IPMG-DF 1952 for any given BD. Details of the IPMG-DF selection procedure are in 1953 Section 5.1. The IPMG-DF for a given BD, say BD-S, has special 1954 functions to perform when it receives (S,G) frames on that BD: 1956 o If the frames are from a non-OISM PE-S: 1958 * The IPMG-DF forwards them to OISM PEs that do not attach to 1959 BD-S but have interest in (S,G). 1961 Note that OISM PEs that do attach to BD-S will have received 1962 the frames on the BUM tunnel from the non-OISM PE-S. 1964 * The IPMG-DF forwards them to non-OISM PEs that have interest in 1965 (S,G) on ACs that do not belong to BD-S. 1967 Note that if a non-OISM PE has multiple BDs other than BD-S 1968 with interest in (S,G), it will receive one copy of the frame 1969 for each such BD. This is necessary because the non-OISM PEs 1970 cannot move IP multicast traffic from one BD to another. 1972 o If the frames are from an OISM PE, the IPMG-DF forwards them to 1973 non-OISM PEs that have interest in (S,G) on ACs that do not belong 1974 to BD-S. 1976 If a non-OISM PE has interest in (S,G) on an AC belonging to BD-S, 1977 it will have received a copy of the (S,G) frame, encapsulated for 1978 BD-S, from the OISM PE-S. (See Section 3.2.2.) If the non-OISM 1979 PE has interest in (S,G) on one or more ACs belonging to 1980 BD-R1,...,BD-Rk where the BD-Ri are distinct from BD-S, the 1981 IPMG-DF needs to send it a copy of the frame for BD-Ri. 1983 If an IPMG receives a frame on a BD for which it is not the IPMG-DF, 1984 it just follows normal OISM procedures. 1986 This section specifies several sets of procedures: 1988 o the procedures that the IPMG-DF for a given BD needs to follow 1989 when receiving, on that BD, an IP multicast frame from a non-OISM 1990 PE; 1992 o the procedures that the IPMG-DF for a given BD needs to follow 1993 when receiving, on that BD, an IP multicast frame from an OISM PE; 1995 o the procedures that an OISM PE needs to follow when receiving, on 1996 a given BD, an IP multicast frame from a non-OISM PE, when the 1997 OISM PE is not the IPMG-DF for that BD. 1999 To enable OISM/non-OISM interworking in a given Tenant Domain, the 2000 Tenant Domain MUST have some EVPN-PEs that can function as IPMGs. An 2001 IPMG must be configured with the SBD. It must also be configured 2002 with every BD of the Tenant Domain that exists on any of the non-OISM 2003 PEs of that domain. (Operationally, it may be simpler to configure 2004 the IPMG with all the BDs of the Tenant Domain.) 2006 A non-OISM PE of course only needs to be configured with BDs for 2007 which it has ACs. An OISM PE that is not an IPMG only needs to be 2008 configured with the SBD and with the BDs for which it has ACs. 2010 An IPMG MUST originate a wildcard SMET route (with (C-*,C-*) in the 2011 NLRI) for each BD in the Tenant Domain. This will cause it to 2012 receive all the IP multicast traffic that is sourced in the Tenant 2013 Domain. Note that non-OISM nodes that do not support [IGMP-Proxy] 2014 will send all the multicast traffic from a given BD to all PEs 2015 attached to that BD, even if those PEs do not originate an SMET 2016 route. 2018 The interworking procedures vary somewhat depending upon whether 2019 packets are transmitted from PE to PE via Ingress Replication (IR) or 2020 via Point-to-Multipoint (P2MP) tunnels. We do not consider the use 2021 of BIER in this section, due to the low likelihood of there being a 2022 non-OISM PE that supports BIER. 2024 5.1. IPMG Designated Forwarder 2026 Every PE that is eligible for selection as an IPMG-DF for a 2027 particular BD originates both an IMET route for that BD and an 2028 SBD-IMET route. As stated in Section 5, these SBD-IMET routes carry 2029 a Multicast Flags EC with the IPMG Flag set. 2031 These SBD-IMET routes SHOULD also carry a DF Election EC. The DF 2032 Election EC and its use is specified in ([DF-Election-Framework]). 2033 When the route is originated, the AC-DF bit in the DF Election EC 2034 SHOULD be set to zero. This bit is not used when selecting an 2035 IPMSG-DF, i.e., it MUST be ignored by the receiver of an SBD-IMET 2036 route. 2038 In the context of a given Tenant Domain, to select the IPMG-DF for a 2039 particular BD, say BD1, the IPMGs of the Tenant Domain perform the 2040 following procedure: 2042 o From the set of received SBD-IMET routes for the given tenant 2043 domain, determine the candidate set of PEs that support IPMG 2044 functionality for that domain. 2046 o Eliminate from that candidate set any PEs from which an IMET route 2047 for BD1 has not been received. 2049 o Select a DF Election algorithm as specified in 2050 [DF-Election-Framework]. Some of the possible algorithms can be 2051 found, e.g., in [DF-Election-Framework], [RFC7432], and 2052 [EVPN-DF-WEIGHTED]. 2054 o Apply the DF Election Algorithm (see [DF-Election-Framework]) to 2055 the candidate set of PEs. The "winner' becomes the IPMG-DF for 2056 BD1. 2058 Note that even if a given PE supports MEG (Section 6.1.2) and/or PEG 2059 (Section 6.1.4) functionality, as well as IPMG functionality, its 2060 SBD-IMET routes carry only one DF Election EC. 2062 5.2. Ingress Replication 2064 The procedures of this section are used when Ingress Replication is 2065 used to transmit packets from one PE to another. 2067 When a non-OISM PE-S transmits a multicast frame from BD-S to another 2068 PE, PE-R, PE-S will use the encapsulation specified in the BD-S IMET 2069 route that was originated by PE-R. This encapsulation will include 2070 the label that appears in the "MPLS label" field of the PMSI Tunnel 2071 attribute (PTA) of the IMET route. If the tunnel type is VXLAN, the 2072 "label" is actually a Virtual Network Identifier (VNI); for other 2073 tunnel types, the label is an MPLS label. In either case, we will 2074 speak of the transmitted frames as carrying a label that was assigned 2075 to a particular BD by the PE-R to which the frame is being 2076 transmitted. 2078 To support OISM/non-OISM interworking, an OISM PE-R MUST originate, 2079 for each of its BDs, both an IMET route and an S-PMSI (C-*,C-*) A-D 2080 route. Note that even when IR is being used, interworking between 2081 OISM and non-OISM PEs requires the OISM PEs to follow the rules of 2082 Section 3.2.5.2, as modified below. 2084 Non-OISM PEs will not understand S-PMSI A-D routes. So when a 2085 non-OISM PE-S transmits an IP multicast frame with a particular 2086 source BD to an IPMG, it encapsulates the frame using the label 2087 specified in that IPMG's BD-S IMET route. (This is just the 2088 procedure of [RFC7432].) 2090 The (C-*,C-*) S-PMSI A-D route originated by a given OISM PE will 2091 have a PTA that specifies IR. 2093 o If MPLS tunneling is being used, the MPLS label field SHOULD 2094 contain a non-zero value, and the LIR flag SHOULD be zero. (The 2095 case where the MPLS label field is zero or the LIR flag is set is 2096 outside the scope of this document.) 2098 o If the tunnel encapsulation is VXLAN, the MPLS label field MUST 2099 contain a non-zero value, and the LIR flag MUST be zero. 2101 When an OISM PE-S transmits an IP multicast frame to an IPMG, it will 2102 use the label specified in that IPMG's (C-*,C-*) S-PMSI A-D route. 2104 When a PE originates both an IMET route and a (C-*,C-*) S-PMSI A-D 2105 route, the values of the MPLS label field in the respective PTAs must 2106 be distinct. Further, each MUST map uniquely (in the context of the 2107 originating PE) to the route's BD. 2109 As a result, an IPMG receiving an MPLS-encapsulated IP multicast 2110 frame can always tell by the label whether the frame's ingress PE is 2111 an OISM PE or a non-OISM PE. When an IPMG receives a VXLAN- 2112 encapsulated IP multicast frame it may need to determine the identity 2113 of the ingress PE from the outer IP encapsulation; it can then 2114 determine whether the ingress PE is an OISM PE or a non-OISM PE by 2115 looking the IMET route from that PE. 2117 Suppose an IPMG receives an IP multicast frame from another EVPN-PE 2118 in the Tenant Domain, and the IPMG is not the IPMG-DF for the frame's 2119 source BD. Then the IPMG performs only the ordinary OISM functions; 2120 it does not perform the IPMG-specific functions for that frame. In 2121 the remainder of this section, when we discuss the procedures applied 2122 by an IPMG when it receives an IP multicast frame, we are presuming 2123 that the source BD of the frame is a BD for which the IPMG is the 2124 IPMG-DF. 2126 We have two basic cases to consider: (1) a frame's ingress PE is a 2127 non-OISM node, and (2) a frame's ingress PE is an OISM node. 2129 5.2.1. Ingress PE is non-OISM 2131 In this case, a non-OISM PE, PE-S, has received an (S,G) multicast 2132 frame over an AC that is attached to a particular BD, BD-S. By 2133 virtue of normal EVPN procedures, PE-S has sent a copy of the frame 2134 to every PE-R (both OISM and non-OISM) in the Tenant Domain that is 2135 attached to BD-S. If the non-OISM node supports [IGMP-Proxy], only 2136 PEs that have expressed interest in (S,G) receive the frame. The 2137 IPMG will have expressed interest via a (C-*,C-*) SMET route and thus 2138 receives the frame. 2140 Any OISM PE (including an IPMG) receiving the frame will apply normal 2141 OISM procedures. As a result it will deliver the frame to any of its 2142 local ACs (in BD-S or in any other BD) that have interest in (S,G). 