idnits 2.17.00 (12 Aug 2021) /tmp/idnits13168/draft-ietf-l2vpn-evpn-01.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack a Security Considerations section. ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) ** The document seems to lack separate sections for Informative/Normative References. All references will be assumed normative when checking for downward references. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (July 14, 2012) is 3598 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) == Missing Reference: 'RFC2119' is mentioned on line 151, but not defined == Missing Reference: 'BGP-VPLS-MH' is mentioned on line 274, but not defined == Missing Reference: 'EVPN-REQ' is mentioned on line 337, but not defined == Missing Reference: 'RFC4271' is mentioned on line 416, but not defined == Missing Reference: 'RFC4760' is mentioned on line 425, but not defined == Missing Reference: 'MPLS-ENCAPS' is mentioned on line 1097, but not defined == Missing Reference: 'MLDP' is mentioned on line 1669, but not defined == Unused Reference: 'RFC4761' is defined on line 1872, but no explicit reference was found in the text == Unused Reference: 'RFC4762' is defined on line 1876, but no explicit reference was found in the text == Unused Reference: 'VPLS-MULTIHOMING' is defined on line 1880, but no explicit reference was found in the text == Unused Reference: 'PIM-SNOOPING' is defined on line 1884, but no explicit reference was found in the text == Unused Reference: 'IGMP-SNOOPING' is defined on line 1887, but no explicit reference was found in the text == Unused Reference: 'EVPN-SEGMENT-ROUTE' is defined on line 1896, but no explicit reference was found in the text == Outdated reference: A later version (-01) exists of draft-sajassi-raggarwa-l2vpn-evpn-req-00 -- Possible downref: Normative reference to a draft: ref. 'E-VPN-REQ' == Outdated reference: draft-ietf-l2vpn-vpls-mcast has been published as RFC 7117 == Outdated reference: A later version (-07) exists of draft-ietf-l2vpn-vpls-multihoming-00 == Outdated reference: A later version (-07) exists of draft-ietf-l2vpn-vpls-pim-snooping-01 ** Downref: Normative reference to an Informational draft: draft-ietf-l2vpn-vpls-pim-snooping (ref. 'PIM-SNOOPING') ** Downref: Normative reference to an Informational RFC: RFC 4541 (ref. 'IGMP-SNOOPING') == Outdated reference: A later version (-01) exists of draft-sajassi-l2vpn-evpn-segment-route-00 Summary: 5 errors (**), 0 flaws (~~), 19 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group A. Sajassi 3 INTERNET-DRAFT Cisco 4 Category: Standards Track 5 R. Aggarwal 6 N. Bitar Arktan 7 Verizon 8 W. Henderickx 9 S. Boutros F. Balus 10 K. Patel Alcatel-Lucent 11 S. Salam 12 Cisco Aldrin Isaac 13 Bloomberg 14 J. Drake 15 R. Shekhar J. Uttaro 16 Juniper Networks AT&T 18 Expires: January 14, 2012 July 14, 2012 20 BGP MPLS Based Ethernet VPN 21 draft-ietf-l2vpn-evpn-01 23 Status of this Memo 25 This Internet-Draft is submitted to IETF in full conformance with the 26 provisions of BCP 78 and BCP 79. 28 Internet-Drafts are working documents of the Internet Engineering 29 Task Force (IETF), its areas, and its working groups. Note that 30 other groups may also distribute working documents as 31 Internet-Drafts. 33 Internet-Drafts are draft documents valid for a maximum of six months 34 and may be updated, replaced, or obsoleted by other documents at any 35 time. It is inappropriate to use Internet-Drafts as reference 36 material or to cite them other than as "work in progress." 38 The list of current Internet-Drafts can be accessed at 39 http://www.ietf.org/1id-abstracts.html 41 The list of Internet-Draft Shadow Directories can be accessed at 42 http://www.ietf.org/shadow.html 44 Copyright and License Notice 46 Copyright (c) 2012 IETF Trust and the persons identified as the 47 document authors. All rights reserved. 49 This document is subject to BCP 78 and the IETF Trust's Legal 50 Provisions Relating to IETF Documents 51 (http://trustee.ietf.org/license-info) in effect on the date of 52 publication of this document. Please review these documents 53 carefully, as they describe your rights and restrictions with respect 54 to this document. Code Components extracted from this document must 55 include Simplified BSD License text as described in Section 4.e of 56 the Trust Legal Provisions and are provided without warranty as 57 described in the Simplified BSD License. 59 Abstract 61 This document describes procedures for BGP MPLS based Ethernet VPNs 62 (E-VPN). 64 Table of Contents 66 1. Specification of requirements . . . . . . . . . . . . . . . . . 5 67 2. Contributors . . . . . . . . . . . . . . . . . . . . . . . . . 5 68 3. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 5 69 4. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . 5 70 5. BGP MPLS Based E-VPN Overview . . . . . . . . . . . . . . . . . 6 71 6. Ethernet Segment . . . . . . . . . . . . . . . . . . . . . . . 7 72 7. Ethernet Tag . . . . . . . . . . . . . . . . . . . . . . . . . 8 73 7.1 VLAN Based Service Interface . . . . . . . . . . . . . . . . 9 74 7.2 VLAN Bundle Service Interface . . . . . . . . . . . . . . . 9 75 7.2.1 Port Based Service Interface . . . . . . . . . . . . . . 9 76 7.3 VLAN Aware Bundle Service Interface . . . . . . . . . . . . 9 77 8. BGP E-VPN NLRI . . . . . . . . . . . . . . . . . . . . . . . . 10 78 8.1. Ethernet Auto-Discovery Route . . . . . . . . . . . . . . . 10 79 8.2. MAC Advertisement Route . . . . . . . . . . . . . . . . . 11 80 8.3. Inclusive Multicast Ethernet Tag Route . . . . . . . . . . 11 81 8.4 Ethernet Segment Route . . . . . . . . . . . . . . . . . . . 12 82 8.5 ESI MPLS Label Extended Community . . . . . . . . . . . . . 12 83 8.6 ES-Import Extended Community . . . . . . . . . . . . . . . . 13 84 8.7 MAC Mobility Extended Community . . . . . . . . . . . . . . 13 85 9. Multi-homing Functions . . . . . . . . . . . . . . . . . . . . 13 86 9.1 Multi-homed Ethernet Segment Auto-Discovery . . . . . . . . 13 87 9.1.1 Constructing the Ethernet Segment Route . . . . . . . . 14 88 9.2 Fast Convergence . . . . . . . . . . . . . . . . . . . . . . 14 89 9.2.1 Constructing the Ethernet A-D Route per Ethernet 90 Segment . . . . . . . . . . . . . . . . . . . . . . . . 15 91 9.2.1.1. Ethernet A-D Route Targets . . . . . . . . . . . . 15 92 9.3 Split Horizon . . . . . . . . . . . . . . . . . . . . . . . 16 93 9.3.1 ESI MPLS Label Assignment . . . . . . . . . . . . . . . 16 94 9.3.1.1 Ingress Replication . . . . . . . . . . . . . . . . 16 95 9.3.1.2. P2MP MPLS LSPs . . . . . . . . . . . . . . . . . . 17 96 9.3.1.3. MP2MP LSPs . . . . . . . . . . . . . . . . . . . . 18 98 9.4 Aliasing . . . . . . . . . . . . . . . . . . . . . . . . . . 18 99 9.4.1 Constructing the Ethernet A-D Route per EVI . . . . . . 18 100 9.4.1.1 Ethernet A-D Route Targets . . . . . . . . . . . . . 19 101 9.5 Designated Forwarder Election . . . . . . . . . . . . . . . 20 102 9.5.1 Default DF Election Procedure . . . . . . . . . . . . . 21 103 9.5.2 DF Election with Service Carving . . . . . . . . . . . . 21 104 10. Determining Reachability to Unicast MAC Addresses . . . . . . 22 105 10.1. Local Learning . . . . . . . . . . . . . . . . . . . . . . 23 106 10.2. Remote learning . . . . . . . . . . . . . . . . . . . . . 23 107 10.2.1. Constructing the BGP E-VPN MAC Address Advertisement . 23 108 11. ARP and ND . . . . . . . . . . . . . . . . . . . . . . . . . . 25 109 12. Handling of Multi-Destination Traffic . . . . . . . . . . . . 26 110 12.1. Construction of the Inclusive Multicast Ethernet Tag 111 Route . . . . . . . . . . . . . . . . . . . . . . . . . . 26 112 12.2. P-Tunnel Identification . . . . . . . . . . . . . . . . . 27 113 13. Processing of Unknown Unicast Packets . . . . . . . . . . . . 28 114 13.1. Ingress Replication . . . . . . . . . . . . . . . . . . . 28 115 13.2. P2MP MPLS LSPs . . . . . . . . . . . . . . . . . . . . . . 29 116 14. Forwarding Unicast Packets . . . . . . . . . . . . . . . . . . 29 117 14.1. Forwarding packets received from a CE . . . . . . . . . . 29 118 14.2. Forwarding packets received from a remote PE . . . . . . . 30 119 14.2.1. Unknown Unicast Forwarding . . . . . . . . . . . . . . 30 120 14.2.2. Known Unicast Forwarding . . . . . . . . . . . . . . . 31 121 15. Load Balancing of Unicast Frames . . . . . . . . . . . . . . . 31 122 15.1. Load balancing of traffic from an PE to remote CEs . . . . 31 123 15.1.1 Active-Standby Redundancy Mode . . . . . . . . . . . . 31 124 15.1.2 All-Active Redundancy Mode . . . . . . . . . . . . . . 32 125 15.2. Load balancing of traffic between an PE and a local CE . . 33 126 15.2.1. Data plane learning . . . . . . . . . . . . . . . . . 34 127 15.2.2. Control plane learning . . . . . . . . . . . . . . . . 34 128 16. MAC Mobility . . . . . . . . . . . . . . . . . . . . . . . . . 34 129 17. Multicast . . . . . . . . . . . . . . . . . . . . . . . . . . 36 130 17.1. Ingress Replication . . . . . . . . . . . . . . . . . . . 36 131 17.2. P2MP LSPs . . . . . . . . . . . . . . . . . . . . . . . . 36 132 17.3. MP2MP LSPs . . . . . . . . . . . . . . . . . . . . . . . . 36 133 17.3.1. Inclusive Trees . . . . . . . . . . . . . . . . . . . 36 134 17.3.2. Selective Trees . . . . . . . . . . . . . . . . . . . 37 135 17.4. Explicit Tracking . . . . . . . . . . . . . . . . . . . . 38 136 18. Convergence . . . . . . . . . . . . . . . . . . . . . . . . . 38 137 18.1. Transit Link and Node Failures between PEs . . . . . . . . 38 138 18.2. PE Failures . . . . . . . . . . . . . . . . . . . . . . . 38 139 18.2.1. Local Repair . . . . . . . . . . . . . . . . . . . . . 38 140 18.3. PE to CE Network Failures . . . . . . . . . . . . . . . . 39 141 19. LACP State Synchronization . . . . . . . . . . . . . . . . . . 39 142 20. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 40 143 21. References . . . . . . . . . . . . . . . . . . . . . . . . . . 40 144 21. Author's Address . . . . . . . . . . . . . . . . . . . . . . . 41 146 1. Specification of requirements 148 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 149 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 150 document are to be interpreted as described in [RFC2119]. 152 2. Contributors 154 In addition to the authors listed above, the following individuals 155 also contributed to this document: 157 Quaizar Vohra 158 Kireeti Kompella 159 Apurva Mehta 160 Nadeem Mohammad 161 Juniper Networks 163 Clarence Filsfils 164 Dennis Cai 165 Cisco 167 3. Introduction 169 This document describes procedures for BGP MPLS based Ethernet VPNs 170 (E-VPN). The procedures described here are intended to meet the 171 requirements specified in [E-VPN-REQ]. Please refer to [E-VPN-REQ] 172 for the detailed requirements and motivation. E-VPN requires 173 extensions to existing IP/MPLS protocols as described in this 174 document. In addition to these extensions E-VPN uses several building 175 blocks from existing MPLS technologies. 177 4. Terminology 179 CE: Customer Edge device e.g., host or router or switch 181 E-VPN Instance (EVI): An E-VPN routing and forwarding instance on a 182 PE. 184 Ethernet segment identifier (ESI): If a CE is multi-homed to two or 185 more PEs, the set of Ethernet links that attaches the CE to the PEs 186 is an 'Ethernet segment'. Ethernet segments MUST have a unique non- 187 zero identifier, the 'Ethernet Segment Identifier'. 189 Ethernet Tag: An Ethernet Tag identifies a particular broadcast 190 domain, e.g., a VLAN. An E-VPN instance consists of one or more 191 broadcast domains. Ethernet tag(s) are assigned to the broadcast 192 domains of a given E-VPN instance by the provider of that E-VPN, and 193 each PE in that E-VPN instance performs a mapping between broadcast 194 domain identifier(s) understood by each of its attached CEs and the 195 corresponding Ethernet tag. 197 Link Aggregation Control Protocol (LACP): 199 Multipoint to Multipoint (MP2MP): 201 Point to Multipoint (P2MP): 203 Point to Point (P2P): 205 5. BGP MPLS Based E-VPN Overview 207 This section provides an overview of E-VPN. 209 An E-VPN comprises CEs that are connected to PEs that form the edge 210 of the MPLS infrastructure. A CE may be a host, a router or a switch. 211 The PEs provide virtual Layer 2 bridged connectivity between the CEs. 212 There may be multiple E-VPNs in the provider's network. 