idnits 2.17.1 draft-ietf-bess-evpn-pref-df-07.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == The document seems to lack the recommended RFC 2119 boilerplate, even if it appears to use RFC 2119 keywords -- however, there's a paragraph with a matching beginning. Boilerplate error? (The document does seem to have the reference to RFC 2119 which the ID-Checklist requires). -- The document date (March 12, 2021) is 1134 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-15) exists of draft-ietf-bess-evpn-virtual-eth-segment-06 Summary: 0 errors (**), 0 flaws (~~), 3 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 BESS Workgroup J. Rabadan, Ed. 3 Internet-Draft S. Sathappan 4 Intended status: Standards Track Nokia 5 Expires: September 13, 2021 T. Przygienda 6 W. Lin 7 J. Drake 8 Juniper Networks 9 A. Sajassi 10 S. Mohanty 11 Cisco Systems 12 March 12, 2021 14 Preference-based EVPN DF Election 15 draft-ietf-bess-evpn-pref-df-07 17 Abstract 19 The Designated Forwarder (DF) in Ethernet Virtual Private Networks 20 (EVPN) is defined as the PE responsible for sending Broadcast, 21 Unknown unicast and Broadcast traffic (BUM) to a multi-homed device/ 22 network in the case of an all-active multi-homing Ethernet Segment 23 (ES), or BUM and unicast in the case of single-active multi-homing. 24 The DF is selected out of a candidate list of PEs that advertise the 25 same Ethernet Segment Identifier (ESI) to the EVPN network, according 26 to the Default DF Election algorithm. While the Default Algorithm 27 provides an efficient and automated way of selecting the DF across 28 different Ethernet Tags in the ES, there are some use cases where a 29 more 'deterministic' and user-controlled method is required. At the 30 same time, Service Providers require an easy way to force an on- 31 demand DF switchover in order to carry out some maintenance tasks on 32 the existing DF or control whether a new active PE can preempt the 33 existing DF PE. 35 This document proposes a DF Election algorithm that meets the 36 requirements of determinism and operation control. 38 Status of This Memo 40 This Internet-Draft is submitted in full conformance with the 41 provisions of BCP 78 and BCP 79. 43 Internet-Drafts are working documents of the Internet Engineering 44 Task Force (IETF). Note that other groups may also distribute 45 working documents as Internet-Drafts. The list of current Internet- 46 Drafts is at https://datatracker.ietf.org/drafts/current/. 48 Internet-Drafts are draft documents valid for a maximum of six months 49 and may be updated, replaced, or obsoleted by other documents at any 50 time. It is inappropriate to use Internet-Drafts as reference 51 material or to cite them other than as "work in progress." 53 This Internet-Draft will expire on September 13, 2021. 55 Copyright Notice 57 Copyright (c) 2021 IETF Trust and the persons identified as the 58 document authors. All rights reserved. 60 This document is subject to BCP 78 and the IETF Trust's Legal 61 Provisions Relating to IETF Documents 62 (https://trustee.ietf.org/license-info) in effect on the date of 63 publication of this document. Please review these documents 64 carefully, as they describe your rights and restrictions with respect 65 to this document. Code Components extracted from this document must 66 include Simplified BSD License text as described in Section 4.e of 67 the Trust Legal Provisions and are provided without warranty as 68 described in the Simplified BSD License. 70 Table of Contents 72 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 73 1.1. Problem Statement . . . . . . . . . . . . . . . . . . . . 3 74 1.2. Solution requirements . . . . . . . . . . . . . . . . . . 3 75 2. Requirements Language and Terminology . . . . . . . . . . . . 4 76 3. EVPN BGP Attributes Extensions . . . . . . . . . . . . . . . 5 77 4. Solution description . . . . . . . . . . . . . . . . . . . . 6 78 4.1. Use of the Highest-Preference Algorithm . . . . . . . . . 7 79 4.2. Use of the Lowest-Preference Algorithm . . . . . . . . . 9 80 4.3. Use of the Highest-Preference algorithm in [RFC7432] 81 Ethernet Segments . . . . . . . . . . . . . . . . . . . . 