idnits 2.17.1 draft-ietf-bess-evpn-vpws-08.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == Using lowercase 'not' together with uppercase 'MUST', 'SHALL', 'SHOULD', or 'RECOMMENDED' is not an accepted usage according to RFC 2119. Please use uppercase 'NOT' together with RFC 2119 keywords (if that is what you mean). Found 'SHOULD not' in this paragraph: In a multihoming all-active scenario, there is no DF election, and all the PEs in the ES that are active and ready to forward traffic to/from the CE will set the P bit to 1. A remote PE will do per-flow load balancing to the PEs that send P=1 for the same Ethernet Tag and ESI. B bit in control flags SHOULD not be set in the multihoming all-active scenario and MUST be ignored by receiving PE(s) if set. == Using lowercase 'not' together with uppercase 'MUST', 'SHALL', 'SHOULD', or 'RECOMMENDED' is not an accepted usage according to RFC 2119. Please use uppercase 'NOT' together with RFC 2119 keywords (if that is what you mean). Found 'MUST not' in this paragraph: If a network uses entropy labels per [RFC6790] then the C Bit MUST not be set to 1 and control word MUST NOT be used when sending EVPN-encapsulated packets over a P2P LSP. -- The document date (February 7, 2017) is 2629 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: 'RFCXXXX' is mentioned on line 544, but not defined == Unused Reference: 'RFC7153' is defined on line 566, but no explicit reference was found in the text ** Downref: Normative reference to an Informational RFC: RFC 4664 Summary: 1 error (**), 0 flaws (~~), 5 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 INTERNET-DRAFT Sami Boutros 3 Intended Status: Standard Track VMware 4 Ali Sajassi 5 Samer Salam 6 Cisco Systems 7 John Drake 8 Juniper Networks 9 J. Rabadan 10 Nokia 12 Expires: August 11, 2017 February 7, 2017 14 VPWS support in EVPN 15 draft-ietf-bess-evpn-vpws-08.txt 17 Abstract 19 This document describes how EVPN can be used to support Virtual 20 Private Wire Service (VPWS) in MPLS/IP networks. EVPN enables the 21 following characteristics for VPWS: single-active as well as all- 22 active multi-homing with flow-based load-balancing, eliminates the 23 need for traditional way of PW signaling, and provides fast 24 protection convergence upon node or link failure. 26 Status of this Memo 28 This Internet-Draft is submitted to IETF in full conformance with the 29 provisions of BCP 78 and BCP 79. 31 Internet-Drafts are working documents of the Internet Engineering 32 Task Force (IETF), its areas, and its working groups. Note that 33 other groups may also distribute working documents as 34 Internet-Drafts. 36 Internet-Drafts are draft documents valid for a maximum of six months 37 and may be updated, replaced, or obsoleted by other documents at any 38 time. It is inappropriate to use Internet-Drafts as reference 39 material or to cite them other than as "work in progress." 41 The list of current Internet-Drafts can be accessed at 42 http://www.ietf.org/1id-abstracts.html 44 The list of Internet-Draft Shadow Directories can be accessed at 45 http://www.ietf.org/shadow.html 47 Copyright and License Notice 48 Copyright (c) 2017 IETF Trust and the persons identified as the 49 document authors. All rights reserved. 51 This document is subject to BCP 78 and the IETF Trust's Legal 52 Provisions Relating to IETF Documents 53 (http://trustee.ietf.org/license-info) in effect on the date of 54 publication of this document. Please review these documents 55 carefully, as they describe your rights and restrictions with respect 56 to this document. Code Components extracted from this document must 57 include Simplified BSD License text as described in Section 4.e of 58 the Trust Legal Provisions and are provided without warranty as 59 described in the Simplified BSD License. 61 Table of Contents 63 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 64 1.1 Terminology . . . . . . . . . . . . . . . . . . . . . . . . 4 65 1.2 Requirements . . . . . . . . . . . . . . . . . . . . . . . . 5 66 2 Service interface . . . . . . . . . . . . . . . . . . . . . . . 6 67 2.1 VLAN-Based Service Interface . . . . . . . . . . . . . . . . 