idnits 2.17.1 draft-ietf-idr-rfc7752bis-04.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 has examples using IPv4 documentation addresses according to RFC6890, but does not use any IPv6 documentation addresses. Maybe there should be IPv6 examples, too? Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (September 4, 2020) is 1301 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) -- Possible downref: Non-RFC (?) normative reference: ref. 'ISO10589' ** Obsolete normative reference: RFC 7752 (Obsoleted by RFC 9552) == Outdated reference: A later version (-13) exists of draft-ietf-idr-bgpls-inter-as-topology-ext-08 -- Obsolete informational reference (is this intentional?): RFC 5316 (Obsoleted by RFC 9346) Summary: 1 error (**), 0 flaws (~~), 2 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Inter-Domain Routing K. Talaulikar, Ed. 3 Internet-Draft Cisco Systems 4 Obsoletes: 7752 (if approved) September 4, 2020 5 Intended status: Standards Track 6 Expires: March 8, 2021 8 Distribution of Link-State and Traffic Engineering Information Using BGP 9 draft-ietf-idr-rfc7752bis-04 11 Abstract 13 In a number of environments, a component external to a network is 14 called upon to perform computations based on the network topology and 15 current state of the connections within the network, including 16 Traffic Engineering (TE) information. This is information typically 17 distributed by IGP routing protocols within the network. 19 This document describes a mechanism by which link-state and TE 20 information can be collected from networks and shared with external 21 components using the BGP routing protocol. This is achieved using a 22 new BGP Network Layer Reachability Information (NLRI) encoding 23 format. The mechanism is applicable to physical and virtual IGP 24 links. The mechanism described is subject to policy control. 26 Applications of this technique include Application-Layer Traffic 27 Optimization (ALTO) servers and Path Computation Elements (PCEs). 29 This document obsoletes RFC 7752 by completely replacing that 30 document. It makes a number of small changes and clarifications to 31 the previous specification. 33 Status of This Memo 35 This Internet-Draft is submitted in full conformance with the 36 provisions of BCP 78 and BCP 79. 38 Internet-Drafts are working documents of the Internet Engineering 39 Task Force (IETF). Note that other groups may also distribute 40 working documents as Internet-Drafts. The list of current Internet- 41 Drafts is at https://datatracker.ietf.org/drafts/current/. 43 Internet-Drafts are draft documents valid for a maximum of six months 44 and may be updated, replaced, or obsoleted by other documents at any 45 time. It is inappropriate to use Internet-Drafts as reference 46 material or to cite them other than as "work in progress." 48 This Internet-Draft will expire on March 8, 2021. 50 Copyright Notice 52 Copyright (c) 2020 IETF Trust and the persons identified as the 53 document authors. All rights reserved. 55 This document is subject to BCP 78 and the IETF Trust's Legal 56 Provisions Relating to IETF Documents 57 (https://trustee.ietf.org/license-info) in effect on the date of 58 publication of this document. Please review these documents 59 carefully, as they describe your rights and restrictions with respect 60 to this document. Code Components extracted from this document must 61 include Simplified BSD License text as described in Section 4.e of 62 the Trust Legal Provisions and are provided without warranty as 63 described in the Simplified BSD License. 65 Table of Contents 67 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 68 1.1. Requirements Language . . . . . . . . . . . . . . . . . . 5 69 2. Motivation and Applicability . . . . . . . . . . . . . . . . 5 70 2.1. MPLS-TE with PCE . . . . . . . . . . . . . . . . . . . . 5 71 2.2. ALTO Server Network API . . . . . . . . . . . . . . . . . 7 72 3. BGP Speaker Roles for BGP-LS . . . . . . . . . . . . . . . . 7 73 4. Carrying Link-State Information in BGP . . . . . . . . . . . 9 74 4.1. TLV Format . . . . . . . . . . . . . . . . . . . . . . . 9 75 4.2. The Link-State NLRI . . . . . . . . . . . . . . . . . . . 10 76 4.2.1. Node Descriptors . . . . . . . . . . . . . . . . . . 14 77 4.2.2. Link Descriptors . . . . . . . . . . . . . . . . . . 18 78 4.2.3. Prefix Descriptors . . . . . . . . . . . . . . . . . 21 79 4.3. The BGP-LS Attribute . . . . . . . . . . . . . . . . . . 23 80 4.3.1. Node Attribute TLVs . . . . . . . . . . . . . . . . . 23 81 4.3.2. Link Attribute TLVs . . . . . . . . . . . . . . . . . 27 82 4.3.3. Prefix Attribute TLVs . . . . . . . . . . . . . . . . 32 83 4.4. Private Use . . . . . . . . . . . . . . . . . . . . . . . 35 84 4.5. BGP Next-Hop Information . . . . . . . . . . . . . . . . 36 85 4.6. Inter-AS Links . . . . . . . . . . . . . . . . . . . . . 36 86 4.7. Handling of Unreachable IGP Nodes . . . . . . . . . . . . 37 87 4.8. Router-ID Anchoring Example: ISO Pseudonode . . . . . . . 38 88 4.9. Router-ID Anchoring Example: OSPF Pseudonode . . . . . . 39 89 4.10. Router-ID Anchoring Example: OSPFv2 to IS-IS Migration . 40 90 5. Link to Path Aggregation . . . . . . . . . . . . . . . . . . 41 91 5.1. Example: No Link Aggregation . . . . . . . . . . . . . . 41 92 5.2. Example: ASBR to ASBR Path Aggregation . . . . . . . . . 42 93 5.3. Example: Multi-AS Path Aggregation . . . . . . . . . . . 42 94 6. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 43 95 6.1. Guidance for Designated Experts . . . . . . . . . . . . . 44 96 7. Manageability Considerations . . . . . . . . . . . . . . . . 44 97 7.1. Operational Considerations . . . . . . . . . . . . . . . 44 98 7.1.1. Operations . . . . . . . . . . . . . . . . . . . . . 44 99 7.1.2. Installation and Initial Setup . . . . . . . . . . . 45 100 7.1.3. Migration Path . . . . . . . . . . . . . . . . . . . 45 101 7.1.4. Requirements on Other Protocols and Functional 102 Components . . . . . . . . . . . . . . . . . . . . . 45 103 7.1.5. Impact on Network Operation . . . . . . . . . . . . . 45 104 7.1.6. Verifying Correct Operation . . . . . . . . . . . . . 45 105 7.2. Management Considerations . . . . . . . . . . . . . . . . 46 106 7.2.1. Management Information . . . . . . . . . . . . . . . 46 107 7.2.2. Fault Management . . . . . . . . . . . . . . . . . . 46 108 7.2.3. Configuration Management . . . . . . . . . . . . . . 48 109 7.2.4. Accounting Management . . . . . . . . . . . . . . . . 49 110 7.2.5. Performance Management . . . . . . . . . . . . . . . 49 111 7.2.6. Security Management . . . . . . . . . . . . . . . . . 49 112 8. TLV/Sub-TLV Code Points Summary . . . . . . . . . . . . . . . 49 113 9. Security Considerations . . . . . . . . . . . . . . . . . . . 51 114 10. Contributors . . . . . . . . . . . . . . . . . . . . . . . . 52 115 11. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 52 116 12. References . . . . . . . . . . . . . . . . . . . . . . . . . 53 117 12.1. Normative References . . . . . . . . . . . . . . . . . . 53 118 12.2. Informative References . . . . . . . . . . . . . . . . . 56 119 Appendix A. Changes from RFC 7752 . . . . . . . . . . . . . . . 57 120 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 59 122 1. Introduction 124 The contents of a Link-State Database (LSDB) or of an IGP's Traffic 125 Engineering Database (TED) describe only the links and nodes within 126 an IGP area. Some applications, such as end-to-end Traffic 127 Engineering (TE), would benefit from visibility outside one area or 128 Autonomous System (AS) in order to make better decisions. 130 The IETF has defined the Path Computation Element (PCE) [RFC4655] as 131 a mechanism for achieving the computation of end-to-end TE paths that 132 cross the visibility of more than one TED or that require CPU- 133 intensive or coordinated computations. The IETF has also defined the 134 ALTO server [RFC5693] as an entity that generates an abstracted 135 network topology and provides it to network-aware applications. 137 Both a PCE and an ALTO server need to gather information about the 138 topologies and capabilities of the network in order to be able to 139 fulfill their function. 141 This document describes a mechanism by which link-state and TE 142 information can be collected from networks and shared with external 143 components using the BGP routing protocol [RFC4271]. This is 144 achieved using a new BGP Network Layer Reachability Information 145 (NLRI) encoding format. The mechanism is applicable to physical and 146 virtual links. The mechanism described is subject to policy control. 148 A router maintains one or more databases for storing link-state 149 information about nodes and links in any given area. Link attributes 150 stored in these databases include: local/remote IP addresses, local/ 151 remote interface identifiers, link metric and TE metric, link 152 bandwidth, reservable bandwidth, per Class-of-Service (CoS) class 153 reservation state, preemption, and Shared Risk Link Groups (SRLGs). 154 The router's BGP process can retrieve topology from these LSDBs and 155 distribute it to a consumer, either directly or via a peer BGP 156 speaker (typically a dedicated Route Reflector), using the encoding 157 specified in this document. 159 An illustration of the collection of link-state and TE information 160 and its distribution to consumers is shown in the Figure 1 below. 162 +-----------+ 163 | Consumer | 164 +-----------+ 165 ^ 166 | 167 +-----------+ +-----------+ 168 | BGP | | BGP | 169 | Speaker |<----------->| Speaker | +-----------+ 170 | RR1 | | RRm | | Consumer | 171 +-----------+ +-----------+ +-----------+ 172 ^ ^ ^ ^ 173 | | | | 174 +-----+ +---------+ +---------+ | 175 | | | | 176 +-----------+ +-----------+ +-----------+ 177 | BGP | | BGP | | BGP | 178 | Speaker | | Speaker | . . . | Speaker | 179 | R1 | | R2 | | Rn | 180 +-----------+ +-----------+ +-----------+ 181 ^ ^ ^ 182 | | | 183 IGP IGP IGP 185 Figure 1: Collection of Link-State and TE Information 187 A BGP speaker may apply configurable policy to the information that 188 it distributes. Thus, it may distribute the real physical topology 189 from the LSDB or the TED. Alternatively, it may create an abstracted 190 topology, where virtual, aggregated nodes are connected by virtual 191 paths. Aggregated nodes can be created, for example, out of multiple 192 routers in a Point of Presence (POP). Abstracted topology can also 193 be a mix of physical and virtual nodes and physical and virtual 194 links. Furthermore, the BGP speaker can apply policy to determine 195 when information is updated to the consumer so that there is a 196 reduction of information flow from the network to the consumers. 197 Mechanisms through which topologies can be aggregated or virtualized 198 are outside the scope of this document. 200 This document obsoletes [RFC7752] by completely replacing that 201 document. It makes a number of small changes and clarifications to 202 the previous specification as documented in Appendix A. 204 1.1. Requirements Language 206 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 207 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 208 "OPTIONAL" in this document are to be interpreted as described in BCP 209 14 [RFC2119] [RFC8174] when, and only when, they appear in all 210 capitals, as shown here. 212 2. Motivation and Applicability 214 This section describes use cases from which the requirements can be 215 derived. 217 2.1. MPLS-TE with PCE 219 As described in [RFC4655], a PCE can be used to compute MPLS-TE paths 220 within a "domain" (such as an IGP area) or across multiple domains 221 (such as a multi-area AS or multiple ASes). 223 o Within a single area, the PCE offers enhanced computational power 224 that may not be available on individual routers, sophisticated 225 policy control and algorithms, and coordination of computation 226 across the whole area. 228 o If a router wants to compute a MPLS-TE path across IGP areas, then 229 its own TED lacks visibility of the complete topology. That means 230 that the router cannot determine the end-to-end path and cannot 231 even select the right exit router (Area Border Router (ABR)) for 232 an optimal path. This is an issue for large-scale networks that 233 need to segment their core networks into distinct areas but still 234 want to take advantage of MPLS-TE. 236 Previous solutions used per-domain path computation [RFC5152]. The 237 source router could only compute the path for the first area because 238 the router only has full topological visibility for the first area 239 along the path, but not for subsequent areas. Per-domain path 240 computation uses a technique called "loose-hop-expansion" [RFC3209] 241 and selects the exit ABR and other ABRs or AS Border Routers (ASBRs) 242 using the IGP-computed shortest path topology for the remainder of 243 the path. This may lead to sub-optimal paths, makes alternate/back- 244 up path computation hard, and might result in no TE path being found 245 when one really does exist. 247 The PCE presents a computation server that may have visibility into 248 more than one IGP area or AS, or may cooperate with other PCEs to 249 perform distributed path computation. The PCE obviously needs access 250 to the TED for the area(s) it serves, but [RFC4655] does not describe 251 how this is achieved. Many implementations make the PCE a passive 252 participant in the IGP so that it can learn the latest state of the 253 network, but this may be sub-optimal when the network is subject to a 254 high degree of churn or when the PCE is responsible for multiple 255 areas. 257 The following figure shows how a PCE can get its TED information 258 using the mechanism described in this document. 260 +----------+ +---------+ 261 | ----- | | BGP | 262 | | TED |<-+-------------------------->| Speaker | 263 | ----- | TED synchronization | | 264 | | | mechanism: +---------+ 265 | | | BGP with Link-State NLRI 266 | v | 267 | ----- | 268 | | PCE | | 269 | ----- | 270 +----------+ 271 ^ 272 | Request/ 273 | Response 274 v 275 Service +----------+ Signaling +----------+ 276 Request | Head-End | Protocol | Adjacent | 277 -------->| Node |<------------>| Node | 278 +----------+ +----------+ 280 Figure 2: External PCE Node Using a TED Synchronization Mechanism 282 The mechanism in this document allows the necessary TED information 283 to be collected from the IGP within the network, filtered according 284 to configurable policy, and distributed to the PCE as necessary. 286 2.2. ALTO Server Network API 288 An ALTO server [RFC5693] is an entity that generates an abstracted 289 network topology and provides it to network-aware applications over a 290 web-service-based API. Example applications are peer-to-peer (P2P) 291 clients or trackers, or Content Distribution Networks (CDNs). The 292 abstracted network topology comes in the form of two maps: a Network 293 Map that specifies allocation of prefixes to Partition Identifiers 294 (PIDs), and a Cost Map that specifies the cost between PIDs listed in 295 the Network Map. For more details, see [RFC7285]. 