2144 An OISM PE that is also the IPMG-DF for a particular BD, say BD-S, 2145 has additional procedures that it applies to frames received on BD-S 2146 from non-OISM PEs: 2148 1. When the IPMG-DF for BD-S receives an (S,G) frame from a 2149 non-OISM node, it MUST forward a copy of the frame to every OISM 2150 PE that is NOT attached to BD-S but has interest in (S,G). The 2151 copy sent to a given OISM PE-R must carry the label that PE-R 2152 has assigned to the SBD in an S-PMSI A-D route. The IPMG MUST 2153 NOT do any IP processing of the frame's IP payload. TTL 2154 decrement and other IP processing will be done by PE-R, per the 2155 normal OISM procedures. There is no need for the IPMG to 2156 include an ESI label in the frame's tunnel encapsulation, 2157 because it is already known that the frame's source BD has no 2158 presence on PE-R. There is also no need for the IPMG to modify 2159 the frame's MAC SA. 2161 2. In addition, when the IPMG-DF for BD-S receives an (S,G) frame 2162 from a non-OISM node, it may need to forward copies of the frame 2163 to other non-OISM nodes. Before it does so, it MUST decapsulate 2164 the (S,G) packet, and do the IP processing (e.g., TTL 2165 decrement). Suppose PE-R is a non-OISM node that has an AC to 2166 BD-R, where BD-R is not the same as BD-S, and that AC has 2167 interest in (S,G). The IPMG must then encapsulate the (S,G) 2168 packet (after the IP processing has been done) in an ethernet 2169 header. The MAC SA field will have the MAC address of the 2170 IPMG's IRB interface to BD-R. The IPMG then sends the frame to 2171 PE-R. The tunnel encapsulation will carry the label that PE-R 2172 advertised in its IMET route for BD-R. There is no need to 2173 include an ESI label, as the source and destination BDs are 2174 known to be different. 2176 Note that if a non-OISM PE-R has several BDs (other than BD-S) 2177 with local ACs that have interest in (S,G), the IPMG will send 2178 it one copy for each such BD. This is necessary because the 2179 non-OISM PE cannot move packets from one BD to another. 2181 There may be deployment scenarios in which every OISM PE is 2182 configured with every BD that is present on any non-OISM PE. In such 2183 scenarios, the procedures of item 1 above will not actually result in 2184 the transmission of any packets. Hence if it is known a priori that 2185 this deployment scenario exists for a given tenant domain, the 2186 procedures of item 1 above can be disabled. 2188 5.2.2. Ingress PE is OISM 2190 In this case, an OISM PE, PE-S, has received an (S,G) multicast frame 2191 over an AC that attaches to a particular BD, BD-S. 2193 By virtue of receiving all the IMET routes about BD-S, PE-S will know 2194 all the PEs attached to BD-S. By virtue of normal OISM procedures: 2196 o PE-S will send a copy of the frame to every OISM PE-R (including 2197 the IPMG) in the Tenant Domain that is attached to BD-S and has 2198 interest in (S,G). The copy sent to a given PE-R carries the 2199 label that that the PE-R has assigned to BD-S in its (C-*,C-*) 2200 S-PMSI A-D route. 2202 o PE-S will also transmit a copy of the (S,G) frame to every OISM 2203 PE-R that has interest in (S,G) but is not attached to BD-S. The 2204 copy will contain the label that the PE-R has assigned to the SBD. 2205 (As in Section 5.2.1, an IPMG is assumed to have indicated 2206 interest in all multicast flows.) 2208 o PE-S will also transmit a copy of the (S,G) frame to every 2209 non-OISM PE-R that is attached to BD-S. It does this using the 2210 label advertised by that PE-R in its IMET route for BD-S. 2212 The PE-Rs follow their normal procedures. An OISM PE that receives 2213 the (S,G) frame on BD-S applies the OISM procedures to deliver the 2214 frame to its local ACs, as necessary. A non-OISM PE that receives 2215 the (S,G) frame on BD-S delivers the frame only to its local BD-S 2216 ACs, as necessary. 2218 Suppose that a non-OISM PE-R has interest in (S,G) on a BD, BD-R, 2219 that is different than BD-S. If the non-OISM PE-R is attached to 2220 BD-S, the OISM PE-S will send forward it the original (S,G) multicast 2221 frame, but the non-OISM PE-R will not be able to send the frame to 2222 ACs that are not in BD-S. If PE-R is not even attached to BD-S, the 2223 OISM PE-S will not send it a copy of the frame at all, because PE-R 2224 is not attached to the SBD. In these cases, the IPMG needs to relay 2225 the (S,G) multicast traffic from OISM PE-S to non-OISM PE-R. 2227 When the IPMG-DF for BD-S receives an (S,G) frame from an OISM PE-S, 2228 it has to forward it to every non-OISM PE-R that that has interest in 2229 (S,G) on a BD-R that is different than BD-S. The IPMG MUST 2230 decapsulate the IP multicast packet, do the IP processing, re- 2231 encapsulate it for BD-R (changing the MAC SA to the IPMG's own MAC 2232 address on BD-R), and send a copy of the frame to PE-R. Note that a 2233 given non-OISM PE-R will receive multiple copies of the frame, if it 2234 has multiple BDs on which there is interest in the frame. 2236 5.3. P2MP Tunnels 2238 When IR is used to distribute the multicast traffic among the 2239 EVPN-PEs, the procedures of Section 5.2 ensure that there will be no 2240 duplicate delivery of multicast traffic. That is, no egress PE will 2241 ever send a frame twice on any given AC. If P2MP tunnels are being 2242 used to distribute the multicast traffic, it is necessary have 2243 additional procedures to prevent duplicate delivery. 2245 At the present time, it is not clear that there will be a use case in 2246 which OISM nodes need to interwork with non-OISM nodes that use P2MP 2247 tunnels. If it is determined that there is such a use case, 2248 procedures for it will be included in a future revision of this 2249 document. 2251 6. Traffic to/from Outside the EVPN Tenant Domain 2253 In this section, we discuss scenarios where a multicast source 2254 outside a given EVPN Tenant Domain sends traffic to receivers inside 2255 the domain (as well as, possibly, to receivers outside the domain). 2256 This requires the OISM procedures to interwork with various layer 3 2257 multicast routing procedures. 2259 We assume in this section that the Tenant Domain is not being used as 2260 an intermediate transit network for multicast traffic; that is, we do 2261 not consider the case where the Tenant Domain contains multicast 2262 routers that will receive traffic from sources outside the domain and 2263 forward the traffic to receivers outside the domain. The transit 2264 scenario is considered in Section 7. 2266 We can divide the non-transit scenarios into two classes: 2268 1. One or more of the EVPN PE routers provide the functionality 2269 needed to interwork with layer 3 multicast routing procedures. 2271 2. A single BD in the Tenant Domain contains external multicast 2272 routers ("tenant multicast routers"), and those tenant multicast 2273 routers are used to interwork, on behalf of the entire Tenant 2274 Domain, with layer 3 multicast routing procedures. 2276 6.1. Layer 3 Interworking via EVPN OISM PEs 2278 6.1.1. General Principles 2280 Sometimes it is necessary to interwork an EVPN Tenant Domain with an 2281 external layer 3 multicast domain (the "external domain"). This is 2282 needed to allow EVPN tenant systems to receive multicast traffic from 2283 sources ("external sources") outside the EVPN Tenant Domain. It is 2284 also needed to allow receivers ("external receivers") outside the 2285 EVPN Tenant Domain to receive traffic from sources inside the Tenant 2286 Domain. 2288 In order to allow interworking between an EVPN Tenant Domain and an 2289 external domain, one or more OISM PEs must be "L3 Gateways". An L3 2290 Gateway participates both in the OISM procedures and in the L3 2291 multicast routing procedures of the external domain. 2293 An L3 Gateway that has interest in receiving (S,G) traffic must be 2294 able to determine the best route to S. If an L3 Gateway has interest 2295 in (*,G), it must be able to determine the best route to G's RP. In 2296 these interworking scenarios, the L3 Gateway must be running a layer 2297 3 unicast routing protocol. Via this protocol, it imports unicast 2298 routes (either IP routes or VPN-IP routes) from routers other than 2299 EVPN PEs. And since there may be multicast sources inside the EVPN 2300 Tenant Domain, the EVPN PEs also need to export, either as IP routes 2301 or as VPN-IP routes (depending upon the external domain), unicast 2302 routes to those sources. 2304 When selecting the best route to a multicast source or RP, an L3 2305 Gateway might have a choice between an EVPN route and an IP/VPN-IP 2306 route. When such a choice exists, the L3 Gateway SHOULD always 2307 prefer the EVPN route. This will ensure that when traffic originates 2308 in the Tenant Domain and has a receiver in the Tenant Domain, the 2309 path to that receiver will remain within the EVPN Tenant Domain, even 2310 if the source is also reachable via a routed path. This also 2311 provides protection against sub-optimal routing that might occur if 2312 two EVPN PEs export IP/VPN-IP routes and each imports the other's IP/ 2313 VPN-IP routes. 2315 Section 4.2 discusses the way layer 3 multicast states are 2316 constructed by OISM PEs. These layer 3 multicast states have IRB 2317 interfaces as their IIF and OIF list entries, and are the basis for 2318 interworking OISM with other layer 3 multicast procedures such as 2319 MVPN or PIM. From the perspective of the layer 3 multicast 2320 procedures running in a given L3 Gateway, an EVPN Tenant Domain is a 2321 set of IRB interfaces. 2323 When interworking an EVPN Tenant Domain with an external domain, the 2324 L3 Gateway's layer 3 multicast states will not only have IRB 2325 interfaces as IIF and OIF list entries, but also other "interfaces" 2326 that lead outside the Tenant Domain. For example, when interworking 2327 with MVPN, the multicast states may have MVPN tunnels as well as IRB 2328 interfaces as IIF or OIF list members. When interworking with PIM, 2329 the multicast states may have PIM-enabled non-IRB interfaces as IIF 2330 or OIF list members. 2332 As long as a Tenant Domain is not being used as an intermediate 2333 transit network for IP multicast traffic, it is not necessary to 2334 enable PIM on its IRB interfaces. 2336 In general, an L3 Gateway has the following responsibilities: 2338 o It exports, to the external domain, unicast routes to those 2339 multicast sources in the EVPN Tenant Domain that are locally 2340 attached to the L3 Gateway. 2342 o It imports, from the external domain, unicast routes to multicast 2343 sources that are in the external domain. 