6 68 2.2 VLAN Bundle Service Interface . . . . . . . . . . . . . . . 6 69 2.2.1 Port-Based Service Interface . . . . . . . . . . . . . . 7 70 2.3 VLAN-Aware Bundle Service Interface . . . . . . . . . . . . 7 71 3. BGP Extensions . . . . . . . . . . . . . . . . . . . . . . . . 7 72 3.1 EVPN Layer 2 attributes extended community . . . . . . . . . 8 73 4 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 74 5 EVPN Comparison to PW Signaling . . . . . . . . . . . . . . . . 11 75 6 Failure Scenarios . . . . . . . . . . . . . . . . . . . . . . . 11 76 6.1 Single-Homed CEs . . . . . . . . . . . . . . . . . . . . . . 11 77 6.2 Multi-Homed CEs . . . . . . . . . . . . . . . . . . . . . . 12 78 7 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . 12 79 8 Security Considerations . . . . . . . . . . . . . . . . . . . . 12 80 9 IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 12 81 10 References . . . . . . . . . . . . . . . . . . . . . . . . . . 12 82 10.1 Normative References . . . . . . . . . . . . . . . . . . . 12 83 10.2 Informative References . . . . . . . . . . . . . . . . . . 13 84 Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 85 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 13 87 1 Introduction 89 This document describes how EVPN can be used to support Virtual 90 Private Wire Service (VPWS) in MPLS/IP networks. The use of EVPN 91 mechanisms for VPWS (EVPN-VPWS) brings the benefits of EVPN to p2p 92 services. These benefits include single-active redundancy as well as 93 all-active redundancy with flow-based load-balancing. Furthermore, 94 the use of EVPN for VPWS eliminates the need for traditional way of 95 PW signaling for p2p Ethernet services, as described in section 4. 97 [RFC7432] has the ability to forward customer traffic to/from a given 98 customer Attachment Circuit (AC), without any MAC lookup. This 99 capability is ideal in providing p2p services (aka VPWS services). 100 [MEF] defines Ethernet Virtual Private Line (EVPL) service as p2p 101 service between a pair of ACs (designated by VLANs) and Ethernet 102 Private Line (EPL) service, in which all traffic flows are between a 103 single pair of ports, that in EVPN terminology would mean a single 104 pair of Ethernet Segments ES(es). EVPL can be considered as a VPWS 105 with only two ACs. In delivering an EVPL service, the traffic 106 forwarding capability of EVPN based on the exchange of a pair of 107 Ethernet Auto-discovery (A-D) routes is used; whereas, for more 108 general VPWS as per [RFC4664], traffic forwarding capability of EVPN 109 based on the exchange of a group of Ethernet AD routes (one Ethernet 110 AD route per AC/ES) is used. In a VPWS service, the traffic from an 111 originating Ethernet Segment can be forwarded only to a single 112 destination Ethernet Segment; hence, no MAC lookup is needed and the 113 MPLS label associated with the per EVPN instance (EVI) Ethernet A-D 114 route can be used in forwarding user traffic to the destination AC. 116 Both services are supported by using the per EVI Ethernet A-D route 117 which contains an Ethernet Segment Identifier, in which the customer 118 ES is encoded, and an Ethernet Tag, in which the VPWS service 119 instance identifier is encoded. I.e., for both EPL and EVPL 120 services, a specific VPWS service instance is identified by a pair of 121 per EVI Ethernet A-D routes which together identify the VPWS service 122 instance endpoints and the VPWS service instance. In the control 123 plane the VPWS service instance is identified using the VPWS service 124 instance identifiers advertised by each PE and in the data plane the 125 value of the MPLS label advertised by one PE is used by the other PE 126 to send traffic for that VPWS service instance. As with the Ethernet 127 Tag in standard EVPN, the VPWS service instance identifier has 128 uniqueness within an EVPN instance. 130 Unlike EVPN where Ethernet Tag ID in EVPN routes are set to zero for 131 Port-based, vlan-based, and vlan-bundle interface mode and it is set 132 to non-zero Ethernet tag ID for vlan-aware bundle mode, in EVPN-VPWS, 133 for all the four interface modes, Ethernet tag ID in the Ethernet A-D 134 route MUST be set to a non-zero value in all the service interface 135 types. 