297 ALTO abstract network topologies can be auto-generated from the 298 physical topology of the underlying network. The generation would 299 typically be based on policies and rules set by the operator. Both 300 prefix and TE data are required: prefix data is required to generate 301 ALTO Network Maps, and TE (topology) data is required to generate 302 ALTO Cost Maps. Prefix data is carried and originated in BGP, and TE 303 data is originated and carried in an IGP. The mechanism defined in 304 this document provides a single interface through which an ALTO 305 server can retrieve all the necessary prefix and network topology 306 data from the underlying network. Note that an ALTO server can use 307 other mechanisms to get network data, for example, peering with 308 multiple IGP and BGP speakers. 310 The following figure shows how an ALTO server can get network 311 topology information from the underlying network using the mechanism 312 described in this document. 314 +--------+ 315 | Client |<--+ 316 +--------+ | 317 | ALTO +--------+ BGP with +---------+ 318 +--------+ | Protocol | ALTO | Link-State NLRI | BGP | 319 | Client |<--+------------| Server |<----------------| Speaker | 320 +--------+ | | | | | 321 | +--------+ +---------+ 322 +--------+ | 323 | Client |<--+ 324 +--------+ 326 Figure 3: ALTO Server Using Network Topology Information 328 3. BGP Speaker Roles for BGP-LS 330 In the illustration shown in Figure 1, the BGP Speakers can be seen 331 playing different roles in the distribution of information using BGP- 332 LS. This section introduces terms that explain the different roles 333 of the BGP Speakers which are then used through the rest of this 334 document. 336 o BGP-LS Producer: The BGP Speakers R1, R2, ... Rn, originate link- 337 state information from their underlying link-state IGP protocols 338 into BGP-LS. If R1 and R2 are in the same IGP area, then likely 339 they are originating the same link-state information into BGP-LS. 340 R1 may also source information from sources other than IGP, e.g. 341 its local node information. The term BGP-LS Producer refers to 342 the BGP Speaker that is originating link-state information into 343 BGP. 345 o BGP-LS Consumer: The BGP Speakers RR1 and Rn are handing off the 346 BGP-LS information that they have collected to a consumer 347 application. The BGP protocol implementation and the consumer 348 application may be on the same or different nodes. The term BGP- 349 LS Consumer refers to the consumer application/process and not the 350 BGP Speaker. This document only covers the BGP implementation. 351 The consumer application and the design of interface between BGP 352 and consumer application may be implementation specific and 353 outside the scope of this document. 355 o BGP-LS Propagator: The BGP Speaker RRm propagates the BGP-LS 356 information between the BGP Speaker Rn and the BGP Speaker RR1. 357 The BGP implementation on RRm is doing the propagation of BGP-LS 358 updates and performing BGP best path calculations. Similarly, the 359 BGP Speaker RR1 is receiving BGP-LS information from R1, R2 and 360 RRm and propagating the information to the BGP-LS Consumer after 361 performing BGP best path calculations. The term BGP-LS Propagator 362 refers to the BGP Speaker that is performing BGP protocol 363 processing on the link-state information. 365 The above roles are not mutually exclusive. The same BGP Speaker may 366 be the producer for some link-state information and propagator for 367 some other link-state information while also providing this 368 information to a consumer application. Nothing precludes a BGP 369 implementation performing some of the validation and processing on 370 behalf of the BGP-LS Consumer as long as it does not impact the 371 semantics of its role as BGP-LS Propagator as described in this 372 document. 374 The rest of this document refers to the role when describing 375 procedures that are specific to that role. When the role is not 376 specified, then the said procedure applies to all BGP Speakers. 378 4. Carrying Link-State Information in BGP 380 This specification contains two parts: definition of a new BGP NLRI 381 that describes links, nodes, and prefixes comprising IGP link-state 382 information and definition of a new BGP path attribute (BGP-LS 383 Attribute) that carries link, node, and prefix properties and 384 attributes, such as the link and prefix metric or auxiliary Router- 385 IDs of nodes, etc. 387 It is desirable to keep the dependencies on the protocol source of 388 this attribute to a minimum and represent any content in an IGP- 389 neutral way, such that applications that want to learn about a link- 390 state topology do not need to know about any OSPF or IS-IS protocol 391 specifics. 393 This section mainly describes the procedures at a BGP-LS Producer 394 that originate link-state information into BGP-LS. 396 4.1. TLV Format 398 Information in the new Link-State NLRIs and the BGP-LS Attribute is 399 encoded in Type/Length/Value triplets. The TLV format is shown in 400 Figure 4 and applies to both the NLRI and the BGP-LS Attribute 401 encodings. 403 0 1 2 3 404 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 1 405 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 406 | Type | Length | 407 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 408 // Value (variable) // 409 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 411 Figure 4: TLV Format 413 The Length field defines the length of the value portion in octets 414 (thus, a TLV with no value portion would have a length of zero). The 415 TLV is not padded to 4-octet alignment. Unknown and unsupported 416 types MUST be preserved and propagated within both the NLRI and the 417 BGP-LS Attribute. The presence of unrecognized or unexpected TLVs 418 MUST NOT result in the NLRI or the BGP-LS Attribute being considered 419 as malformed. 421 In order to compare NLRIs with unknown TLVs, all TLVs within the NLRI 422 MUST be ordered in ascending order by TLV Type. If there are 423 multiple TLVs of the same type within a single NLRI, then the TLVs 424 sharing the same type MUST be in ascending order based on the value 425 field. Comparison of the value fields is performed by treating the 426 entire field as an opaque hexadecimal string. Standard string 427 comparison rules apply. NLRIs having TLVs which do not follow the 428 above ordering rules MUST be considered as malformed by a BGP-LS 429 Propagator. This ensures that multiple copies of the same NLRI from 430 multiple BGP-LS Producers and the ambiguity arising there from is 431 prevented. 433 All TLVs within the NLRI that are not specified as mandatory are 434 considered optional. All TLVs within the BGP-LS Attribute are 435 considered optional unless specified otherwise. 437 The TLVs within the BGP-LS Attribute MAY be ordered in ascending 438 order by TLV type. BGP-LS Attribute with unordered TLVs MUST NOT be 439 considered malformed. 441 4.2. The Link-State NLRI 443 The MP_REACH_NLRI and MP_UNREACH_NLRI attributes are BGP's containers 444 for carrying opaque information. This specification defines three 445 Link-State NLRI types that describes either a node, a link, and a 446 prefix. 448 All non-VPN link, node, and prefix information SHALL be encoded using 449 AFI 16388 / SAFI 71. VPN link, node, and prefix information SHALL be 450 encoded using AFI 16388 / SAFI 72. 452 In order for two BGP speakers to exchange Link-State NLRI, they MUST 453 use BGP Capabilities Advertisement to ensure that they are both 454 capable of properly processing such NLRI. This is done as specified 455 in [RFC4760], by using capability code 1 (multi-protocol BGP), with 456 AFI 16388 / SAFI 71 for BGP-LS, and AFI 16388 / SAFI 72 for 457 BGP-LS-VPN. 459 New Link-State NLRI Types may be introduced in the future. Since 460 supported NLRI type values within the address family are not 461 expressed in the Multiprotocol BGP (MP-BGP) capability [RFC4760], it 462 is possible that a BGP speaker has advertised support for Link-State 463 but does not support a particular Link-State NLRI type. In order to 464 allow introduction of new Link-State NLRI types seamlessly in the 465 future, without the need for upgrading all BGP speakers in the 466 propagation path (e.g. a route reflector), this document deviates 467 from the default handling behavior specified by [RFC7606] for Link- 468 State address-family. An implementation MUST handle unrecognized 469 Link-State NLRI types as opaque objects and MUST preserve and 470 propagate them. 472 The format of the Link-State NLRI is shown in the following figures. 474 0 1 2 3 475 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 1 476 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 477 | NLRI Type | Total NLRI Length | 478 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 479 | | 480 // Link-State NLRI (variable) // 481 | | 482 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 484 Figure 5: Link-State AFI 16388 / SAFI 71 NLRI Format 486 0 1 2 3 487 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 1 488 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 489 | NLRI Type | Total NLRI Length | 490 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 491 | | 492 + Route Distinguisher + 493 | | 494 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 495 | | 496 // Link-State NLRI (variable) // 497 | | 498 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 500 Figure 6: Link-State VPN AFI 16388 / SAFI 72 NLRI Format 502 The Total NLRI Length field contains the cumulative length, in 503 octets, of the rest of the NLRI, not including the NLRI Type field or 504 itself. For VPN applications, it also includes the length of the 505 Route Distinguisher. 507 +-------------+---------------------------+ 508 | Type | NLRI Type | 509 +-------------+---------------------------+ 510 | 1 | Node NLRI | 511 | 2 | Link NLRI | 512 | 3 | IPv4 Topology Prefix NLRI | 513 | 4 | IPv6 Topology Prefix NLRI | 514 | 65000-65535 | Private Use | 515 +-------------+---------------------------+ 517 Table 1: NLRI Types 519 Route Distinguishers are defined and discussed in [RFC4364]. 521 The Node NLRI (NLRI Type = 1) is shown in the following figure. 523 0 1 2 3 524 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 1 525 +-+-+-+-+-+-+-+-+ 526 | Protocol-ID | 527 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 528 | Identifier | 529 | (64 bits) | 530 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 531 // Local Node Descriptors (variable) // 532 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 534 Figure 7: The Node NLRI Format 536 The Link NLRI (NLRI Type = 2) is shown in the following figure. 538 0 1 2 3 539 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 1 540 +-+-+-+-+-+-+-+-+ 541 | Protocol-ID | 542 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 543 | Identifier | 544 | (64 bits) | 545 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 546 // Local Node Descriptors (variable) // 547 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 548 // Remote Node Descriptors (variable) // 549 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 550 // Link Descriptors (variable) // 551 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 553 Figure 8: The Link NLRI Format 555 The IPv4 and IPv6 Prefix NLRIs (NLRI Type = 3 and Type = 4) use the 556 same format, as shown in the following figure. 558 0 1 2 3 559 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 1 560 +-+-+-+-+-+-+-+-+ 561 | Protocol-ID | 562 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 563 | Identifier | 564 | (64 bits) | 565 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 566 // Local Node Descriptors (variable) // 567 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 568 // Prefix Descriptors (variable) // 569 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 571 Figure 9: The IPv4/IPv6 Topology Prefix NLRI Format 573 The Protocol-ID field can contain one of the following values: 575 +-------------+----------------------------------+ 576 | Protocol-ID | NLRI information source protocol | 577 +-------------+----------------------------------+ 578 | 1 | IS-IS Level 1 | 579 | 2 | IS-IS Level 2 | 580 | 3 | OSPFv2 | 581 | 4 | Direct | 582 | 5 | Static configuration | 583 | 6 | OSPFv3 | 584 | 200-255 | Private Use | 585 +-------------+----------------------------------+ 587 Table 2: Protocol Identifiers 589 The 'Direct' and 'Static configuration' protocol types SHOULD be used 590 when BGP-LS is sourcing local information. For all information 591 derived from other protocols, the corresponding Protocol-ID MUST be 592 used. If BGP-LS has direct access to interface information and wants 593 to advertise a local link, then the Protocol-ID 'Direct' SHOULD be 594 used. For modeling virtual links, such as described in Section 5, 595 the Protocol-ID 'Static configuration' SHOULD be used. 597 A router MAY run multiple protocol instances of OSPF or ISIS where by 598 it becomes a border router between multiple IGP domains. Both OSPF 599 and IS-IS MAY also run multiple routing protocol instances over the 600 same link. See [RFC8202] and [RFC6549]. These instances define 601 independent IGP routing domains. The 64-bit Identifier field carries 602 a BGP-LS Instance Identifier (Instance-ID) that is used to identify 603 the IGP routing domain where the NLRI belongs. The NLRIs 604 representing link-state objects (nodes, links, or prefixes) from the 605 same IGP routing instance MUST have the same Identifier field value. 607 NLRIs with different Identifier field values MUST be considered to be 608 from different IGP routing instances. The Identifier field value 0 609 is RECOMMENDED to be used when there is only a single protocol 610 instance in the network where BGP-LS is operational. 612 An implementation which supports multiple IGP instances MUST support 613 the configuration of unique BGP-LS Instance-IDs at the routing 614 protocol instance level. The network operator MUST assign consistent 615 BGP-LS Instance-ID values on all BGP-LS Producers within a given IGP 616 domain. Unique BGP-LS Instance-ID values MUST be assigned to routing 617 protocol instances operating in different IGP domains. This allows 618 the BGP-LS Consumer to build an accurate segregated multi-domain 619 topology based on the Identifier field even when the topology is 620 advertised via BGP-LS by multiple BGP-LS Producers in the network. 622 When the above described semantics and recommendations are not 623 followed, a BGP-LS Consumer may see duplicate link-state objects for 624 the same node, link or prefix when there are multiple BGP-LS 625 Producers deployed. This may also result in the BGP-LS Consumers 626 getting an inaccurate network-wide topology. 