2345 o It executes the procedures necessary to draw externally sourced 2346 multicast traffic that is of interest to locally attached 2347 receivers in the EVPN Tenant Domain. When such traffic is 2348 received, the traffic is sent down the IRB interfaces of the BDs 2349 on which the locally attached receivers reside. 2351 One of the L3 Gateways in a given Tenant Domain becomes the "DR" for 2352 the SBD. (See Section 6.1.2.4.) This L3 gateway has the following 2353 additional responsibilities: 2355 o It exports, to the external domain, unicast routes to multicast 2356 sources that in the EVPN Tenant Domain that are not locally 2357 attached to any L3 gateway. 2359 o It imports, from the external domain, unicast routes to multicast 2360 sources that are in the external domain. 2362 o It executes the procedures necessary to draw externally sourced 2363 multicast traffic that is of interest to receivers in the EVPN 2364 Tenant Domain that are not locally attached to an L3 gateway. 2365 When such traffic is received, the traffic is sent down the SBD 2366 IRB interface. OISM procedures already described in this document 2367 will then ensure that the IP multicast traffic gets distributed 2368 throughout the Tenant Domain to any EVPN PEs that have interest in 2369 it. Thus to an OISM PE that is not an L3 gateway the externally 2370 sourced traffic will appear to have been sourced on the SBD. 2372 In order for this to work, some special care is needed when an L3 2373 gateway creates or modifies a layer 3 (*,G) multicast state. Suppose 2374 group G has both external sources (sources outside the EVPN Tenant 2375 Domain) and internal sources (sources inside the EVPN tenant domain). 2376 Section 4.2 states that when there are internal sources, the SBD IRB 2377 interface must not be added to the OIF list of the (*,G) state. 2378 Traffic from internal sources will already have been delivered to all 2379 the EVPN PEs that have interest in it. However, if the OIF list of 2380 the (*,G) state does not contain its SBD IRB interface, then traffic 2381 from external sources will not get delivered to other EVPN PEs. 2383 One way of handling this is the following. When a L3 gateway 2384 receives (S,G) traffic from other than an IRB interface, and the 2385 traffic corresponds to a layer 3 (*,G) state, the L3 gateway can 2386 create (S,G) state. The IIF will be set to the external interface 2387 over which the traffic is expected. The OIF list will contain the 2388 SBD IRB interface, as well as the IRB interfaces of any other BDs 2389 attached to the PEG DR that have locally attached receivers with 2390 interest in the (S,G) traffic. The (S,G) state will ensure that the 2391 external traffic is sent down the SBD IRB interface. The following 2392 text will assume this procedure; however other implementation 2393 techniques may also be possible. 2395 If a particular BD is attached to several L3 Gateways, one of the L3 2396 Gateways becomes the DR for that BD. (See Section 6.1.2.4.) If the 2397 interworking scenario requires FHR functionality, it is generally the 2398 DR for a particular BD that is responsible for performing that 2399 functionality on behalf of the source hosts on that BD. (E.g., if 2400 the interworking scenario requires that PIM Register messages be sent 2401 by a FHR, the DR for a given BD would send the PIM Register messages 2402 for sources on that BD.) Note though that the DR for the SBD does 2403 not perform FHR functionality on behalf of external sources. 2405 An optional alternative is to have each L3 gateway perform FHR 2406 functionality for locally attached sources. Then the DR would only 2407 have to perform FHR functionality on behalf of sources that are 2408 locally attached to itself AND sources that are not attached to any 2409 L3 gateway. 2411 N.B.: If it is possible that more than one BD contains a tenant 2412 multicast router, then a PE receiving an SMET route for that BD MUST 2413 NOT reconstruct IGMP Join Reports from the SMET route, and MUST NOT 2414 transmit any such IGMP Join Reports on its local ACs attaching to 2415 that BD. Otherwise, multicast traffic may be duplicated. 2417 6.1.2. Interworking with MVPN 2419 In this section, we specify the procedures necessary to allow EVPN 2420 PEs running OISM procedures to interwork with L3VPN PEs that run BGP- 2421 based MVPN ([RFC6514]) procedures. More specifically, the procedures 2422 herein allow a given EVPN Tenant Domain to become part of an L3VPN/ 2423 MVPN, and support multicast flows where either: 2425 o The source of a given multicast flow is attached to an ethernet 2426 segment whose BD is part of an EVPN Tenant Domain, and one or more 2427 receivers of the flow are attached to the network via L3VPN/MVPN. 2428 (Other receivers may be attached to the network via EVPN.) 2430 o The source of a given multicast flow is attached to the network 2431 via L3VPN/MVPN, and one or more receivers of the flow are attached 2432 to an ethernet segment that is part of an EVPN tenant domain. 2433 (Other receivers may be attached via L3VPN/MVPN.) 2435 In this interworking model, existing L3VPN/MVPN PEs are unaware that 2436 certain sources or receivers are part of an EVPN Tenant Domain. The 2437 existing L3VPN/MVPN nodes run only their standard procedures and are 2438 entirely unaware of EVPN. Interworking is achieved by having some or 2439 all of the EVPN PEs function as L3 Gateways running L3VPN/MVPN 2440 procedures, as detailed in the following sub-sections. 2442 In this section, we assume that there are no tenant multicast routers 2443 on any of the EVPN-attached ethernet segments. (There may of course 2444 be multicast routers in the L3VPN.) Consideration of the case where 2445 there are tenant multicast routers is deferred till Section 7.) 2447 To support MVPN/EVPN interworking, we introduce the notion of an 2448 MVPN/EVPN Gateway, or MEG. 2450 A MEG is an L3 Gateway (see Section 6.1.1), hence is both an OISM PE 2451 and an L3VPN/MVPN PE. For a given EVPN Tenant Domain it will have an 2452 IP-VRF. If the Tenant Domain is part of an L3VPN/MVPN, the IP-VRF 2453 also serves as an L3VPN VRF ([RFC4364]). The IRB interfaces of the 2454 IP-VRF are considered to be "VRF interfaces" of the L3VPN VRF. The 2455 L3VPN VRF may also have other local VRF interfaces that are not EVPN 2456 IRB interfaces. 2458 The VRF on the MEG will import VPN-IP routes ([RFC4364]) from other 2459 L3VPN Provider Edge (PE) routers. It will also export VPN-IP routes 2460 to other L3VPN PE routers. In order to do so, it must be 2461 appropriately configured with the Route Targets used in the L3VPN to 2462 control the distribution of the VPN-IP routes. These Route Targets 2463 will in general be different than the Route Targets used for 2464 controlling the distribution of EVPN routes, as there is no need to 2465 distribute EVPN routes to L3VPN-only PEs and no reason to distribute 2466 L3VPN/MVPN routes to EVPN-only PEs. 2468 Note that the RDs in the imported VPN-IP routes will not necessarily 2469 conform to the EVPN rules (as specified in [RFC7432]) for creating 2470 RDs. Therefore a MEG MUST NOT expect the RDs of the VPN-IP routes to 2471 be of any particular format other than what is required by the L3VPN/ 2472 MVPN specifications. 2474 The VPN-IP routes that a MEG exports to L3VPN are subnet routes and/ 2475 or host routes for the multicast sources that are part of the EVPN 2476 tenant domain. The exact set of routes that need to be exported is 2477 discussed in Section 6.1.2.2. 2479 Each IMET route originated by a MEG SHOULD carry a Multicast Flags 2480 Extended Community with the "MEG" flag set, indicating that the 2481 originator of the IMET route is a MEG. However, PE1 will consider 2482 PE2 to be a MEG if PE1 imports at least one IMET route from PE2 that 2483 carries the Multicast Flags EC with the MEG flag set. 2485 All the MEGs of a given Tenant Domain attach to the SBD of that 2486 domain, and one of them is selected to be the SBD's Designated Router 2487 (the "MEG SBD-DR") for the domain. The selection procedure is 2488 discussed in Section 6.1.2.4. 2490 In this model of operation, MVPN procedures and EVPN procedures are 2491 largely independent. In particular, there is no assumption that MVPN 2492 and EVPN use the same kind of tunnels. Thus no special procedures 2493 are needed to handle the common scenarios where, e.g., EVPN uses 2494 VXLAN tunnels but MVPN uses MPLS P2MP tunnels, or where EVPN uses 2495 Ingress Replication but MVPN uses MPLS P2MP tunnels. 2497 Similarly, no special procedures are needed to prevent duplicate data 2498 delivery on ethernet segments that are multi-homed. 2500 The MEG does have some special procedures (described below) for 2501 interworking between EVPN and MVPN; these have to do with selection 2502 of the Upstream PE for a given multicast source, with the exporting 2503 of VPN-IP routes, and with the generation of MVPN C-multicast routes 2504 triggered by the installation of SMET routes. 2506 6.1.2.1. MVPN Sources with EVPN Receivers 2508 6.1.2.1.1. Identifying MVPN Sources 2510 Consider a multicast source S. It is possible that a MEG will import 2511 both an EVPN unicast route to S and a VPN-IP route (or an ordinary IP 2512 route), where the prefix length of each route is the same. In order 2513 to draw (S,G) multicast traffic for any group G, the MEG SHOULD use 2514 the EVPN route rather than the VPN-IP or IP route to determine the 2515 "Upstream PE" (see section 5 of [RFC6513]). 2517 Doing so ensures that when an EVPN tenant system desires to receive a 2518 multicast flow from another EVPN tenant system, the traffic from the 2519 source to that receiver stays within the EVPN domain. This prevents 2520 problems that might arise if there is a unicast route via L3VPN to S, 2521 but no multicast routers along the routed path. This also prevents 2522 problem that might arise as a result of the fact that the MEGs will 2523 import each others' VPN-IP routes. 2525 In the Section 6.1.2.1.2, we describe the procedures to be used when 2526 the selected route to S is a VPN-IP route. 