137 In terms of route advertisement and MPLS label lookup behavior, EVPN- 138 VPWS resembles the vlan-aware bundle mode of [RFC7432] such that when 139 a PE advertises per EVI Ethernet A-D route, the VPWS service instance 140 serves as a 24-bit normalized Ethernet tag ID. The value of the MPLS 141 label in this route represents both the EVI and the VPWS service 142 instance, so that upon receiving an MPLS encapsulated packet, the 143 disposition PE can identify the egress AC from the lookup of the MPLS 144 label alone and perform any required tag translation. For EVPL 145 service, the Ethernet frames transported over an MPLS/IP network 146 SHOULD remain tagged with the originating Vlan-ID (VID) and any VID 147 translation MUST be performed at the disposition PE. For EPL service, 148 the Ethernet frames are transported as is and the tags are not 149 altered. 151 The MPLS label value in the Ethernet A-D route can be set to the 152 VXLAN Network Identifier (VNI) for VxLAN encap, and this VNI may have 153 a global scope or local scope per PE and may also be made equal to 154 the VPWS service instance identifier set in the Ethernet A-D route. 156 The Ethernet Segment identifier encoded in the Ethernet A-D per EVI 157 route is not used to identify the service, however it can be used for 158 flow-based load-balancing and mass withdraw functions as per 159 [RFC7432] baseline. 161 As with standard EVPN, the Ethernet A-D per ES route is used for fast 162 convergence upon link or node failure and the Ethernet Segment route 163 is used for auto-discovery of the PEs attached to a given multi-homed 164 CE and to synchronize state between them. 166 1.1 Terminology 168 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 169 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 170 document are to be interpreted as described in RFC 2119 [RFC2119]. 172 MAC: Media Access Control 174 MPLS: Multi Protocol Label Switching. 176 OAM: Operations, Administration and Maintenance. 178 PE: Provide Edge Node. 180 CE: Customer Edge device e.g., host or router or switch. 182 EVPL: Ethernet Virtual Private Line. 184 EPL: Ethernet Private Line. 186 EP-LAN: Ethernet Private LAN. 188 EVP-LAN: Ethernet Virtual Private LAN. 190 S-VLAN: Service VLAN identifier. 192 C-VLAN: Customer VLAN identifier. 194 VPWS: Virtual Private Wire Service. 196 EVI: EVPN Instance. 198 ES: Ethernet Segment on a PE refers to the link attached to it, this 199 link can be part of a set of links attached to different PEs in multi 200 homed cases, or could be a single link in single homed cases. 202 ESI: Ethernet Segment Identifier. 204 Single-Active Mode: When a device or a network is multi-homed to two 205 or more PEs and when only a single PE in such redundancy group can 206 forward traffic to/from the multi-homed device or network for a given 207 VLAN, then such multi-homing or redundancy is referred to as "Single- 208 Active". 210 All-Active: When a device is multi-homed to two or more PEs and when 211 all PEs in such redundancy group can forward traffic to/from the 212 multi-homed device for a given VLAN, then such multi-homing or 213 redundancy is referred to as "All-Active". 215 VPWS Service Instance: It is represented by a pair of EVPN service 216 labels associated with a pair of endpoints. Each label is downstream 217 assigned and advertised by the disposition PE through an Ethernet A-D 218 per-EVI route. The downstream label identifies the endpoint on the 219 disposition PE. A VPWS service instance can be associated with only 220 one VPWS service identifier. 222 1.2 Requirements 224 1. EPL service access circuit MUST map to the whole Ethernet port. 226 2. EVPL service access circuit MUST map to an individual VLAN or 227 double tagged combination on a given trunk port, 228 without any direct dependency on any other VLANs on the same trunk. 229 Other VLANs on the same trunk MAY also be used for EVPL services, but 230 MAY also be associated with other services. 232 3. If multiple VLANs on the same trunk are associated with EVPL 233 services, the respective remote endpoints of these EVPLs MAY be 234 dispersed across any number of PEs, i.e. different VLANs MAY lead to 235 different destinations. 