628 When adding, removing or modifying a TLV/sub-TLV from a Link-State 629 NLRI, the BGP-LS Producer MUST withdraw the old NLRI by including it 630 in the MP_UNREACH_NLRI. Not doing so can result in duplicate and in- 631 consistent link-state objects hanging around in the BGP-LS table. 633 Each Node Descriptor, Link Descriptor and Prefix Descriptor consists 634 of one or more TLVs, as described in the following sections. These 635 Descriptor TLVs are applicable for the Node, Link and Prefix NLRI 636 Types for the protocols listed in Table 2. Documents extending BGP- 637 LS specifications with new NLRI Types and/or protocols MUST specify 638 the NLRI Descriptors for them. 640 4.2.1. Node Descriptors 642 Each link is anchored by a pair of Router-IDs that are used by the 643 underlying IGP, namely, a 48-bit ISO System-ID for IS-IS and a 32-bit 644 Router-ID for OSPFv2 and OSPFv3. An IGP may use one or more 645 additional auxiliary Router-IDs, mainly for Traffic Engineering 646 purposes. For example, IS-IS may have one or more IPv4 and IPv6 TE 647 Router-IDs [RFC5305] [RFC6119]. These auxiliary Router-IDs MUST be 648 included in the node attribute described in Section 4.3.1 and MAY be 649 included in link attribute described in Section 4.3.2. The 650 advertisement of the TE Router-IDs help a BGP-LS Consumer to 651 correlate multiple link-state objects (e.g. in different IGP 652 instances or areas/levels) to the same node in the network. 654 It is desirable that the Router-ID assignments inside the Node 655 Descriptor are globally unique. However, there may be Router-ID 656 spaces (e.g., ISO) where no global registry exists, or worse, Router- 657 IDs have been allocated following the private-IP allocation described 658 in RFC 1918 [RFC1918]. BGP-LS uses the Autonomous System (AS) Number 659 to disambiguate the Router-IDs, as described in Section 4.2.1.1. 661 4.2.1.1. Globally Unique Node/Link/Prefix Identifiers 663 One problem that needs to be addressed is the ability to identify an 664 IGP node globally (by "globally", we mean within the BGP-LS database 665 collected by all BGP-LS speakers that talk to each other). This can 666 be expressed through the following two requirements: 668 (A) The same node MUST NOT be represented by two keys (otherwise, 669 one node will look like two nodes). 671 (B) Two different nodes MUST NOT be represented by the same key 672 (otherwise, two nodes will look like one node). 674 We define an "IGP domain" to be the set of nodes (hence, by extension 675 links and prefixes) within which each node has a unique IGP 676 representation by using the combination of Area-ID, Router-ID, 677 Protocol-ID, Multi-Topology ID, and Instance-ID. The problem is that 678 BGP may receive node/link/prefix information from multiple 679 independent "IGP domains", and we need to distinguish between them. 680 Moreover, we can't assume there is always one and only one IGP domain 681 per AS. During IGP transitions, it may happen that two redundant 682 IGPs are in place. 684 The mapping of the Instance-ID to the Identifier field as described 685 earlier along with a set of sub-TLVs described in Section 4.2.1.4, 686 allows specification of a flexible key for any given node/link 687 information such that global uniqueness of the NLRI is ensured. 689 4.2.1.2. Local Node Descriptors 691 The Local Node Descriptors TLV contains Node Descriptors for the node 692 anchoring the local end of the link. This is a mandatory TLV in all 693 three types of NLRIs (node, link, and prefix). The Type is 256. The 694 length of this TLV is variable. The value contains one or more Node 695 Descriptor Sub-TLVs defined in Section 4.2.1.4. 697 0 1 2 3 698 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 1 699 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 700 | Type | Length | 701 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 702 | | 703 // Node Descriptor Sub-TLVs (variable) // 704 | | 705 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 707 Figure 10: Local Node Descriptors TLV Format 709 4.2.1.3. Remote Node Descriptors 711 The Remote Node Descriptors TLV contains Node Descriptors for the 712 node anchoring the remote end of the link. This is a mandatory TLV 713 for Link NLRIs. The type is 257. The length of this TLV is 714 variable. The value contains one or more Node Descriptor Sub-TLVs 715 defined in Section 4.2.1.4. 717 0 1 2 3 718 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 1 719 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 720 | Type | Length | 721 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 722 | | 723 // Node Descriptor Sub-TLVs (variable) // 724 | | 725 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 727 Figure 11: Remote Node Descriptors TLV Format 729 4.2.1.4. Node Descriptor Sub-TLVs 731 The Node Descriptor Sub-TLV type code points and lengths are listed 732 in the following table: 734 +--------------------+--------------------------------+----------+ 735 | Sub-TLV Code Point | Description | Length | 736 +--------------------+--------------------------------+----------+ 737 | 512 | Autonomous System | 4 | 738 | 513 | BGP-LS Identifier (deprecated) | 4 | 739 | 514 | OSPF Area-ID | 4 | 740 | 515 | IGP Router-ID | Variable | 741 +--------------------+--------------------------------+----------+ 743 Table 3: Node Descriptor Sub-TLVs 745 The sub-TLV values in Node Descriptor TLVs are defined as follows: 747 Autonomous System: Opaque value (32-bit AS Number). This is an 748 optional TLV. The value SHOULD be set to the AS Number associated 749 with the BGP process originating the link-state information. An 750 implementation MAY provide a configuration option on the BGP-LS 751 Producer to use a value different. 753 BGP-LS Identifier: Opaque value (32-bit ID). This is an optional 754 TLV. In conjunction with Autonomous System Number (ASN), uniquely 755 identifies the BGP-LS domain. The combination of ASN and BGP-LS 756 ID MUST be globally unique. All BGP-LS speakers within an IGP 757 flooding-set (set of IGP nodes within which an LSP/LSA is flooded) 758 MUST use the same ASN, BGP-LS ID tuple. If an IGP domain consists 759 of multiple flooding-sets, then all BGP-LS speakers within the IGP 760 domain SHOULD use the same ASN, BGP-LS ID tuple. 762 Area-ID: Used to identify the 32-bit area to which the information 763 advertised in the NLRI belongs. This is a mandatory TLV when 764 originating information from OSPF that is derived from area-scope 765 LSAs. The Area Identifier allows different NLRIs of the same 766 router to be discriminated on a per area basis. It is not used 767 for NLRIs when carrying information that is derived from AS-scope 768 LSAs as it is not associated with a specific area. 770 IGP Router-ID: Opaque value. This is a mandatory TLV when 771 originating information from IS-IS, OSPF, direct or static. For 772 an IS-IS non-pseudonode, this contains a 6-octet ISO Node-ID (ISO 773 system-ID). For an IS-IS pseudonode corresponding to a LAN, this 774 contains the 6-octet ISO Node-ID of the Designated Intermediate 775 System (DIS) followed by a 1-octet, nonzero PSN identifier (7 776 octets in total). For an OSPFv2 or OSPFv3 non-pseudonode, this 777 contains the 4-octet Router-ID. For an OSPFv2 pseudonode 778 representing a LAN, this contains the 4-octet Router-ID of the 779 Designated Router (DR) followed by the 4-octet IPv4 address of the 780 DR's interface to the LAN (8 octets in total). Similarly, for an 781 OSPFv3 pseudonode, this contains the 4-octet Router-ID of the DR 782 followed by the 4-octet interface identifier of the DR's interface 783 to the LAN (8 octets in total). The TLV size in combination with 784 the protocol identifier enables the decoder to determine the type 785 of the node. For Direct or Static configuration, the value SHOULD 786 be taken from an IPv4 or IPv6 address (e.g. loopback interface) 787 configured on the node. 789 There can be at most one instance of each sub-TLV type present in 790 any Node Descriptor. The sub-TLVs within a Node Descriptor MUST 791 be arranged in ascending order by sub-TLV type. This needs to be 792 done in order to compare NLRIs, even when an implementation 793 encounters an unknown sub-TLV. Using stable sorting, an 794 implementation can do binary comparison of NLRIs and hence allow 795 incremental deployment of new key sub-TLVs. 797 The BGP-LS Identifier was introduced by [RFC7752] and it's use is 798 being deprecated by this document. Implementations MUST continue to 799 support this sub-TLV for backward compatibility. The default value 800 of 0 is RECOMMENDED to be use when a BGP-LS Producer includes this 801 sub-TLV when originating information into BGP-LS. Implementations 802 MAY provide an option to configure this value for backward 803 compatibility reasons. The use of the Instance-ID in the Identifier 804 field is the RECOMMENDED way of segregation of different IGP domains 805 in BGP-LS. 807 4.2.2. Link Descriptors 809 The Link Descriptor field is a set of Type/Length/Value (TLV) 810 triplets. The format of each TLV is shown in Section 4.1. The Link 811 Descriptor TLVs uniquely identify a link among multiple parallel 812 links between a pair of anchor routers. A link described by the Link 813 Descriptor TLVs actually is a "half-link", a unidirectional 814 representation of a logical link. In order to fully describe a 815 single logical link, two originating routers advertise a half-link 816 each, i.e., two Link NLRIs are advertised for a given point-to-point 817 link. 819 A BGP-LS Consumer should not consider a link between two nodes as 820 being available unless it has received the two Link NLRIs 821 corresponding to the half-link representation of that link from both 822 the nodes. This check is similar to the 'two way connectivity check' 823 that is performed by link-state IGPs and is also required to be done 824 by BGP-LS Consumers of link-state topology. 826 A BGP-LS Producer MAY supress the advertisement of a Link NLRI, 827 corresponding to a half link, from a link-state IGP unless it has 828 verified that the link is being reported in the IS-IS LSP or OSPF 829 Router LSA by both the nodes connected by that link. This 'two way 830 connectivity check' is performed by link-state IGPs during their 831 computation and may be leveraged before passing information for any 832 half-link that is reported from these IGPs in to BGP-LS. This 833 ensures that only those Link State IGP adjacencies which are 834 established get reported via Link NLRIs. Such a 'two way 835 connectivity check' may be also required in certain cases (e.g. with 836 OSPF) to obtain the proper link identifiers of the remote node. 838 The format and semantics of the Value fields in most Link Descriptor 839 TLVs correspond to the format and semantics of Value fields in IS-IS 840 Extended IS Reachability sub-TLVs, defined in [RFC5305], [RFC5307], 841 and [RFC6119]. Although the encodings for Link Descriptor TLVs were 842 originally defined for IS-IS, the TLVs can carry data sourced by 843 either IS-IS or OSPF. 845 The following TLVs are defined as Link Descriptors in the Link NLRI: 847 +-----------+---------------------+--------------+------------------+ 848 | TLV Code | Description | IS-IS | Reference | 849 | Point | | TLV/Sub-TLV | (RFC/Section) | 850 +-----------+---------------------+--------------+------------------+ 851 | 258 | Link Local/Remote | 22/4 | [RFC5307] / 1.1 | 852 | | Identifiers | | | 853 | 259 | IPv4 interface | 22/6 | [RFC5305] / 3.2 | 854 | | address | | | 855 | 260 | IPv4 neighbor | 22/8 | [RFC5305] / 3.3 | 856 | | address | | | 857 | 261 | IPv6 interface | 22/12 | [RFC6119] / 4.2 | 858 | | address | | | 859 | 262 | IPv6 neighbor | 22/13 | [RFC6119] / 4.3 | 860 | | address | | | 861 | 263 | Multi-Topology | --- | Section 4.2.2.1 | 862 | | Identifier | | | 863 +-----------+---------------------+--------------+------------------+ 865 Table 4: Link Descriptor TLVs 867 The information about a link present in the LSA/LSP originated by the 868 local node of the link determines the set of TLVs in the Link 869 Descriptor of the link. 871 If interface and neighbor addresses, either IPv4 or IPv6, are 872 present, then the IP address TLVs MUST be included and the Link 873 Local/Remote Identifiers TLV MUST NOT be included in the Link 874 Descriptor. The Link Local/Remote Identifiers TLV MAY be included 875 in the link attribute when available. IPv6 link-local addresses 876 MUST NOT be carried in the IPv6 address TLVs as descriptors of a 877 link as they are not considered unique. 879 If interface and neighbor addresses are not present and the link 880 local/remote identifiers are present, then the Link Local/Remote 881 Identifiers TLV MUST be included in the Link Descriptor. The Link 882 Local/Remote Identifiers MUST be included in the Link Descriptor 883 also in the case of links having only IPv6 link-local addressing 884 on them. 886 The Multi-Topology Identifier TLV MUST be included in Link 887 Descriptor if the underlying IGP link object is associated with a 888 non-default topology. 890 The TLVs/sub-TLVs corresponding to the interface addresses and/or the 891 local/remote identfiers may not always be signaled in the IGPs unless 892 their advertisement is enabled specifically. In such cases, a BGP-LS 893 Producer may not be able to generate valid Link NLRIs for such link 894 advertisements from the IGPs. 896 4.2.2.1. Multi-Topology ID 898 The Multi-Topology ID (MT-ID) TLV carries one or more IS-IS or OSPF 899 Multi-Topology IDs for a link, node, or prefix. 901 Semantics of the IS-IS MT-ID are defined in Section 7.1 and 7.2 of 902 RFC 5120 [RFC5120]. Semantics of the OSPF MT-ID are defined in 903 Section 3.7 of RFC 4915 [RFC4915]. If the value in the MT-ID TLV is 904 derived from OSPF, then the upper 5 bits of the MT-ID field MUST be 905 set to 0. 907 The format of the MT-ID TLV is shown in the following figure. 909 0 1 2 3 910 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 1 911 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 912 | Type | Length=2*n | 913 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 914 |R R R R| Multi-Topology ID 1 | .... // 915 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 916 // .... |R R R R| Multi-Topology ID n | 917 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 919 Figure 12: Multi-Topology ID TLV Format 921 where Type is 263, Length is 2*n, and n is the number of MT-IDs 922 carried in the TLV. 924 The MT-ID TLV MAY be present in a Link Descriptor, a Prefix 925 Descriptor, or the BGP-LS attribute of a Node NLRI. In a Link or 926 Prefix Descriptor, only a single MT-ID TLV containing the MT-ID of 927 the topology where the link or the prefix is reachable is allowed. 928 In case one wants to advertise multiple topologies for a given Link 929 Descriptor or Prefix Descriptor, multiple NLRIs MUST be generated 930 where each NLRI contains a single unique MT-ID. When used in the 931 Link or Prefix Descriptor TLV for IS-IS, the Bits R are reserved and 932 MUST be set to 0 (as per Section 7.2 of RFC 5120 [RFC5120]) when 933 originated and ignored on receipt. 