2528 6.1.2.1.2. Joining a Flow from an MVPN Source 2530 Consider a tenant system, R, on a particular BD, BD-R. Suppose R 2531 wants to receive (S,G) multicast traffic, where source S is not 2532 attached to any PE in the EVPN Tenant Domain, but is attached to an 2533 MVPN PE. 2535 o Suppose R is on a singly homed ethernet segment of BD-R, and that 2536 segment is attached to PE1, where PE1 is a MEG. PE1 learns via 2537 IGMP/MLD listening that R is interested in (S,G). PE1 determines 2538 from its VRF that there is no route to S within the Tenant Domain 2539 (i.e., no EVPN RT-2 route with S's IP address), but that there is 2540 a route to S via L3VPN (i.e., the VRF contains a subnet or host 2541 route to S that was received as a VPN-IP route). PE1 thus 2542 originates (if it hasn't already) an MVPN C-multicast Source Tree 2543 Join(S,G) route. The route is constructed according to normal 2544 MVPN procedures. 2546 The layer 2 multicast state is constructed as specified in 2547 Section 4.1. 2549 In the layer 3 multicast state, the IIF is the appropriate MVPN 2550 tunnel, and the IRB interface to BD-R is added to the OIF list. 2552 When PE1 receives (S,G) traffic from the appropriate MVPN tunnel, 2553 it performs IP processing of the traffic, and then sends the 2554 traffic down its IRB interface to BD-R. Following normal OISM 2555 procedures, the (S,G) traffic will be encapsulated for ethernet 2556 and sent out the AC to which R is attached. 2558 o Suppose R is on a singly homed ethernet segment of BD-R, and that 2559 segment is attached to PE1, where PE1 is an OISM PE but is NOT a 2560 MEG. PE1 learns via IGMP/MLD listening that R is interested in 2561 (S,G). PE1 follows normal OISM procedures, originating an SBD- 2562 SMET route for (S,G); this route will be received by all the MEGs 2563 of the Tenant Domain, including the MEG SBD-DR. The MEG SBD-DR 2564 can determine from PE1's IMET routes whether PE1 is itself a MEG. 2565 If PE1 is not a MEG, the MEG SBD-DR will originate (if it hasn't 2566 already) an MVPN C-multicast Source Tree Join(S,G) route. This 2567 will cause the MEG SBD-DR to receive (S,G) traffic on an MVPN 2568 tunnel. 2570 The layer 2 multicast state is constructed as specified in 2571 Section 4.1. 2573 In the layer 3 multicast state, the IIF is the appropriate MVPN 2574 tunnel, and the IRB interface to the SBD is added to the OIF list. 2576 When the MEG SBD-DR receives (S,G) traffic on an MVPN tunnel, it 2577 performs IP processing of the traffic, and the sends the traffic 2578 down its IRB interface to the SBD. Following normal OISM 2579 procedures, the traffic will be encapsulated for ethernet and 2580 delivered to all PEs in the Tenant Domain that have interest in 2581 (S,G), including PE1. 2583 o If R is on a multi-homed ethernet segment of BD-R, one of the PEs 2584 attached to the segment will be its DF (following normal EVPN 2585 procedures), and the DF will know (via IGMP/MLD listening or the 2586 procedures of [IGMP-Proxy]) that a tenant system reachable via one 2587 of its local ACs to BD-R is interested in (S,G) traffic. The DF 2588 is responsible for originating an SBD-SMET route for (S,G), 2589 following normal OISM procedures. If the DF is a MEG, it MUST 2590 originate the corresponding MVPN C-multicast Source Tree Join(S,G) 2591 route; if the DF is not a MEG, the MEG SBD-DR SBD MUST originate 2592 the C-multicast route when it receives the SMET route. 2594 Optionally, if the non-DF is a MEG, it MAY originate the 2595 corresponding MVPN C-multicast Source Tree Join(S,G) route. This 2596 will cause the traffic to flow to both the DF and the non-DF, but 2597 only the DF will forward the traffic out an AC. This allows for 2598 quicker recovery if the DF's local AC to R fails. 2600 o If R is attached to a non-OISM PE, it will receive the traffic via 2601 an IPMG, as specified in Section 5. 2603 If an EVPN-attached receiver is interested in (*,G) traffic, and if 2604 it is possible for there to be sources of (*,G) traffic that are 2605 attached only to L3VPN nodes, the MEGs will have to know the group- 2606 to-RP mappings. That will enable them to originate MVPN C-multicast 2607 Shared Tree Join(*,G) routes and to send them towards the RP. (Since 2608 we are assuming in this section that there are no tenant multicast 2609 routers attached to the EVPN Tenant Domain, the RP must be attached 2610 via L3VPN. Alternatively, the MEG itself could be configured to 2611 function as an RP for group G.) 2613 The layer 2 multicast states are constructed as specified in 2614 Section 4.1. 2616 In the layer 3 (*,G) multicast state, the IIF is the appropriate MVPN 2617 tunnel. A MEG will add to the (*,G) OIF list its IRB interfaces for 2618 any BDs containing locally attached receivers. If there are 2619 receivers attached to other EVPN PEs, then whenever (S,G) traffic 2620 from an external source matches a (*,G) state, the MEG will create 2621 (S,G) state, with the MVPN tunnel as the IIF, the OIF list copied 2622 from the (*,G) state, and the SBD IRB interface added to the OIF 2623 list. (Please see the discussion in Section 6.1.1 regarding the 2624 inclusion of the SBD IRB interface in a (*,G) state; the SBD IRB 2625 interface is used in the OIF list only for traffic from external 2626 sources.) 2628 Normal MVPN procedures will then result in the MEG getting the (*,G) 2629 traffic from all the multicast sources for G that are attached via 2630 L3VPN. This traffic arrives on MVPN tunnels. When the MEG removes 2631 the traffic from these tunnels, it does the IP processing. If there 2632 are any receivers on a given BD, BD-R, that are attached via local 2633 EVPN ACs, the MEG sends the traffic down its BD-R IRB interface. If 2634 there are any other EVPN PEs that are interested in the (*,G) 2635 traffic, the MEG sends the traffic down the SBD IRB interface. 2636 Normal OISM procedures then distribute the traffic as needed to other 2637 EVPN-PEs. 2639 6.1.2.2. EVPN Sources with MVPN Receivers 2641 6.1.2.2.1. General procedures 2643 Consider the case where an EVPN tenant system S is sending IP 2644 multicast traffic to group G, and there is a receiver R for the (S,G) 2645 traffic that is attached to the L3VPN, but not attached to the EVPN 2646 Tenant Domain. (We assume in this document that the L3VPN/MVPN-only 2647 nodes will not have any special procedures to deal with the case 2648 where a source is inside an EVPN domain.) 2650 In this case, an L3VPN PE through which R can be reached has to send 2651 an MVPN C-multicast Join(S,G) route to one of the MEGs that is 2652 attached to the EVPN Tenant Domain. For this to happen, the L3VPN PE 2653 must have imported a VPN-IP route for S (either a host route or a 2654 subnet route) from a MEG. 2656 If a MEG determines that there is multicast source transmitting on 2657 one of its ACs, the MEG SHOULD originate a VPN-IP host route for that 2658 source. This determination SHOULD be made by examining the IP 2659 multicast traffic that arrives on the ACs. (It MAY be made by 2660 provisioning.) A MEG SHOULD NOT export a VPN-IP host route for any 2661 IP address that is not known to be a multicast source (unless it has 2662 some other reason for exporting such a route). The VPN-IP host route 2663 for a given multicast source MUST be withdrawn if the source goes 2664 silent for a configurable period of time, or if it can be determined 2665 that the source is no longer reachable via a local AC. 2667 A MEG SHOULD also originate a VPN-IP subnet route for each of the BDs 2668 in the Tenant Domain. 2670 VPN-IP routes exported by a MEG must carry any attributes or extended 2671 communities that are required by L3VPN and MVPN. In particular, a 2672 VPN-IP route exported by a MEG must carry a VRF Route Import Extended 2673 Community corresponding to the IP-VRF from which it is imported, and 2674 a Source AS Extended Community. 2676 As a result, if S is attached to a MEG, the L3VPN nodes will direct 2677 their MVPN C-multicast Join routes to that MEG. Normal MVPN 2678 procedures will cause the traffic to be delivered to the L3VPN nodes. 2679 The layer 3 multicast state for (S,G) will have the MVPN tunnel on 2680 its OIF list. The IIF will be the IRB interface leading to the BD 2681 containing S. 2683 If S is not attached to a MEG, the L3VPN nodes will direct their 2684 C-multicast Join routes to whichever MEG appears to be on the best 2685 route to S's subnet. Upon receiving the C-multicast Join, that MEG 2686 will originate an EVPN SMET route for (S,G). As a result, the MEG 2687 will receive the (S,G) traffic at layer 2 via the OISM procedures. 2688 The (S,G) traffic will be sent up the appropriate IRB interface, and 2689 the layer 3 MVPN procedures will ensure that the traffic is delivered 2690 to the L3VPN nodes that have requested it. The layer 3 multicast 2691 state for (S,G) will have the MVPN tunnel in the OIF list, and the 2692 IIF will be one of the following: 2694 o If S belongs to a BD that is attached to the MEG, the IIF will be 2695 the IRB interface to that BD; 2697 o Otherwise the IIF will be the SBD IRB interface. 2699 Note that this works even if S is attached to a non-OISM PE, per the 2700 procedures of Section 5. 2702 6.1.2.2.2. Any-Source Multicast (ASM) Groups 2704 Suppose the MEG SBD-DR learns that one of the PEs in its Tenant 2705 Domain is interested in (*,G), traffic, where G is an Any-Source 2706 Multicast (ASM) group. If there are no tenant multicast routers, the 2707 MEG SBD-DR SHOULD perform the "First Hop Router" (FHR) functionality 2708 for group G on behalf of the Tenant Domain, as described in 2709 [RFC7761]. This means that the MEG SBD-DR must know the identity of 2710 the Rendezvous Point (RP) for each group, must send Register messages 2711 to the Rendezvous Point, etc. 2713 If the MEG SBD-DR is to be the FHR for the Tenant Domain, it must see 2714 all the multicast traffic that is sourced from within the domain and 2715 destined to an ASM group address. The MEG can ensure this by 2716 originating an SBD-SMET route for (*,*). 2718 (As a possible optimization, an SBD-SMET route for (*, "any ASM 2719 group") may be defined in a future revision of this draft.) 