237 4. The VLAN tag on the access trunk MUST only have PE-local 238 significance. The VLAN tag on the remote end could be different, and 239 could also be double tagged when the other side is single tagged. 241 5. Also, multiple EVPL service VLANs on the same trunk MAY belong to 242 the same EVPN instance (EVI), or they MAY belong to different EVIs. 243 This should be purely an administrative choice of the network 244 operator. 246 6. A given PE MAY have thousands of EVPLs configured. It MUST be 247 possible to configure multiple EVPL services within the same EVI. 249 7. Local access circuits configured to belong to a given EVPN 250 instance MAY also belong to different physical access trunks. 252 8. EP-LAN and EVP-LAN MAY be possible on the same system and also 253 ESIs can be shared between EVPL and EVP-LANs. 255 2 Service interface 257 2.1 VLAN-Based Service Interface 259 With this service interface, a VPWS instance identifier corresponds 260 to only a single VLAN on a specific interface. Therefore, there is a 261 one-to-one mapping between a VID on this interface and the VPWS 262 service instance identifier. The PE provides the cross-connect 263 functionality between MPLS LSP identified by the VPWS service 264 instance identifier and a specific . If the VLAN is 265 represented by different VIDs on different PEs. (e.g., a different 266 VID per Ethernet segment per PE), then each PE needs to perform VID 267 translation for frames destined to its Ethernet segment. In such 268 scenarios, the Ethernet frames transported over an MPLS/IP network 269 SHOULD remain tagged with the originating VID, and a VID translation 270 MUST be supported in the data path and MUST be performed on the 271 disposition PE. 273 2.2 VLAN Bundle Service Interface 275 With this service interface, a VPWS service instance identifier 276 corresponds to multiple VLANs on a specific interface. The PE 277 provides the cross-connect functionality between MPLS label 278 identified by the VPWS service instance identifier and a group of 279 VLANs on a specific interface. For this service interface, each VLAN 280 is presented by a single VID which means no VLAN translation is 281 allowed. The receiving PE, can direct the traffic based on EVPN label 282 alone to a specific port. The transmitting PE can cross connect 283 traffic from a group of VLANs on a specific port to the MPLS label. 284 The MPLS-encapsulated frames MUST remain tagged with the originating 285 VID. 287 2.2.1 Port-Based Service Interface 289 This service interface is a special case of the VLAN bundle service 290 interface, where all of the VLANs on the port are mapped to the same 291 VPWS service instance identifier. The procedures are identical to 292 those described in Section 2.2. 294 2.3 VLAN-Aware Bundle Service Interface 296 Contrary to EVPN, in EVPN-VPWS this service interface maps to VLAN- 297 based service interface (defined in section 2.1) and thus this 298 service interface is not used in EVPN-VPWS. In other words, if one 299 tries to define data-plane and control plane behavior for this 300 service interface, he would realize that it is the same as that of 301 VLAN-based service. 303 3. BGP Extensions 305 This document specifies the use of the per EVI Ethernet A-D route to 306 signal VPWS services. The Ethernet Segment Identifier field is set to 307 the customer ES and the Ethernet Tag ID 32-bit field MUST be set to 308 the 24-bit VPWS service instance identifier value. For both EPL and 309 EVPL services, for a given VPWS service instance the pair of PEs 310 instantiating that VPWS service instance will each advertise a per 311 EVI Ethernet A-D route with its VPWS service instance identifier and 312 will each be configured with the other PE's VPWS service instance 313 identifier. When each PE has received the other PE's per EVI Ethernet 314 A-D route the VPWS service instance is instantiated. It should be 315 noted that the same VPWS service instance identifier may be 316 configured on both PEs. 318 The Route-Target (RT) extended community with which the per EVI 319 Ethernet A-D route is tagged identifies the EVPN instance in which 320 the VPWS service instance is configured. It is the operator's choice 321 as to how many and which VPWS service instances are configured in a 322 given EVPN instance. However, a given EVPN instance MUST NOT be 323 configured with both VPWS service instances and standard EVPN multi- 324 point services. 