214 The PEs may be connected by an MPLS LSP infrastructure which provides 215 the benefits of MPLS technology such as fast-reroute, resiliency, 216 etc. The PEs may also be connected by an IP infrastructure in which 217 case IP/GRE tunneling or other IP tunneling can be used between the 218 PEs. The detailed procedures in this version of this document are 219 specified only for MPLS LSPs as the tunneling technology. However 220 these procedures are designed to be extensible to IP tunneling as the 221 PSN tunneling technology. 223 In an E-VPN, MAC learning between PEs occurs not in the data plane 224 (as happens with traditional bridging) but in the control plane. 225 Control plane learning offers greater control over the MAC learning 226 process, such as restricting who learns what, and the ability to 227 apply policies. Furthermore, the control plane chosen for 228 advertising MAC reachability information is multi-protocol (MP) BGP 229 (similar to IP VPNs (RFC 4364)). This provides greater scalability 230 and the ability to preserve the "virtualization" or isolation of 231 groups of interacting agents (hosts, servers, virtual machines) from 232 each other. In E-VPN, PEs advertise the MAC addresses learned from 233 the CEs that are connected to them, along with an MPLS label, to 234 other PEs in the control plane using MP-BGP. Control plane learning 235 enables load balancing of traffic to and from CEs that are multi- 236 homed to multiple PEs. This is in addition to load balancing across 237 the MPLS core via multiple LSPs between the same pair of PEs. In 238 other words it allows CEs to connect to multiple active points of 239 attachment. It also improves convergence times in the event of 240 certain network failures. 242 However, learning between PEs and CEs is done by the method best 243 suited to the CE: data plane learning, IEEE 802.1x, LLDP, 802.1aq, 244 ARP, management plane or other protocols. 246 It is a local decision as to whether the Layer 2 forwarding table on 247 an PE is populated with all the MAC destination addresses known to 248 the control plane, or whether the PE implements a cache based scheme. 249 For instance the MAC forwarding table may be populated only with the 250 MAC destinations of the active flows transiting a specific PE. 252 The policy attributes of E-VPN are very similar to those of IP-VPN. 253 An EVI requires a Route-Distinguisher (RD) and one or more Route- 254 Targets (RTs). A CE attaches to an E-VPN instance (EVI) on an PE, on 255 an Ethernet interface which may be configured for one or more 256 Ethernet Tags, e.g., VLANs. Some deployment scenarios guarantee 257 uniqueness of VLANs across E-VPNs: all points of attachment of a 258 given EVI use the same VLAN, and no other EVI uses this VLAN. This 259 document refers to this case as a "Unique VLAN E-VPN" and describes 260 simplified procedures to optimize for it. 262 6. Ethernet Segment 264 If a CE is multi-homed to two or more PEs, the set of Ethernet links 265 constitutes an "Ethernet Segment". An Ethernet segment may appear to 266 the CE as a Link Aggregation Group (LAG). Ethernet segments have an 267 identifier, called the "Ethernet Segment Identifier" (ESI) which is 268 encoded as a ten octets integer. A single-homed CE is considered to 269 be attached to an Ethernet segment with ESI 0. Otherwise, an Ethernet 270 segment MUST have a unique non-zero ESI. The ESI can be assigned 271 using various mechanisms: 272 1. The ESI may be configured. For instance when E-VPNs are used to 273 provide a VPLS service the ESI is fairly analogous to the Multi- 274 homing site ID in [BGP-VPLS-MH]. 276 2. If IEEE 802.1AX LACP is used between the PEs and CEs, then 277 the ESI is determined from LACP by concatenating the following 278 parameters: 280 + CE LACP System Identifier comprised of two octets of System 281 Priority and six octets of System MAC address, where the 282 System Priority is encoded in the most significant two octets. 283 The CE LACP identifier MUST be encoded in the high order eight 284 octets of the ESI. 286 + CE LACP two octets Port Key. The CE LACP port key MUST be 287 encoded in the low order two octets of the ESI. 289 As far as the CE is concerned, it would treat the multiple PEs 290 that it is connected to as the same switch. This allows the CE 291 to aggregate links that are attached to different PEs in the 292 same bundle. 294 3. If LLDP is used between the PEs and CEs that are hosts, then 295 the ESI is determined by LLDP. The ESI will be specified in a 296 following version. 298 4. In the case of indirectly connected hosts via a bridged LAN 299 between the CEs and the PEs, the ESI is determined based on the 300 Layer 2 bridge protocol as follows: If MST is used in the bridged 301 LAN then the value of the ESI is derived by listening to BPDUs on 302 the Ethernet segment. To achieve this the PE is not required to 303 run MST. However the PE must learn the Root Bridge MAC address 304 and Bridge Priority of the root of the Internal Spanning Tree 305 (IST) by listening to the BPDUs. The ESI is constructed as 306 follows: 308 {Bridge Priority (16 bits) , Root Bridge MAC Address (48 bits)} 310 7. Ethernet Tag 312 An Ethernet Tag identifies a particular broadcast domain, e.g. a 313 VLAN, in an EVI. An EVI consists of one or more broadcast domains. 314 Ethernet Tags are assigned to the broadcast domains of a given EVI by 315 the provider of the E-VPN service. Each PE, in a given EVI, performs 316 a mapping between the Ethernet Tag and the corresponding broadcast 317 domain identifier(s) understood by each of its attached CEs (e.g. CE 318 VLAN Identifiers or CE-VIDs). 320 If the broadcast domain identifiers(s) are understood consistently by 321 all of the CEs in an EVI, the broadcast domain identifier(s) MAY be 322 used as the corresponding Ethernet Tag(s). In other words, the 323 Ethernet Tag ID assigned by the provider is numerically equal to the 324 broadcast domain identifier (e.g., CE-VID = Ethernet Tag). 326 Further, some deployment scenarios guarantee uniqueness of broadcast 327 domain identifiers across all EVIs; all points of attachment of a 328 given EVI use the same broadcast domain identifier(s) and no other 329 EVI uses these broadcast domain identifier(s). This allows the RT(s) 330 for each EVI to be derived automatically, as described in section 331 9.4.1.1.1 "Auto-Derivation from the Ethernet Tag ID". 333 The following subsections discuss the relationship between Ethernet 334 Tags, EVIs and broadcast domain identifiers as well as the setting of 335 the Ethernet Tag Identifier, in the various E-VPN BGP routes (defined 336 in section 8), for the different types of service interfaces 337 described in [EVPN-REQ]. 339 7.1 VLAN Based Service Interface 341 With this service interface, there is a one-to-one mapping between 342 the broadcast domain identifier understood by a CE on a port (e.g. 343 CE-VID) and an EVI. Furthermore, there is a single bridge domain per 344 PE for the EVI. Different CEs connected to different PE ports MAY use 345 different broadcast domain identifiers (e.g. CE-VIDs) for the same 346 EVI. If said identifiers are different, the frames SHOULD remain 347 tagged with the originating CE's broadcast domain identifier (e.g. 348 CE-VID). When the CE broadcast domain identifiers are not consistent, 349 a tag translation function MUST be supported in the data path and 350 MUST be performed on the disposition PE. The Ethernet Tag Identifier 351 in all E-VPN routes MUST be set to 0. 353 7.2 VLAN Bundle Service Interface 355 With this service interface, there is a many-to-one mapping between 356 the broadcast domain identifier understood by a CE on a port (e.g. 357 CE-VID) and an EVI. Furthermore, there is a single bridge domain per 358 PE for the EVI. Different CEs connected to different PE ports MUST 359 use the same broadcast domain identifiers (e.g. CE-VIDs) for the same 360 EVI. The MPLS encapsulated frames MUST remain tagged with the 361 originating CE's broadcast domain identifier (e.g. CE-VID). Tag 362 translation is NOT permitted. The Ethernet Tag Identifier in all E- 363 VPN routes MUST be set to 0. 365 7.2.1 Port Based Service Interface 367 This service interface is a special case of the VLAN Bundle service 368 interface, where all of the VLANs on the port are part of the same 369 service and map to the same bundle. The procedures are identical to 370 those described in section 7.2. 372 7.3 VLAN Aware Bundle Service Interface 374 With this service interface, there is a many-to-one mapping between 375 the broadcast domain identifier understood by a CE on a port (e.g. 376 CE-VID) and an EVI. Furthermore, there are multiple bridge domains 377 per PE for the EVI: one broadcast domain per CE broadcast domain 378 identifier. In the case where the CE broadcast domain identifiers are 379 not consistent for different CEs, a normalized Ethernet Tag MUST be 380 carried in the MPLS encapsulated frames and a tag translation 381 function MUST be supported in the data path. This translation MUST be 382 performed on both the imposition as well as the disposition PEs. The 383 Ethernet Tag Identifier in all E-VPN routes MUST be set to the 384 normalized Ethernet Tag assigned by the E-VPN provider. 386 8. BGP E-VPN NLRI 388 This document defines a new BGP NLRI, called the E-VPN NLRI. 390 Following is the format of the E-VPN NLRI: 392 +-----------------------------------+ 393 | Route Type (1 octet) | 394 +-----------------------------------+ 395 | Length (1 octet) | 396 +-----------------------------------+ 397 | Route Type specific (variable) | 398 +-----------------------------------+ 400 The Route Type field defines encoding of the rest of the E-VPN NLRI 401 (Route Type specific E-VPN NLRI). 403 The Length field indicates the length in octets of the Route Type 404 specific field of E-VPN NLRI. 406 This document defines the following Route Types: 408 + 1 - Ethernet Auto-Discovery (A-D) route 409 + 2 - MAC advertisement route 410 + 3 - Inclusive Multicast Route 411 + 4 - Ethernet Segment Route 413 The detailed encoding and procedures for these route types are 414 described in subsequent sections. 416 The E-VPN NLRI is carried in BGP [RFC4271] using BGP Multiprotocol 417 Extensions [RFC4760] with an AFI of TBD and an SAFI of E-VPN (To be 418 assigned by IANA). The NLRI field in the 419 MP_REACH_NLRI/MP_UNREACH_NLRI attribute contains the E-VPN NLRI 420 (encoded as specified above). 422 In order for two BGP speakers to exchange labeled E-VPN NLRI, they 423 must use BGP Capabilities Advertisement to ensure that they both are 424 capable of properly processing such NLRI. This is done as specified 425 in [RFC4760], by using capability code 1 (multiprotocol BGP) with an 426 AFI of TBD and an SAFI of E-VPN. 428 8.1. Ethernet Auto-Discovery Route 430 A Ethernet A-D route type specific E-VPN NLRI consists of the 431 following: 433 +---------------------------------------+ 434 | RD (8 octets) | 435 +---------------------------------------+ 436 |Ethernet Segment Identifier (10 octets)| 437 +---------------------------------------+ 438 | Ethernet Tag ID (4 octets) | 439 +---------------------------------------+ 440 | MPLS Label (3 octets) | 441 +---------------------------------------+ 443 For procedures and usage of this route please see section 9.2 "Fast 444 Convergence" and section 9.4 "Aliasing". 446 8.2. MAC Advertisement Route 448 A MAC advertisement route type specific E-VPN NLRI consists of the 449 following: 451 +---------------------------------------+ 452 | RD (8 octets) | 453 +---------------------------------------+ 454 |Ethernet Segment Identifier (10 octets)| 455 +---------------------------------------+ 456 | Ethernet Tag ID (4 octets) | 457 +---------------------------------------+ 458 | MAC Address Length (1 octet) | 459 +---------------------------------------+ 460 | MAC Address (6 octets) | 461 +---------------------------------------+ 462 | IP Address Length (1 octet) | 463 +---------------------------------------+ 464 | IP Address (4 or 16 octets) | 465 +---------------------------------------+ 466 | MPLS Label (n * 3 octets) | 467 +---------------------------------------+ 469 For procedures and usage of this route please see section 10 470 "Determining Reachability to Unicast MAC Addresses" and section 15 471 "Load Balancing of Unicast Packets". 