2721 In some deployment scenarios, it may be preferred that the MEG that 2722 receives the (S,G) traffic over an AC be the one provides the FHR 2723 functionality. This behavior is OPTIONAL. If this option is used, 2724 it MUST be ensured that the MEG DR does not provide the FHR 2725 functionality for (S,G) traffic that is attached to another MEG; FHR 2726 functionality for (S,G) traffic from a particular source S MUST be 2727 provided by only a single router. 2729 Other deployment scenarios are also possible. For example, one might 2730 want to configure the MEGs to themselves be RPs. In this case, the 2731 RPs would have to exchange with each other information about which 2732 sources are active. The method exchanging such information is 2733 outside the scope of this document. 2735 6.1.2.2.3. Source on Multihomed Segment 2737 Suppose S is attached to a segment that is all-active multi-homed to 2738 PEl and PE2. If S is transmitting to two groups, say G1 and G2, it 2739 is possible that PE1 will receive the (S,G1) traffic from S while PE2 2740 receives the (S,G2) traffic from S. 2742 This creates an issue for MVPN/EVPN interworking, because there is no 2743 way to cause L3VPN/MVPN nodes to select PE1 as the ingress PE for 2744 (S,G1) traffic while selecting PE2 as the ingress PE for (S,G2) 2745 traffic. 2747 However, the following procedure ensures that the IP multicast 2748 traffic will still flow, even if the L3VPN/MVPN nodes picks the 2749 "wrong" EVPN-PE as the Upstream PE for (say) the (S,G1) traffic. 2751 Suppose S is on an ethernet segment, belonging to BD1, that is 2752 multi-homed to both PE1 and PE2, where PE1 is a MEG. And suppose 2753 that IP multicast traffic from S to G travels over the AC that 2754 attaches the segment to PE2 . If PE1 receives a C-multicast Source 2755 Tree Join (S,G) route, it MUST originate an SMET route for (S,G). 2756 Normal OISM procedures will then cause PE2 to send the (S,G) traffic 2757 to PE1 on an EVPN IP multicast tunnel. Normal OISM procedures will 2758 also cause PE1 to send the (S,G) traffic up its BD1 IRB interface. 2759 Normal MVPN procedures will then cause PE1 to forward the traffic on 2760 an MVPN tunnel. In this case, the routing is not optimal, but the 2761 traffic does flow correctly. 2763 6.1.2.3. Obtaining Optimal Routing of Traffic Between MVPN and EVPN 2765 The routing of IP multicast traffic between MVPN nodes and EVPN nodes 2766 will be optimal as long as there is a MEG along the optimal route. 2767 There are various deployment strategies that can be used to obtain 2768 optimal routing between MVPN and EVPN. 2770 In one such scenario, a Tenant Domain will have a small number of 2771 strategically placed MEGs. For example, a Data Center may have a 2772 small number of MEGs that connect it to a wide-area network. Then 2773 the optimal route into or out of the Data Center would be through the 2774 MEGs. 2776 In this scenario, the MEGs do not need to originate VPN-IP host 2777 routes for the multicast sources, they only need to originate VPN-IP 2778 subnet routes. The internal structure of the EVPN is completely 2779 hidden from the MVPN node. EVPN actions such as MAC Mobility and 2780 Mass Withdrawal ([RFC7432]) have zero impact on the MVPN control 2781 plane. 2783 While this deployment scenario provides the most optimal routing and 2784 has the least impact on the installed based of MVPN nodes, it does 2785 complicate network planning considerations. 2787 Another way of providing routing that is close to optimal is to turn 2788 each EVPN PE into a MEG. Then routing of MVPN-to-EVPN traffic is 2789 optimal. However, routing of EVPN-to-MVPN traffic is not guaranteed 2790 to be optimal when a source host is on a multi-homed ethernet segment 2791 (as discussed in Section 6.1.2.2.) 2793 The obvious disadvantage of this method is that it requires every 2794 EVPN PE to be a MEG. 2796 The procedures specified in this document allow an operator to add 2797 MEG functionality to any subset of his EVPN OISM PEs. This allows an 2798 operator to make whatever trade-offs he deems appropriate between 2799 optimal routing and MEG deployment. 2801 6.1.2.4. Selecting the MEG SBD-DR 2803 Every PE that is eligible for selection as the MEG SBD-DR originates 2804 an SBD-IMET route. As stated in Section 5, these SBD-IMET routes 2805 carry a Multicast Flags EC with the MEG Flag set. 2807 These SBD-IMET routes SHOULD also carry a DF Election EC. The DF 2808 Election EC and its use is specified in ([DF-Election-Framework]). 2809 When the route is originated, the AC-DF bit in the DF Election EC 2810 SHOULD be set to zero. This bit is not used when selecting a MEG 2811 SBD-DR, i.e., it MUST be ignored by the receiver of an SBD-IMET 2812 route. 2814 In the context of a given Tenant Domain, to select the MEG SBD-DR, 2815 the MEGs of the Tenant Domain perform the following procedure: 2817 o From the set of received SBD-IMET routes for the given tenant 2818 domain, determine he candidate set of PEs that support MEG 2819 functionality for that domain. 2821 o Select a DF Election algorithm as specified in 2822 [DF-Election-Framework]. Some of the possible algorithms can be 2823 found, e.g., in [RFC7432], [DF-Election-Framework], and 2824 [EVPN-DF-WEIGHTED]. 2826 o Apply the DF Election Algorithm (see [DF-Election-Framework]) to 2827 the candidate set of PEs. The "winner" becomes the MEG SBD-DR. 2829 Note that if a given PE supports IPMG (Section 6.1.2) or PEG 2830 (Section 6.1.4) functionality as well as MEG functionality, its 2831 SBD-IMET routes carry only one DF Election EC. 2833 6.1.3. Interworking with 'Global Table Multicast' 2835 If multicast service to the outside sources and/or receivers is 2836 provided via the BGP-based "Global Table Multicast" (GTM) procedures 2837 of [RFC7716], the procedures of Section 6.1.2 can easily be adapted 2838 for EVPN/GTM interworking. The way to adapt the MVPN procedures to 2839 GTM is explained in [RFC7716]. 2841 6.1.4. Interworking with PIM 2843 As we have been discussing, there may be receivers in an EVPN tenant 2844 domain that are interested in multicast flows whose sources are 2845 outside the EVPN Tenant Domain. Or there may be receivers outside an 2846 EVPN Tenant Domain that are interested in multicast flows whose 2847 sources are inside the Tenant Domain. 2849 If the outside sources and/or receivers are part of an MVPN, 2850 interworking procedures are covered in Section 6.1.2. 2852 There are also cases where an external source or receiver are 2853 attached via IP, and the layer 3 multicast routing is done via PIM. 2854 In this case, the interworking between the "PIM domain" and the EVPN 2855 tenant domain is done at L3 Gateways that perform "PIM/EVPN Gateway" 2856 (PEG) functionality. A PEG is very similar to a MEG, except that its 2857 layer 3 multicast routing is done via PIM rather than via BGP. 2859 If external sources or receivers for a given group are attached to a 2860 PEG via a layer 3 interface, that interface should be treated as a 2861 VRF interface attached to the Tenant Domain's L3VPN VRF. The layer 3 2862 multicast routing instance for that Tenant Domain will either run PIM 2863 on the VRF interface or will listen for IGMP/MLD messages on that 2864 interface. If the external receiver is attached elsewhere on an IP 2865 network, the PE has to enable PIM on its interfaces to the backbone 2866 network. In both cases, the PE needs to perform PEG functionality, 2867 and its IMET routes must carry the Muliticast Flags EC with the PEG 2868 flag set. 2870 For each BD on which there is a multicast source or receiver, one of 2871 the PEGs will becomes the PEG DR. DR selection can be done using the 2872 same procedures specified in Section 6.1.2.4, except with "PEG" 2873 substituted for "MEG". 2875 As long as there are no tenant multicast routers within the EVPN 2876 Tenant Domain, the PEGs do not need to run PIM on their IRB 2877 interfaces. 2879 6.1.4.1. Source Inside EVPN Domain 2881 If a PEG receives a PIM Join(S,G) from outside the EVPN tenant 2882 domain, it may find it necessary to create (S,G) state. The PE needs 2883 to determine whether S is within the Tenant Domain. If S is not 2884 within the EVPN Tenant Domain, the PE carries out normal layer 3 2885 multicast routing procedures. If S is within the EVPN tenant domain, 2886 the IIF of the (S,G) state is set as follows: 2888 o if S is on a BD that is attached to the PE, the IIF is the PE's 2889 IRB interface to that BD; 2891 o if S is not on a BD that is attached to the PE, the IIF is the 2892 PE's IRB interface to the SBD. 2894 When the PE creates such an (S,G) state, it MUST originate (if it 2895 hasn't already) an SBD-SMET route for (S,G). This will cause it to 2896 pull the (S,G) traffic via layer 2. When the traffic arrives over an 2897 EVPN tunnel, it gets sent up an IRB interface where the layer 3 2898 multicast routing determines the packet's disposition. The SBD-SMET 2899 route is withdrawn when the (S,G) state no longer exists (unless 2900 there is some other reason for not withdrawing it). 2902 If there are no tenant multicast routers with the EVPN tenant domain, 2903 there cannot be an RP in the Tenant Domain, so a PEG does not have to 2904 handle externally arriving PIM Join(*,G) messages. 