935 In the BGP-LS attribute of a Node NLRI, one MT-ID TLV containing the 936 array of MT-IDs of all topologies where the node is reachable is 937 allowed. When used in the Node Attribute TLV for IS-IS, the Bits R 938 are set as per Section 7.1 of RFC 5120 [RFC5120]. 940 4.2.3. Prefix Descriptors 942 The Prefix Descriptor field is a set of Type/Length/Value (TLV) 943 triplets. Prefix Descriptor TLVs uniquely identify an IPv4 or IPv6 944 prefix originated by a node. The following TLVs are defined as 945 Prefix Descriptors in the IPv4/IPv6 Prefix NLRI: 947 +-------------+---------------------+----------+--------------------+ 948 | TLV Code | Description | Length | Reference | 949 | Point | | | (RFC/Section) | 950 +-------------+---------------------+----------+--------------------+ 951 | 263 | Multi-Topology | variable | Section 4.2.2.1 | 952 | | Identifier | | | 953 | 264 | OSPF Route Type | 1 | Section 4.2.3.1 | 954 | 265 | IP Reachability | variable | Section 4.2.3.2 | 955 | | Information | | | 956 +-------------+---------------------+----------+--------------------+ 958 Table 5: Prefix Descriptor TLVs 960 The Multi-Topology Identifier TLV MUST be included in Prefix 961 Descriptor if the underlying IGP prefix object is associated with a 962 non-default topology. 964 4.2.3.1. OSPF Route Type 966 The OSPF Route Type TLV is a mandatory TLV corresponding to Prefix 967 NLRIs originated from OSPF. It is used to identify the OSPF route 968 type of the prefix. An OSPF prefix MAY be advertised in the OSPF 969 domain with multiple route types. The Route Type TLV allows the 970 discrimination of these advertisements. The format of the OSPF Route 971 Type TLV is shown in the following figure. 973 0 1 2 3 974 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 1 975 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 976 | Type | Length | 977 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 978 | Route Type | 979 +-+-+-+-+-+-+-+-+ 981 Figure 13: OSPF Route Type TLV Format 983 where the Type and Length fields of the TLV are defined in Table 5. 984 The OSPF Route Type field values are defined in the OSPF protocol and 985 can be one of the following: 987 o Intra-Area (0x1) 989 o Inter-Area (0x2) 991 o External 1 (0x3) 993 o External 2 (0x4) 995 o NSSA 1 (0x5) 997 o NSSA 2 (0x6) 999 4.2.3.2. IP Reachability Information 1001 The IP Reachability Information TLV is a mandatory TLV for IPv4 & 1002 IPv6 Prefix NLRI types. The TLV contains one IP address prefix (IPv4 1003 or IPv6) originally advertised in the IGP topology. Its purpose is 1004 to glue a particular BGP service NLRI by virtue of its BGP next hop 1005 to a given node in the LSDB. A router SHOULD advertise an IP Prefix 1006 NLRI for each of its BGP next hops. The format of the IP 1007 Reachability Information TLV is shown in the following figure: 1009 0 1 2 3 1010 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 1011 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1012 | Type | Length | 1013 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1014 | Prefix Length | IP Prefix (variable) // 1015 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1017 Figure 14: IP Reachability Information TLV Format 1019 The Type and Length fields of the TLV are defined in Table 5. The 1020 following two fields determine the reachability information of the 1021 address family. The Prefix Length field contains the length of the 1022 prefix in bits. The IP Prefix field contains the most significant 1023 octets of the prefix, i.e., 1 octet for prefix length 1 up to 8, 2 1024 octets for prefix length 9 to 16, 3 octets for prefix length 17 up to 1025 24, 4 octets for prefix length 25 up to 32, etc. 1027 4.3. The BGP-LS Attribute 1029 The BGP-LS Attribute is an optional, non-transitive BGP attribute 1030 that is used to carry link, node, and prefix parameters and 1031 attributes. It is defined as a set of Type/Length/Value (TLV) 1032 triplets, described in the following section. This attribute SHOULD 1033 only be included with Link-State NLRIs. This attribute MUST be 1034 ignored for all other address families. 1036 The Node Attribute TLVs, Link Attribute TLVs and Prefix Attribute 1037 TLVs are sets of TLVs that may be encoded in the BGP-LS Attribute 1038 associated with a Node NLRI, Link NLRI and Prefix NLRI respectively. 1040 The BGP-LS Attribute may potentially grow large in size depending on 1041 the amount of link-state information associated with a single Link- 1042 State NLRI. The BGP specification [RFC4271] mandates a maximum BGP 1043 message size of 4096 octets. It is RECOMMENDED that an 1044 implementation support [RFC8654] in order to accommodate larger size 1045 of information within the BGP-LS Attribute. BGP-LS Producers MUST 1046 ensure that they limit the TLVs included in the BGP-LS Attribute to 1047 ensure that a BGP update message for a single Link-State NLRI does 1048 not cross the maximum limit for a BGP message. The determination of 1049 the types of TLVs to be included MAY be made by the BGP-LS Producer 1050 based on the BGP-LS Consumer applications requirement and is outside 1051 the scope of this document. When a BGP-LS Propagator finds that it 1052 is exceeding the maximum BGP message size due to addition or update 1053 of some other BGP Attribute (e.g. AS_PATH), it MUST consider the 1054 BGP-LS Attribute to be malformed and handle the propagation as 1055 described in Section 7.2.2. 1057 4.3.1. Node Attribute TLVs 1059 The following Node Attribute TLVs are defined for the BGP-LS 1060 Attribute associated with a Node NLRI: 1062 +-------------+----------------------+----------+-------------------+ 1063 | TLV Code | Description | Length | Reference | 1064 | Point | | | (RFC/Section) | 1065 +-------------+----------------------+----------+-------------------+ 1066 | 263 | Multi-Topology | variable | Section 4.2.2.1 | 1067 | | Identifier | | | 1068 | 1024 | Node Flag Bits | 1 | Section 4.3.1.1 | 1069 | 1025 | Opaque Node | variable | Section 4.3.1.5 | 1070 | | Attribute | | | 1071 | 1026 | Node Name | variable | Section 4.3.1.3 | 1072 | 1027 | IS-IS Area | variable | Section 4.3.1.2 | 1073 | | Identifier | | | 1074 | 1028 | IPv4 Router-ID of | 4 | [RFC5305] / 4.3 | 1075 | | Local Node | | | 1076 | 1029 | IPv6 Router-ID of | 16 | [RFC6119] / 4.1 | 1077 | | Local Node | | | 1078 +-------------+----------------------+----------+-------------------+ 1080 Table 6: Node Attribute TLVs 1082 4.3.1.1. Node Flag Bits TLV 1084 The Node Flag Bits TLV carries a bit mask describing node attributes. 1085 The value is a 1 octet length bit array of flags, where each bit 1086 represents a node operational state or attribute. 1088 0 1 2 3 1089 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 1090 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1091 | Type | Length | 1092 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1093 |O|T|E|B|R|V| Rsvd| 1094 +-+-+-+-+-+-+-+-+-+ 1096 Figure 15: Node Flag Bits TLV Format 1098 The bits are defined as follows: 1100 +-----------------+-------------------------+------------+ 1101 | Bit | Description | Reference | 1102 +-----------------+-------------------------+------------+ 1103 | 'O' | Overload Bit | [ISO10589] | 1104 | 'T' | Attached Bit | [ISO10589] | 1105 | 'E' | External Bit | [RFC2328] | 1106 | 'B' | ABR Bit | [RFC2328] | 1107 | 'R' | Router Bit | [RFC5340] | 1108 | 'V' | V6 Bit | [RFC5340] | 1109 | Reserved (Rsvd) | Reserved for future use | | 1110 +-----------------+-------------------------+------------+ 1112 Table 7: Node Flag Bits Definitions 1114 4.3.1.2. IS-IS Area Identifier TLV 1116 An IS-IS node can be part of one or more IS-IS areas. Each of these 1117 area addresses is carried in the IS-IS Area Identifier TLV. If 1118 multiple area addresses are present, multiple TLVs are used to encode 1119 them. The IS-IS Area Identifier TLV may be present in the BGP-LS 1120 attribute only when advertised in the Link-State Node NLRI. 1122 0 1 2 3 1123 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 1124 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1125 | Type | Length | 1126 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1127 // Area Identifier (variable) // 1128 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1130 Figure 16: IS-IS Area Identifier TLV Format 1132 4.3.1.3. Node Name TLV 1134 The Node Name TLV is optional. Its structure and encoding has been 1135 borrowed from [RFC5301]. The Value field identifies the symbolic 1136 name of the router node. This symbolic name can be the Fully 1137 Qualified Domain Name (FQDN) for the router, it can be a subset of 1138 the FQDN (e.g., a hostname), or it can be any string operators want 1139 to use for the router. The use of FQDN or a subset of it is strongly 1140 RECOMMENDED. The maximum length of the Node Name TLV is 255 octets. 1142 The Value field is encoded in 7-bit ASCII. If a user interface for 1143 configuring or displaying this field permits Unicode characters, that 1144 user interface is responsible for applying the ToASCII and/or 1145 ToUnicode algorithm as described in [RFC5890] to achieve the correct 1146 format for transmission or display. 1148 [RFC5301] describes an IS-IS-specific extension and [RFC5642] 1149 describes an OSPF extension for advertisement of Node Name which MAY 1150 encoded in the Node Name TLV. 1152 0 1 2 3 1153 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 1154 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1155 | Type | Length | 1156 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1157 // Node Name (variable) // 1158 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1160 Figure 17: Node Name Format 1162 4.3.1.4. Local IPv4/IPv6 Router-ID TLVs 1164 The local IPv4/IPv6 Router-ID TLVs are used to describe auxiliary 1165 Router-IDs that the IGP might be using, e.g., for TE and migration 1166 purposes such as correlating a Node-ID between different protocols. 1167 If there is more than one auxiliary Router-ID of a given type, then 1168 each one is encoded in its own TLV. 1170 4.3.1.5. Opaque Node Attribute TLV 1172 The Opaque Node Attribute TLV is an envelope that transparently 1173 carries optional Node Attribute TLVs advertised by a router. An 1174 originating router shall use this TLV for encoding information 1175 specific to the protocol advertised in the NLRI header Protocol-ID 1176 field or new protocol extensions to the protocol as advertised in the 1177 NLRI header Protocol-ID field for which there is no protocol-neutral 1178 representation in the BGP Link-State NLRI. The primary use of the 1179 Opaque Node Attribute TLV is to bridge the document lag between, 1180 e.g., a new IGP link-state attribute being defined and the protocol- 1181 neutral BGP-LS extensions being published. A router, for example, 1182 could use this extension in order to advertise the native protocol's 1183 Node Attribute TLVs, such as the OSPF Router Informational 1184 Capabilities TLV defined in [RFC7770] or the IGP TE Node Capability 1185 Descriptor TLV described in [RFC5073]. 1187 0 1 2 3 1188 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 1189 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1190 | Type | Length | 1191 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1192 // Opaque node attributes (variable) // 1193 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1195 Figure 18: Opaque Node Attribute Format 1197 4.3.2. Link Attribute TLVs 1199 Link Attribute TLVs are TLVs that may be encoded in the BGP-LS 1200 attribute with a Link NLRI. Each 'Link Attribute' is a Type/Length/ 1201 Value (TLV) triplet formatted as defined in Section 4.1. The format 1202 and semantics of the Value fields in some Link Attribute TLVs 1203 correspond to the format and semantics of the Value fields in IS-IS 1204 Extended IS Reachability sub-TLVs, defined in [RFC5305] and 1205 [RFC5307]. Other Link Attribute TLVs are defined in this document. 1206 Although the encodings for Link Attribute TLVs were originally 1207 defined for IS-IS, the TLVs can carry data sourced by either IS-IS or 1208 OSPF. 1210 The following Link Attribute TLVs are defined for the BGP-LS 1211 Attribute associated with a Link NLRI: 1213 +-----------+---------------------+--------------+------------------+ 1214 | TLV Code | Description | IS-IS | Reference | 1215 | Point | | TLV/Sub-TLV | (RFC/Section) | 1216 +-----------+---------------------+--------------+------------------+ 1217 | 1028 | IPv4 Router-ID of | 134/--- | [RFC5305] / 4.3 | 1218 | | Local Node | | | 1219 | 1029 | IPv6 Router-ID of | 140/--- | [RFC6119] / 4.1 | 1220 | | Local Node | | | 1221 | 1030 | IPv4 Router-ID of | 134/--- | [RFC5305] / 4.3 | 1222 | | Remote Node | | | 1223 | 1031 | IPv6 Router-ID of | 140/--- | [RFC6119] / 4.1 | 1224 | | Remote Node | | | 1225 | 1088 | Administrative | 22/3 | [RFC5305] / 3.1 | 1226 | | group (color) | | | 1227 | 1089 | Maximum link | 22/9 | [RFC5305] / 3.4 | 1228 | | bandwidth | | | 1229 | 1090 | Max. reservable | 22/10 | [RFC5305] / 3.5 | 1230 | | link bandwidth | | | 1231 | 1091 | Unreserved | 22/11 | [RFC5305] / 3.6 | 1232 | | bandwidth | | | 1233 | 1092 | TE Default Metric | 22/18 | Section 4.3.2.3 | 1234 | 1093 | Link Protection | 22/20 | [RFC5307] / 1.2 | 1235 | | Type | | | 1236 | 1094 | MPLS Protocol Mask | --- | Section 4.3.2.2 | 1237 | 1095 | IGP Metric | --- | Section 4.3.2.4 | 1238 | 1096 | Shared Risk Link | --- | Section 4.3.2.5 | 1239 | | Group | | | 1240 | 1097 | Opaque Link | --- | Section 4.3.2.6 | 1241 | | Attribute | | | 1242 | 1098 | Link Name | --- | Section 4.3.2.7 | 1243 +-----------+---------------------+--------------+------------------+ 1245 Table 8: Link Attribute TLVs 1247 4.3.2.1. IPv4/IPv6 Router-ID TLVs 1249 The local/remote IPv4/IPv6 Router-ID TLVs are used to describe 1250 auxiliary Router-IDs that the IGP might be using, e.g., for TE 1251 purposes. All auxiliary Router-IDs of both the local and the remote 1252 node MUST be included in the link attribute of each Link NLRI. If 1253 there is more than one auxiliary Router-ID of a given type, then 1254 multiple TLVs are used to encode them. 1256 4.3.2.2. MPLS Protocol Mask TLV 1258 The MPLS Protocol Mask TLV carries a bit mask describing which MPLS 1259 signaling protocols are enabled. The length of this TLV is 1. The 1260 value is a bit array of 8 flags, where each bit represents an MPLS 1261 Protocol capability. 1263 Generation of the MPLS Protocol Mask TLV is only valid for and SHOULD 1264 only be used with originators that have local link insight, for 1265 example, the Protocol-IDs 'Static configuration' or 'Direct' as per 1266 Table 2. The MPLS Protocol Mask TLV MUST NOT be included in NLRIs 1267 with the other Protocol-IDs listed in Table 2. 