473 8.3. Inclusive Multicast Ethernet Tag Route 475 An Inclusive Multicast Ethernet Tag route type specific E-VPN NLRI 476 consists of the following: 478 +---------------------------------------+ 479 | RD (8 octets) | 480 +---------------------------------------+ 481 | Ethernet Tag ID (4 octets) | 482 +---------------------------------------+ 483 | IP Address Length (1 octet) | 484 +---------------------------------------+ 485 | Originating Router's IP Addr | 486 | (4 or 16 octets) | 487 +---------------------------------------+ 489 For procedures and usage of this route please see section 12 490 "Handling of Multi-Destination Traffic", section 13 "Processing of 491 Unknown Unicast Traffic" and section 17 "Multicast". 493 8.4 Ethernet Segment Route 495 The Ethernet Segment Route is encoded in the E-VPN NLRI using the 496 Route Type value of 4. The Route Type Specific field of the NLRI is 497 formatted as follows: 499 +---------------------------------------+ 500 | RD (8 octets) | 501 +---------------------------------------+ 502 |Ethernet Segment Identifier (10 octets)| 503 +---------------------------------------+ 505 For procedures and usage of this route please see section 9.5 506 "Designated Forwarder Election". 508 8.5 ESI MPLS Label Extended Community 510 This extended community is a new transitive extended community. It 511 may be advertised along with Ethernet Auto-Discovery routes and it 512 enables split-horizon procedures for multi-homed sites as described 513 in section 9.3 "Split Horizon". 515 Each ESI MPLS Label Extended Community is encoded as a 8-octet value 516 as follows: 518 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 519 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 520 | 0x44 | Sub-Type | Flags (One Octet) |Reserved=0 | 521 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 522 | Reserved = 0| ESI MPLS label | 523 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 525 The low order bit of the flags octet is defined as the "Active- 526 Standby" bit and may be set to 1. The other bits must be set to 0. 528 8.6 ES-Import Extended Community 530 This is a new transitive extended community carried with the Ethernet 531 Segment route. When used, it enables all the PEs connected to the 532 same multi-homed site to import the Ethernet Segment routes. The 533 value is derived automatically from the ESI by encoding the 6-byte 534 MAC address portion of the ESI in the ES-Import Extended Community. 535 The format of this extended community is as follows: 537 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 538 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 539 | 0x44 | Sub-Type | ES-Import | 540 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 541 | ES-Import Cont'd | 542 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 544 For procedures and usage of this attribute, please see section 9.1 545 "Redundancy Group Discovery". 547 8.7 MAC Mobility Extended Community 549 This extended community is a new transitive extended community. It 550 may be advertised along with MAC Advertisement routes. The procedures 551 for using this Extended Community are described in section 16 "MAC 552 Mobility". 554 The MAC Mobility Extended Community is encoded as a 8-octet value as 555 follows: 557 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 558 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 559 | 0x44 | Sub-Type | Reserved=0 | 560 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 561 | Sequence Number | 562 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 564 9. Multi-homing Functions 566 This section discusses the functions, procedures and associated BGP 567 routes used to support multi-homing in E-VPN. This covers both multi- 568 homed device (MHD) as well as multi-homed network (MHN) scenarios. 570 9.1 Multi-homed Ethernet Segment Auto-Discovery 572 PEs connected to the same Ethernet segment can automatically discover 573 each other with minimal to no configuration through the exchange of 574 the Ethernet Segment route. 576 9.1.1 Constructing the Ethernet Segment Route 578 The Route-Distinguisher (RD) MUST be a Type 1 RD [RFC4364]. The value 579 field comprises an IP address of the MES (typically, the loopback 580 address) followed by 0's. 582 The Ethernet Segment Identifier MUST be set to the ten octet ESI 583 identifier described in section 6. 585 The BGP advertisement that advertises the Ethernet Segment route MUST 586 also carry an ES-Import extended community attribute, as defined in 587 section 8.6. 589 The Ethernet Segment Route filtering MUST be done such that the 590 Ethernet Segment Route is imported only by the PEs that are multi- 591 homed to the same Ethernet Segment. To that end, each PE that is 592 connected to a particular Ethernet segment constructs an import 593 filtering rule to import a route that carries the ES-Import extended 594 community, constructed from the ESI. 596 Note that the new ES-Import extended community is not the same as the 597 Route Target Extended Community. The Ethernet Segment route carries 598 this new ES-Import extended community. The PEs apply filtering on 599 this new extended community. As a result the Ethernet Segment route 600 is imported only by the PEs that are connected to the same Ethernet 601 segment. 603 9.2 Fast Convergence 605 In E-VPN, MAC address reachability is learnt via the BGP control- 606 plane over the MPLS network. As such, in the absence of any fast 607 protection mechanism, the network convergence time is a function of 608 the number of MAC Advertisement routes that must be withdrawn by the 609 PE encountering a failure. For highly scaled environments, this 610 scheme yields slow convergence. 612 To alleviate this, E-VPN defines a mechanism to efficiently and 613 quickly signal, to remote PE nodes, the need to update their 614 forwarding tables upon the occurrence of a failure in connectivity to 615 an Ethernet segment. This is done by having each PE advertise an 616 Ethernet A-D Route per Ethernet segment for each locally attached 617 segment (refer to section 9.2.1 below for details on how this route 618 is constructed). Upon a failure in connectivity to the attached 619 segment, the PE withdraws the corresponding Ethernet A-D route. This 620 triggers all PEs that receive the withdrawal to update their next-hop 621 adjacencies for all MAC addresses associated with the Ethernet 622 segment in question. If no other PE had advertised an Ethernet A-D 623 route for the same segment, then the PE that received the withdrawal 624 simply invalidates the MAC entries for that segment. Otherwise, the 625 PE updates the next-hop adjacencies to point to the backup PE(s). 627 9.2.1 Constructing the Ethernet A-D Route per Ethernet Segment 629 This section describes procedures to construct the Ethernet A-D route 630 when a single such route is advertised by an PE for a given Ethernet 631 Segment. This flavor of the Ethernet A-D route is used for fast 632 convergence (as discussed above) as well as for advertising the ESI 633 MPLS label used for split-horizon filtering (as discussed in section 634 9.2). Support of this route flavor is MANDATORY. 636 Route-Distinguisher (RD) MUST be a Type 1 RD [RFC4364]. The value 637 field comprises an IP address of the PE (typically, the loopback 638 address) followed by 0. The reason for such encoding is that the RD 639 cannot be that of a given EVI since the ESI can span across one or 640 more EVIs. 642 The Ethernet Segment Identifier MUST be a ten octet entity as 643 described in section "Ethernet Segment". This document does not 644 specify the use of the Ethernet A-D route when the Segment Identifier 645 is set to 0. 647 The Ethernet Tag ID MUST be set to 0. 649 The MPLS label in the NLRI MUST be set to 0. 651 The "ESI MPLS Label Extended Community" MUST be included in the 652 route. If all-active multi-homing is desired, then the "Active- 653 Standby" bit in the flags of the ESI MPLS Label Extended Community 654 MUST be set to 0 and the MPLS label in that extended community MUST 655 be set to a valid MPLS label value. The MPLS label in this Extended 656 Community is referred to as an "ESI label". This label MUST be a 657 downstream assigned MPLS label if the advertising PE is using ingress 658 replication for receiving multicast, broadcast or unknown unicast 659 traffic from other PEs. If the advertising PE is using P2MP MPLS LSPs 660 for sending multicast, broadcast or unknown unicast traffic, then 661 this label MUST be an upstream assigned MPLS label. The usage of this 662 label is described in section 9.2. 664 If the Ethernet Segment is connected to more than one PE and active- 665 standby multi-homing is desired, then the "Active-Standby" bit in the 666 flags of the ESI MPLS Label Extended Community MUST be set to 1. 668 9.2.1.1. Ethernet A-D Route Targets 669 The Ethernet A-D route MUST carry one or more Route Target (RT) 670 attributes. These RTs MUST be the set of RTs associated with all the 671 EVIs to which the Ethernet Segment, corresponding to the Ethernet A-D 672 route, belongs. 674 9.3 Split Horizon 676 Consider a CE that is multi-homed to two or more PEs on an Ethernet 677 segment ES1. If the CE sends a multicast, broadcast or unknown 678 unicast packet to a particular PE, say PE1, then PE1 will forward 679 that packet to all or subset of the other PEs in the EVI. In this 680 case the PEs, other than PE1, that the CE is multi-homed to MUST drop 681 the packet and not forward back to the CE. This is referred to as 682 "split horizon" filtering in this document. 684 In order to achieve this split horizon function, every multicast, 685 broadcast or unknown unicast packet is encapsulated with an MPLS 686 label that identifies the Ethernet segment of origin (i.e. the 687 segment from which the frame entered the E-VPN network). This label 688 is referred to as the ESI MPLS label, and is distributed using the 689 "Ethernet A-D route per Ethernet Segment" as per the procedures in 690 section 9.1.1 above. This route is imported by the PEs connected to 691 the Ethernet Segment and also by the PEs that have at least one EVI 692 in common with the Ethernet Segment in the route. As described in 693 section 9.1.1, the route MUST carry an ESI MPLS Label Extended 694 Community with a valid ESI MPLS label. The disposition PEs rely on 695 the value of the ESI MPLS label to determine whether or not a flooded 696 frame is allowed to egress a specific Ethernet segment. 698 9.3.1 ESI MPLS Label Assignment 700 The following subsections describe the assignment procedures for the 701 ESI MPLS label, which differ depending on the type of tunnels being 702 used to deliver multi-destination packets in the E-VPN network. 704 9.3.1.1 Ingress Replication 706 An PE that is using ingress replication for sending broadcast, 707 multicast or unknown unicast traffic, distributes to other PEs, that 708 belong to the Ethernet segment, a downstream assigned "ESI MPLS 709 label" in the Ethernet A-D route. This label MUST be programmed in 710 the platform label space by the advertising PE. Further the 711 forwarding entry for this label must result in NOT forwarding packets 712 received with this label onto the Ethernet segment that the label was 713 distributed for. 715 Consider PE1 and PE2 that are multi-homed to CE1 on ES1. Further 716 consider that PE1 is using P2P or MP2P LSPs to send packets to PE2. 718 Consider that PE1 receives a multicast, broadcast or unknown unicast 719 packet from CE1 on VLAN1 on ESI1. In this scenario, PE2 distributes 720 an Inclusive Multicast Ethernet Tag route for VLAN1 in the associated 721 EVI. So, when PE1 sends a multicast, broadcast or unknown unicast 722 packet, that it receives from CE1, it MUST first push onto the MPLS 723 label stack the ESI label that PE2 has distributed for ESI1. It MUST 724 then push on the MPLS label distributed by PE2 in the Inclusive 725 Multicast Ethernet Tag route for VLAN1. The resulting packet is 726 further encapsulated in the P2P or MP2P LSP label stack required to 727 transmit the packet to PE2. When PE2 receives this packet it 728 determines the set of ESIs to replicate the packet to from the top 729 MPLS label, after any P2P or MP2P LSP labels have been removed. If 730 the next label is the ESI label assigned by PE2 for ESI1, then PE2 731 MUST NOT forward the packet onto ESI1. 