9 82 4.4. The Non-Revertive Capability . . . . . . . . . . . . . . 10 83 5. Security Considerations . . . . . . . . . . . . . . . . . . . 14 84 6. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 14 85 7. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 14 86 8. Contributors . . . . . . . . . . . . . . . . . . . . . . . . 15 87 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 15 88 9.1. Normative References . . . . . . . . . . . . . . . . . . 15 89 9.2. Informative References . . . . . . . . . . . . . . . . . 16 90 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 16 92 1. Introduction 94 1.1. Problem Statement 96 [RFC7432] defines the Designated Forwarder (DF) in EVPN networks as 97 the PE responsible for sending broadcast, multicast and unknown 98 unicast traffic (BUM) to a multi-homed device/network in the case of 99 an all-active multi-homing ES or BUM and unicast traffic to a multi- 100 homed device or network in case of single-active multi-homing. The 101 DF is selected out of a candidate list of PEs that advertise the 102 Ethernet Segment Identifier (ESI) to the EVPN network and according 103 to the DF Election Algorithm, or DF Alg as per [RFC8584]. 105 While the Default DF Alg [RFC7432] or HRW [RFC8584] provide an 106 efficient and automated way of selecting the DF across different 107 Ethernet Tags in the ES, there are some use-cases where a more 108 'deterministic' and user-controlled method is required. At the same 109 time, Service Providers require an easy way to force an on-demand DF 110 switchover in order to carry out some maintenance tasks on the 111 existing DF or control whether a new active PE can preempt the 112 existing DF PE. 114 This document proposes a new DF Alg and capability to address the 115 above needs. 117 1.2. Solution requirements 119 The procedures described in this document meet the following 120 requirements: 122 a. The solution provides an administrative preference option so that 123 the user can control in what order the candidate PEs may become 124 DF, assuming they are all operationally ready to take over as DF. 126 b. This extension works for [RFC7432] Ethernet Segments and virtual 127 ES, as defined in [I-D.ietf-bess-evpn-virtual-eth-segment]. 129 c. The user may force a PE to preempt the existing DF for a given 130 Ethernet Tag without re-configuring all the PEs in the ES. 132 d. The solution allows an option to NOT preempt the current DF, even 133 if the former DF PE comes back up after a failure. This is also 134 known as "non-revertive" behavior, as opposed to the [RFC7432] DF 135 election procedures that are always revertive. 137 e. The solution works for single-active and all-active multi-homing 138 Ethernet Segments. 140 2. Requirements Language and Terminology 142 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 143 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 144 "OPTIONAL" in this document are to be interpreted as described in 145 BCP14 [RFC2119] [RFC8174] when, and only when, they appear in all 146 capitals, as shown here. 148 o AC - Attachment Circuit. An AC has an Ethernet Tag associated to 149 it. 151 o BUM - refers to the Broadcast, Unknown unicast and Multicast 152 traffic. 154 o DF, NDF and BDF - Designated Forwarder, Non-Designated Forwarder 155 and Backup Designated Forwarder. 157 o DF Alg or simply Alg - refers to Designated Forwarder Election 158 Algorithm. 160 o HRW - Highest Random Weight, as per [RFC8584]. 162 o ES, vES and ESI - Ethernet Segment, virtual Ethernet Segment and 163 Ethernet Segment Identifier. 165 o EVI - EVPN Instance. 167 o ISID - refers to Service Instance Identifiers in Provider Backbone 168 Bridging (PBB) networks. 170 o MAC-VRF - A Virtual Routing and Forwarding table for Media Access 171 Control (MAC) addresses on a PE. 173 o BD - Broadcast Domain. An EVI may be comprised of one (VLAN-Based 174 or VLAN Bundle services) or multiple (VLAN-Aware Bundle services) 175 Broadcast Domains. 177 o EVC - Ethernet Virtual Circuit. 179 o DP - refers to the "Don't Preempt me" capability in the DF 180 Election extended community. 