326 3.1 EVPN Layer 2 attributes extended community 328 This draft proposes a new extended community, defined below, to be 329 included with the per EVI Ethernet A-D route. This attribute is 330 mandatory if multihoming is enabled. 332 +------------------------------------+ 333 | Type(0x06)/Sub-type(0x04)(2 octet)| 334 +------------------------------------+ 335 | Control Flags (2 octets) | 336 +------------------------------------+ 337 | L2 MTU (2 octets) | 338 +------------------------------------+ 339 | Reserved (2 octets) | 340 +------------------------------------+ 342 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 343 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 344 | MBZ |C|P|B| (MBZ = MUST Be Zero) 345 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 347 The following bits in the Control Flags are defined; the remaining 348 bits MUST be set to zero when sending and MUST be ignored when 349 receiving this community. 351 Name Meaning 353 P If set to 1 in multihoming single-active scenarios, it 354 indicates that the advertising PE is the Primary PE. 355 MUST be set to 1 for multihoming all-active scenarios by 356 all active PE(s). 358 B If set to 1 in multihoming single-active scenarios, it 359 indicates that the advertising PE is the Backup PE. 361 C If set to 1, a Control word [RFC4448] MUST be present 362 when sending EVPN packets to this PE. 364 L2 MTU (Maximum Transmission Unit) is a 2-octet value indicating the 365 MTU in octets. 367 A received L2 MTU=0 means no MTU checking against local MTU is 368 needed. A received non-zero MTU MUST be checked against local MTU and 369 if there is a mismatch, the local PE MUST NOT add the remote PE as 370 the EVPN destination for the corresponding VPWS service instance. 372 The usage of the Per ES Ethernet A-D route is unchanged from its 373 usage in [RFC7432], i.e. the "Single-Active" bit in the flags of the 374 ESI Label extended community will indicate if single-active or all- 375 active redundancy is used for this ES. 377 In multihoming scenarios, both B and P flags MUST NOT be both set. A 378 PE that receives an update with both B and P flags set MUST treat the 379 route as a withdrawal. If the PE receives a route with both B and P 380 unset, it MUST NOT forward any traffic to the sender PE. 382 In a multihoming all-active scenario, there is no DF election, and 383 all the PEs in the ES that are active and ready to forward traffic 384 to/from the CE will set the P bit to 1. A remote PE will do per-flow 385 load balancing to the PEs that send P=1 for the same Ethernet Tag and 386 ESI. B bit in control flags SHOULD not be set in the multihoming all- 387 active scenario and MUST be ignored by receiving PE(s) if set. 389 In multihoming single-active scenario, the DF election will determine 390 who the primary and the backup PEs are, and only those PEs will set 391 the P bit and B bit respectively. A remote PE will forward the 392 traffic to the primary PE and switch over to the backup PE as soon as 393 it receives an Ethernet A-D route withdrawal from the primary PE in 394 the Ethernet Segment. 396 In multihoming single-active scenario, during transient situations, a 397 remote PE receiving P=1 from more than one PE will select the last 398 advertising PE as the primary PE when forwarding traffic. A remote PE 399 receiving B=1 from more than one PE will select only one backup PE. A 400 remote PE MUST receive P=1 from at least one PE before forwarding 401 traffic. 