1269 0 1 2 3 1270 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 1271 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1272 | Type | Length | 1273 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1274 |L|R| Reserved | 1275 +-+-+-+-+-+-+-+-+ 1277 Figure 19: MPLS Protocol Mask TLV 1279 The following bits are defined: 1281 +------------+------------------------------------------+-----------+ 1282 | Bit | Description | Reference | 1283 +------------+------------------------------------------+-----------+ 1284 | 'L' | Label Distribution Protocol (LDP) | [RFC5036] | 1285 | 'R' | Extension to RSVP for LSP Tunnels | [RFC3209] | 1286 | | (RSVP-TE) | | 1287 | 'Reserved' | Reserved for future use | | 1288 +------------+------------------------------------------+-----------+ 1290 Table 9: MPLS Protocol Mask TLV Codes 1292 4.3.2.3. TE Default Metric TLV 1294 The TE Default Metric TLV carries the Traffic Engineering metric for 1295 this link. The length of this TLV is fixed at 4 octets. If a source 1296 protocol uses a metric width of less than 32 bits, then the high- 1297 order bits of this field MUST be padded with zero. 1299 0 1 2 3 1300 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 1301 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1302 | Type | Length | 1303 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1304 | TE Default Link Metric | 1305 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1307 Figure 20: TE Default Metric TLV Format 1309 4.3.2.4. IGP Metric TLV 1311 The IGP Metric TLV carries the metric for this link. The length of 1312 this TLV is variable, depending on the metric width of the underlying 1313 protocol. IS-IS small metrics have a length of 1 octet. Since the 1314 ISIS small metrics are of 6 bit size, the two most significant bits 1315 MUST be set to 0 and MUST be ignored by receiver. OSPF link metrics 1316 have a length of 2 octets. IS-IS wide metrics have a length of 3 1317 octets. 1319 0 1 2 3 1320 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 1321 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1322 | Type | Length | 1323 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1324 // IGP Link Metric (variable length) // 1325 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1327 Figure 21: IGP Metric TLV Format 1329 4.3.2.5. Shared Risk Link Group TLV 1331 The Shared Risk Link Group (SRLG) TLV carries the Shared Risk Link 1332 Group information (see Section 2.3 ("Shared Risk Link Group 1333 Information") of [RFC4202]). It contains a data structure consisting 1334 of a (variable) list of SRLG values, where each element in the list 1335 has 4 octets, as shown in Figure 22. The length of this TLV is 4 * 1336 (number of SRLG values). 1338 0 1 2 3 1339 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 1340 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1341 | Type | Length | 1342 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1343 | Shared Risk Link Group Value | 1344 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1345 // ............ // 1346 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1347 | Shared Risk Link Group Value | 1348 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1350 Figure 22: Shared Risk Link Group TLV Format 1352 The SRLG TLV for OSPF-TE is defined in [RFC4203]. In IS-IS, the SRLG 1353 information is carried in two different TLVs: the IPv4 (SRLG) TLV 1354 (Type 138) defined in [RFC5307] and the IPv6 SRLG TLV (Type 139) 1355 defined in [RFC6119]. In Link-State NLRI, both IPv4 and IPv6 SRLG 1356 information are carried in a single TLV. 1358 4.3.2.6. Opaque Link Attribute TLV 1360 The Opaque Link Attribute TLV is an envelope that transparently 1361 carries optional Link Attribute TLVs advertised by a router. An 1362 originating router shall use this TLV for encoding information 1363 specific to the protocol advertised in the NLRI header Protocol-ID 1364 field or new protocol extensions to the protocol as advertised in the 1365 NLRI header Protocol-ID field for which there is no protocol-neutral 1366 representation in the BGP Link-State NLRI. The primary use of the 1367 Opaque Link Attribute TLV is to bridge the document lag between, 1368 e.g., a new IGP link-state attribute being defined and the 'protocol- 1369 neutral' BGP-LS extensions being published. 1371 0 1 2 3 1372 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 1373 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1374 | Type | Length | 1375 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1376 // Opaque link attributes (variable) // 1377 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1379 Figure 23: Opaque Link Attribute TLV Format 1381 4.3.2.7. Link Name TLV 1383 The Link Name TLV is optional. The Value field identifies the 1384 symbolic name of the router link. This symbolic name can be the FQDN 1385 for the link, it can be a subset of the FQDN, or it can be any string 1386 operators want to use for the link. The use of FQDN or a subset of 1387 it is strongly RECOMMENDED. The maximum length of the Link Name TLV 1388 is 255 octets. 1390 The Value field is encoded in 7-bit ASCII. If a user interface for 1391 configuring or displaying this field permits Unicode characters, that 1392 user interface is responsible for applying the ToASCII and/or 1393 ToUnicode algorithm as described in [RFC5890] to achieve the correct 1394 format for transmission or display. 1396 How a router derives and injects link names is outside of the scope 1397 of this document. 1399 0 1 2 3 1400 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 1401 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1402 | Type | Length | 1403 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1404 // Link Name (variable) // 1405 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1407 Figure 24: Link Name TLV Format 1409 4.3.3. Prefix Attribute TLVs 1411 Prefixes are learned from the IGP topology (IS-IS or OSPF) with a set 1412 of IGP attributes (such as metric, route tags, etc.) that are 1413 advertised in the BGP-LS Attribute with Prefix NLRI types 3 and 4. 1415 The following Prefix Attribute TLVs are defined for the BGP-LS 1416 Attribute associated with a Prefix NLRI: 1418 +---------------+-----------------------+----------+----------------+ 1419 | TLV Code | Description | Length | Reference | 1420 | Point | | | | 1421 +---------------+-----------------------+----------+----------------+ 1422 | 1152 | IGP Flags | 1 | Section 4.3.3. | 1423 | | | | 1 | 1424 | 1153 | IGP Route Tag | 4*n | [RFC5130] | 1425 | 1154 | IGP Extended Route | 8*n | [RFC5130] | 1426 | | Tag | | | 1427 | 1155 | Prefix Metric | 4 | [RFC5305] | 1428 | 1156 | OSPF Forwarding | 4 | [RFC2328] | 1429 | | Address | | | 1430 | 1157 | Opaque Prefix | variable | Section 4.3.3. | 1431 | | Attribute | | 6 | 1432 +---------------+-----------------------+----------+----------------+ 1434 Table 10: Prefix Attribute TLVs 1436 4.3.3.1. IGP Flags TLV 1438 The IGP Flags TLV contains IS-IS and OSPF flags and bits originally 1439 assigned to the prefix. The IGP Flags TLV is encoded as follows: 1441 0 1 2 3 1442 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 1443 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1444 | Type | Length | 1445 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1446 |D|N|L|P| Resvd.| 1447 +-+-+-+-+-+-+-+-+ 1449 Figure 25: IGP Flag TLV Format 1451 The Value field contains bits defined according to the table below: 1453 +----------+---------------------------+-----------+ 1454 | Bit | Description | Reference | 1455 +----------+---------------------------+-----------+ 1456 | 'D' | IS-IS Up/Down Bit | [RFC5305] | 1457 | 'N' | OSPF "no unicast" Bit | [RFC5340] | 1458 | 'L' | OSPF "local address" Bit | [RFC5340] | 1459 | 'P' | OSPF "propagate NSSA" Bit | [RFC5340] | 1460 | Reserved | Reserved for future use. | | 1461 +----------+---------------------------+-----------+ 1463 Table 11: IGP Flag Bits Definitions 1465 4.3.3.2. IGP Route Tag TLV 1467 The IGP Route Tag TLV carries original IGP Tags (IS-IS [RFC5130] or 1468 OSPF) of the prefix and is encoded as follows: 1470 0 1 2 3 1471 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 1472 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1473 | Type | Length | 1474 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1475 // Route Tags (one or more) // 1476 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1478 Figure 26: IGP Route Tag TLV Format 1480 Length is a multiple of 4. 1482 The Value field contains one or more Route Tags as learned in the IGP 1483 topology. 1485 4.3.3.3. Extended IGP Route Tag TLV 1487 The Extended IGP Route Tag TLV carries IS-IS Extended Route Tags of 1488 the prefix [RFC5130] and is encoded as follows: 1490 0 1 2 3 1491 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 1492 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1493 | Type | Length | 1494 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1495 // Extended Route Tag (one or more) // 1496 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1498 Figure 27: Extended IGP Route Tag TLV Format 1500 Length is a multiple of 8. 1502 The Extended Route Tag field contains one or more Extended Route Tags 1503 as learned in the IGP topology. 1505 4.3.3.4. Prefix Metric TLV 1507 The Prefix Metric TLV is an optional attribute and may only appear 1508 once. If present, it carries the metric of the prefix as known in 1509 the IGP topology as described in Section 4 of [RFC5305] (and 1510 therefore represents the reachability cost to the prefix). If not 1511 present, it means that the prefix is advertised without any 1512 reachability. 1514 0 1 2 3 1515 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 1516 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1517 | Type | Length | 1518 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1519 | Metric | 1520 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1522 Figure 28: Prefix Metric TLV Format 1524 Length is 4. 1526 4.3.3.5. OSPF Forwarding Address TLV 1528 The OSPF Forwarding Address TLV [RFC2328] [RFC5340] carries the OSPF 1529 forwarding address as known in the original OSPF advertisement. 1530 Forwarding address can be either IPv4 or IPv6. 1532 0 1 2 3 1533 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 1534 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1535 | Type | Length | 1536 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1537 // Forwarding Address (variable) // 1538 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1540 Figure 29: OSPF Forwarding Address TLV Format 1542 Length is 4 for an IPv4 forwarding address, and 16 for an IPv6 1543 forwarding address. 1545 4.3.3.6. Opaque Prefix Attribute TLV 1547 The Opaque Prefix Attribute TLV is an envelope that transparently 1548 carries optional Prefix Attribute TLVs advertised by a router. An 1549 originating router shall use this TLV for encoding information 1550 specific to the protocol advertised in the NLRI header Protocol-ID 1551 field or new protocol extensions to the protocol as advertised in the 1552 NLRI header Protocol-ID field for which there is no protocol-neutral 1553 representation in the BGP Link-State NLRI. The primary use of the 1554 Opaque Prefix Attribute TLV is to bridge the document lag between, 1555 e.g., a new IGP link-state attribute being defined and the protocol- 1556 neutral BGP-LS extensions being published. 1558 The format of the TLV is as follows: 1560 0 1 2 3 1561 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 1562 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1563 | Type | Length | 1564 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1565 // Opaque Prefix Attributes (variable) // 1566 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1568 Figure 30: Opaque Prefix Attribute TLV Format 1570 Type is as specified in Table 10. Length is variable. 1572 4.4. Private Use 1574 TLVs for Vendor Private use are supported using the code point range 1575 reserved as indicated in Section 6. For such TLV use in the NLRI or 1576 BGP-LS Attribute, the format as described in Section 4.1 is to be 1577 used and a 4 octet field MUST be included as the first field in the 1578 value to carry the Enterprise Code. For a private use NLRI Type, a 4 1579 octet field MUST be included as the first field in the NLRI 1580 immediately following the Total NLRI Length field of the Link-State 1581 NLRI format as described in Section 4.2 to carry the Enterprise Code. 1582 The Enterprise Codes are listed at . This enables use vendor specific extensions 1584 without conflicts. 1586 Multiple instances of private-use TLVs MAY appear in the BGP-LS 1587 Attribute. 1589 4.5. BGP Next-Hop Information 1591 BGP link-state information for both IPv4 and IPv6 networks can be 1592 carried over either an IPv4 BGP session or an IPv6 BGP session. If 1593 an IPv4 BGP session is used, then the next hop in the MP_REACH_NLRI 1594 SHOULD be an IPv4 address. Similarly, if an IPv6 BGP session is 1595 used, then the next hop in the MP_REACH_NLRI SHOULD be an IPv6 1596 address. Usually, the next hop will be set to the local endpoint 1597 address of the BGP session. The next-hop address MUST be encoded as 1598 described in [RFC4760]. The Length field of the next-hop address 1599 will specify the next-hop address family. If the next-hop length is 1600 4, then the next hop is an IPv4 address; if the next-hop length is 1601 16, then it is a global IPv6 address; and if the next-hop length is 1602 32, then there is one global IPv6 address followed by a link-local 1603 IPv6 address. The link-local IPv6 address should be used as 1604 described in [RFC2545]. For VPN Subsequent Address Family Identifier 1605 (SAFI), as per custom, an 8-byte Route Distinguisher set to all zero 1606 is prepended to the next hop. 1608 The BGP Next Hop attribute is used by each BGP-LS speaker to validate 1609 the NLRI it receives. In case identical NLRIs are sourced by 1610 multiple BGP-LS Producers, the BGP Next Hop attribute is used to 1611 tiebreak as per the standard BGP path decision process. This 1612 specification doesn't mandate any rule regarding the rewrite of the 1613 BGP Next Hop attribute. 1615 4.6. Inter-AS Links 1617 The main source of TE information is the IGP, which is not active on 1618 inter-AS links. In some cases, the IGP may have information of 1619 inter-AS links [RFC5392] [RFC5316]. In other cases, an 1620 implementation SHOULD provide a means to inject inter-AS links into 1621 BGP-LS. The exact mechanism used to provision the inter-AS links is 1622 outside the scope of this document and are described in 1623 [I-D.ietf-idr-bgpls-inter-as-topology-ext]. 