2906 The PEG DR for a particular BD MUST act as the a First Hop Router for 2907 that BD. It will examine all (S,G) traffic on the BD, and whenever G 2908 is an ASM group, the PEG DR will send Register messages to the RP for 2909 G. This means that the PEG DR will need to pull all the (S,G) 2910 traffic originating on a given BD, by originating an SMET (*,*) route 2911 for that BD. If a PEG DR is the DR for all the BDS, in SHOULD 2912 originate just an SBD-SMET (*,*) route rather than an SMET (*,*) 2913 route for each BD. 2915 The rules for exporting IP routes to multicast sources are the same 2916 as those specified for MEGs in Section 6.1.2.2, except that the 2917 exported routes will be IP routes rather than VPN-IP routes, and it 2918 is not necessary to attach the VRF Route Import EC or the Source AS 2919 EC. 2921 When a source is on a multi-homed segment, the same issue discussed 2922 in Section 6.1.2.2.3 exists. Suppose S is on an ethernet segment, 2923 belonging to BD1, that is multi-homed to both PE1 and PE2, where PE1 2924 is a PEG. And suppose that IP multicast traffic from S to G travels 2925 over the AC that attaches the segment to PE2. If PE1 receives an 2926 external PIM Join (S,G) route, it MUST originate an SMET route for 2927 (S,G). Normal OISM procedures will cause PE2 to send the (S,G) 2928 traffic to PE1 on an EVPN IP multicast tunnel. Normal OISM 2929 procedures will also cause PE1 to send the (S,G) traffic up its BD1 2930 IRB interface. Normal PIM procedures will then cause PE1 to forward 2931 the traffic along a PIM tree. In this case, the routing is not 2932 optimal, but the traffic does flow correctly. 2934 6.1.4.2. Source Outside EVPN Domain 2936 By means of normal OISM procedures, a PEG learns whether there are 2937 receivers in the Tenant Domain that are interested in receiving (*,G) 2938 or (S,G) traffic. The PEG must determine whether S (or the RP for G) 2939 is outside the EVPN Tenant Domain. If so, and if there is a receiver 2940 on BD1 interested in receiving such traffic, the PEG DR for BD1 is 2941 responsible for originating a PIM Join(S,G) or Join(*,G) control 2942 message. 2944 An alternative would be to allow any PEG that is directly attached to 2945 a receiver to originate the PIM Joins. Then the PEG DR would only 2946 have to originate PIM Joins on behalf of receivers that are not 2947 attached to a PEG. However, if this is done, it is necessary for the 2948 PEGs to run PIM on all their IRB interfaces, so that the PIM Assert 2949 procedures can be used to prevent duplicate delivery to a given BD. 2951 The IIF for the layer 3 (S,G) or (*,G) state is determined by normal 2952 PIM procedures. If a receiver is on BD1, and the PEG DR is attached 2953 to BD1, its IRB interface to BD1 is added to the OIF list. This 2954 ensures that any receivers locally attached to the PEG DR will 2955 receive the traffic. If there are receivers attached to other EVPN 2956 PEs, then whenever (S,G) traffic from an external source matches a 2957 (*,G) state, the PEG will create (S,G) state. The IIF will be set to 2958 whatever external interface the traffic is expected to arrive on 2959 (copied from the (*,G) state), the OIF list is copied from the (*,G) 2960 state, and the SBD IRB interface added to the OIF list. 2962 6.2. Interworking with PIM via an External PIM Router 2964 Section 6.1 describes how to use an OISM PE router as the gateway to 2965 a non-EVPN multicast domain, when the EVPN tenant domain is not being 2966 used as an intermediate transit network for multicast. An 2967 alternative approach is to have one or more external PIM routers 2968 (perhaps operated by a tenant) on one of the BDs of the tenant 2969 domain. We will refer to this BD as the "gateway BD". 2971 In this model: 2973 o The EVPN Tenant Domain is treated as a stub network attached to 2974 the external PIM routers. 2976 o The external PIM routers follow normal PIM procedures, and provide 2977 the FHR and LHR functionality for the entire Tenant Domain. 2979 o The OISM PEs do not run PIM. 2981 o There MUST NOT be more than one gateway BD. 2983 o If an OISM PE not attached to the gateway BD has interest in a 2984 given multicast flow, it conveys that interest, following normal 2985 OISM procedures, by originating an SBD-SMET route for that flow. 2987 o If a PE attached to the gateway BD receives an SBD-SMET, it may 2988 need to generate and transmit a corresponding IGMP/MLD Join out 2989 one or more of its ACs. (Procedures for generating an IGMP/MLD 2990 Join as a result of receiving an SMET route are given in 2991 [IGMP-Proxy].) The PE MUST know which BD is the Gateway BD and 2992 MUST NOT transmit an IGMP/MLD Join to any other BDs. Furthermore, 2993 even if a particular AC is part of that BD, the PE SHOULD NOT 2994 transmit an IGMP/MLD Join on that AC unless that an external PIM 2995 route is attached via that AC. 2997 As a result, IGMP/MLD messages will seen by the external PIM 2998 routers on the gateway BD, and those external PIM routers will 2999 send PIM Join messages externally as required. Traffic of the 3000 given multicast flow will then be received by one of the external 3001 PIM routers, and that traffic will be forwarded by that router to 3002 the gateway BD. 3004 The normal OISM procedures will then cause the given multicast 3005 flow to be tunneled to any PEs of the EVPN Tenant Domain that have 3006 interest in the flow. PEs attached to the gateway BD will see the 3007 flow as originating from the gateway BD, other PEs will see the 3008 flow as originating from the SBD. 3010 o An OISM PE attached to a gateway BD MUST set its layer 2 multicast 3011 state to indicate that each AC to the gateway BD has interest in 3012 all multicast flows. It MUST also originate an SMET route for 3013 (*,*). The procedures for originating SMET routes are discussed 3014 in Section 2.5. 3016 This will cause the OISM PEs attached to the gateway BD to receive 3017 all the IP multicast traffic that is sourced within the EVPN 3018 tenant domain, and to transmit that traffic to the gateway BD, 3019 where the external PIM routers will see it. This enables the 3020 external PIM routers to perform FHR functions on behalf of the 3021 entire Tenant Domain. (Of course, if the gateway BD has a 3022 multi-homed segment, only the PE that is the DF for that segment 3023 will transmit the multicast traffic to the segment.) 3025 7. Using an EVPN Tenant Domain as an Intermediate (Transit) Network for 3026 Multicast traffic 3028 In this section, we consider the scenario where one or more BDs of an 3029 EVPN Tenant Domain are being used to carry IP multicast traffic for 3030 which the source and at least one receiver are not part the tenant 3031 domain. That is, one or more BDs of the Tenant Domain are 3032 intermediate "links" of a larger multicast tree created by PIM. 3034 We define a "tenant multicast router" as a multicast router, running 3035 PIM, that is: 3037 1. attached to one or more BDs of the Tenant Domain, but 3039 2. is not an EVPN PE router. 3041 In order an EVPN Tenant Domain to be used as a transit network for IP 3042 multicast, one or more of its BDs must have tenant multicast routers, 3043 and an OISM PE that attaching to such a BD MUST be provisioned to 3044 enable PIM on its IRB interface to that BD. (This is true even if 3045 none of the tenant routers is on a segment attached to the PE.) 3046 Further, all the OISM PEs (even ones not attached to a BD with tenant 3047 multicast routers) MUST be provisioned to enable PIM on their SBD IRB 3048 interfaces. 3050 If PIM is enabled on a particular BD, the DR Selection procedure of 3051 Section 6.1.2.4 MUST be replaced by the normal PIM DR Election 3052 procedure of [RFC7761]. Note that this may result in one of the 3053 tenant routers being selected as the DR, rather than one of the OISM 3054 PE routers. In this case, First Hop Router and Last Hop Router 3055 functionality will not be performed by any of the EVPN PEs. 3057 A PIM control message on a particular BD is considered to be a 3058 link-local multicast message, and as such is sent transparently from 3059 PE to PE via the BUM tunnel for that BD. This is true whether the 3060 control message was received from an AC, or whether it was received 3061 from the local layer 3 routing instance via an IRB interface. 3063 A PIM Join/Prune message contains three fields that are relevant to 3064 the present discussion: 3066 o Upstream Neighbor 3068 o Group Address (G) 3070 o Source Address (S), omitted in the case of (*,G) Join/Prune 3071 messages. 3073 We will generally speak of a PIM Join as a "Join(S,G)" or a 3074 "Join(*,G)" message, and will use the term "Join(X,G)" to mean 3075 "either Join(S,G) or Join(*,G)". In the context of a Join(X,G), we 3076 will use the term "X" to mean "S in the case of (S,G), or G's RP in 3077 the case of (*,G)". 3079 Suppose BD1 contains two tenant multicast routers, C1 and C2. 3080 Suppose C1 is on a segment attached to PE1, and C2 is on a segment 3081 attached to PE2. When C1 sends a PIM Join(X,G) to BD1, the Upstream 3082 Neighbor field might be set to either PE1, PE2, or C2. C1 chooses 3083 the Upstream Neighbor based on its unicast routing. Typically, it 3084 will choose as the Upstream Neighbor the PIM router on BD1 that is 3085 "closest" (according to the unicast routing) to X. Note that this 3086 will not necessarily be PE1. PE1 may not even be visible to the 3087 unicast routing algorithm used by the tenant routers. Even if it is, 3088 it is unlikely to be the PIM router that is closest to X. So we need 3089 to consider the following two cases: 3091 1. C1 sends a PIM Join(X,G) to BD1, with PE1 as the Upstream 3092 Neighbor. 3094 PE1's PIM routing instance will see the Join arrive on the BD1 3095 IRB interface. If X is not within the Tenant Domain, PE1 3096 handles the Join according to normal PIM procedures. This will 3097 generally result in PE1 selecting an Upstream Neighbor and 3098 sending it a Join(X,G). 3100 If X is within the Tenant Domain, but is attached to some other 3101 PE, PE1 sends (if it hasn't already) an SBD-SMET route for 3102 (X,G). The IIF of the layer 3 (X,G) state will be the SBD IRB 3103 interface, and the OIF list will include the IRB interface to 3104 BD1. 3106 The SBD-SMET route will pull the (X,G) traffic to PE1, and the 3107 (X,G) state will result in the (X,G) traffic being forwarded to 3108 C1. 3110 If X is within the Tenant Domain, but is attached to PE1 itself, 3111 no SBD-SMET route is sent. The IIF of the layer 3 (X,G) state 3112 will be the IRB interface to X's BD, and the OIF list will 3113 include the IRB interface to BD1. 3115 2. C1 sends a PIM Join(X,G) to BD1, with either PE2 or C2 as the 3116 Upstream Neighbor. 