182 o OAM - refers to Operations And Maintenance protocols. 184 o Ethernet A-D per ES route - refers to [RFC7432] route type 1 or 185 Auto-Discovery per Ethernet Segment route. 187 o Ethernet A-D per EVI route - refers to [RFC7432] route type 1 or 188 Auto-Discovery per EVPN Instance route. 190 o Ethernet Tag - used to represent a Broadcast Domain that is 191 configured on a given ES for the purpose of DF election. Note 192 that any of the following may be used to represent a Broadcast 193 Domain: VIDs (including Q-in-Q tags), configured IDs, VNI (VXLAN 194 Network Identifiers), normalized VID, I-SIDs (Service Instance 195 Identifiers), etc., as long as the representation of the broadcast 196 domains is configured consistently across the multi-homed PEs 197 attached to that ES. The Ethernet Tag value MUST be different 198 from zero. 200 3. EVPN BGP Attributes Extensions 202 This solution reuses and extends the DF Election Extended Community 203 defined in [RFC8584] that is advertised along with the ES route: 205 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 1 206 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 207 | Type=0x06 | Sub-Type(0x06)| RSV | DF Alg | Bitmap ~ 208 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 209 ~ Bitmap | Reserved | DF Preference (2 octets) | 210 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 212 Figure 1: DF Election Extended Community 214 Where the following fields are defined as follows: 216 o DF Alg can have the following values: 218 - Alg 0 - Default DF Election algorithm, or modulus-based 219 algorithm as per [RFC7432]. 221 - Alg 1 - HRW algorithm as per [RFC8584]. 223 - Alg 2 - Highest-Preference algorithm (this document). 225 - Alg TBD - Lowest-Preference algorithm (this document). TBD 226 will be replaced by the allocated value at the time of 227 publication. 229 o Bitmap (2 octets) can have the following values: 231 1 1 1 1 1 1 232 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 233 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 234 |D|A| | 235 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 237 Figure 2: Bitmap field in the DF Election Extended Community 239 - Bit 0 (corresponds to Bit 24 of the DF Election Extended 240 Community and it is defined by this document): D bit or 'Don't 241 Preempt' bit (DP hereafter), determines if the PE advertising 242 the ES route requests the remote PEs in the ES not to preempt 243 it as DF. The default value is DP=0, which is compatible with 244 the 'preempt' or 'revertive' behavior in the Default DF Alg 245 [RFC7432]. The DP capability is supported by Alg 2 and Alg 246 TBD, and MAY be used with DF Alg 0 or 1. The procedures of the 247 DP capability for DF Alg 0 or 1 are out of the scope of this 248 document. 250 - Bit 1: AC-DF or AC-Influenced DF Election, as explained in 251 [RFC8584]. When set to 1, it indicates the desire to use AC- 252 Influenced DF Election with the rest of the PEs in the ES. The 253 AC-DF capability bit MAY be set along with the DP capability 254 and DF Alg 2 or Alg TBD. 256 o DF Preference (defined in this document): defines a 2-octet value 257 that indicates the PE preference to become the DF in the ES. The 258 allowed values are within the range 0-65535, and the default value 259 MUST be 32767. This value is the midpoint in the allowed 260 Preference range of values, which gives the operator the 261 flexibility of choosing a significant number of values, above or 262 below the default Preference. The DF Preference field is specific 263 to DF Alg 2 and DF Alg TBD, and does not represent any Preference 264 value for other Algs. If the DF Alg is different than Alg 2 or 265 Alg TBD, these two octets can be encoded differently. 267 4. Solution description 269 Figure 3 illustrates an example that will be used in the description 270 of the solution. 