1625 4.7. Handling of Unreachable IGP Nodes 1627 The origination and propagation of IGP link-state information via BGP 1628 needs to provide a consistent and true view of the topology of the 1629 IGP domain. BGP-LS provides an abstraction of the protocol specifics 1630 and BGP-LS Consumers may be varied types of applications. While the 1631 information propagated via BGP-LS from a link-state routing protocol 1632 is sourced from that protocol's LSDB, it does not serve as a true 1633 reflection of the originating router's LSDB since it does not include 1634 the LSA/LSP sequence number information. The sequence numbers are 1635 not included since a single NLRI update may be put together with 1636 information that is coming from multiple LSAs/LSPs. 1638 Consider an OSPF network as shown in Figure 31, where R2 and R3 are 1639 the BGP-LS Producers and also the OSPF Area Border Routers (ABRs). 1640 The link between R2 and R3 is in area 0 while the other links shown 1641 are in area 1. 1643 A BGP-LS Consumer talks to a BGP route-reflector (RR) R0 which is 1644 aggregating the BGP-LS feed from the BGP-LS Producers R2 and R3. 1645 Here R2 and R3 provide a redundant topology feed via BGP-LS to R0. 1646 Normally, R0 would receive two identical copies of all the Link-State 1647 NLRIs from both R2 and R3 and it would pick one of them (say R2) 1648 based on the standard BGP best path decision process. 1650 Consumer 1651 ^ 1652 | 1653 R0 1654 (BGP Route Reflector) 1655 / \ 1656 / \ 1657 a1 / a0 \ a1 1658 R1 ------ R2 -------- R3 ------ R4 1659 a1 | | a1 1660 | | 1661 R5 ---------------------------- R6 1662 a1 1664 Figure 31: Incorrect Reporting due to BGP Path Selection 1666 Consider a scenario where the link between R5 and R6 is lost (thereby 1667 partitioning the area 1) and its impact on the OSPF LSDB at R2 and 1668 R3. 1670 Now, R5 will remove the link 5-6 from its Router LSA and this updated 1671 LSA is available at R2. R2 also has a stale copy of R6's Router LSA 1672 which still has the link 6-5 in it. Based on this view in its LSDB, 1673 R2 will advertise only the half-link 6-5 that it derives from R6's 1674 stale Router LSA. 1676 At the same time, R6 has removed the link 6-5 from its Router LSA and 1677 this updated LSA is available at R3. Similarly, R3 also has a stale 1678 copy of R5's Router LSA having the link 5-6 in it. Based on it's 1679 LSDB, R3 will advertise only the half-link 5-6 that it has derived 1680 from R5's stale Router LSA. 1682 Now, the BGP-LS Consumer receives both the Link NLRIs corresponding 1683 to the half-links from R2 and R3 via R0. When viewed together, it 1684 would not detect or realize that the area 1 is actually partitioned. 1685 Also if R2 continues to report Link-State NLRIs corresponding to the 1686 stale copy of Router LSA of R4 and R6 nodes then R0 would prefer them 1687 over the valid Link-State NLRIs for R4 and R6 that it is receiving 1688 from R3 based on its BGP decision process. This would result in the 1689 BGP-LS Consumer getting stale and inaccurate topology information. 1690 This problems scenario is avoided if R2 were to not advertise the 1691 link-state information corresponding to R4 and R6 and if R3 were to 1692 not advertise similarly for R1 and R5. 1694 A BGP-LS Producer SHOULD withdraw all link-state objects advertised 1695 by it in BGP when the node that originated its corresponding LSP/LSAs 1696 is determined to have become unreachable in the IGP. An 1697 implementation MAY continue to advertise link-state objects 1698 corresponding to unreachable nodes in a deployment use-case where the 1699 BGP-LS Consumer is interested in receiving a topology feed 1700 corresponding to a complete IGP LSDB view. In such deployments, it 1701 is expected that the problem described above is mitigated by the BGP- 1702 LS Consumer via appropriate handling of such a topology feed in 1703 addition to the use of either a direct BGP peering with the producer 1704 nodes or mechanisms such as [RFC7911] when using RR. Details of 1705 these mechanisms are outside the scope of this draft. 1707 If the BGP-LS Producer does withdraw link-state objects associated 1708 with an IGP node based on failure of reachability check for that 1709 node, then it MUST re-advertise those link-state objects after that 1710 node becomes reachable again in the IGP domain. 1712 4.8. Router-ID Anchoring Example: ISO Pseudonode 1714 Encoding of a broadcast LAN in IS-IS provides a good example of how 1715 Router-IDs are encoded. Consider Figure 32. This represents a 1716 Broadcast LAN between a pair of routers. The "real" (non-pseudonode) 1717 routers have both an IPv4 Router-ID and IS-IS Node-ID. The 1718 pseudonode does not have an IPv4 Router-ID. Node1 is the DIS for the 1719 LAN. Two unidirectional links (Node1, Pseudonode1) and (Pseudonode1, 1720 Node2) are being generated. 1722 The Link NLRI of (Node1, Pseudonode1) is encoded as follows. The IGP 1723 Router-ID TLV of the local Node Descriptor is 6 octets long and 1724 contains the ISO-ID of Node1, 1920.0000.2001. The IGP Router-ID TLV 1725 of the remote Node Descriptor is 7 octets long and contains the ISO- 1726 ID of Pseudonode1, 1920.0000.2001.02. The BGP-LS attribute of this 1727 link contains one local IPv4 Router-ID TLV (TLV type 1028) containing 1728 192.0.2.1, the IPv4 Router-ID of Node1. 1730 The Link NLRI of (Pseudonode1, Node2) is encoded as follows. The IGP 1731 Router-ID TLV of the local Node Descriptor is 7 octets long and 1732 contains the ISO-ID of Pseudonode1, 1920.0000.2001.02. The IGP 1733 Router-ID TLV of the remote Node Descriptor is 6 octets long and 1734 contains the ISO-ID of Node2, 1920.0000.2002. The BGP-LS attribute 1735 of this link contains one remote IPv4 Router-ID TLV (TLV type 1030) 1736 containing 192.0.2.2, the IPv4 Router-ID of Node2. 1738 +-----------------+ +-----------------+ +-----------------+ 1739 | Node1 | | Pseudonode1 | | Node2 | 1740 |1920.0000.2001.00|--->|1920.0000.2001.02|--->|1920.0000.2002.00| 1741 | 192.0.2.1 | | | | 192.0.2.2 | 1742 +-----------------+ +-----------------+ +-----------------+ 1744 Figure 32: IS-IS Pseudonodes 1746 4.9. Router-ID Anchoring Example: OSPF Pseudonode 1748 Encoding of a broadcast LAN in OSPF provides a good example of how 1749 Router-IDs and local Interface IPs are encoded. Consider Figure 33. 1750 This represents a Broadcast LAN between a pair of routers. The 1751 "real" (non-pseudonode) routers have both an IPv4 Router-ID and an 1752 Area Identifier. The pseudonode does have an IPv4 Router-ID, an IPv4 1753 Interface Address (for disambiguation), and an OSPF Area. Node1 is 1754 the DR for the LAN; hence, its local IP address 10.1.1.1 is used as 1755 both the Router-ID and Interface IP for the pseudonode keys. Two 1756 unidirectional links, (Node1, Pseudonode1) and (Pseudonode1, Node2), 1757 are being generated. 1759 The Link NLRI of (Node1, Pseudonode1) is encoded as follows: 1761 o Local Node Descriptor 1763 TLV #515: IGP Router-ID: 11.11.11.11 1765 TLV #514: OSPF Area-ID: ID:0.0.0.0 1767 o Remote Node Descriptor 1769 TLV #515: IGP Router-ID: 11.11.11.11:10.1.1.1 1771 TLV #514: OSPF Area-ID: ID:0.0.0.0 1773 The Link NLRI of (Pseudonode1, Node2) is encoded as follows: 1775 o Local Node Descriptor 1777 TLV #515: IGP Router-ID: 11.11.11.11:10.1.1.1 1779 TLV #514: OSPF Area-ID: ID:0.0.0.0 1781 o Remote Node Descriptor 1783 TLV #515: IGP Router-ID: 33.33.33.34 1785 TLV #514: OSPF Area-ID: ID:0.0.0.0 1787 10.1.1.1/24 10.1.1.2/24 1788 +-------------+ +-------------+ +-------------+ 1789 | Node1 | | Pseudonode1 | | Node2 | 1790 | 11.11.11.11 |--->| 11.11.11.11 |--->| 33.33.33.34 | 1791 | | | 10.1.1.1 | | | 1792 | Area 0 | | Area 0 | | Area 0 | 1793 +-------------+ +-------------+ +-------------+ 1795 Figure 33: OSPF Pseudonodes 1797 The LAN subnet 10.1.1.0/24 is not included in the Router LSA of Node1 1798 or Node2. The Network LSA for this LAN advertised by the DR Node1 1799 contains the subnet mask for the LAN along with the DR address. A 1800 Prefix NLRI corresponding to the LAN subnet is advertised with the 1801 Pseudonode1 used as the Local node using the DR address and the 1802 subnet mask from the Network LSA. 1804 4.10. Router-ID Anchoring Example: OSPFv2 to IS-IS Migration 1806 Graceful migration from one IGP to another requires coordinated 1807 operation of both protocols during the migration period. Such a 1808 coordination requires identifying a given physical link in both IGPs. 1809 The IPv4 Router-ID provides that "glue", which is present in the Node 1810 Descriptors of the OSPF Link NLRI and in the link attribute of the 1811 IS-IS Link NLRI. 1813 Consider a point-to-point link between two routers, A and B, that 1814 initially were OSPFv2-only routers and then IS-IS is enabled on them. 1816 Node A has IPv4 Router-ID and ISO-ID; node B has IPv4 Router-ID, IPv6 1817 Router-ID, and ISO-ID. Each protocol generates one Link NLRI for the 1818 link (A, B), both of which are carried by BGP-LS. The OSPFv2 Link 1819 NLRI for the link is encoded with the IPv4 Router-ID of nodes A and B 1820 in the local and remote Node Descriptors, respectively. The IS-IS 1821 Link NLRI for the link is encoded with the ISO-ID of nodes A and B in 1822 the local and remote Node Descriptors, respectively. In addition, 1823 the BGP-LS attribute of the IS-IS Link NLRI contains the TLV type 1824 1028 containing the IPv4 Router-ID of node A, TLV type 1030 1825 containing the IPv4 Router-ID of node B, and TLV type 1031 containing 1826 the IPv6 Router-ID of node B. In this case, by using IPv4 Router-ID, 1827 the link (A, B) can be identified in both the IS-IS and OSPF 1828 protocol. 1830 5. Link to Path Aggregation 1832 Distribution of all links available in the global Internet is 1833 certainly possible; however, it not desirable from a scaling and 1834 privacy point of view. Therefore, an implementation may support a 1835 link to path aggregation. Rather than advertising all specific links 1836 of a domain, an ASBR may advertise an "aggregate link" between a non- 1837 adjacent pair of nodes. The "aggregate link" represents the 1838 aggregated set of link properties between a pair of non-adjacent 1839 nodes. The actual methods to compute the path properties (of 1840 bandwidth, metric, etc.) are outside the scope of this document. The 1841 decision whether to advertise all specific links or aggregated links 1842 is an operator's policy choice. To highlight the varying levels of 1843 exposure, the following deployment examples are discussed. 1845 5.1. Example: No Link Aggregation 1847 Consider Figure 34. Both AS1 and AS2 operators want to protect their 1848 inter-AS {R1, R3}, {R2, R4} links using RSVP-FRR LSPs. If R1 wants 1849 to compute its link-protection LSP to R3, it needs to "see" an 1850 alternate path to R3. Therefore, the AS2 operator exposes its 1851 topology. All BGP-TE-enabled routers in AS1 "see" the full topology 1852 of AS2 and therefore can compute a backup path. Note that the 1853 computing router decides if the direct link between {R3, R4} or the 1854 {R4, R5, R3} path is used. 1856 AS1 : AS2 1857 : 1858 R1-------R3 1859 | : | \ 1860 | : | R5 1861 | : | / 1862 R2-------R4 1863 : 1864 : 1866 Figure 34: No Link Aggregation 1868 5.2. Example: ASBR to ASBR Path Aggregation 1870 The brief difference between the "no-link aggregation" example and 1871 this example is that no specific link gets exposed. Consider 1872 Figure 35. The only link that gets advertised by AS2 is an 1873 "aggregate" link between R3 and R4. This is enough to tell AS1 that 1874 there is a backup path. However, the actual links being used are 1875 hidden from the topology. 1877 AS1 : AS2 1878 : 1879 R1-------R3 1880 | : | 1881 | : | 1882 | : | 1883 R2-------R4 1884 : 1885 : 1887 Figure 35: ASBR Link Aggregation 1889 5.3. Example: Multi-AS Path Aggregation 1891 Service providers in control of multiple ASes may even decide to not 1892 expose their internal inter-AS links. Consider Figure 36. AS3 is 1893 modeled as a single node that connects to the border routers of the 1894 aggregated domain. 1896 AS1 : AS2 : AS3 1897 : : 1898 R1-------R3----- 1899 | : : \ 1900 | : : vR0 1901 | : : / 1902 R2-------R4----- 1903 : : 1904 : : 1906 Figure 36: Multi-AS Aggregation 1908 6. IANA Considerations 1910 IANA has assigned address family number 16388 (BGP-LS) in the 1911 "Address Family Numbers" registry with [RFC7752] as a reference. 1913 IANA has assigned SAFI values 71 (BGP-LS) and 72 (BGP-LS-VPN) in the 1914 "SAFI Values" sub-registry under the "Subsequent Address Family 1915 Identifiers (SAFI) Parameters" registry. 1917 IANA has assigned value 29 (BGP-LS Attribute) in the "BGP Path 1918 Attributes" sub-registry under the "Border Gateway Protocol (BGP) 1919 Parameters" registry. 1921 IANA has created a new "Border Gateway Protocol - Link State (BGP-LS) 1922 Parameters" registry at . All of the following registries are BGP-LS specific and 1924 are accessible under this registry: 1926 o "BGP-LS NLRI-Types" registry 1928 Value 0 is reserved. The maximum value is 65535. The range 1929 65000-65535 is for Private Use. The registry has been populated 1930 with the values shown in Table 1. Allocations within the registry 1931 under the "Expert Review" policy require documentation of the 1932 proposed use of the allocated value and approval by the Designated 1933 Expert assigned by the IESG (see [RFC8126]). 1935 o "BGP-LS Protocol-IDs" registry 1937 Value 0 is reserved. The maximum value is 255. The range 200-255 1938 is for Private Use. The registry has been populated with the 1939 values shown in Table 2. Allocations within the registry under 1940 the "Expert Review" policy require documentation of the proposed 1941 use of the allocated value and approval by the Designated Expert 1942 assigned by the IESG (see [RFC8126]). 1944 o "BGP-LS Well-Known Instance-IDs" registry 1946 This registry was setup via [RFC7752] and is no longer required. 1947 It may be retained as deprecated. 1949 o "BGP-LS Node Descriptor, Link Descriptor, Prefix Descriptor, and 1950 Attribute TLVs" registry 1952 Values 0-255 are reserved. Values 256-65535 will be used for code 1953 points. The range 65000-65535 is for Private Use. The registry 1954 has been populated with the values shown in Table 12. Allocations 1955 within the registry under the "Expert Review" policy require 1956 documentation of the proposed use of the allocated value and 1957 approval by the Designated Expert assigned by the IESG (see 1958 [RFC8126]). 1960 6.1. Guidance for Designated Experts 1962 In all cases of review by the Designated Expert (DE) described here, 1963 the DE is expected to ascertain the existence of suitable 1964 documentation (a specification) as described in [RFC8126]. The DE is 1965 also expected to check the clarity of purpose and use of the 1966 requested code points. Additionally, the DE must verify that any 1967 request for one of these code points has been made available for 1968 review and comment within the IETF: the DE will post the request to 1969 the IDR Working Group mailing list (or a successor mailing list 1970 designated by the IESG). If the request comes from within the IETF, 1971 it should be documented in an Internet-Draft. Lastly, the DE must 1972 ensure that any other request for a code point does not conflict with 1973 work that is active or already published within the IETF. 1975 7. Manageability Considerations 1977 This section is structured as recommended in [RFC5706]. 