733 9.3.1.2. P2MP MPLS LSPs 735 An PE that is using P2MP LSPs for sending broadcast, multicast or 736 unknown unicast traffic, distributes to other PEs, that belong to the 737 Ethernet segment or have an E-VPN in common with the Ethernet 738 Segment, an upstream assigned "ESI MPLS label" in the Ethernet A-D 739 route. This label is upstream assigned by the PE that advertises the 740 route. This label MUST be programmed by the other PEs, that are 741 connected to the ESI advertised in the route, in the context label 742 space for the advertising PE. Further the forwarding entry for this 743 label must result in NOT forwarding packets received with this label 744 onto the Ethernet segment that the label was distributed for. This 745 label MUST also be programmed by the other PEs, that import the route 746 but are not connected to the ESI advertised in the route, in the 747 context label space for the advertising PE. Further the forwarding 748 entry for this label must be a POP with no other associated action. 750 Consider PE1 and PE2 that are multi-homed to CE1 on ES1. Also 751 consider PE3 that is in the same EVI as one of the EVIs to which ES1 752 belongs. Further, assume that PE1 is using P2MP MPLS LSPs to send 753 broadcast, multicast or uknown unicast packets. When PE1 sends a 754 multicast, broadcast or unknown unicast packet, that it receives from 755 CE1, it MUST first push onto the MPLS label stack the ESI label that 756 it has assigned for the ESI that the packet was received on. The 757 resulting packet is further encapsulated in the P2MP MPLS label stack 758 necessary to transmit the packet to the other PEs. Penultimate hop 759 popping MUST be disabled on the P2MP LSPs used in the MPLS transport 760 infrastructure for E-VPN. When PE2 receives this packet, it de- 761 capsulates the top MPLS label and forwards the packet using the 762 context label space determined by the top label. If the next label is 763 the ESI label assigned by PE1 to ESI1, then PE2 MUST NOT forward the 764 packet onto ESI1. When PE3 receives this packet, it de-capsulates the 765 top MPLS label and forwards the packet using the context label space 766 determined by the top label. If the next label is the ESI label 767 assigned by PE1 to ESI1 and PE3 is not connected to ESI1, then PE3 768 MUST pop the label and flood the packet over all local ESIs in the 769 EVI. 771 9.3.1.3. MP2MP LSPs 773 The procedures for ESI MPLS Label assignment and usage for MP2MP LSPs 774 will be described in a future version. 776 9.4 Aliasing 778 In the case where a CE is multi-homed to multiple PE nodes, using a 779 LAG with all-active redundancy, it is possible that only a single PE 780 learns a set of the MAC addresses associated with traffic transmitted 781 by the CE. This leads to a situation where remote PE nodes receive 782 MAC advertisement routes, for these addresses, from a single PE even 783 though multiple PEs are connected to the multi-homed segment. As a 784 result, the remote PEs are not able to effectively load-balance 785 traffic among the PE nodes connected to the multi-homed Ethernet 786 segment. This could be the case, for e.g. when the PEs perform data- 787 path learning on the access, and the load-balancing function on the 788 CE hashes traffic from a given source MAC address to a single PE. 789 Another scenario where this occurs is when the PEs rely on control 790 plane learning on the access (e.g. using ARP), since ARP traffic will 791 be hashed to a single link in the LAG. 793 To alleviate this issue, E-VPN introduces the concept of 'Aliasing'. 794 Aliasing refers to the ability of an PE to signal that it has 795 reachability to a given locally attached Ethernet segment, even when 796 it has learnt no MAC addresses from that segment. The Ethernet A-D 797 route per EVI is used to that end. Remote PEs which receive MAC 798 advertisement routes with non-zero ESI SHOULD consider the advertised 799 MAC address as reachable via all PEs which have advertised 800 reachability to the relevant Segment using Ethernet A-D routes with 801 the same ESI (and Ethernet Tag if applicable). 803 9.4.1 Constructing the Ethernet A-D Route per EVI 805 This section describes procedures to construct the Ethernet A-D route 806 when one or more such routes are advertised by an PE for a given EVI. 807 This flavor of the Ethernet A-D route is used for aliasing, and 808 support of this route flavor is OPTIONAL. 810 Route-Distinguisher (RD) MUST be set to the RD of the EVI that is 811 advertising the NLRI. An RD MUST be assigned for a given EVI on an 812 PE. This RD MUST be unique across all EVIs on an PE. It is 813 RECOMMENDED to use the Type 1 RD [RFC4364]. The value field comprises 814 an IP address of the PE (typically, the loopback address) followed by 815 a number unique to the PE. This number may be generated by the PE. 816 Or in the Unique VLAN E-VPN case, the low order 12 bits may be the 12 817 bit VLAN ID, with the remaining high order 4 bits set to 0. 819 The Ethernet Segment Identifier MUST be a ten octet entity as 820 described in section "Ethernet Segment Identifier". This document 821 does not specify the use of the Ethernet A-D route when the Segment 822 Identifier is set to 0. 824 The Ethernet Tag ID is the identifier of an Ethernet Tag on the 825 Ethernet segment. This value may be a 12 bit VLAN ID, in which case 826 the low order 12 bits are set to the VLAN ID and the high order 20 827 bits are set to 0. Or it may be another Ethernet Tag used by the E- 828 VPN. It MAY be set to the default Ethernet Tag on the Ethernet 829 segment or to the value 0. 831 Note that the above allows the Ethernet A-D route to be advertised 832 with one of the following granularities: 834 + One Ethernet A-D route for a given tuple 835 per EVI. This is applicable when the PE uses MPLS-based 836 disposition. 838 + One Ethernet A-D route per (where the Ethernet 839 Tag ID is set to 0). This is applicable when the PE uses 840 MAC-based disposition, or when the PE uses MPLS-based 841 disposition when no VLAN translation is required. 843 The usage of the MPLS label is described in the section on "Load 844 Balancing of Unicast Packets". 846 The Next Hop field of the MP_REACH_NLRI attribute of the route MUST 847 be set to the IPv4 or IPv6 address of the advertising PE. 849 9.4.1.1 Ethernet A-D Route Targets 851 The Ethernet A-D route MUST carry one or more Route Target (RT) 852 attributes. RTs may be configured (as in IP VPNs), or may be derived 853 automatically. 855 If an PE uses Route Target Constrain [RT-CONSTRAIN], the PE SHOULD 856 advertise all such RTs using Route Target Constrains. The use of RT 857 Constrains allows each Ethernet A-D route to reach only those PEs 858 that are configured to import at least one RT from the set of RTs 859 carried in the Ethernet A-D route. 861 9.4.1.1.1 Auto-Derivation from the Ethernet Tag ID 862 The following is the procedure for deriving the RT attribute 863 automatically from the Ethernet Tag ID associated with the 864 advertisement: 866 + The Global Administrator field of the RT MUST 867 be set to the Autonomous System (AS) number that the PE 868 belongs to. 870 + The Local Administrator field of the RT contains a 4 871 octets long number that encodes the Ethernet Tag-ID. If the 872 Ethernet Tag-ID is a two octet VLAN ID then it MUST be 873 encoded in the lower two octets of the Local Administrator 874 field and the higher two octets MUST be set to zero. 876 For the "Unique VLAN E-VPN" this results in auto-deriving the RT from 877 the Ethernet Tag, e.g., VLAN ID for that E-VPN. 879 9.5 Designated Forwarder Election 881 Consider a CE that is a host or a router that is multi-homed directly 882 to more than one PE in an E-VPN on a given Ethernet segment. One or 883 more Ethernet Tags may be configured on the Ethernet segment. In this 884 scenario only one of the PEs, referred to as the Designated Forwarder 885 (DF), is responsible for certain actions: 887 - Sending multicast and broadcast traffic, on a given Ethernet 888 Tag on a particular Ethernet segment, to the CE. 890 - Flooding unknown unicast traffic (i.e. traffic for 891 which an PE does not know the destination MAC address), 892 on a given Ethernet Tag on a particular Ethernet segment 893 to the CE, if the environment requires flooding of 894 unknown unicast traffic. 895 Note that this behavior, which allows selecting a DF at the 896 granularity of for multicast, broadcast and unknown 897 unicast traffic, is the default behavior in this specification. 898 Optional mechanisms, which will be specified in the future, will 899 allow selecting a DF at the granularity of . 901 Note that a CE always sends packets belonging to a specific flow 902 using a single link towards an PE. For instance, if the CE is a host 903 then, as mentioned earlier, the host treats the multiple links that 904 it uses to reach the PEs as a Link Aggregation Group (LAG). The CE 905 employs a local hashing function to map traffic flows onto links in 906 the LAG. 908 If a bridged network is multi-homed to more than one PE in an E-VPN 909 via switches, then the support of all-active points of attachments, 910 as described in this specification, requires the bridge network to be 911 connected to two or more PEs using a LAG. In this case the reasons 912 for doing DF election are the same as those described above when a CE 913 is a host or a router. 915 If a bridged network does not connect to the PEs using LAG, then only 916 one of the links between the switched bridged network and the PEs 917 must be the active link for a given Ethernet Tag. In this case, the 918 Ethernet A-D route per Ethernet segment MUST be advertised with the 919 "Active-Standby" flag set to one. Procedures for supporting all- 920 active points of attachments, when a bridge network connects to the 921 PEs using LAG, are for further study. 923 The granularity of the DF election MUST be at least the Ethernet 924 segment via which the CE is multi-homed to the PEs. If the DF 925 election is done at the Ethernet segment granularity then a single PE 926 MUST be elected as the DF on the Ethernet segment. 928 If there are one or more EVIs enabled on the Ethernet segment, then 929 the granularity of the DF election SHOULD be the combination of the 930 Ethernet segment and EVI on that Ethernet segment. In this case a 931 single PE MUST be elected as the DF for a particular EVI on that 932 Ethernet segment. 934 The detailed procedures for DF election are described next. 936 9.5.1 Default DF Election Procedure 938 As a PE discovers the other PEs that are connected to the same 939 Ethernet Segment, using the Ethernet Segment routes, it starts 940 building an ordered list based on the originating PE IP addresses. 941 This list is used to select a DF and a backup DF (BDF) on a per 942 Ethernet Segment basis. By default, the PE with the numerically 943 highest IP address is considered the DF for that Ethernet Segment and 944 the next PE in the list is considered the BDF. 946 If the Ethernet Segment is a multi-homed device, then the elected DF 947 is the only PE that must forward flooded multi-destination packets 948 towards the segment. All other PE nodes must not permit multi- 949 destination packets to egress to the segment. In the case where the 950 DF fails, the BDF takes over its functionality. 952 This procedure enables the election of a single DF per Ethernet 953 Segment, for all EVIs enabled on the segment. It is possible to 954 achieve more granular load-balancing of traffic among the PE nodes by 955 employing Service Carving, as discussed in the next section. 957 9.5.2 DF Election with Service Carving 958 With service carving, it is possible to elect multiple DFs per 959 Ethernet Segment (one per EVI) in order to perform load-balancing of 960 multi-destination traffic destined to a given Segment. The load- 961 balancing procedures carve up the EVI space among the PE nodes 962 evenly, in such a way that every PE is the DF for a disjoint set of 963 EVIs. The procedure for service carving is as follows: 965 1. When a PE discovers the ESI of the attached Ethernet Segment, it 966 advertises an Ethernet Segment route with the associated ES-Import 967 extended community attribute. 