272 EVPN network 273 +-------------------+ 274 | +-------+ ENNI Aggregation 275 | <---ESI1,500 | PE1 | /\ +----Network---+ 276 | <-----ESI2,100 | |===||=== | 277 | | |===||== \ vES1 | +----+ 278 +-----+ | | \/ |\----------------+CE1 | 279 CE3--+ PE4 | +-------+ | \ ------------+ | 280 +-----+ | | \ / | +----+ 281 | | | X | 282 | <---ESI1,255 +-----+============ \ | 283 | <-----ESI2,200 | PE2 |========== \ vES2 | +----+ 284 | +-----+ | \ ----------+CE2 | 285 | | | --------------+ | 286 | +-----+ ----------------------+ | 287 | <-----ESI2,300 | PE3 +--/ | | +----+ 288 | +-----+ +--------------+ 289 --------------------+ 291 Figure 3: Preference-based DF Election 293 Figure 3 shows three PEs that are connecting EVCs coming from the 294 Aggregation Network to their EVIs in the EVPN network. CE1 is 295 connected to vES1 - that spans PE1 and PE2 - and CE2 is connected to 296 vES2, that is defined in PE1, PE2 and PE3. 298 If the algorithm chosen for vES1 and vES2 is Alg 2 or Alg TBD, i.e., 299 Highest-Preference or Lowest-Preference, the PEs may become DF 300 irrespective of their IP address and based on an administrative 301 Preference value. The following sections provide some examples of 302 the procedures and how they are applied in the use-case of Figure 3. 304 4.1. Use of the Highest-Preference Algorithm 306 Assuming the operator wants to control - in a flexible way - what PE 307 becomes the DF for a given vES and the order in which the PEs become 308 DF in case of multiple failures, the following procedure may be used: 310 a. vES1 and vES2 are now configurable with three optional parameters 311 that are signaled in the DF Election extended community. These 312 parameters are the Preference, Preemption option (or "Don't 313 Preempt Me" option) and DF Alg. We will represent these 314 parameters as (Pref,DP,Alg). Let's assume vES1 is configured as 315 (500,0,Highest-Pref) in PE1, and (255,0,Highest-Pref) in PE2. 316 vES2 is configured as (100,0,Highest-Pref), (200,0,Highest-Pref) 317 and (300,0,Highest-Pref) in PE1, PE2 and PE3 respectively. 319 b. The PEs will advertise an ES route for each vES, including the 3 320 parameters in the DF Election Extended Community. 322 c. According to [RFC8584], each PE will run the DF election 323 algorithm upon expiration of the DF Wait timer. In this case, 324 each PE runs the Highest-Preference DF Alg for each ES as 325 follows: 327 - The PE will check the DF Alg value in each ES route, and 328 assuming all the ES routes are consistent in this DF Alg and 329 the value is 2 (Highest-Preference), the PE will run the 330 procedure in this section. Otherwise, the procedure will fall 331 back to [RFC7432] Default Alg. 333 - In this Highest-Preference Alg, each PE builds a list of 334 candidate PEs, ordered by Preference. E.g. PE1 will build a 335 list of candidate PEs for vES1 ordered by the Preference, from 336 high to low: PE1>PE2. Hence PE1 will become the DF for vES1. 337 In the same way, PE3 becomes the DF for vES2. 339 d. Assuming some maintenance tasks had to be executed on, E.g., PE3, 340 the operator could set vES2's Preference to E.g., 50 so that PE2 341 is forced to take over as DF for vES2 (irrespective of the DP 342 capability). Once the maintenance task on PE3 is over, the 343 operator could decide to leave the existing preference or 344 configure the old preference back. 346 e. In case of equal Preference in two or more PEs in the ES, the DP 347 bit and the lowest IP of the candidate PEs are used as tie- 348 breakers. After selecting the PEs with the highest Preference 349 value, an implementation MUST first select the PE advertising the 350 DP bit set, and then select the PE with the lowest IP address (if 351 the DP bit selection does not yield a unique candidate). The 352 PE's IP address is the address used in the candidate list and it 353 is derived from the Originating Router's IP address of the ES 354 route. Some examples of the use of the DP bit and IP address 355 tie-breakers follow: 357 - If vES1 parameters were (500,0,Highest-Pref) in PE1 and 358 (500,1,Highest-Pref) in PE2, PE2 would be elected due to the 359 DP bit. 361 - If vES1 parameters were (500,0,Highest-Pref) in PE1 and 362 (500,0,Highest-Pref) in PE2, PE1 would be elected, assuming 363 PE1's IP address is lower than PE2's. 365 f. The Preference is an administrative option that MUST be 366 configured on a per-ES basis from the management plane, but MAY 367 also be dynamically changed based on the use of local policies. 