403 If a network uses entropy labels per [RFC6790] then the C Bit MUST 404 not be set to 1 and control word MUST NOT be used when sending EVPN- 405 encapsulated packets over a P2P LSP. 407 4 Operation 409 The following figure shows an example of a P2P service deployed with 410 EVPN. 411 Ethernet Ethernet 412 Native |<--------- EVPN Instance ----------->| Native 413 Service | | Service 414 (AC) | |<-PSN1->| |<-PSN2->| | (AC) 415 | V V V V V V | 416 | +-----+ +-----+ +-----+ +-----+ | 417 +----+ | | PE1 |======|ASBR1|==|ASBR2|===| PE3 | | +----+ 418 | |-------+-----+ +-----+ +-----+ +-----+-------| | 419 | CE1| | | |CE2 | 420 | |-------+-----+ +-----+ +-----+ +-----+-------| | 421 +----+ | | PE2 |======|ASBR3|==|ASBR4|===| PE4 | | +----+ 422 ^ +-----+ +-----+ +-----+ +-----+ ^ 423 | Provider Edge 1 ^ Provider Edge 2 | 424 | | | 425 | | | 426 | EVPN Inter-provider point | 427 | | 428 |<---------------- Emulated Service -------------------->| 430 Figure 1: EVPN-VPWS Deployement Model 431 iBGP sessions are established between PE1, PE2, ASBR1 and ASBR3, 432 possibly via a BGP route-reflector. Similarly, iBGP sessions are 433 established between PE3, PE4, ASBR2 and ASBR4. eBGP sessions are 434 established among ASBR1, ASBR2, ASBR3, and ASBR4. 436 All PEs and ASBRs are enabled for the EVPN SAFI and exchange per EVI 437 Ethernet A-D routes, one route per VPWS service instance. For inter- 438 AS option B, the ASBRs re-advertise these routes with NEXT_HOP 439 attribute set to their IP addresses as per [RFC4271]. The link 440 between the CE and the PE is either a C-tagged or S-tagged interface, 441 as described in [802.1Q], that can carry a single VLAN tag or two 442 nested VLAN tags and it is configured as a trunk with multiple VLANs, 443 one per VPWS service instance. It should be noted that the VLAN ID 444 used by the customer at either end of a VPWS service instance to 445 identify that service instance may be different and EVPN doesn't 446 perform that translation between the two values. Rather, the MPLS 447 label will identify the VPWS service instance and if translation is 448 needed, it should be done by the Ethernet interface for each service. 450 For single-homed CE, in an advertised per EVI Ethernet A-D route the 451 ESI field is set to 0 and the Ethernet Tag ID is set to the VPWS 452 service instance identifier that identifies the EVPL or EPL service. 454 For a multi-homed CE, in an advertised per EVI Ethernet A-D route the 455 ESI field is set to the CE's ESI and the Ethernet Tag ID is set to 456 the VPWS service instance identifier, which MUST have the same value 457 on all PEs attached to that ES. This allows an ingress PE to perform 458 flow-based load-balancing of traffic flows to all of the PEs attached 459 to that ES. In all cases traffic follows the transport paths, which 460 may be asymmetric. 462 The VPWS service instance identifier encoded in the Ethernet Tag ID 463 in an advertised per EVI Ethernet A-D route MUST either be unique 464 across all ASs, or an ASBR needs to perform a translation when the 465 per EVI Ethernet A-D route is re-advertised by the ASBR from one AS 466 to the other AS. 468 Per ES Ethernet A-D route can be used for mass withdraw to withdraw 469 all per EVI Ethernet A-D routes associated with the multi-home site 470 on a given PE. 472 5 EVPN Comparison to PW Signaling 474 In EVPN, service endpoint discovery and label signaling are done 475 concurrently using BGP. Whereas, with VPWS based on [RFC4448], label 476 signaling is done via LDP and service endpoint discovery is either 477 through manual provisioning or through BGP. 479 In existing implementation of VPWS using pseudowires(PWs), redundancy 480 is limited to single-active mode, while with EVPN implementation of 481 VPWS both single-active and all-active redundancy modes can be 482 supported. 484 In existing implementation with PWs, backup PWs are not used to carry 485 traffic, while with EVPN, traffic can be load-balanced among 486 different PEs multi-homed to a single CE. 488 Upon link or node failure, EVPN can trigger failover with the 489 withdrawal of a single BGP route per EVPL service or multiple EVPL 490 services, whereas with VPWS PW redundancy, the failover sequence 491 requires exchange of two control plane messages: one message to 492 deactivate the group of primary PWs and a second message to activate 493 the group of backup PWs associated with the access link. 495 Finally, EVPN may employ data plane egress link protection mechanisms 496 not available in