969 2. The PE then starts a timer to allow the reception of Ethernet 970 Segment routes from other PE nodes connected to the same Ethernet 971 Segment. 973 3. When the timer expires, each PE builds an ordered list of the IP 974 addresses of all the PE nodes connected to the Ethernet Segment 975 (including itself), in increasing numeric value. Every PE is then 976 given an ordinal indicating its position in the ordered list, 977 starting with 0 as the ordinal for the PE with the numerically lowest 978 IP address. The ordinals are used to determine which PE node will be 979 the DF for a given EVI on the Ethernet Segment using the following 980 rule: Assuming a redundancy group of N PE nodes, the PE with ordinal 981 i is the DF for EVI V when (V mod N) = i. 983 The above procedure results in the entire EVI range being divided up 984 among the PEs in the RG, regardless of whether a given EVI is 985 configured/enabled on the associated Ethernet Segment or not. 987 4. The PE that is elected as a DF for a given EVI will unblock 988 traffic for that EVI only if the EVI is configured/enabled on the 989 Segment. Note that the DF PE unblocks multi-destination traffic in 990 the egress direction towards the Segment. All non-DF PEs continue to 991 drop multi-destination traffic (for the associated EVIs) in the 992 egress direction towards the Segment. 994 In the case of link or port failure, the affected PE withdraws its 995 Ethernet Segment route. This will re-trigger the service carving 996 procedures on all the PEs in the RG. For PE node failure, or upon PE 997 commissioning or decommissioning, the PEs re-trigger the service 998 carving. When a service moves from one PE in the RG to another PE as 999 a result of re-carving, the PE, which ends up being the elected DF 1000 for the service, must trigger a MAC address flush notification 1001 towards the associated Ethernet Segment. This can be done, for e.g. 1002 using IEEE 802.1ak MVRP 'new' declaration. 1004 10. Determining Reachability to Unicast MAC Addresses 1005 PEs forward packets that they receive based on the destination MAC 1006 address. This implies that PEs must be able to learn how to reach a 1007 given destination unicast MAC address. 1009 There are two components to MAC address learning, "local learning" 1010 and "remote learning": 1012 10.1. Local Learning 1014 A particular PE must be able to learn the MAC addresses from the CEs 1015 that are connected to it. This is referred to as local learning. 1017 The PEs in a particular E-VPN MUST support local data plane learning 1018 using standard IEEE Ethernet learning procedures. An PE must be 1019 capable of learning MAC addresses in the data plane when it receives 1020 packets such as the following from the CE network: 1022 - DHCP requests 1024 - ARP request for its own MAC. 1026 - ARP request for a peer. 1028 Alternatively PEs MAY learn the MAC addresses of the CEs in the 1029 control plane or via management plane integration between the PEs and 1030 the CEs. 1032 There are applications where a MAC address that is reachable via a 1033 given PE on a locally attached Segment (e.g. with ESI X) may move 1034 such that it becomes reachable via the same PE or another PE on 1035 another Segment (e.g. with ESI Y). This is referred to as a "MAC 1036 Mobility". Procedures to support this are described in section "MAC 1037 Mobility". 1039 10.2. Remote learning 1041 A particular PE must be able to determine how to send traffic to MAC 1042 addresses that belong to or are behind CEs connected to other PEs 1043 i.e. to remote CEs or hosts behind remote CEs. We call such MAC 1044 addresses as "remote" MAC addresses. 1046 This document requires an PE to learn remote MAC addresses in the 1047 control plane. In order to achieve this, each PE advertises the MAC 1048 addresses it learns from its locally attached CEs in the control 1049 plane, to all the other PEs in the EVI, using MP-BGP and specifically 1050 the MAC Advertisement route. 1052 10.2.1. Constructing the BGP E-VPN MAC Address Advertisement 1053 BGP is extended to advertise these MAC addresses using the MAC 1054 Advertisement route type in the E-VPN NLRI. 1056 The RD MUST be the RD of the EVI that is advertising the NLRI. The 1057 procedures for setting the RD for a given EVI are described in 1058 section 9.4.1. 1060 The Ethernet Segment Identifier is set to the ten octet ESI described 1061 in section "Ethernet Segment". 1063 The Ethernet Tag ID may be zero or may represent a valid Ethernet Tag 1064 ID. This field may be non-zero when there are multiple bridge 1065 domains in the EVI (e.g., the PE needs to perform qualified learning 1066 for the VLANs in that EVI). 1068 When the the Ethernet Tag ID in the NLRI is set to a non-zero value, 1069 for a particular bridge domain, then this Ethernet Tag may either be 1070 the Ethernet tag value associated with the CE, e.g., VLAN ID, or it 1071 may be the Ethernet Tag Identifier, e.g., VLAN ID assigned by the E- 1072 VPN provider and mapped to the CE's Ethernet tag. The latter would be 1073 the case if the CE Ethernet tags, e.g., VLAN ID, for a particular 1074 bridge domain are different on different CEs. 1076 The MAC address length field is typically set to 48. However this 1077 specification enables specifying the MAC address as a prefix; in 1078 which case, the MAC address length field is set to the length of the 1079 prefix. This provides the ability to aggregate MAC addresses if the 1080 deployment environment supports that. The encoding of a MAC address 1081 MUST be the 6-octet MAC address specified by [802.1D-ORIG] [802.1D- 1082 REV]. If the MAC address is advertised as a prefix then the trailing 1083 bits of the prefix MUST be set to 0 to ensure that the entire prefix 1084 is encoded as 6 octets. 1086 The IP Address Length field value is set to the number of octets in 1087 the IP Address field. 1089 The IP Address field is optional. By default, the IP Address Length 1090 field is set to 0 and the IP address field is omitted from the route. 1091 When a valid IP address is included, it is encoded as specified in 1092 section 12. 1094 The MPLS label field carries one or more labels (that corresponds to 1095 the stack of labels [MPLS-ENCAPS]). Each label is encoded as 3 1096 octets, where the high-order 20 bits contain the label value, and the 1097 low order bit contains "Bottom of Stack" (as defined in [MPLS- 1098 ENCAPS]). The MPLS label stack MUST be the downstream assigned E-VPN 1099 MPLS label stack that is used by the PE to forward MPLS-encapsulated 1100 Ethernet frames received from remote PEs, where the destination MAC 1101 address in the Ethernet frame is the MAC address advertised in the 1102 above NLRI. The forwarding procedures are specified in section 1103 "Forwarding Unicast Packets" and "Load Balancing of Unicast Packets". 1105 An PE may advertise the same single E-VPN label for all MAC addresses 1106 in a given EVI. This label assignment methodology is referred to as a 1107 per EVI label assignment. Alternatively, an PE may advertise a unique 1108 E-VPN label per combination. This label 1109 assignment methodology is referred to as a per 1110 label assignment. As a third option, an PE may advertise a unique E- 1111 VPN label per MAC address. All of these methodologies have their 1112 tradeoffs. 1114 Per EVI label assignment requires the least number of E-VPN labels, 1115 but requires a MAC lookup in addition to an MPLS lookup on an egress 1116 PE for forwarding. On the other hand, a unique label per or a unique label per MAC allows an egress PE to 1118 forward a packet that it receives from another PE, to the connected 1119 CE, after looking up only the MPLS labels without having to perform a 1120 MAC lookup. This includes the capability to perform appropriate VLAN 1121 ID translation on egress to the CE. 1123 The Next Hop field of the MP_REACH_NLRI attribute of the route MUST 1124 be set to the IPv4 or IPv6 address of the advertising PE. 1126 The BGP advertisement for the MAC advertisement route MUST also carry 1127 one or more Route Target (RT) attributes. RTs may be configured (as 1128 in IP VPNs), or may be derived automatically from the Ethernet Tag 1129 ID, in the Unique VLAN case, as described in section "Ethernet A-D 1130 Route per E-VPN". 1132 It is to be noted that this document does not require PEs to create 1133 forwarding state for remote MACs when they are learnt in the control 1134 plane. When this forwarding state is actually created is a local 1135 implementation matter. 1137 11. ARP and ND 1139 The IP address field in the MAC advertisement route may optionally 1140 carry one of the IP addresses associated with the MAC address. This 1141 provides an option which can be used to minimize the flooding of ARP 1142 or Neighbor Discovery (ND) messages over the MPLS network and to 1143 remote CEs. This option also minimizes ARP (or ND) message processing 1144 on end-stations/hosts connected to the E-VPN network. An PE may learn 1145 the IP address associated with a MAC address in the control or 1146 management plane between the CE and the PE. Or, it may learn this 1147 binding by snooping certain messages to or from a CE. When an PE 1148 learns the IP address associated with a MAC address, of a locally 1149 connected CE, it may advertise this address to other PEs by including 1150 it in the MAC Advertisement route. The IP Address may be an IPv4 1151 address encoded using four octets, or an IPv6 address encoded using 1152 sixteen octets. The IP Address length field MUST be set to 32 for an 1153 IPv4 address or to 128 for an IPv6 address. 1155 If there are multiple IP addresses associated with a MAC address, 1156 then multiple MAC advertisement routes MUST be generated, one for 1157 each IP address. For instance, this may be the case when there are 1158 both an IPv4 and an IPv6 address associated with the MAC address. 1159 When the IP address is dissociated with the MAC address, then the MAC 1160 advertisement route with that particular IP address MUST be 1161 withdrawn. 1163 When an PE receives an ARP request for an IP address from a CE, and 1164 if the PE has the MAC address binding for that IP address, the PE 1165 SHOULD perform ARP proxy and respond to the ARP request. 1167 Further detailed procedures will be specified in a later version. 1169 12. Handling of Multi-Destination Traffic 1171 Procedures are required for a given PE to send broadcast or multicast 1172 traffic, received from a CE encapsulated in a given Ethernet Tag in 1173 an EVI, to all the other PEs that span that Ethernet Tag in the EVI. 1174 In certain scenarios, described in section "Processing of Unknown 1175 Unicast Packets", a given PE may also need to flood unknown unicast 1176 traffic to other PEs. 1178 The PEs in a particular E-VPN may use ingress replication, P2MP LSPs 1179 or MP2MP LSPs to send unknown unicast, broadcast or multicast traffic 1180 to other PEs. 1182 Each PE MUST advertise an "Inclusive Multicast Ethernet Tag Route" to 1183 enable the above. The following subsection provides the procedures to 1184 construct the Inclusive Multicast Ethernet Tag route. Subsequent 1185 subsections describe in further detail its usage. 1187 12.1. Construction of the Inclusive Multicast Ethernet Tag Route 1189 The RD MUST be the RD of the EVI that is advertising the NLRI. The 1190 procedures for setting the RD for a given E-VPN are described in 1191 section 9.4.1. 1193 The Ethernet Tag ID is the identifier of the Ethernet Tag. It MAY be 1194 set to 0 or to a valid Ethernet Tag value. 1196 The Originating Router's IP address MUST be set to an IP address of 1197 the PE. This address SHOULD be common for all the EVIs on the PE 1198 (e.,g., this address may be PE's loopback address). 1200 The Next Hop field of the MP_REACH_NLRI attribute of the route MUST 1201 be set to the same IP address as the one carried in the Originating 1202 Router's IP Address field. 