368 For instance, on PE1, ES1's Preference can be lowered from 500 to 369 100 in case the bandwidth on the ENNI port is decreased a 50% 370 (that could happen if e.g. the 2-port LAG between PE1 and the 371 Aggregation Network loses one port). Policies MAY also trigger 372 dynamic Preference changes based on the PE's bandwidth 373 availability in the core, specific ports going operationally 374 down, etc. The definition of the actual local policies is out of 375 scope of this document. The default Preference value is 32767. 377 The Highest-Preference Alg MAY be used along with the AC-DF 378 capability. Assuming all the PEs in the ES are configured 379 consistently with Highest-Preference Alg and AC-DF capability, a 380 given PE in the ES is not considered as candidate for DF Election 381 until its corresponding Ethernet A-D per ES and Ethernet A-D per EVI 382 routes are not received, as described in [RFC8584]. 384 The procedures in this document can be used in [RFC7432] based ES or 385 vES as in [I-D.ietf-bess-evpn-virtual-eth-segment], and including 386 EVPN networks as in [RFC8214], [RFC7623] or [RFC8365]. 388 4.2. Use of the Lowest-Preference Algorithm 390 In addition to the Highest-Preference Alg described in Section 4.1 391 this document defines the Lowest-Preference Alg. In this case, and 392 using the example of vES1 in Figure 3, if the Lowest-Preference Alg 393 is configured in all the PEs in the ES, PE2 will be the DF due to its 394 lower Preference. 396 All the procedures described in Section 4.1 apply to the Lowest- 397 Preference Alg, only replacing the Highest-Preference tie-breaker 398 with the Lowest-Preference tie-breaker. The Highest-Preference and 399 Lowest-Preference Algs are different Algs, therefore if two PEs 400 configured for Highest-Preference and Lowest-Preference respectively, 401 are attached to the same ES, the operational DF Election Alg will 402 fall back to the Default Alg. 404 4.3. Use of the Highest-Preference algorithm in [RFC7432] Ethernet 405 Segments 407 While the Highest-Preference (or Lowest-Preference for that matter) 408 DF Alg described in Section 4.1 is typically used in virtual ES 409 scenarios where there is normally an individual Ethernet Tag per vES, 410 the existing [RFC7432] definition of an ES allows potentially up to 411 thousands of Ethernet Tags on the same ES. If this is the case, if 412 Highest-Preference (or Lowest-Preference) Alg is configured in all 413 the PEs of the ES, the same PE will be the elected DF for all the 414 Ethernet Tags of the ES. A potential way to achive a more granular 415 load balancing is decribed below. 417 The ES is configured with an administrative Preference value and 418 E.g., Highest-Preference Alg, but then a range of Ethernet Tags can 419 be defined to use the Lowest-Preference depending on the desired 420 behavior. With this option, the PE will build a list of candidate 421 PEs ordered by Preference, however the DF for a given Ethernet Tag 422 will be determined by the local configuration. 424 For instance: 426 o Assuming ES3 is defined in PE1 and PE2, PE1 may be configured as 427 (500,0,Highest-Preference) for ES3 and PE2 as (100,0,Highest- 428 Preference). 430 o In addition, assuming VLAN-based service interfaces and that the 431 PEs are attached to all Ethernet Tags in the range 1-4000, both 432 PE1 and PE2 will be configured with (Ethernet Tag-range,low), 433 E.g., (2001-4000, low). 