3118 PE1's PIM routing instance will see the Join arrive on the BD1 3119 IRB interface. If neither X nor Upstream Neighbor is within the 3120 tenant domain, PE1 handles the Join according to normal PIM 3121 procedures. This will NOT result in PE1 sending a Join(X,G). 3123 If either X or Upstream Neighbor is within the Tenant Domain, 3124 PE1 sends (if it hasn't already) an SBD-SMET route for (X,G). 3125 The IIF of the layer 3 (X,G) state will be the SBD IRB 3126 interface, and the OIF list will include the IRB interface to 3127 BD1. 3129 The SBD-SMET route will pull the (X,G) traffic to PE1, and the 3130 (X,G) state will result in the (X,G) traffic being forwarded to 3131 C1. 3133 8. IANA Considerations 3135 IANA is requested to assign new flags in the "Multicast Flags 3136 Extended Community Flags" registry. These flags are: 3138 o IPMG 3140 o MEG 3142 o PEG 3144 o OISM SBD 3146 o OISM-supported 3148 9. Security Considerations 3150 This document uses protocols and procedures defined in the normative 3151 references, and inherits the security considerations of those 3152 references. 3154 This document adds flags or Extended Communities (ECs) to a number of 3155 BGP routes, in order to signal that particular nodes support the 3156 OISM, IPMG, MEG, and/or PEG functionalities that are defined in this 3157 document. Incorrect addition, removal, or modification of those 3158 flags and/or ECs will cause the procedures defined herein to 3159 malfunction, in which case loss or diversion of data traffic is 3160 possible. 3162 10. Acknowledgements 3164 The authors thank Vikram Nagarajan and Princy Elizabeth for their 3165 work on Section 6.2. The authors also benefited tremendously from 3166 discussions with Aldrin Isaac on EVPN multicast optimizations. 3168 11. References 3170 11.1. Normative References 3172 [DF-Election-Framework] 3173 Rabadan, J., Mohanty, S., Sajassi, A., Drake, J., Nagaraj, 3174 K., and S. Sathappan, "Framework for EVPN Designated 3175 Forwarder Election Extensibility", internet-draft draft- 3176 ietf-bess-evpn-df-election-framework-07.txt, December 3177 2018. 3179 [EVPN-AR] Rabadan, J., Ed., "Optimized Ingress Replication solution 3180 for EVPN", internet-draft ietf-bess-evpn-optimized-ir- 3181 06.txt, October 2018. 3183 [EVPN-BUM] 3184 Zhang, Z., Lin, W., Rabadan, J., and K. Patel, "Updates on 3185 EVPN BUM Procedures", internet-draft ietf-bess-evpn-bum- 3186 procedure-updates-05.txt, December 2018. 3188 [EVPN-IRB] 3189 Sajassi, A., Salam, S., Thoria, S., Drake, J., and J. 3190 Rabadan, "Integrated Routing and Bridging in EVPN", 3191 internet-draft draft-ietf-bess-evpn-inter-subnet- 3192 forwarding-05.txt, July 2018. 3194 [EVPN_IP_Prefix] 3195 Rabadan, J., Henderickx, W., Drake, J., Lin, W., and A. 3196 Sajassi, "IP Prefix Advertisement in EVPN", internet- 3197 draft ietf-bess-evpn-prefix-advertisement-11.txt, May 3198 2018. 3200 [IGMP-Proxy] 3201 Sajassi, A., Thoria, S., Patel, K., Yeung, D., Drake, J., 3202 and W. Lin, "IGMP and MLD Proxy for EVPN", internet-draft 3203 draft-ietf-bess-evpn-igmp-mld-proxy-02.txt, June 2018. 3205 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 3206 Requirement Levels", BCP 14, RFC 2119, 3207 DOI 10.17487/RFC2119, March 1997, 3208 . 3210 [RFC2236] Fenner, W., "Internet Group Management Protocol, Version 3211 2", RFC 2236, DOI 10.17487/RFC2236, November 1997, 3212 . 3214 [RFC2710] Deering, S., Fenner, W., and B. Haberman, "Multicast 3215 Listener Discovery (MLD) for IPv6", RFC 2710, 3216 DOI 10.17487/RFC2710, October 1999, 3217 . 3219 [RFC3032] Rosen, E., Tappan, D., Fedorkow, G., Rekhter, Y., 3220 Farinacci, D., Li, T., and A. Conta, "MPLS Label Stack 3221 Encoding", RFC 3032, DOI 10.17487/RFC3032, January 2001, 3222 . 3224 [RFC4360] Sangli, S., Tappan, D., and Y. Rekhter, "BGP Extended 3225 Communities Attribute", RFC 4360, DOI 10.17487/RFC4360, 3226 February 2006, . 3228 [RFC6625] Rosen, E., Ed., Rekhter, Y., Ed., Hendrickx, W., and R. 3229 Qiu, "Wildcards in Multicast VPN Auto-Discovery Routes", 3230 RFC 6625, DOI 10.17487/RFC6625, May 2012, 3231 . 3233 [RFC7153] Rosen, E. and Y. Rekhter, "IANA Registries for BGP 3234 Extended Communities", RFC 7153, DOI 10.17487/RFC7153, 3235 March 2014, . 3237 [RFC7432] Sajassi, A., Ed., Aggarwal, R., Bitar, N., Isaac, A., 3238 Uttaro, J., Drake, J., and W. Henderickx, "BGP MPLS-Based 3239 Ethernet VPN", RFC 7432, DOI 10.17487/RFC7432, February 3240 2015, . 3242 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 3243 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 3244 May 2017, . 3246 11.2. Informative References 3248 [EVPN-BIER] 3249 Zhang, Z., Przygienda, T., Sajassi, A., and J. Rabadan, 3250 "EVPN BUM Using BIER", internet-draft ietf-bier-evpn- 3251 01.txt, April 2018. 3253 [EVPN-DF-WEIGHTED] 3254 Rabadan, J., Sathappan, S., Przygienda, T., Lin, W., 3255 Drake, J., Sajassi, A., and S. Mohanty, "Preference-based 3256 EVPN DF Election", internet-draft ietf-bess-evpn-pref-df- 3257 03.txt, December 2018. 3259 [RFC4364] Rosen, E. and Y. Rekhter, "BGP/MPLS IP Virtual Private 3260 Networks (VPNs)", RFC 4364, DOI 10.17487/RFC4364, February 3261 2006, . 3263 [RFC6513] Rosen, E., Ed. and R. Aggarwal, Ed., "Multicast in MPLS/ 3264 BGP IP VPNs", RFC 6513, DOI 10.17487/RFC6513, February 3265 2012, . 3267 [RFC6514] Aggarwal, R., Rosen, E., Morin, T., and Y. Rekhter, "BGP 3268 Encodings and Procedures for Multicast in MPLS/BGP IP 3269 VPNs", RFC 6514, DOI 10.17487/RFC6514, February 2012, 3270 . 3272 [RFC7606] Chen, E., Ed., Scudder, J., Ed., Mohapatra, P., and K. 3273 Patel, "Revised Error Handling for BGP UPDATE Messages", 3274 RFC 7606, DOI 10.17487/RFC7606, August 2015, 3275 . 3277 [RFC7716] Zhang, J., Giuliano, L., Rosen, E., Ed., Subramanian, K., 3278 and D. Pacella, "Global Table Multicast with BGP Multicast 3279 VPN (BGP-MVPN) Procedures", RFC 7716, 3280 DOI 10.17487/RFC7716, December 2015, 3281 . 3283 [RFC7761] Fenner, B., Handley, M., Holbrook, H., Kouvelas, I., 3284 Parekh, R., Zhang, Z., and L. Zheng, "Protocol Independent 3285 Multicast - Sparse Mode (PIM-SM): Protocol Specification 3286 (Revised)", STD 83, RFC 7761, DOI 10.17487/RFC7761, March 3287 2016, . 3289 [RFC8296] Wijnands, IJ., Ed., Rosen, E., Ed., Dolganow, A., 3290 Tantsura, J., Aldrin, S., and I. Meilik, "Encapsulation 3291 for Bit Index Explicit Replication (BIER) in MPLS and Non- 3292 MPLS Networks", RFC 8296, DOI 10.17487/RFC8296, January 3293 2018, . 3295 Appendix A. Integrated Routing and Bridging 3297 This Appendix provides a short tutorial on the interaction of routing 3298 and bridging. First it shows the traditional model, where bridging 3299 and routing are performed in separate boxes. Then it shows the model 3300 specified in [EVPN-IRB], where a single box contains both routing and 3301 bridging functions. The latter model is presupposed in the body of 3302 this document. 3304 Figure 1 shows a "traditional" router that only does routing and has 3305 no L2 bridging capabilities. There are two LANs, LAN1 and LAN2. 3306 LAN1 is realized by switch1, LAN2 by switch2. The router has an 3307 interface, "lan1" that attaches to LAN1 (via switch1) and an 3308 interface "lan2" that attachs to LAN2 (via switch2). Each intreface 3309 is configured, as an IP interface, with an IP address and a subnet 3310 mask. 3312 +-------+ +--------+ +-------+ 3313 | | lan1| |lan2 | | 3314 H1 -----+Switch1+--------+ Router1+--------+Switch2+------H3 3315 | | | | | | 3316 H2 -----| | | | | | 3317 +-------+ +--------+ +-------+ 3318 |_________________| |__________________| 3319 LAN1 LAN2 3321 Figure 1: Conventional Router with LAN Interfaces 3323 IP traffic (unicast or multicast) that remains within a single subnet 3324 never reaches the router. For instance, if H1 emits an ethernet 3325 frame with H2's MAC address in the ethernet destination address 3326 field, the frame will go from H1 to Switch1 to H2, without ever 3327 reaching the router. Since the frame is never seen by a router, the 3328 IP datagram within the frame remains entirely unchanged; e.g., its 3329 TTL is not decremented. The ethernet Source and Destination MAC 3330 addresses are not changed either. 3332 If H1 wants to send a unicast IP datagram to H3, which is on a 3333 different subnet, H1 has to be configured with the IP address of a 3334 "default router". Let's assume that H1 is configured with an IP 3335 address of Router1 as its default router address. H1 compares H3's 3336 IP address with its own IP address and IP subnet mask, and determines 3337 that H3 is on a different subnet. So the packet has to be routed. 3338 H1 uses ARP to map Router1's IP address to a MAC address on LAN1. H1 3339 then encapsulates the datagram in an ethernet frame, using router1's 3340 MAC address as the destination MAC address, and sends the frame to 3341 Router1. 3343 Router1 then receives the frame over its lan1 interface. Router1 3344 sees that the frame is addressed to it, so it removes the ethernet 3345 encapsulation and processes the IP datagram. The datagram is not 3346 addressed to Router1, so it must be forwarded further. Router1 does 3347 a lookup of the datagram's IP destination field, and determines that 3348 the destination (H3) can be reached via Router1's lan2 interface. 3349 Router1 now performs the IP processing of the datagram: it decrements 3350 the IP TTL, adjusts the IP header checksum (if present), may fragment 3351 the packet is necessary, etc. Then the datagram (or its fragments) 3352 are encapsulated in an ethernet header, with Router1's MAC address on 3353 LAN2 as the MAC Source Address, and H3's MAC address on LAN2 (which 3354 Router1 determines via ARP) as the MAC Destination Address. Finally 3355 the packet is sent out the lan2 interface. 3357 If H1 has an IP multicast datagram to send (i.e., an IP datagram 3358 whose Destination Address field is an IP Multicast Address), it 3359 encapsulates it in an ethernet frame whose MAC Destination Address is 3360 computed from the IP Destination Address. 