1204 The BGP advertisement for the Inclusive Multicast Ethernet Tag route 1205 MUST also carry one or more Route Target (RT) attributes. The 1206 assignment of RTs described in the section on "Constructing the BGP 1207 E-VPN MAC Address Advertisement" MUST be followed. 1209 12.2. P-Tunnel Identification 1211 In order to identify the P-Tunnel used for sending broadcast, unknown 1212 unicast or multicast traffic, the Inclusive Multicast Ethernet Tag 1213 route MUST carry a "PMSI Tunnel Attribute" as specified in [BGP 1214 MVPN]. 1216 Depending on the technology used for the P-tunnel for the E-VPN on 1217 the PE, the PMSI Tunnel attribute of the Inclusive Multicast Ethernet 1218 Tag route is constructed as follows. 1220 + If the PE that originates the advertisement uses a 1221 P-Multicast tree for the P-tunnel for E-VPN, the PMSI 1222 Tunnel attribute MUST contain the identity of the tree 1223 (note that the PE could create the identity of the 1224 tree prior to the actual instantiation of the tree). 1226 + An PE that uses a P-Multicast tree for the P-tunnel MAY 1227 aggregate two or more Ethernet Tags in the same or different 1228 EVIs present on the PE onto the same tree. In this case, in 1229 addition to carrying the identity of the tree, the PMSI Tunnel 1230 attribute MUST carry an MPLS upstream assigned label which 1231 the PE has bound uniquely to the Ethernet Tag for the EVI 1232 associated with this update (as determined by its RTs). 1234 If the PE has already advertised Inclusive Multicast 1235 Ethernet Tag routes for two or more Ethernet Tags that it 1236 now desires to aggregate, then the PE MUST re-advertise 1237 those routes. The re-advertised routes MUST be the same 1238 as the original ones, except for the PMSI Tunnel attribute 1239 and the label carried in that attribute. 1241 + If the PE that originates the advertisement uses ingress 1242 replication for the P-tunnel for E-VPN, the route MUST 1243 include the PMSI Tunnel attribute with the Tunnel Type set to 1244 Ingress Replication and Tunnel Identifier set to a routable 1245 address of the PE. The PMSI Tunnel attribute MUST carry a 1246 downstream assigned MPLS label. This label is used to 1247 demultiplex the broadcast, multicast or unknown unicast E-VPN 1248 traffic received over a MP2P tunnel by the PE. 1250 + The Leaf Information Required flag of the PMSI Tunnel 1251 attribute MUST be set to zero, and MUST be ignored on receipt. 1253 13. Processing of Unknown Unicast Packets 1255 The procedures in this document do not require the PEs to flood 1256 unknown unicast traffic to other PEs. If PEs learn CE MAC addresses 1257 via a control plane protocol, the PEs can then distribute MAC 1258 addresses via BGP, and all unicast MAC addresses will be learnt prior 1259 to traffic to those destinations. 1261 However, if a destination MAC address of a received packet is not 1262 known by the PE, the PE may have to flood the packet. Flooding must 1263 take into account "split horizon forwarding" as follows: The 1264 principles behind the following procedures are borrowed from the 1265 split horizon forwarding rules in VPLS solutions [RFC 4761, RFC 1266 4762]. When an PE capable of flooding (say PEx) receives a broadcast 1267 or multicast Ethernet frame, or one with an unknown destination MAC 1268 address, it must flood the frame. If the frame arrived from an 1269 attached CE, PEx must send a copy of the frame to every other 1270 attached CE participating in the EVI, on a different ESI than the one 1271 it received the frame on, as long as the PE is the DF for the egress 1272 ESI. In addition, the PE must flood the frame to all other PEs 1273 participating in the EVI. If, on the other hand, the frame arrived 1274 from another PE (say PEy), PEx must send a copy of the packet only to 1275 attached CEs as long as it is the DF for the egress ESI. PEx MUST NOT 1276 send the frame to other PEs, since PEy would have already done so. 1277 Split horizon forwarding rules apply to broadcast and multicast 1278 packets, as well as packets to an unknown MAC address. 1280 Whether or not to flood packets to unknown destination MAC addresses 1281 should be an administrative choice, depending on how learning happens 1282 between CEs and PEs. 1284 The PEs in a particular E-VPN may use ingress replication using RSVP- 1285 TE P2P LSPs or LDP MP2P LSPs for sending broadcast, multicast and 1286 unknown unicast traffic to other PEs. Or they may use RSVP-TE P2MP or 1287 LDP P2MP or LDP MP2MP LSPs for sending such traffic to other PEs. 1289 13.1. Ingress Replication 1291 If ingress replication is in use, the P-Tunnel attribute, carried in 1292 the Inclusive Multicast Ethernet Tag routes for the EVI, specifies 1293 the downstream label that the other PEs can use to send unknown 1294 unicast, multicast or broadcast traffic for the EVI to this 1295 particular PE. 1297 The PE that receives a packet with this particular MPLS label MUST 1298 treat the packet as a broadcast, multicast or unknown unicast packet. 1299 Further if the MAC address is a unicast MAC address, the PE MUST 1300 treat the packet as an unknown unicast packet. 1302 13.2. P2MP MPLS LSPs 1304 The procedures for using P2MP LSPs are very similar to VPLS 1305 procedures [VPLS-MCAST]. The P-Tunnel attribute used by an PE for 1306 sending unknown unicast, broadcast or multicast traffic for a 1307 particular EVI is advertised in the Inclusive Ethernet Tag Multicast 1308 route as described in section "Handling of Multi-Destination 1309 Traffic". 1311 The P-Tunnel attribute specifies the P2MP LSP identifier. This is the 1312 equivalent of an Inclusive tree in [VPLS-MCAST]. Note that multiple 1313 Ethernet Tags, which may be in different EVIs, may use the same P2MP 1314 LSP, using upstream labels [VPLS-MCAST]. This is the equivalent of an 1315 Aggregate Inclusive tree in [VPLS-MCAST]. When P2MP LSPs are used for 1316 flooding unknown unicast traffic, packet re-ordering is possible. 1318 The PE that receives a packet on the P2MP LSP specified in the PMSI 1319 Tunnel Attribute MUST treat the packet as a broadcast, multicast or 1320 unknown unicast packet. Further if the MAC address is a unicast MAC 1321 address, the PE MUST treat the packet as an unknown unicast packet. 1323 14. Forwarding Unicast Packets 1325 14.1. Forwarding packets received from a CE 1327 When an PE receives a packet from a CE, on a given Ethernet Tag, it 1328 must first look up the source MAC address of the packet. In certain 1329 environments the source MAC address MAY be used to authenticate the 1330 CE and determine that traffic from the host can be allowed into the 1331 network. Source MAC lookup MAY also be used for local MAC address 1332 learning. 1334 If the PE decides to forward the packet, the destination MAC address 1335 of the packet must be looked up. If the PE has received MAC address 1336 advertisements for this destination MAC address from one or more 1337 other PEs or learned it from locally connected CEs, it is considered 1338 as a known MAC address. Otherwise, the MAC address is considered as 1339 an unknown MAC address. 1341 For known MAC addresses the PE forwards this packet to one of the 1342 remote PEs or to a locally attached CE. When forwarding to a remote 1343 PE, the packet is encapsulated in the E-VPN MPLS label advertised by 1344 the remote PE, for that MAC address, and in the MPLS LSP label stack 1345 to reach the remote PE. 1347 If the MAC address is unknown and if the administrative policy on the 1348 PE requires flooding of unknown unicast traffic then: 1350 - The PE MUST flood the packet to other PEs. The 1351 PE MUST first encapsulate the packet in the ESI MPLS 1352 label as described in section 9.3. 1353 If ingress replication is used, the packet MUST be replicated 1354 one or more times to each remote PE with the outermost 1355 label being an MPLS label determined as follows: This 1356 is the MPLS label advertised by the remote PE in a PMSI 1357 Tunnel Attribute in the Inclusive Multicast Ethernet Tag 1358 route for an combination. The Ethernet 1359 Tag in the route must be the same as the Ethernet Tag 1360 associated with the interface on which the ingress PE 1361 receives the packet. If P2MP LSPs are being used the packet 1362 MUST be sent on the P2MP LSP that the PE is the root of for 1363 the Ethernet Tag in the EVI. If the same P2MP LSP is used 1364 for all Ethernet Tags, then all the PEs in the EVI MUST 1365 be the leaves of the P2MP LSP. If a distinct P2MP LSP is 1366 used for a given Ethernet Tag in the EVI, then only the 1367 PEs in the Ethernet Tag MUST be the leaves of the P2MP 1368 LSP. The packet MUST be encapsulated in the P2MP LSP label 1369 stack. 1371 If the MAC address is unknown then, if the administrative policy on 1372 the PE does not allow flooding of unknown unicast traffic: 1374 - The PE MUST drop the packet. 1376 14.2. Forwarding packets received from a remote PE 1377 14.2.1. Unknown Unicast Forwarding 1379 When an PE receives an MPLS packet from a remote PE then, after 1380 processing the MPLS label stack, if the top MPLS label ends up being 1381 a P2MP LSP label associated with an EVI or the downstream label 1382 advertised in the P-Tunnel attribute, and after performing the split 1383 horizon procedures described in section "Split Horizon": 1385 - If the PE is the designated forwarder of unknown unicast, broadcast 1386 or multicast traffic, on a particular set of ESIs for the Ethernet 1387 Tag, the default behavior is for the PE to flood the packet on these 1388 ESIs. In other words, the default behavior is for the PE to assume 1389 that the destination MAC address is unknown unicast, broadcast or 1390 multicast and it is not required to perform a destination MAC address 1391 lookup. As an option, the PE may perform a destination MAC lookup to 1392 flood the packet to only a subset of the CE interfaces in the 1393 Ethernet Tag. For instance the PE may decide to not flood an unknown 1394 unicast packet on certain Ethernet segments even if it is the DF on 1395 the Ethernet segment, based on administrative policy. 1397 - If the PE is not the designated forwarder on any of the ESIs for 1398 the Ethernet Tag, the default behavior is for it to drop the packet. 1400 14.2.2. Known Unicast Forwarding 1402 If the top MPLS label ends up being an E-VPN label that was 1403 advertised in the unicast MAC advertisements, then the PE either 1404 forwards the packet based on CE next-hop forwarding information 1405 associated with the label or does a destination MAC address lookup to 1406 forward the packet to a CE. 1408 15. Load Balancing of Unicast Frames 1410 This section specifies the load balancing procedures for sending 1411 known unicast frames to a multi-homed CE. 1413 15.1. Load balancing of traffic from an PE to remote CEs 1415 Whenever a remote PE imports a MAC advertisement for a given in an EVI, it MUST examine all imported Ethernet A-D 1417 routes for that ESI in order to determine the load-balancing 1418 characteristics of the Ethernet segment. 1420 15.1.1 Active-Standby Redundancy Mode 1422 For a given ESI, if the remote PE has imported an Ethernet A-D route 1423 per Ethernet Segment from at least one PE, where the "Active-Standby" 1424 flag in the ESI MPLS Label Extended Community is set, then the remote 1425 PE MUST deduce that the Ethernet segment is operating in Active- 1426 Standby redundancy mode. As such, the MAC address will be reachable 1427 only via the PE announcing the associated MAC Advertisement route - 1428 this is referred to as the primary PE. The set of other PE nodes 1429 advertising Ethernet A-D routes per Ethernet Segment for the same ESI 1430 serve as backup paths, in case the active PE encounters a failure. 1431 These are referred to as the backup PEs. 1433 If the primary PE encounters a failure, it MAY withdraw its Ethernet 1434 A-D route for the affected segment prior to withdrawing the entire 1435 set of MAC Advertisement routes. In the case where only a single 1436 other backup PE in the network had advertised an Ethernet A-D route 1437 for the same ESI, the remote PE can then use the Ethernet A-D route 1438 withdrawal as a trigger to update its forwarding entries, for the 1439 associated MAC addresses, to point towards the backup PE. As the 1440 backup PE starts learning the MAC addresses over its attached 1441 Ethernet segment, it will start sending MAC Advertisement routes 1442 while the failed PE withdraws its own. This mechanism minimizes the 1443 flooding of traffic during fail-over events. 