435 o This will result in PE1 being DF for Ethernet Tags 1-2000 (since 436 they use the default Highest-Preference Alg) and PE2 being DF for 437 Ethernet Tags 2001-4000, due to the local policy overriding the 438 Highest-Preference Alg. 440 For Ethernet Segments attached to three or more PEs, any other logic 441 that provides a fair distribution of the DF function among the PEs is 442 valid, as long as that logic is consistent in all the PEs in the ES. 443 It is important to note that, when a local policy overrides the 444 Highest-Preference or Lowest-Preference signaled by all the PEs in 445 the ES, this local policy MUST be consistent in all the PEs of the 446 ES. If the local policy is inconsistent for a given Ethernet Tag in 447 the ES, black-holes or packet duplication may occur on that Ethernet 448 Tag. 450 4.4. The Non-Revertive Capability 452 As discussed in Section 1.2 (d), a capability to NOT preempt the 453 existing DF for a given Ethernet Tag is required and therefore added 454 to the DF Election extended community. This option will allow a non- 455 revertive behavior in the DF election. 457 Note that, when a given PE in an ES is taken down for maintenance 458 operations, before bringing it back, the Preference may be changed in 459 order to provide a non-revertive behavior. The DP bit and the 460 mechanism explained in this section will be used for those cases when 461 a former DF comes back up without any controlled maintenance 462 operation, and the non-revertive option is desired in order to avoid 463 service impact. 465 In Figure 3, we assume that based on the Highest-Preference Alg, PE3 466 is the DF for ESI2. 468 If PE3 has a link, EVC or node failure, PE2 would take over as DF. 469 If/when PE3 comes back up again, PE3 will take over, causing some 470 unnecessary packet loss in the ES. 472 The following procedure avoids preemption upon failure recovery 473 (please refer to Figure 3). The procedure supports a non-revertive 474 mode that can be used along with: 476 o Highest-Preference Alg 478 o Highest-Preference Alg, where a local policy overrides the 479 Highest-Preference tie-breaker for a range of Ethernet Tags 481 o Lowest-Preference Alg 483 The procedure is described assuming Highest-Preference Alg in the ES, 484 where local policy overrides the tie-breaker for a given Ethernet 485 Tag, since this is the most complex case. The other two cases above 486 are a sub-set of this one and the differences will be explained 487 later. 489 1. A "Don't Preempt Me" capability is defined on a per-PE/per-ES 490 basis, as described in Section 3. If "Don't Preempt Me" is 491 disabled (default behavior), the advertised DP bit will be 0. If 492 "Don't Preempt Me" is enabled, the ES route will be advertised 493 with DP=1 ("Don't Preempt Me"). All the PEs in an ES SHOULD be 494 consistent in their configuration of the DP capability, however 495 this document does not enforce the consistency across all the 496 PEs. In case of inconsistency in the support of the DP 497 capability in the PEs of the same ES, non-revertive behavior is 498 not guaranteed. However, PEs supporting this capability will 499 still attempt this procedure. 501 2. We assume we want to avoid 'preemption' in all the PEs in the ES, 502 the three PEs are configured with the "Don't Preempt Me" 503 capability. In this example, we assume ESI2 is configured as 504 'DP=enabled' in the three PEs. 506 3. We also assume vES2 is attached to Ethernet Tag-1 and Ethernet 507 Tag-2. vES2 uses Highest-Preference as DF Alg and a local policy 508 is configured in the three PEs to use Lowest-Preference for 509 Ethernet Tag-2. When vES2 is enabled in the three PEs, the PEs 510 will exchange the ES routes and select PE3 as DF for Ethernet 511 Tag-1 (due to the Highest-Preference), and PE1 as DF for Ethernet 512 Tag-2 (due to the Lowest-Preference). 514 4. If PE3's vES2 goes down (due to EVC failure - detected by OAM, or 515 port failure or node failure), PE2 will become the DF for 516 Ethernet Tag-1. No changes will occur for Ethernet Tag-2. 518 5. When PE3's vES2 comes back up, PE3 will start a boot-timer (if 519 booting up) or hold-timer (if the port or EVC recovers). That 520 timer will allow some time for PE3 to receive the ES routes from 521 PE1 and PE2. This timer is applied between the INIT and the 522 DF_WAIT states in the DF Election Finite State Machine described 523 in [RFC8584]. PE3 will then: 525 - Select two "reference-PEs" among the ES routes in the vES, the 526 "Highest-PE" and the "Lowest-PE": 528 * The Highest-PE is the PE with higher Preference, using the 529 DP bit first (with DP=1 being better) and, after that, the 530 lower PE-IP address as tie-breakers. PE3 will select PE2 531 as Highest-PE over PE1, since, when comparing (Pref,DP,PE- 532 IP), (200,1,PE2-IP) wins over (100,1,PE1-IP). 