1979 7.1. Operational Considerations 1981 7.1.1. Operations 1983 Existing BGP operational procedures apply. No new operation 1984 procedures are defined in this document. It is noted that the NLRI 1985 information present in this document carries purely application-level 1986 data that has no immediate impact on the corresponding forwarding 1987 state computed by BGP. As such, any churn in reachability 1988 information has a different impact than regular BGP updates, which 1989 need to change the forwarding state for an entire router. It is 1990 expected that the distribution of this NLRI SHOULD be handled by 1991 dedicated route reflectors in most deployments providing a level of 1992 isolation and fault containment between different NLRI types. In the 1993 event of dedicated route reflectors not being available, other 1994 alternate mechanisms like separation of BGP instances or separate BGP 1995 sessions (e.g. using different addresses for peering) for Link-State 1996 information distribution SHOULD be used. 1998 7.1.2. Installation and Initial Setup 2000 Configuration parameters defined in Section 7.2.3 SHOULD be 2001 initialized to the following default values: 2003 o The Link-State NLRI capability is turned off for all neighbors. 2005 o The maximum rate at which Link-State NLRIs will be advertised/ 2006 withdrawn from neighbors is set to 200 updates per second. 2008 7.1.3. Migration Path 2010 The proposed extension is only activated between BGP peers after 2011 capability negotiation. Moreover, the extensions can be turned on/ 2012 off on an individual peer basis (see Section 7.2.3), so the extension 2013 can be gradually rolled out in the network. 2015 7.1.4. Requirements on Other Protocols and Functional Components 2017 The protocol extension defined in this document does not put new 2018 requirements on other protocols or functional components. 2020 7.1.5. Impact on Network Operation 2022 Frequency of Link-State NLRI updates could interfere with regular BGP 2023 prefix distribution. A network operator MAY use a dedicated Route- 2024 Reflector infrastructure to distribute Link-State NLRIs. 2026 Distribution of Link-State NLRIs SHOULD be limited to a single admin 2027 domain, which can consist of multiple areas within an AS or multiple 2028 ASes. 2030 7.1.6. Verifying Correct Operation 2032 Existing BGP procedures apply. In addition, an implementation SHOULD 2033 allow an operator to: 2035 o List neighbors with whom the speaker is exchanging Link-State 2036 NLRIs. 2038 7.2. Management Considerations 2040 7.2.1. Management Information 2042 The IDR working group has documented and continues to document parts 2043 of the Management Information Base and YANG models for managing and 2044 monitoring BGP speakers and the sessions between them. It is 2045 currently believed that the BGP session running BGP-LS is not 2046 substantially different from any other BGP session and can be managed 2047 using the same data models. 2049 7.2.2. Fault Management 2051 This section describes the fault management actions, as described in 2052 [RFC7606] , that are to be performed for handling of BGP update 2053 messages for BGP-LS. 2055 A Link-State NLRI MUST NOT be considered as malformed or invalid 2056 based on the inclusion/exclusion of TLVs or contents of the TLV 2057 fields (i.e. semantic errors), as described in Section 4.1 and 2058 Section 4.2. 2060 A BGP-LS Speaker MUST perform the following syntactic validation of 2061 the Link-State NLRI to determine if it is malformed. 2063 o Does the sum of all TLVs found in the BGP MP_REACH_NLRI attribute 2064 correspond to the BGP MP_REACH_NLRI length? 2066 o Does the sum of all TLVs found in the BGP MP_UNREACH_NLRI 2067 attribute correspond to the BGP MP_UNREACH_NLRI length? 2069 o Does the sum of all TLVs found in a Link-State NLRI correspond to 2070 the Total NLRI Length field of all its Descriptors? 2072 o Is the length of the TLVs and, when the TLV is recognized then, 2073 its sub-TLVs in the NLRI valid? 2075 o Has the syntactic correctness of the NLRI fields been verified as 2076 per [RFC7606]? 2078 o Has the rule regarding ordering of TLVs been followed as described 2079 in Section 4.1? 2081 When the error determined allows for the router to skip the malformed 2082 NLRI(s) and continue processing of the rest of the update message 2083 (e.g. when the TLV ordering rule is violated), then it MUST handle 2084 such malformed NLRIs as 'Treat-as-withdraw'. In other cases, where 2085 the error in the NLRI encoding results in the inability to process 2086 the BGP update message (e.g. length related encoding errors), then 2087 the router SHOULD handle such malformed NLRIs as 'AFI/SAFI disable' 2088 when other AFI/SAFI besides BGP-LS are being advertised over the same 2089 session. Alternately, the router MUST perform 'session reset' when 2090 the session is only being used for BGP-LS or when it 'AFI/SAFI 2091 disable' action is not possible. 2093 A BGP-LS Attribute MUST NOT be considered as malformed or invalid 2094 based on the inclusion/exclusion of TLVs or contents of the TLV 2095 fields (i.e. semantic errors), as described in Section 4.1 and 2096 Section 4.3. 2098 A BGP-LS Speaker MUST perform the following syntactic validation of 2099 the BGP-LS Attribute to determine if it is malformed. 2101 o Does the sum of all TLVs found in the BGP-LS Attribute correspond 2102 to the BGP-LS Attribute length? 2104 o Has the syntactic correctness of the Attributes (including BGP-LS 2105 Attribute) been verified as per [RFC7606]? 2107 o Is the length of each TLV and, when the TLV is recognized then, 2108 its sub-TLVs in the BGP-LS Attribute valid? 2110 When the error determined allows for the router to skip the malformed 2111 BGP-LS Attribute and continue processing of the rest of the update 2112 message (e.g. when the BGP-LS Attribute length and the total Path 2113 Attribute Length are correct but some TLV/sub-TLV length within the 2114 BGP-LS Attribute is invalid), then it MUST handle such malformed BGP- 2115 LS Attribute as 'Attribute Discard'. In other cases, where the error 2116 in the BGP-LS Attribute encoding results in the inability to process 2117 the BGP update message then the handling is the same as described 2118 above for the malformed NLRI. 2120 Note that the 'Attribute Discard' action results in the loss of all 2121 TLVs in the BGP-LS Attribute and not the removal of a specific 2122 malformed TLV. The removal of specific malformed TLVs may give a 2123 wrong indication to a BGP-LS Consumer of that specific information 2124 being deleted or not available. 2126 When a BGP Speaker receives an update message with Link-State NLRI(s) 2127 in the MP_REACH_NLRI but without the BGP-LS Attribute, it is most 2128 likely an indication that a BGP Speaker preceding it has performed 2129 the 'Attribute Discard' fault handling. An implementation SHOULD 2130 preserve and propagate the Link-State NLRIs in such an update message 2131 so that the BGP-LS Consumers can detect the loss of link-state 2132 information for that object and not assume its deletion/withdraw. 2133 This also makes it possible for a network operator to trace back to 2134 the BGP-LS Propagator which actually detected a fault with the BGP-LS 2135 Attribute. 2137 An implementation SHOULD log an error for any errors found during 2138 syntax validation for further analysis. 2140 A BGP-LS Propagator SHOULD NOT perform semantic validation of the 2141 Link-State NLRI or the BGP-LS Attribute to determine if it is 2142 malformed or invalid. Some types of semantic validation that are not 2143 to be performed by a BGP-LS Propagator are as follows (and this is 2144 not to be considered as an exhaustive list): 2146 o is a mandatory TLV present or not? 2148 o is the length of a fixed length TLV correct or the length of a 2149 variable length TLV a valid/permissible? 2151 o are the values of TLV fields valid or permissible? 2153 o are the inclusion and use of TLVs/sub-TLVs with specific Link- 2154 State NLRI types valid? 2156 Each TLV MAY indicate the valid and permissible values and their 2157 semantics that can to be used only by a BGP-LS Consumer for its 2158 semantic validation. However, the handling of any errors may be 2159 specific to the particular application and outside the scope of this 2160 document. A BGP-LS Consumer should ignore unrecognized and 2161 unexpected TLV types in both the NLRI and BGP-LS Attribute portions 2162 and not consider their presence as an error. 2164 7.2.3. Configuration Management 2166 An implementation SHOULD allow the operator to specify neighbors to 2167 which Link-State NLRIs will be advertised and from which Link-State 2168 NLRIs will be accepted. 2170 An implementation SHOULD allow the operator to specify the maximum 2171 rate at which Link-State NLRIs will be advertised/withdrawn from 2172 neighbors. 2174 An implementation SHOULD allow the operator to specify the maximum 2175 number of Link-State NLRIs stored in a router's Routing Information 2176 Base (RIB). 2178 An implementation SHOULD allow the operator to create abstracted 2179 topologies that are advertised to neighbors and create different 2180 abstractions for different neighbors. 2182 An implementation SHOULD allow the operator to configure a 64-bit 2183 Instance-ID. 2185 An implementation SHOULD allow the operator to configure ASN and BGP- 2186 LS identifiers (refer Section 4.2.1.4). 2188 An implementation SHOULD allow the operator to configure the maximum 2189 size of the BGP-LS Attribute that may be used on a BGP-LS Producer. 2191 7.2.4. Accounting Management 2193 Not Applicable. 2195 7.2.5. Performance Management 2197 An implementation SHOULD provide the following statistics: 2199 o Total number of Link-State NLRI updates sent/received 2201 o Number of Link-State NLRI updates sent/received, per neighbor 2203 o Number of errored received Link-State NLRI updates, per neighbor 2205 o Total number of locally originated Link-State NLRIs 2207 These statistics should be recorded as absolute counts since system 2208 or session start time. An implementation MAY also enhance this 2209 information by recording peak per-second counts in each case. 2211 7.2.6. Security Management 2213 An operator SHOULD define an import policy to limit inbound updates 2214 as follows: 2216 o Drop all updates from peers that are only serving BGP-LS 2217 Consumers. 2219 An implementation MUST have the means to limit inbound updates. 2221 8. TLV/Sub-TLV Code Points Summary 2223 This section contains the global table of all TLVs/sub-TLVs defined 2224 in this document. 2226 +-----------+---------------------+--------------+------------------+ 2227 | TLV Code | Description | IS-IS TLV/ | Reference | 2228 | Point | | Sub-TLV | (RFC/Section) | 2229 +-----------+---------------------+--------------+------------------+ 2230 | 256 | Local Node | --- | Section 4.2.1.2 | 2231 | | Descriptors | | | 2232 | 257 | Remote Node | --- | Section 4.2.1.3 | 2233 | | Descriptors | | | 2234 | 258 | Link Local/Remote | 22/4 | [RFC5307] / 1.1 | 2235 | | Identifiers | | | 2236 | 259 | IPv4 interface | 22/6 | [RFC5305] / 3.2 | 2237 | | address | | | 2238 | 260 | IPv4 neighbor | 22/8 | [RFC5305] / 3.3 | 2239 | | address | | | 2240 | 261 | IPv6 interface | 22/12 | [RFC6119] / 4.2 | 2241 | | address | | | 2242 | 262 | IPv6 neighbor | 22/13 | [RFC6119] / 4.3 | 2243 | | address | | | 2244 | 263 | Multi-Topology ID | --- | Section 4.2.2.1 | 2245 | 264 | OSPF Route Type | --- | Section 4.2.3 | 2246 | 265 | IP Reachability | --- | Section 4.2.3 | 2247 | | Information | | | 2248 | 512 | Autonomous System | --- | Section 4.2.1.4 | 2249 | 513 | BGP-LS Identifier | --- | Section 4.2.1.4 | 2250 | | (deprecated) | | | 2251 | 514 | OSPF Area-ID | --- | Section 4.2.1.4 | 2252 | 515 | IGP Router-ID | --- | Section 4.2.1.4 | 2253 | 1024 | Node Flag Bits | --- | Section 4.3.1.1 | 2254 | 1025 | Opaque Node | --- | Section 4.3.1.5 | 2255 | | Attribute | | | 2256 | 1026 | Node Name | variable | Section 4.3.1.3 | 2257 | 1027 | IS-IS Area | variable | Section 4.3.1.2 | 2258 | | Identifier | | | 2259 | 1028 | IPv4 Router-ID of | 134/--- | [RFC5305] / 4.3 | 2260 | | Local Node | | | 2261 | 1029 | IPv6 Router-ID of | 140/--- | [RFC6119] / 4.1 | 2262 | | Local Node | | | 2263 | 1030 | IPv4 Router-ID of | 134/--- | [RFC5305] / 4.3 | 2264 | | Remote Node | | | 2265 | 1031 | IPv6 Router-ID of | 140/--- | [RFC6119] / 4.1 | 2266 | | Remote Node | | | 2267 | 1088 | Administrative | 22/3 | [RFC5305] / 3.1 | 2268 | | group (color) | | | 2269 | 1089 | Maximum link | 22/9 | [RFC5305] / 3.4 | 2270 | | bandwidth | | | 2271 | 1090 | Max. reservable | 22/10 | [RFC5305] / 3.5 | 2272 | | link bandwidth | | | 2273 | 1091 | Unreserved | 22/11 | [RFC5305] / 3.6 | 2274 | | bandwidth | | | 2275 | 1092 | TE Default Metric | 22/18 | Section 4.3.2.3 | 2276 | 1093 | Link Protection | 22/20 | [RFC5307] / 1.2 | 2277 | | Type | | | 2278 | 1094 | MPLS Protocol Mask | --- | Section 4.3.2.2 | 2279 | 1095 | IGP Metric | --- | Section 4.3.2.4 | 2280 | 1096 | Shared Risk Link | --- | Section 4.3.2.5 | 2281 | | Group | | | 2282 | 1097 | Opaque Link | --- | Section 4.3.2.6 | 2283 | | Attribute | | | 2284 | 1098 | Link Name | --- | Section 4.3.2.7 | 2285 | 1152 | IGP Flags | --- | Section 4.3.3.1 | 2286 | 1153 | IGP Route Tag | --- | [RFC5130] | 2287 | 1154 | IGP Extended Route | --- | [RFC5130] | 2288 | | Tag | | | 2289 | 1155 | Prefix Metric | --- | [RFC5305] | 2290 | 1156 | OSPF Forwarding | --- | [RFC2328] | 2291 | | Address | | | 2292 | 1157 | Opaque Prefix | --- | Section 4.3.3.6 | 2293 | | Attribute | | | 2294 +-----------+---------------------+--------------+------------------+ 2296 Table 12: Summary Table of TLV/Sub-TLV Code Points 2298 9. Security Considerations 2300 Procedures and protocol extensions defined in this document do not 2301 affect the BGP security model. See the Security Considerations 2302 section of [RFC4271] for a discussion of BGP security. Also refer to 2303 [RFC4272] and [RFC6952] for analysis of security issues for BGP. 2305 In the context of the BGP peerings associated with this document, a 2306 BGP speaker MUST NOT accept updates from a peer that is only 2307 providing information to a BGP-LS Consumer. That is, a participating 2308 BGP speaker should be aware of the nature of its relationships for 2309 link-state relationships and should protect itself from peers sending 2310 updates that either represent erroneous information feedback loops or 2311 are false input. Such protection can be achieved by manual 2312 configuration of consumer peers at the BGP speaker. 2314 An operator SHOULD employ a mechanism to protect a BGP speaker 2315 against DDoS attacks from BGP-LS Consumers. The principal attack a 2316 consumer may apply is to attempt to start multiple sessions either 2317 sequentially or simultaneously. Protection can be applied by 2318 imposing rate limits. 