1445 15.1.2 All-Active Redundancy Mode 1447 If for the given ESI, none of the Ethernet A-D routes per Ethernet 1448 Segment imported by the remote PE have the "Active-Standby" flag set 1449 in the ESI MPLS Label Extended Community, then the remote PE MUST 1450 treat the Ethernet segment as operating in all-active redundancy 1451 mode. The remote PE would then treat the MAC address as reachable via 1452 all of the PE nodes from which it has received both an Ethernet A-D 1453 route per Ethernet Segment as well as an Ethernet A-D route per EVI 1454 for the ESI in question. The remote PE MUST use the MAC advertisement 1455 and eligible Ethernet A-D routes to construct the set of next-hops 1456 that it can use to send the packet to the destination MAC. Each next- 1457 hop comprises an MPLS label stack that is to be used by the egress PE 1458 to forward the packet. This label stack is determined as follows: 1460 -If the next-hop is constructed as a result of a MAC route then this 1461 label stack MUST be used. However, if the MAC route doesn't exist, 1462 then the next-hop and MPLS label stack is constructed as a result of 1463 the Ethernet A-D routes. Note that the following description applies 1464 to determining the label stack for a particular next-hop to reach a 1465 given PE, from which the remote PE has received and imported Ethernet 1466 A-D routes that have the matching ESI and Ethernet Tag as the one 1467 present in the MAC advertisement. The Ethernet A-D routes mentioned 1468 in the following description refer to the ones imported from this 1469 given PE. 1471 -If an Ethernet A-D route per Ethernet Segment for that ESI exists, 1472 together with an Ethernet A-D route per EVI, then the label from that 1473 latter route must be used. 1475 The following example explains the above. 1477 Consider a CE (CE1) that is dual-homed to two PEs (PE1 and PE2) on a 1478 LAG interface (ES1), and is sending packets with MAC address MAC1 on 1479 VLAN1. A remote PE, say PE3, is able to learn that MAC1 is reachable 1480 via PE1 and PE2. Both PE1 and PE2 may advertise MAC1 in BGP if they 1481 receive packets with MAC1 from CE1. If this is not the case, and if 1482 MAC1 is advertised only by PE1, PE3 still considers MAC1 as reachable 1483 via both PE1 and PE2 as both PE1 and PE2 advertise a Ethernet A-D 1484 route per ESI for ESI1 as well as an Ethernet A-D route per EVI for 1485 . 1487 The MPLS label stack to send the packets to PE1 is the MPLS LSP stack 1488 to get to PE1 and the E-VPN label advertised by PE1 for CE1's MAC. 1490 The MPLS label stack to send packets to PE2 is the MPLS LSP stack to 1491 get to PE2 and the MPLS label in the Ethernet A-D route advertised by 1492 PE2 for , if PE2 has not advertised MAC1 in BGP. 1494 We will refer to these label stacks as MPLS next-hops. 1496 The remote PE (PE3) can now load balance the traffic it receives from 1497 its CEs, destined for CE1, between PE1 and PE2. PE3 may use N-Tuple 1498 flow information to hash traffic into one of the MPLS next-hops for 1499 load balancing of IP traffic. Alternatively PE3 may rely on the 1500 source MAC addresses for load balancing. 1502 Note that once PE3 decides to send a particular packet to PE1 or PE2 1503 it can pick one out of multiple possible paths to reach the 1504 particular remote PE using regular MPLS procedures. For instance, if 1505 the tunneling technology is based on RSVP-TE LSPs, and PE3 decides to 1506 send a particular packet to PE1, then PE3 can choose from multiple 1507 RSVP-TE LSPs that have PE1 as their destination. 1509 When PE1 or PE2 receive the packet destined for CE1 from PE3, if the 1510 packet is a unicast MAC packet it is forwarded to CE1. If it is a 1511 multicast or broadcast MAC packet then only one of PE1 or PE2 must 1512 forward the packet to the CE. Which of PE1 or PE2 forward this packet 1513 to the CE is determined based on which of the two is the DF. 1515 If the connectivity between the multi-homed CE and one of the PEs 1516 that it is attached to fails, the PE MUST withdraw the Ethernet Tag 1517 A-D routes, that had been previously advertised, for the Ethernet 1518 Segment to the CE. When the MAC entry on the PE ages out, the PE MUST 1519 withdraw the MAC address from BGP. Note that to aid convergence, the 1520 Ethernet Tag A-D routes MAY be withdrawn before the MAC routes. This 1521 enables the remote PEs to remove the MPLS next-hop to this particular 1522 PE from the set of MPLS next-hops that can be used to forward traffic 1523 to the CE. For further details and procedures on withdrawal of E-VPN 1524 route types in the event of PE to CE failures please section "PE to 1525 CE Network Failures". 1527 15.2. Load balancing of traffic between an PE and a local CE 1529 A CE may be configured with more than one interface connected to 1530 different PEs or the same PE for load balancing, using a technology 1531 such as LAG. The PE(s) and the CE can load balance traffic onto these 1532 interfaces using one of the following mechanisms. 1534 15.2.1. Data plane learning 1536 Consider that the PEs perform data plane learning for local MAC 1537 addresses learned from local CEs. This enables the PE(s) to learn a 1538 particular MAC address and associate it with one or more interfaces, 1539 if the technology between the PE and the CE supports multi-pathing. 1540 The PEs can now load balance traffic destined to that MAC address on 1541 the multiple interfaces. 1543 Whether the CE can load balance traffic that it generates on the 1544 multiple interfaces is dependent on the CE implementation. 1546 15.2.2. Control plane learning 1548 The CE can be a host that advertises the same MAC address using a 1549 control protocol on both interfaces. This enables the PE(s) to learn 1550 the host's MAC address and associate it with one or more interfaces. 1551 The PEs can now load balance traffic destined to the host on the 1552 multiple interfaces. The host can also load balance the traffic it 1553 generates onto these interfaces and the PE that receives the traffic 1554 employs E-VPN forwarding procedures to forward the traffic. 1556 16. MAC Mobility 1558 It is possible for a given host or end-station (as defined by its MAC 1559 address) to move from one Ethernet segment to another; this is 1560 referred to as 'MAC Mobility' or 'MAC move' and it is different from 1561 the multi-homing situation in which a given MAC address is reachable 1562 via multiple PEs for the same Ethernet segment. In a MAC move, there 1563 would be two sets of MAC Advertisement routes, one set with the new 1564 Ethernet segment and one set with the previous Ethernet segment, and 1565 the MAC address would appear to be reachable via each of these 1566 segments. 1568 In order to allow all of the PEs in the E-VPN to correctly determine 1569 the current location of the MAC address, all advertisements of it 1570 being reachable via the previous Ethernet segment MUST be withdrawn 1571 by the PEs, for the previous Ethernet segment, that had advertised 1572 it. 1574 If local learning is performed using the data plane, these PEs will 1575 not be able to detect that the MAC address has moved to another 1576 Ethernet segment and the receipt of MAC Advertisement routes, with 1577 the MAC Mobility extended community attribute, from other PEs serves 1578 as the trigger for these PEs to withdraw their advertisements. If 1579 local learning is performed using the control or management planes, 1580 these interactions serve as the trigger for these PEs to withdraw 1581 their advertisements. 1583 In a situation where there are multiple moves of a given MAC, 1584 possibly between the same two Ethernet segments, there may be 1585 multiple withdrawals and re-advertisements. In order to ensure that 1586 all PEs in the E-VPN receive all of these correctly through the 1587 intervening BGP infrastructure, it is necessary to introduce a 1588 sequence number into the MAC Mobility extended community attribute. 1590 Since the sequence number is an unsigned 32 bit integer, all sequence 1591 number comparisons must be performed modulo 2**32. This unsigned 1592 arithmetic preserves the relationship of sequence numbers as they 1593 cycle from 2**32 - 1 to 0. 1595 Every MAC mobility event for a given MAC address will contain a 1596 sequence number that is set using the following rules: 1598 - A PE advertising a MAC address for the first time advertises it 1599 with no MAC Mobility extended community attribute. 1601 - A PE detecting a locally attached MAC address for which it had 1602 previously received a MAC Advertisement route with a different 1603 Ethernet segment identifier advertises the MAC address in a MAC 1604 Advertisement route tagged with a MAC Mobility extended community 1605 attribute with a sequence number one greater than the sequence number 1606 in the MAC mobility attribute of the received MAC Advertisement 1607 route. In the case of the first mobility event for a given MAC 1608 address, where the received MAC Advertisement route does not carry a 1609 MAC Mobility attribute, the value of the sequence number in the 1610 received route is assumed to be 0 for purpose of this processing. 1612 - A PE detecting a locally attached MAC address for which it had 1613 previously received a MAC Advertisement route with the same Ethernet 1614 segment identifier advertises it with: 1615 i. no MAC Mobility extended community attribute, if the received 1616 route did not carry said attribute. 1618 ii. a MAC Mobility extended community attribute with the sequence 1619 number equal to the sequence number in the received MAC 1620 Advertisement route, if the received route is tagged with a MAC 1621 Mobility extended community attribute. 1623 A PE receiving a MAC Advertisement route for a MAC address with a 1624 different Ethernet segment identifier and a higher sequence number 1625 than that which it had previously advertised, withdraws its MAC 1626 Advertisement route. If two (or more) PEs advertise the same MAC 1627 address with same sequence number but different Ethernet segment 1628 identifiers, a PE that receives these routes selects the route 1629 advertised by the PE with lowest IP address as the best route. 1631 17. Multicast 1633 The PEs in a particular E-VPN may use ingress replication or P2MP 1634 LSPs to send multicast traffic to other PEs. 1636 17.1. Ingress Replication 1638 The PEs may use ingress replication for flooding unknown unicast, 1639 multicast or broadcast traffic as described in section "Handling of 1640 Multi-Destination Traffic". A given unknown unicast or broadcast 1641 packet must be sent to all the remote PEs. However a given multicast 1642 packet for a multicast flow may be sent to only a subset of the PEs. 1643 Specifically a given multicast flow may be sent to only those PEs 1644 that have receivers that are interested in the multicast flow. 1645 Determining which of the PEs have receivers for a given multicast 1646 flow is done using explicit tracking described below. 1648 17.2. P2MP LSPs 1650 An PE may use an "Inclusive" tree for sending an unknown unicast, 1651 broadcast or multicast packet or a "Selective" tree. This terminology 1652 is borrowed from [VPLS-MCAST]. 1654 A variety of transport technologies may be used in the SP network. 1655 For inclusive P-Multicast trees, these transport technologies include 1656 point-to-multipoint LSPs created by RSVP-TE or mLDP. For selective P- 1657 Multicast trees, only unicast PE-PE tunnels (using MPLS or IP/GRE 1658 encapsulation) and P2MP LSPs are supported, and the supported P2MP 1659 LSP signaling protocols are RSVP-TE, and mLDP. 1661 17.3. MP2MP LSPs 1663 The root of the MP2MP LDP LSP advertises the Inclusive Multicast Tag 1664 route with the PMSI Tunnel attribute set to the MP2MP Tunnel 1665 identifier. This advertisement is then sent to all PEs in the E-VPN. 1666 Upon receiving the Inclusive Multicast Tag routes with a PMSI Tunnel 1667 attribute that contains the MP2MP Tunnel identifier, the receiving 1668 PEs initiate the setup of the MP2MP tunnel towards the root using the 1669 procedures in [MLDP]. 