534 * The Lowest-PE is the PE with lower Preference, using the DP 535 bit first (with DP=1 being better) and, after that, the 536 lower PE-IP address as tie-breakers. PE3 will select PE1 537 as Lowest-PE over PE2, since (100,1,PE1-IP) wins over 538 (200,1,PE2-IP). 540 * Note that if there were only one remote PE in the ES, 541 Lowest and Highest PE would be the same PE. 543 - Check its own administrative Pref and compares it with the one 544 of the Highest-PE and Lowest-PE that have DP=1 in their ES 545 routes. Depending on this comparison PE3 will send the ES 546 route with a (Pref,DP) that may be different from its 547 administrative (Pref,DP): 549 * If PE3's Pref value is higher than the Highest-PE's, PE3 550 will send the ES route with an 'in-use' operational Pref 551 equal to the Highest-PE's and DP=0. 553 * If PE3's Pref value is lower than the Lowest-PE's, PE3 will 554 send the ES route with an 'in-use' operational Preference 555 equal to the Lowest-PE's and DP=0. 557 * If PE3's Pref value is neither higher nor lower than the 558 Highest-PE's or the Lowest-PE's respectively, PE3 will send 559 the ES route with its administrative (Pref,DP)=(300,1). 561 * In this example, PE3's administrative Pref=300 is higher 562 than the Highest-PE with DP=1, that is, PE2 (Pref=200). 563 Hence PE3 will inherit PE2's preference and send the ES 564 route with an operational 'in-use' (Pref,DP)=(200,0). 566 - Note that, a PE will always send DP=0 as long as the 567 advertised Pref is the 'in-use' operational Pref (as opposed 568 to the 'administrative' Pref). 570 - This ES route update sent by PE3, with (200,0,PE3-IP), will 571 not cause any DF switchover for any Ethernet Tag. PE2 will 572 continue being DF for Ethernet Tag-1. This is because the DP 573 bit will be used as a tie-breaker in the DF election. That 574 is, if a PE has two candidate PEs with the same Pref, it will 575 pick up the one with DP=1. There are no DF changes for 576 Ethernet Tag-2 either. 578 6. For any subsequent update/withdraw in the ES, the PEs will go 579 through the process described in (5) to select Highest and 580 Lowest-PEs. For instance, if PE2 fails, upon receiving PE2's ES 581 route withdrawal, PE3 and PE1 will go through the selection of 582 new Highest and Lowest-PEs (considering their own active ES 583 route) and then they will run the DF Election. 585 - If a PE selects itself as new Highest or Lowest-PE and it was 586 not before, the PE will then compare its operational 'in-use' 587 Pref with its administrative Pref. If different, the PE will 588 send an ES route update with its administrative Pref and DP 589 values. In the example, PE3 will be the new Highest-PE, 590 therefore it will send an ES route update with 591 (Pref,DP)=(300,1). 593 - After running the DF Election, PE3 will become the new DF for 594 Ethernet Tag-1. No changes will occur for Ethernet Tag-2. 596 If the ES uses Highest-Preference Alg (for all the Ethernet Tags, no 597 local policy), the PEs only need to select the "Highest-PE" as the 598 "reference-PE" (i.e., no need to select the "Lowest-PE"). If the ES 599 uses Lowest-Preference Alg for all the Ethernet Tags, the PEs only 600 need to select the "Lowest-PE" as the "reference-PE". The rest of 601 the procedure remains the same. 603 Note that, irrespective of the DP bit, when a PE or ES comes back and 604 the PE advertises a DF Election Alg different than the one configured 605 in the rest of the PEs in the ES, all the PEs in the ES MUST fall 606 back to the Default [RFC7432] Alg. 608 This document does not modify the use of the P and B bits in the 609 Ethernet A-D per EVI routes [RFC8214] advertised by the PEs in the ES 610 after running the DF Election, irrespective of the revertive or non- 611 revertive behavior in the PE. 613 5. Security Considerations 615 This document describes a DF Election Algorithm that provides 616 absolute control (by configuration) over what PE is the DF for a 617 given Ethernet Tag. While this control is desired in many situations, 618 a malicious user that gets access to the configuration of a PE in the 619 ES may change the behavior of the network. In other DF Algs such as 620 HRW, the DF Election is more automated and cannot be determined by 621 configuration. 