3362 If H2 is a receiver for that multicast address, H2 will receive a 3363 copy of the frame, unchanged, from H1. The MAC Source Address in the 3364 ethernet encapsulation does not change, the IP TTL field does not get 3365 decremented, etc. 3367 If H3 is a receiver for that multicast address, the datagram must be 3368 routed to H3. In order for this to happen, Router1 must be 3369 configured as a multicast router, and it must accept traffic sent to 3370 ethernet multicast addresses. Router1 will receive H1's multicast 3371 frame on its lan1 interface, will remove the ethernet encapsulation, 3372 and will determine how to dispatch the IP datagram based on Router1's 3373 multicast forwarding states. If Router1 knows that there is a 3374 receiver for the multicast datagram on LAN2, makes a copy of the 3375 datagram, decrements the TTL (and performs any other necessary IP 3376 processing), then encapsulates the datagram in ethernet frame for 3377 LAN2. The MAC Source Address for this frame will be Router1's MAC 3378 Source Address on LAN2. The MAC Destination Address is computed from 3379 the IP Destination Address. Finally, the frame is sent out Router1's 3380 LAN2 interface. 3382 Figure 2 shows an Integrated Router/Bridge that supports the routing/ 3383 bridging integration model of [EVPN-IRB]. 3385 +------------------------------------------+ 3386 | Integrated Router/Bridge | 3388 +-------+ +--------+ +-------+ 3389 | | IRB1| L3 |IRB2 | | 3390 H1 -----+ BD1 +--------+Routing +--------+ BD2 +------H3 3391 | | |Instance| | | 3392 H2 -----| | | | | | 3393 +-------+ +--------+ +-------+ 3394 |___________________| |____________________| 3395 LAN1 LAN2 3397 Figure 2: Integrated Router/Bridge 3399 In Figure 2, a single box consists of one or more "L3 Routing 3400 Instances". The routing/forwarding tables of a given routing 3401 instance is known as an IP-VRF ([EVPN-IRB]). In the context of EVPN, 3402 it is convenient to think of each routing instance as representing 3403 the routing of a particular tenant. Each IP-VRF is attached to one 3404 or more interfaces. 3406 When several EVPN PEs have a routing instance of the same tenant 3407 domain, those PEs advertise IP routes to the attached hosts. This is 3408 done as specified in [EVPN-IRB]. 3410 The integrated router/bridge shown in Figure 2 also attaches to a 3411 number of "Broadcast Domains" (BDs). Each BD performs the functions 3412 that are performed by the bridges in Figure 1. To the L3 routing 3413 instance, each BD appears to be a LAN. The interface attaching a 3414 particular BD to a particular IP-VRF is known as an "IRB Interface". 3415 From the perspective of L3 routing, each BD is a subnet. Thus each 3416 IRB interface is configured with a MAC address (which is the router's 3417 MAC address on the corresponding LAN), as well as an IP address and 3418 subnet mask. 3420 The integrated router/bridge shown in Figure 2 may have multiple ACs 3421 to each BD. These ACs are visible only to the bridging function, not 3422 to the routing instance. To the L3 routing instance, there is just 3423 one "interface" to each BD. 3425 If the L3 routing instance represents the IP routing of a particular 3426 tenant, the BDs attached to that routing instance are BDs belonging 3427 to that same tenant. 3429 Bridging and routing now proceed exactly as in the case of Figure 1, 3430 except that BD1 replaces Switch1, BD2 replaces Switch2, interface 3431 IRB1 replaces interface lan1, and interface IRB2 replaces interface 3432 lan2. 3434 It is important to understand that an IRB interface connects an L3 3435 routing instance to a BD, NOT to a "MAC-VRF". (See [RFC7432] for the 3436 definition of "MAC-VRF".) A MAC-VRF may contain several BDs, as long 3437 as no MAC address appears in more than one BD. From the perspective 3438 of the L3 routing instance, each individual BD is an individual IP 3439 subnet; whether each BD has its own MAC-VRF or not is irrelevant to 3440 the L3 routing instance. 3442 Figure 3 illustrates IRB when a pair of BDs (subnets) are attached to 3443 two different PE routers. In this example, each BD has two segments, 3444 and one segment of each BD is attached to one PE router. 3446 +------------------------------------------+ 3447 | Integrated Router/Bridges | 3449 +-------+ +--------+ +-------+ 3450 | | IRB1| |IRB2 | | 3451 H1 -----+ BD1 +--------+ PE1 +--------+ BD2 +------H3 3452 |(Seg-1)| |(L3 Rtg)| |(Seg-1)| 3453 H2 -----| | | | | | 3454 +-------+ +--------+ +-------+ 3455 |___________________| | |____________________| 3456 LAN1 | LAN2 3457 | 3458 | 3459 +-------+ +--------+ +-------+ 3460 | | IRB1| |IRB2 | | 3461 H4 -----+ BD1 +--------+ PE2 +--------+ BD2 +------H5 3462 |(Seg-2)| |(L3 Rtg)| |(Seg-2)| 3463 | | | | | | 3464 +-------+ +--------+ +-------+ 3466 Figure 3: Integrated Router/Bridges with Distributed Subnet 3468 If H1 needs to send an IP packet to H4, it determines from its IP 3469 address and subnet mask that H4 is on the same subnet as H1. 3470 Although H1 and H4 are not attached to the same PE router, EVPN 3471 provides ethernet communication among all hosts that are on the same 3472 BD. H1 thus uses ARP to find H4's MAC address, and sends an ethernet 3473 frame with H4's MAC address in the Destination MAC address field. 3474 The frame is received at PE1, but since the Destination MAC address 3475 is not PE1's MAC address, PE1 assumes that the frame is to remain on 3476 BD1. Therefore the packet inside the frame is NOT decapsulated, and 3477 is NOT send up the IRB interface to PE1's routing instance. Rather, 3478 standard EVPN intra-subnet procedures (as detailed in [RFC7432] are 3479 used to deliver the frame to PE2, which then sends it to H4. 3481 If H1 needs to send an IP packet to H5, it determines from its IP 3482 address and subnet mask that H5 is NOT on the same subnet as H1. 3483 Assuming that H1 has been configured with the IP address of PE1 as 3484 its default router, H1 sends the packet in an ethernet frame with 3485 PE1's MAC address in its Destination MAC Address field. PE1 receives 3486 the frame, and sees that the frame is addressed to it. PE1 thus 3487 sends the frame up its IRB1 interface to the L3 routing instance. 3488 Appropriate IP processing is done (e.g., TTL decrement). The L3 3489 routing instance determines that the "next hop" for H5 is PE2, so the 3490 packet is encapsulated (e.g., in MPLS) and sent across the backbone 3491 to PE2's routing instance. PE2 will see that the packet's 3492 destination, H5, is on BD2 segment-2, and will send the packet down 3493 its IRB2 interface. This causes the IP packet to be encapsulated in 3494 an ethernet frame with PE2's MAC address (on BD2) in the Source 3495 Address field and H5's MAC address in the Destination Address field. 3497 Note that if H1 has an IP packet to send to H3, the forwarding of the 3498 packet is handled entirely within PE1. PE1's routing instance sees 3499 the packet arrive on its IRB1 interface, and then transmits the 3500 packet by sending it down its IRB2 interface. 3502 Often, all the hosts in a particular Tenant Domain will be 3503 provisioned with the same value of the default router IP address. 3504 This IP address can be assigned, as an "anycast address", to all the 3505 EVPN PEs attached to that Tenant Domain. Thus although all hosts are 3506 provisioned with the same "default router address", the actual 3507 default router for a given host will be one of the PEs that is 3508 attached to the same ethernet segment as the host. This provisioning 3509 method ensures that IP packets from a given host are handled by the 3510 closest EVPN PE that supports IRB. 3512 In the topology of Figure 3, one could imagine that H1 is configured 3513 with a default router address that belongs to PE2 but not to PE1. 3514 Inter-subnet routing would still work, but IP packets from H1 to H3 3515 would then follow the non-optimal path H1-->PE1-->PE2-->PE1-->H3. 3516 Sending traffic on this sort of path, where it leaves a router and 3517 then comes back to the same router, is sometimes known as 3518 "hairpinning". Similarly, if PE2 supports IRB but PE1 dos not, the 3519 same non-optimal path from H1 to H3 would have to be followed. To 3520 avoid hairpinning, each EVPN PE needs to support IRB. 3522 It is worth pointing out the way IRB interfaces interact with 3523 multicast traffic. Referring again to Figure 3, suppose PE1 and PE2 3524 are functioning as IP multicast routers. Suppose also that H3 3525 transmits a multicast packet, and both H1 and H4 are interested in 3526 receiving that packet. PE1 will receive the packet from H3 via its 3527 IRB2 interface. The ethernet encapsulation from BD2 is removed, the 3528 IP header processing is done, and the packet is then reencapsulated 3529 for BD1, with PE1's MAC address in the MAC Source Address field. 3530 Then the packet is sent down the IRB1 interface. Layer 2 procedures 3531 (as defined in [RFC7432] would then be used to deliver a copy of the 3532 packet locally to H1, and remotely to H4. 3534 Please be aware that his document modifies the semantics, described 3535 in the previous paragraph, of sending/receiving multicast traffic on 3536 an IRB interface. This is explained in Section 1.5.1 and subsequent 3537 sections. 3539 Authors' Addresses 3541 Wen Lin 3542 Juniper Networks, Inc. 3543 10 Technology Park Drive 3544 Westford, Massachusetts 01886 3545 United States 3547 EMail: wlin@juniper.net 3549 Zhaohui Zhang 3550 Juniper Networks, Inc. 3551 10 Technology Park Drive 3552 Westford, Massachusetts 01886 3553 United States 3555 EMail: zzhang@juniper.net 3557 John Drake 3558 Juniper Networks, Inc. 3559 1194 N. Mathilda Ave 3560 Sunnyvale, CA 94089 3561 United States 3563 EMail: jdrake@juniper.net 3565 Eric C. Rosen (editor) 3566 Juniper Networks, Inc. 3567 10 Technology Park Drive 3568 Westford, Massachusetts 01886 3569 United States 3571 EMail: erosen52@gmail.com 3572 Jorge Rabadan 3573 Nokia 3574 777 E. Middlefield Road 3575 Mountain View, CA 94043 3576 United States 3578 EMail: jorge.rabadan@nokia.com 3580 Ali Sajassi 3581 Cisco Systems 3582 170 West Tasman Drive 3583 San Jose, CA 95134 3584 United States 3586 EMail: sajassi@cisco.com