1671 17.3.1. Inclusive Trees 1673 An Inclusive Tree allows the use of a single multicast distribution 1674 tree, referred to as an Inclusive P-Multicast tree, in the SP network 1675 to carry all the multicast traffic from a specified set of EVIs on a 1676 given PE. A particular P-Multicast tree can be set up to carry the 1677 traffic originated by sites belonging to a single E-VPN, or to carry 1678 the traffic originated by sites belonging to different E-VPNs. The 1679 ability to carry the traffic of more than one E-VPN on the same tree 1680 is termed 'Aggregation'. The tree needs to include every PE that is a 1681 member of any of the E-VPNs that are using the tree. This implies 1682 that an PE may receive multicast traffic for a multicast stream even 1683 if it doesn't have any receivers that are interested in receiving 1684 traffic for that stream. 1686 An Inclusive P-Multicast tree as defined in this document is a P2MP 1687 tree. A P2MP tree is used to carry traffic only for E-VPN CEs that 1688 are connected to the PE that is the root of the tree. 1690 The procedures for signaling an Inclusive Tree are the same as those 1691 in [VPLS-MCAST] with the VPLS-AD route replaced with the Inclusive 1692 Multicast Ethernet Tag route. The P-Tunnel attribute [VPLS-MCAST] for 1693 an Inclusive tree is advertised in the Inclusive Multicast route as 1694 described in section "Handling of Multi-Destination Traffic". Note 1695 that an PE can "aggregate" multiple inclusive trees for different 1696 EVIs on the same P2MP LSP using upstream labels. The procedures for 1697 aggregation are the same as those described in [VPLS-MCAST], with 1698 VPLS A-D routes replaced by E-VPN Inclusive Multicast routes. 1700 17.3.2. Selective Trees 1702 A Selective P-Multicast tree is used by an PE to send IP multicast 1703 traffic for one or more specific IP multicast streams, originated by 1704 CEs connected to the PE, that belong to the same or different E-VPNs, 1705 to a subset of the PEs that belong to those E-VPNs. Each of the PEs 1706 in the subset should be on the path to a receiver of one or more 1707 multicast streams that are mapped onto the tree. The ability to use 1708 the same tree for multicast streams that belong to different E-VPNs 1709 is termed an PE the ability to create separate SP multicast trees for 1710 specific multicast streams, e.g. high bandwidth multicast streams. 1711 This allows traffic for these multicast streams to reach only those 1712 PE routers that have receivers in these streams. This avoids flooding 1713 other PE routers in the E-VPN. 1715 An SP can use both Inclusive P-Multicast trees and Selective P- 1716 Multicast trees or either of them for a given E-VPN on an PE, based 1717 on local configuration. 1719 The granularity of a selective tree is where S is an 1720 IP multicast source address and G is an IP multicast group address or 1721 G is a multicast MAC address. Wildcard sources and wildcard groups 1722 are supported. Selective trees require explicit tracking as described 1723 below. 1725 A E-VPN PE advertises a selective tree using a E-VPN selective A-D 1726 route. The procedures are the same as those in [VPLS-MCAST] with S- 1727 PMSI A-D routes in [VPLS-MCAST] replaced by E-VPN Selective A-D 1728 routes. The information elements of the E-VPN selective A-D route 1729 are similar to those of the VPLS S-PMSI A-D route with the following 1730 differences. A E-VPN Selective A-D route includes an optional 1731 Ethernet Tag field. Also an E-VPN selective A-D route may encode a 1732 MAC address in the Group field. The encoding details of the E-VPN 1733 selective A-D route will be described in the next revision. 1735 Selective trees can also be aggregated on the same P2MP LSP using 1736 aggregation as described in [VPLS-MCAST]. 1738 17.4. Explicit Tracking 1740 [VPLS-MCAST] describes procedures for explicit tracking that rely on 1741 Leaf A-D routes. The same procedures are used for explicit tracking 1742 in this specification with VPLS Leaf A-D routes replaced with E-VPN 1743 Leaf A-D routes. These procedures allow a root PE to request 1744 multicast membership information for a given (S, G), from leaf PEs. 1745 Leaf PEs rely on IGMP snooping or PIM snooping between the PE and the 1746 CE to determine the multicast membership information. Note that the 1747 procedures in [VPLS-MCAST] do not describe how explicit tracking is 1748 performed if the CEs are enabled with join suppression. The 1749 procedures for this case will be described in a future version. 1751 18. Convergence 1753 This section describes failure recovery from different types of 1754 network failures. 1756 18.1. Transit Link and Node Failures between PEs 1758 The use of existing MPLS Fast-Reroute mechanisms can provide failure 1759 recovery in the order of 50ms, in the event of transit link and node 1760 failures in the infrastructure that connects the PEs. 1762 18.2. PE Failures 1764 Consider a host host1 that is dual homed to PE1 and PE2. If PE1 1765 fails, a remote PE, PE3, can discover this based on the failure of 1766 the BGP session. This failure detection can be in the sub-second 1767 range if BFD is used to detect BGP session failure. PE3 can update 1768 its forwarding state to start sending all traffic for host1 to only 1769 PE2. It is to be noted that this failure recovery is potentially 1770 faster than what would be possible if data plane learning were to be 1771 used. As in that case PE3 would have to rely on re-learning of MAC 1772 addresses via PE2. 1774 18.2.1. Local Repair 1775 It is possible to perform local repair in the case of PE failures. 1776 Details will be specified in the future. 1778 18.3. PE to CE Network Failures 1780 When an Ethernet segment connected to an PE fails or when a Ethernet 1781 Tag is decommissioned on an Ethernet segment, then the PE MUST 1782 withdraw the Ethernet A-D route(s) announced for the that are impacted by the failure or decommissioning. In 1784 addition, the PE MUST also withdraw the MAC advertisement routes that 1785 are impacted by the failure or decommissioning. 1787 The Ethernet A-D routes should be used by an implementation to 1788 optimize the withdrawal of MAC advertisement routes. When an PE 1789 receives a withdrawal of a particular Ethernet A-D route from an PE 1790 it SHOULD consider all the MAC advertisement routes, that are learned 1791 from the same as in the Ethernet A-D route, from 1792 the advertising PE, as having been withdrawn. This optimizes the 1793 network convergence times in the event of PE to CE failures. 1795 19. LACP State Synchronization 1797 This section requires review and discussion amongst the authors and 1798 will be revised in the next version. 1800 To support CE multi-homing with multi-chassis Ethernet bundles, the 1801 PEs connected to a given CE should synchronize [802.1AX] LACP state 1802 amongst each other. This ensures that the PEs can present a single 1803 LACP bundle to the CE. This is required for initial system bring-up 1804 and upon any configuration change. 1806 This includes at least the following LACP specific configuration 1807 parameters: 1809 - System Identifier (MAC Address): uniquely identifies a LACP 1810 speaker. 1811 - System Priority: determines which LACP speaker's port 1812 priorities are used in the Selection logic. 1813 - Aggregator Identifier: uniquely identifies a bundle within 1814 a LACP speaker. 1815 - Aggregator MAC Address: identifies the MAC address of the 1816 bundle. 1817 - Aggregator Key: used to determine which ports can join an 1818 Aggregator. 1819 - Port Number: uniquely identifies an interface within a LACP 1820 speaker. 1821 - Port Key: determines the set of ports that can be bundled. 1822 - Port Priority: determines a port's precedence level to join 1823 a bundle in case the number of eligible ports exceeds the 1824 maximum number of links allowed in a bundle. 1826 Furthermore, the PEs should also synchronize operational (run-time) 1827 data, in order for the LACP Selection logic state-machines to 1828 execute. This operational data includes the following LACP 1829 operational parameters, on a per port basis: 1831 - Partner System Identifier: this is the CE System MAC address. 1832 - Partner System Priority: the CE LACP System Priority 1833 - Partner Port Number: CE's AC port number. 1834 - Partner Port Priority: CE's AC Port Priority. 1835 - Partner Key: CE's key for this AC. 1836 - Partner State: CE's LACP State for the AC. 1837 - Actor State: PE's LACP State for the AC. 1838 - Port State: PE's AC port status. 1840 The above state needs to be communicated between PEs forming a multi- 1841 chassis bundle during LACP initial bringup, upon any configuration 1842 change and upon the occurrence of a failure. 1844 It should be noted that the above configuration and operational state 1845 is localized in scope and is only relevant to PEs which connect to 1846 the same multi-homed CE over a given Ethernet bundle. 1848 Furthermore, the communication of state changes, upon failures, must 1849 occur with minimal latency, in order to minimize the switchover time 1850 and consequent service disruption. The protocol details for 1851 synchronizing the LACP state will be described in the following 1852 version. 1854 20. Acknowledgements 1856 We would like to thank Yakov Rekhter, Pedro Marques, Kaushik Ghosh, 1857 Nischal Sheth, Robert Raszuk, Amit Shukla and Nadeem Mohammed for 1858 discussions that helped shape this document. We would also like to 1859 thank Han Nguyen for his comments and support of this work. We would 1860 also like to thank Steve Kensil for his review. 1862 21. References 1863 [E-VPN-REQ] A. Sajassi, R. Aggarwal et. al., "Requirements for 1864 Ethernet VPN", draft-sajassi-raggarwa-l2vpn-evpn-req- 1865 00.txt 1867 [RFC4364] "BGP/MPLS IP VPNs", Rosen, Rekhter, et. al., February 2006 1869 [VPLS-MCAST] "Multicast in VPLS". R. Aggarwal et.al., draft-ietf- 1870 l2vpn-vpls-mcast-04.txt 1872 [RFC4761] Kompella, K. and Y. Rekhter, "Virtual Private LAN Service 1873 (VPLS) Using BGP for Auto-Discovery and Signaling", RFC 1874 4761, January 2007. 1876 [RFC4762] Lasserre, M. and V. Kompella, "Virtual Private LAN Service 1877 (VPLS) Using Label Distribution Protocol (LDP) Signaling", 1878 RFC 4762, January 2007. 1880 [VPLS-MULTIHOMING] "BGP based Multi-homing in Virtual Private LAN 1881 Service", K. Kompella et. al., draft-ietf-l2vpn-vpls- 1882 multihoming-00.txt 1884 [PIM-SNOOPING] "PIM Snooping over VPLS", V. Hemige et. al., draft- 1885 ietf-l2vpn-vpls-pim-snooping-01 1887 [IGMP-SNOOPING] "Considerations for Internet Group Management 1888 Protocol (IGMP) and Multicast Listener Discovery (MLD) 1889 Snooping Switches", M. Christensen et. al., RFC4541, 1891 [RT-CONSTRAIN] P. Marques et. al., "Constrained Route Distribution 1892 for Border Gateway Protocol/MultiProtocol Label Switching 1893 (BGP/MPLS) Internet Protocol (IP) Virtual Private Networks 1894 (VPNs)", RFC 4684, November 2006 1896 [EVPN-SEGMENT-ROUTE] A. Sajassi et. al., "E-VPN Ethernet Segment 1897 Route", draft-sajassi-l2vpn-evpn-segment-route-00.txt, 1898 work in progress. 1900 21. Author's Address 1902 Rahul Aggarwal 1903 Email: raggarwa_1@yahoo.com 1905 Ali Sajassi 1906 Cisco 1907 170 West Tasman Drive 1908 San Jose, CA 95134, US 1909 Email: sajassi@cisco.com 1911 Wim Henderickx 1912 Alcatel-Lucent 1913 e-mail: wim.henderickx@alcatel-lucent.com 1914 Aldrin Isaac 1915 Bloomberg 1916 Email: aisaac71@bloomberg.net 1918 James Uttaro 1919 AT&T 1920 200 S. Laurel Avenue 1921 Middletown, NJ 07748 1922 USA 1923 Email: uttaro@att.com 1925 Nabil Bitar 1926 Verizon Communications 1927 Email : nabil.n.bitar@verizon.com 1929 Ravi Shekhar 1930 Juniper Networks 1931 1194 N. Mathilda Ave. 1932 Sunnyvale, CA 94089 US 1933 Email: rshekhar@juniper.net 1935 Florin Balus 1936 Alcatel-Lucent 1937 e-mail: Florin.Balus@alcatel-lucent.com 1939 Keyur Patel 1940 Cisco 1941 170 West Tasman Drive 1942 San Jose, CA 95134, US 1943 Email: keyupate@cisco.com 1945 Sami Boutros 1946 Cisco 1947 170 West Tasman Drive 1948 San Jose, CA 95134, US 1949 Email: sboutros@cisco.com 1951 Samer Salam 1952 Cisco 1953 595 Burrard Street, Suite 2123 1954 Vancouver, BC V7X 1J1, Canada 1955 Email: ssalam@cisco.com