VPWS. This can be done by the primary PE (on local 497 AC down) using the label advertised in the per EVI Ethernet A-D route 498 by the backup PE to encapsulate the traffic and direct it to backup 499 PE. 501 6 Failure Scenarios 503 On a link or port failure between the CE and the PE for both single 504 and multi-homed CEs, unlike [RFC7432] the PE MUST withdraw all the 505 associated Ethernet A-D routes for the VPWS service instances on the 506 failed port or link. 508 6.1 Single-Homed CEs 510 Unlike [RFC7432], EVPN-VPWS uses Ethernet A-D route advertisements 511 for single-homed Ethernet Segments. Therefore, upon a link/port 512 failure of this single-homed Ethernet Segment, the PE MUST withdraw 513 the associated per EVI Ethernet A-D routes. 515 6.2 Multi-Homed CEs 517 For a faster convergence in multi-homed scenarios with either Single- 518 Active Redundancy or All-active redundancy, mass withdraw technique 519 is used. A PE previously advertising a per ES Ethernet A-D route, can 520 withdraw this route signaling to the remote PEs to switch all the 521 VPWS service instances associated with this multi-homed ES to the 522 backup PE 524 7 Acknowledgements 526 The authors would like to acknowledge Jeffrey Zhang, Wen Lin, Nitin 527 Singh, Senthil Sathappan and Vinod Prabhu for their feedback and 528 contributions to this document. 530 8 Security Considerations 532 The mechanisms in this document use EVPN control plane as defined in 533 [RFC7432]. Security considerations described in [RFC7432] are equally 534 applicable. 536 This document uses MPLS and IP-based tunnel technologies to support 537 data plane transport. Security considerations described in [RFC7432] 538 and in [ietf-evpn-overlay] are equally applicable. 540 9 IANA Considerations 542 IANA has allocated the following EVPN Extended Community sub-type: 543 SUB-TYPE VALUE NAME Reference 544 0x04 EVPN Layer 2 attributes [RFCXXXX] 545 10 References 547 10.1 Normative References 549 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 550 Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, March 551 1997, . 553 [RFC7432] Sajassi, A., Ed., Aggarwal, R., Bitar, N., Isaac, A., 554 Uttaro, J., Drake, J., and W. Henderickx, "BGP MPLS-Based Ethernet 555 VPN", RFC 7432, DOI 10.17487/RFC7432, February 2015, . 558 [RFC4448] Martini, L., Rosen, E., El-Aawar, N., and G. Heron, 559 "Encapsulation Methods for Transport of Ethernet over MPLS Networks", 560 RFC 4448, April 2006. 562 [RFC6790] Kompella, K., Drake, J., Amante, S., Henderickx, W., and L. 564 Yong, "The Use of Entropy Labels in MPLS Forwarding", November 2012. 566 [RFC7153] Rosen, E. and Y. Rekhter, "IANA Registries for BGP Extended 567 Communities", RFC 7153, March 2014, . 570 [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A Border 571 Gateway Protocol 4 (BGP-4)", RFC 4271, January 2006, . 574 [RFC4664] Andersson, L., Ed., and E. Rosen, Ed., "Framework for 575 Layer 2 Virtual Private Networks (L2VPNs)", RFC 4664, September 2006, 576 . 578 10.2 Informative References 580 [MEF] Metro Ethernet Forum, "Ethernet Services Definitions - Phase 581 2", Technical Specification MEF 6.1, April 2008, 582 . 585 [ietf-evpn-overlay] Sajassi-Drake et al., "A Network Virtualization 586 Overlay Solution using EVPN", draft-ietf-bess-evpn-overlay-07.txt, 587 work in progress, December, 2016 589 Contributors 591 In addition to the authors listed on the front page, the following 592 co-authors have also contributed to this document: 594 Daniel Voyer Bell Canada 596 Authors' Addresses 598 Sami Boutros 599 VMware, Inc. 600 Email: sboutros@vmware.com 602 Ali Sajassi 603 Cisco 604 Email: sajassi@cisco.com 606 Samer Salam 607 Cisco 608 Email: ssalam@cisco.com 609 John Drake 610 Juniper Networks 611 Email: jdrake@juniper.net 613 Jeff Tantsura 614 Individual 615 Email: jefftant@gmail.com 617 Dirk Steinberg 618 Steinberg Consulting 619 Email: dws@steinbergnet.net 621 Patrice Brissette 622 Cisco 623 Email: pbrisset@cisco.com 625 Thomas Beckhaus 626 Deutsche Telecom 627 Email: Thomas.Beckhaus@telekom.de 629 Jorge Rabadan 630 Nokia 631 Email: jorge.rabadan@nokia.com 633 Ryan Bickhart 634 Juniper Networks 635 Email: rbickhart@juniper.net