623 The non-revertive capability described in this document may be seen 624 as a security improvement over the regular EVPN revertive DF 625 Election: an intentional link (or node) "flapping" on a PE will only 626 cause service disruption once, when the PE goes to NDF state. 628 The document also describes how a local policy can override the 629 Highest-Preference Alg for a range of Ethernet Tags in the ES. If 630 the local policy is not consistent across all PEs in the ES and there 631 is an Ethernet Tag that ends up with an inconsistent use of Highest- 632 Preference or Lowest-Preference in different PEs, black-holing or 633 packet duplication may occur for that Ethernet Tag. 635 6. IANA Considerations 637 This document solicits the allocation of the following values: 639 o DF Alg = 2 in the [RFC8584] "DF Alg" registry, with name "Highest- 640 Preference Algorithm". 642 o DF Alg = TBD in the same "DF Alg" registry, with name "Lowest- 643 Preference Algorithm". 645 o Bit 0 in the [RFC8584] DF Election Capabilities registry, with 646 name "D (Don't Preempt) Capability" for Non-revertive ES. 648 7. Acknowledgments 650 The authors would like to thank Kishore Tiruveedhula for his review 651 and comments. Also thank you to Luc Andre Burdet and Stephane 652 Litkowski for their thorough review and suggestions for a new DF Alg 653 for lowest-preference. 655 8. Contributors 657 In addition to the authors listed, the following individuals also 658 contributed to this document: 660 Kiran Nagaraj, Nokia 662 Vinod Prabhu, Nokia 664 Selvakumar Sivaraj, Juniper 666 Sami Boutros, VMWare 668 9. References 670 9.1. Normative References 672 [RFC7432] Sajassi, A., Ed., Aggarwal, R., Bitar, N., Isaac, A., 673 Uttaro, J., Drake, J., and W. Henderickx, "BGP MPLS-Based 674 Ethernet VPN", RFC 7432, DOI 10.17487/RFC7432, February 675 2015, . 677 [RFC8584] Rabadan, J., Ed., Mohanty, S., Ed., Sajassi, A., Drake, 678 J., Nagaraj, K., and S. Sathappan, "Framework for Ethernet 679 VPN Designated Forwarder Election Extensibility", 680 RFC 8584, DOI 10.17487/RFC8584, April 2019, 681 . 683 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 684 Requirement Levels", BCP 14, RFC 2119, 685 DOI 10.17487/RFC2119, March 1997, 686 . 688 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 689 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 690 May 2017, . 692 [I-D.ietf-bess-evpn-virtual-eth-segment] 693 Sajassi, A., Brissette, P., Schell, R., Drake, J., and J. 694 Rabadan, "EVPN Virtual Ethernet Segment", draft-ietf-bess- 695 evpn-virtual-eth-segment-06 (work in progress), March 696 2020. 698 9.2. Informative References 700 [RFC8214] Boutros, S., Sajassi, A., Salam, S., Drake, J., and J. 701 Rabadan, "Virtual Private Wire Service Support in Ethernet 702 VPN", RFC 8214, DOI 10.17487/RFC8214, August 2017, 703 . 705 [RFC8365] Sajassi, A., Ed., Drake, J., Ed., Bitar, N., Shekhar, R., 706 Uttaro, J., and W. Henderickx, "A Network Virtualization 707 Overlay Solution Using Ethernet VPN (EVPN)", RFC 8365, 708 DOI 10.17487/RFC8365, March 2018, 709 . 711 [RFC7623] Sajassi, A., Ed., Salam, S., Bitar, N., Isaac, A., and W. 712 Henderickx, "Provider Backbone Bridging Combined with 713 Ethernet VPN (PBB-EVPN)", RFC 7623, DOI 10.17487/RFC7623, 714 September 2015, . 716 Authors' Addresses 718 J. Rabadan (editor) 719 Nokia 720 777 Middlefield Road 721 Mountain View, CA 94043 722 USA 724 Email: jorge.rabadan@nokia.com 726 S. Sathappan 727 Nokia 729 Email: senthil.sathappan@nokia.com 731 T. Przygienda 732 Juniper Networks 734 Email: prz@juniper.net 736 W. Lin 737 Juniper Networks 739 Email: wlin@juniper.net 740 J. Drake 741 Juniper Networks 743 Email: jdrake@juniper.net 745 A. Sajassi 746 Cisco Systems 748 Email: sajassi@cisco.com 750 S. Mohanty 751 Cisco Systems 753 Email: satyamoh@cisco.com