2320 Additionally, it may be considered that the export of link-state and 2321 TE information as described in this document constitutes a risk to 2322 confidentiality of mission-critical or commercially sensitive 2323 information about the network. BGP peerings are not automatic and 2324 require configuration; thus, it is the responsibility of the network 2325 operator to ensure that only trusted consumers are configured to 2326 receive such information. 2328 10. Contributors 2330 The following persons contributed significant text to RFC7752 and 2331 this document. They should be considered as co-authors. 2333 Hannes Gredler 2334 Rtbrick 2335 Email: hannes@rtbrick.com 2337 Jan Medved 2338 Cisco Systems Inc. 2339 USA 2340 Email: jmedved@cisco.com 2342 Stefano Previdi 2343 Huawei Technologies 2344 Italy 2345 Email: stefano@previdi.net 2347 Adrian Farrel 2348 Old Dog Consulting 2349 Email: adrian@olddog.co.uk 2351 Saikat Ray 2352 Individual 2353 USA 2354 Email: raysaikat@gmail.com 2356 11. Acknowledgements 2358 This document update to the BGP-LS specification [RFC7752] is a 2359 result of feedback and inputs from the discussions in the IDR working 2360 group. It also incorporates certain details and clarifications based 2361 on implementation and deployment experience with BGP-LS. 2363 Cengiz Alaettinoglu and Parag Amritkar brought forward the need to 2364 clarify the advertisement of LAN subnet for OSPF. 2366 We would like to thank Balaji Rajagopalan, Srihari Sangli, Shraddha 2367 Hegde, Andrew Stone, Jeff Tantsura, Acee Lindem, Jie Dong, Aijun Wang 2368 and Nandan Saha for their review and feedback on this document. 2370 We would like to thank Robert Varga for the significant contribution 2371 he gave to RFC7752. 2373 We would like to thank Nischal Sheth, Alia Atlas, David Ward, Derek 2374 Yeung, Murtuza Lightwala, John Scudder, Kaliraj Vairavakkalai, Les 2375 Ginsberg, Liem Nguyen, Manish Bhardwaj, Matt Miller, Mike Shand, 2376 Peter Psenak, Rex Fernando, Richard Woundy, Steven Luong, Tamas 2377 Mondal, Waqas Alam, Vipin Kumar, Naiming Shen, Carlos Pignataro, 2378 Balaji Rajagopalan, Yakov Rekhter, Alvaro Retana, Barry Leiba, and 2379 Ben Campbell for their comments on RFC7752. 2381 12. References 2383 12.1. Normative References 2385 [ISO10589] 2386 International Organization for Standardization, 2387 "Intermediate System to Intermediate System intra-domain 2388 routeing information exchange protocol for use in 2389 conjunction with the protocol for providing the 2390 connectionless-mode network service (ISO 8473)", ISO/ 2391 IEC 10589, November 2002. 2393 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 2394 Requirement Levels", BCP 14, RFC 2119, 2395 DOI 10.17487/RFC2119, March 1997, 2396 . 2398 [RFC2328] Moy, J., "OSPF Version 2", STD 54, RFC 2328, 2399 DOI 10.17487/RFC2328, April 1998, 2400 . 2402 [RFC2545] Marques, P. and F. Dupont, "Use of BGP-4 Multiprotocol 2403 Extensions for IPv6 Inter-Domain Routing", RFC 2545, 2404 DOI 10.17487/RFC2545, March 1999, 2405 . 2407 [RFC3209] Awduche, D., Berger, L., Gan, D., Li, T., Srinivasan, V., 2408 and G. Swallow, "RSVP-TE: Extensions to RSVP for LSP 2409 Tunnels", RFC 3209, DOI 10.17487/RFC3209, December 2001, 2410 . 2412 [RFC4202] Kompella, K., Ed. and Y. Rekhter, Ed., "Routing Extensions 2413 in Support of Generalized Multi-Protocol Label Switching 2414 (GMPLS)", RFC 4202, DOI 10.17487/RFC4202, October 2005, 2415 . 2417 [RFC4203] Kompella, K., Ed. and Y. Rekhter, Ed., "OSPF Extensions in 2418 Support of Generalized Multi-Protocol Label Switching 2419 (GMPLS)", RFC 4203, DOI 10.17487/RFC4203, October 2005, 2420 . 2422 [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A 2423 Border Gateway Protocol 4 (BGP-4)", RFC 4271, 2424 DOI 10.17487/RFC4271, January 2006, 2425 . 2427 [RFC4760] Bates, T., Chandra, R., Katz, D., and Y. Rekhter, 2428 "Multiprotocol Extensions for BGP-4", RFC 4760, 2429 DOI 10.17487/RFC4760, January 2007, 2430 . 2432 [RFC4915] Psenak, P., Mirtorabi, S., Roy, A., Nguyen, L., and P. 2433 Pillay-Esnault, "Multi-Topology (MT) Routing in OSPF", 2434 RFC 4915, DOI 10.17487/RFC4915, June 2007, 2435 . 2437 [RFC5036] Andersson, L., Ed., Minei, I., Ed., and B. Thomas, Ed., 2438 "LDP Specification", RFC 5036, DOI 10.17487/RFC5036, 2439 October 2007, . 2441 [RFC5120] Przygienda, T., Shen, N., and N. Sheth, "M-ISIS: Multi 2442 Topology (MT) Routing in Intermediate System to 2443 Intermediate Systems (IS-ISs)", RFC 5120, 2444 DOI 10.17487/RFC5120, February 2008, 2445 . 2447 [RFC5130] Previdi, S., Shand, M., Ed., and C. Martin, "A Policy 2448 Control Mechanism in IS-IS Using Administrative Tags", 2449 RFC 5130, DOI 10.17487/RFC5130, February 2008, 2450 . 2452 [RFC5301] McPherson, D. and N. Shen, "Dynamic Hostname Exchange 2453 Mechanism for IS-IS", RFC 5301, DOI 10.17487/RFC5301, 2454 October 2008, . 2456 [RFC5305] Li, T. and H. Smit, "IS-IS Extensions for Traffic 2457 Engineering", RFC 5305, DOI 10.17487/RFC5305, October 2458 2008, . 2460 [RFC5307] Kompella, K., Ed. and Y. Rekhter, Ed., "IS-IS Extensions 2461 in Support of Generalized Multi-Protocol Label Switching 2462 (GMPLS)", RFC 5307, DOI 10.17487/RFC5307, October 2008, 2463 . 2465 [RFC5340] Coltun, R., Ferguson, D., Moy, J., and A. Lindem, "OSPF 2466 for IPv6", RFC 5340, DOI 10.17487/RFC5340, July 2008, 2467 . 2469 [RFC5642] Venkata, S., Harwani, S., Pignataro, C., and D. McPherson, 2470 "Dynamic Hostname Exchange Mechanism for OSPF", RFC 5642, 2471 DOI 10.17487/RFC5642, August 2009, 2472 . 2474 [RFC5890] Klensin, J., "Internationalized Domain Names for 2475 Applications (IDNA): Definitions and Document Framework", 2476 RFC 5890, DOI 10.17487/RFC5890, August 2010, 2477 . 2479 [RFC6119] Harrison, J., Berger, J., and M. Bartlett, "IPv6 Traffic 2480 Engineering in IS-IS", RFC 6119, DOI 10.17487/RFC6119, 2481 February 2011, . 2483 [RFC6549] Lindem, A., Roy, A., and S. Mirtorabi, "OSPFv2 Multi- 2484 Instance Extensions", RFC 6549, DOI 10.17487/RFC6549, 2485 March 2012, . 2487 [RFC7606] Chen, E., Ed., Scudder, J., Ed., Mohapatra, P., and K. 2488 Patel, "Revised Error Handling for BGP UPDATE Messages", 2489 RFC 7606, DOI 10.17487/RFC7606, August 2015, 2490 . 2492 [RFC7752] Gredler, H., Ed., Medved, J., Previdi, S., Farrel, A., and 2493 S. Ray, "North-Bound Distribution of Link-State and 2494 Traffic Engineering (TE) Information Using BGP", RFC 7752, 2495 DOI 10.17487/RFC7752, March 2016, 2496 . 2498 [RFC8126] Cotton, M., Leiba, B., and T. Narten, "Guidelines for 2499 Writing an IANA Considerations Section in RFCs", BCP 26, 2500 RFC 8126, DOI 10.17487/RFC8126, June 2017, 2501 . 2503 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2504 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 2505 May 2017, . 2507 [RFC8202] Ginsberg, L., Previdi, S., and W. Henderickx, "IS-IS 2508 Multi-Instance", RFC 8202, DOI 10.17487/RFC8202, June 2509 2017, . 2511 [RFC8654] Bush, R., Patel, K., and D. Ward, "Extended Message 2512 Support for BGP", RFC 8654, DOI 10.17487/RFC8654, October 2513 2019, . 2515 12.2. Informative References 2517 [I-D.ietf-idr-bgpls-inter-as-topology-ext] 2518 Wang, A., Chen, H., Talaulikar, K., and S. Zhuang, "BGP-LS 2519 Extension for Inter-AS Topology Retrieval", draft-ietf- 2520 idr-bgpls-inter-as-topology-ext-08 (work in progress), 2521 April 2020. 2523 [RFC1918] Rekhter, Y., Moskowitz, B., Karrenberg, D., de Groot, G., 2524 and E. Lear, "Address Allocation for Private Internets", 2525 BCP 5, RFC 1918, DOI 10.17487/RFC1918, February 1996, 2526 . 2528 [RFC4272] Murphy, S., "BGP Security Vulnerabilities Analysis", 2529 RFC 4272, DOI 10.17487/RFC4272, January 2006, 2530 . 2532 [RFC4364] Rosen, E. and Y. Rekhter, "BGP/MPLS IP Virtual Private 2533 Networks (VPNs)", RFC 4364, DOI 10.17487/RFC4364, February 2534 2006, . 2536 [RFC4655] Farrel, A., Vasseur, J., and J. Ash, "A Path Computation 2537 Element (PCE)-Based Architecture", RFC 4655, 2538 DOI 10.17487/RFC4655, August 2006, 2539 . 2541 [RFC5073] Vasseur, J., Ed. and J. Le Roux, Ed., "IGP Routing 2542 Protocol Extensions for Discovery of Traffic Engineering 2543 Node Capabilities", RFC 5073, DOI 10.17487/RFC5073, 2544 December 2007, . 2546 [RFC5152] Vasseur, JP., Ed., Ayyangar, A., Ed., and R. Zhang, "A 2547 Per-Domain Path Computation Method for Establishing Inter- 2548 Domain Traffic Engineering (TE) Label Switched Paths 2549 (LSPs)", RFC 5152, DOI 10.17487/RFC5152, February 2008, 2550 . 2552 [RFC5316] Chen, M., Zhang, R., and X. Duan, "ISIS Extensions in 2553 Support of Inter-Autonomous System (AS) MPLS and GMPLS 2554 Traffic Engineering", RFC 5316, DOI 10.17487/RFC5316, 2555 December 2008, . 2557 [RFC5392] Chen, M., Zhang, R., and X. Duan, "OSPF Extensions in 2558 Support of Inter-Autonomous System (AS) MPLS and GMPLS 2559 Traffic Engineering", RFC 5392, DOI 10.17487/RFC5392, 2560 January 2009, . 2562 [RFC5693] Seedorf, J. and E. Burger, "Application-Layer Traffic 2563 Optimization (ALTO) Problem Statement", RFC 5693, 2564 DOI 10.17487/RFC5693, October 2009, 2565 . 2567 [RFC5706] Harrington, D., "Guidelines for Considering Operations and 2568 Management of New Protocols and Protocol Extensions", 2569 RFC 5706, DOI 10.17487/RFC5706, November 2009, 2570 . 2572 [RFC6952] Jethanandani, M., Patel, K., and L. Zheng, "Analysis of 2573 BGP, LDP, PCEP, and MSDP Issues According to the Keying 2574 and Authentication for Routing Protocols (KARP) Design 2575 Guide", RFC 6952, DOI 10.17487/RFC6952, May 2013, 2576 . 2578 [RFC7285] Alimi, R., Ed., Penno, R., Ed., Yang, Y., Ed., Kiesel, S., 2579 Previdi, S., Roome, W., Shalunov, S., and R. Woundy, 2580 "Application-Layer Traffic Optimization (ALTO) Protocol", 2581 RFC 7285, DOI 10.17487/RFC7285, September 2014, 2582 . 2584 [RFC7770] Lindem, A., Ed., Shen, N., Vasseur, JP., Aggarwal, R., and 2585 S. Shaffer, "Extensions to OSPF for Advertising Optional 2586 Router Capabilities", RFC 7770, DOI 10.17487/RFC7770, 2587 February 2016, . 2589 [RFC7911] Walton, D., Retana, A., Chen, E., and J. Scudder, 2590 "Advertisement of Multiple Paths in BGP", RFC 7911, 2591 DOI 10.17487/RFC7911, July 2016, 2592 . 2594 Appendix A. Changes from RFC 7752 2596 This section lists the high-level changes from RFC 7752 and provides 2597 reference to the document sections wherein those have been 2598 introduced. 2600 1. Update the Figure 1 in Section 1 and added Section 3 to 2601 illustrate the different roles of a BGP implementation in 2602 conveying link-state information. 2604 2. In Section 4.1, clarification about the TLV handling aspects 2605 that are applicable to both the NLRI and BGP-LS Attribute parts 2606 and those that are applicable only for the NLRI portion. An 2607 implementation may have missed the part about handling of 2608 unrecognized TLV and so, based on [RFC7606] guidelines, might 2609 discard the unknown NLRI types. This aspect is now 2610 unambiguously clarified in Section 4.2. Also, the TLVs in the 2611 BGP-LS Attribute that are not ordered are not to be considered 2612 as malformed. 2614 3. Clarification of mandatory and optional TLVs in both NLRI and 2615 BGP-LS Attribute portions all through the document. 2617 4. Handling of the growth of the BGP-LS Attribute is covered in 2618 Section 4.3. 2620 5. Clarified that the document describes the NLRI descriptor TLVs 2621 for the protocols and NLRI types specified in this document and 2622 future BGP-LS extensions must describe the same for other 2623 protocols and NLRI types that they introduce. 2625 6. Clarification on the use of Identifier field in the Link-State 2626 NLRI in Section 4.2 is provided. It was defined ambiguously to 2627 refer to only mutli-instance IGP on a single link while it can 2628 also be used for multiple IGP protocol instances on a router. 2629 The IANA registry is accordingly being removed. 2631 7. The BGP-LS Identifier TLV in the Node Descriptors has been 2632 deprecated. Its use was not well specified by [RFC7752] and 2633 there has been some amount of confusion between implementators 2634 on its usage for identification of IGP domains as against the 2635 use of the Identifier doing the same functionality as the 2636 Instance-ID when running multiple instances of IGP routing 2637 protocols. 2639 8. Clarification that the Area-ID TLV is mandatory in the Node 2640 Descriptor for origination of information from OSPF except for 2641 when sourcing information from AS-scope LSAs where this TLV is 2642 not applicable. 2644 9. Moved MT-ID TLV from the Node Descriptor section to under the 2645 Link Descriptor section since it is not a Node Descriptor sub- 2646 TLV. Fixed the ambiguity in the encoding of OSPF MT-ID in this 2647 TLV. Updated the IS-IS specification reference section and 2648 describe the differences in the applicability of the R flags 2649 when MT-ID TLV is used as link descriptor TLV and Prefix 2650 Attribute TLV. MT-ID TLV use is now elevated to SHOULD when it 2651 is enabled in the underlying IGP. 2653 10. Clarified that IPv6 Link-Local Addresses are not advertised in 2654 the Link Descriptor TLVs and the local/remote identifiers are to 2655 be used instead for links with IPv6 link-local addresses only. 2657 11. Update the usage of OSPF Route Type TLV to mandate its use for 2658 OSPF prefixes in Section 4.2.3.1 since this is required for 2659 segregation of intra-area prefixes that are used to reach a node 2660 (e.g. a loopback) from other types of inter-area and external 2661 prefixes. 2663 12. Clarification on the length of the Node Flag Bits TLV to be one 2664 octet. 2666 13. Updated the Node Name TLV in Section 4.3.1.3 with the OSPF 2667 specification. 2669 14. Clarification on the size of the IS-IS Narrow Metric 2670 advertisement via the IGP Metric TLV and the handling of the 2671 unused bits. 2673 15. Clarified the advertisement of the prefix corresponding to the 2674 LAN segment in an OSPF network in Section 4.9. 2676 16. Introduced Private Use TLV code point space and specified their 2677 encoding in Section 4.4. 2679 17. Introduced Section 4.7 where issues related to consistency of 2680 reporting IGP link-state along with their solutions are covered. 2682 18. Handling of large size of BGP-LS Attribute with growth in BGP-LS 2683 information is explained in Section 4.3 along with mitigation of 2684 errors arising out of it. 2686 19. Added recommendation for isolation of BGP-LS sessions from other 2687 BGP route exchange to avoid errors and faults in BGP-LS 2688 affecting the normal BGP routing. 2690 20. Updated the Fault Management section with detailed rules based 2691 on the role in the BGP-LS information propagation flow. 2693 21. Change to the management of BGP-LS IANA registries from 2694 "Specification Required" to "Expert Review" along with updated 2695 guidelines for Designated Experts. 2697 Author's Address 2699 Ketan Talaulikar (editor) 2700 Cisco Systems 2701 India 2703 Email: ketant@cisco.com