idnits 2.17.1 draft-ietf-lsvr-applicability-02.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (May 1, 2019) is 1814 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Outdated reference: A later version (-29) exists of draft-ietf-lsvr-bgp-spf-04 == Outdated reference: A later version (-17) exists of draft-acee-idr-lldp-peer-discovery-04 == Outdated reference: A later version (-18) exists of draft-ietf-lsr-dynamic-flooding-00 == Outdated reference: A later version (-12) exists of draft-ietf-lsvr-l3dl-00 == Outdated reference: A later version (-12) exists of draft-xu-idr-neighbor-autodiscovery-11 -- Obsolete informational reference (is this intentional?): RFC 7752 (Obsoleted by RFC 9552) Summary: 0 errors (**), 0 flaws (~~), 6 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 LSVR K. Patel 3 Internet-Draft Arrcus, Inc. 4 Intended status: Informational A. Lindem 5 Expires: November 2, 2019 Cisco Systems 6 S. Zandi 7 G. Dawra 8 Linkedin 9 May 1, 2019 11 Usage and Applicability of Link State Vector Routing in Data Centers 12 draft-ietf-lsvr-applicability-02.txt 14 Abstract 16 This document discusses the usage and applicability of Link State 17 Vector Routing (LSVR) extensions in data center networks utilizing 18 CLOS or Fat-Tree topologies. The document is intended to provide a 19 simplified guide for the deployment of LSVR extensions. 21 Status of This Memo 23 This Internet-Draft is submitted in full conformance with the 24 provisions of BCP 78 and BCP 79. 26 Internet-Drafts are working documents of the Internet Engineering 27 Task Force (IETF). Note that other groups may also distribute 28 working documents as Internet-Drafts. The list of current Internet- 29 Drafts is at http://datatracker.ietf.org/drafts/current/. 31 Internet-Drafts are draft documents valid for a maximum of six months 32 and may be updated, replaced, or obsoleted by other documents at any 33 time. It is inappropriate to use Internet-Drafts as reference 34 material or to cite them other than as "work in progress." 36 This Internet-Draft will expire on November 2, 2019. 38 Copyright Notice 40 Copyright (c) 2019 IETF Trust and the persons identified as the 41 document authors. All rights reserved. 43 This document is subject to BCP 78 and the IETF Trust's Legal 44 Provisions Relating to IETF Documents 45 (http://trustee.ietf.org/license-info) in effect on the date of 46 publication of this document. Please review these documents 47 carefully, as they describe your rights and restrictions with respect 48 to this document. Code Components extracted from this document must 49 include Simplified BSD License text as described in Section 4.e of 50 the Trust Legal Provisions and are provided without warranty as 51 described in the Simplified BSD License. 53 Table of Contents 55 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 56 2. Requirements Language . . . . . . . . . . . . . . . . . . . . 3 57 3. Recommended Reading . . . . . . . . . . . . . . . . . . . . . 3 58 4. Common Deployment Scenario . . . . . . . . . . . . . . . . . 3 59 5. Justification for BGP SPF Extension . . . . . . . . . . . . . 4 60 6. LSVR Applicability to CLOS Networks . . . . . . . . . . . . . 5 61 6.1. Usage of BGP-LS SPF SAFI . . . . . . . . . . . . . . . . 5 62 6.1.1. Relationship to Other BGP AFI/SAFI Tuples . . . . . . 6 63 6.2. Peering Models . . . . . . . . . . . . . . . . . . . . . 6 64 6.2.1. Sparse Peering Model . . . . . . . . . . . . . . . . 6 65 6.2.2. Bi-Connected Graph Heuristic . . . . . . . . . . . . 7 66 6.3. BGP Spine/Leaf Topology Policy . . . . . . . . . . . . . 7 67 6.4. BGP Peer Discovery Requirements . . . . . . . . . . . . . 8 68 6.5. BGP Peer Discovery . . . . . . . . . . . . . . . . . . . 9 69 6.5.1. BGP Peer Discovery Alternatives . . . . . . . . . . . 9 70 6.5.2. Data Center Interconnect (DCI) Applicability . . . . 9 71 6.6. Non-CLOS/FAT Tree Topology Applicability . . . . . . . . 10 72 7. BGP Policy Applicability . . . . . . . . . . . . . . . . . . 10 73 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 10 74 9. Security Considerations . . . . . . . . . . . . . . . . . . . 10 75 10. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 10 76 11. References . . . . . . . . . . . . . . . . . . . . . . . . . 11 77 11.1. Normative References . . . . . . . . . . . . . . . . . . 11 78 11.2. Informative References . . . . . . . . . . . . . . . . . 11 79 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 13 81 1. Introduction 83 This document complements [I-D.ietf-lsvr-bgp-spf] by discussing the 84 applicability of the technology in a simple and fairly common 85 deployment scenario, which is described in Section 4. 87 After describing the deployment scenario, Section 5 will describe the 88 reasons for BGP modifications for such deployments. 90 Once the control plane routing protocol requirements are described, 91 Section 6 will cover the LSVR protocol enhancements to BGP to meet 92 these requirements and their applicability to Data Center CLOS 93 networks. 95 2. Requirements Language 97 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 98 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 99 "OPTIONAL" in this document are to be interpreted as described in BCP 100 14 [RFC2119] [RFC8174] when, and only when, they appear in all 101 capitals, as shown here. 103 3. Recommended Reading 105 This document assumes knowledge of existing data center networks and 106 data center network topologies [CLOS]. This document also assumes 107 knowledge of data center routing protocols like BGP [RFC4271], BGP- 108 SPF [I-D.ietf-lsvr-bgp-spf], OSPF [RFC2328], as well as, data center 109 OAM protocols like LLDP [RFC4957] and BFD [RFC5580]. 111 4. Common Deployment Scenario 113 Within a Data Center, servers are commonly interconnected the CLOS 114 topology [CLOS]. The CLOS topology is fully non-blocking and the 115 topology is realized using Equal Cost Multi-Path (ECMP). In a CLOS 116 topology, the minimum number of parallel paths between two servers is 117 determined by the width of a tier-1 stage as shown in the figure 1. 119 The following example illustrates multi-stage CLOS topology. 121 Tier-1 122 +-----+ 123 |NODE | 124 +->| 12 |--+ 125 | +-----+ | 126 Tier-2 | | Tier-2 127 +-----+ | +-----+ | +-----+ 128 +------------>|NODE |--+->|NODE |--+--|NODE |-------------+ 129 | +-----| 9 |--+ | 10 | +--| 11 |-----+ | 130 | | +-----+ +-----+ +-----+ | | 131 | | | | 132 | | +-----+ +-----+ +-----+ | | 133 | +-----+---->|NODE |--+ |NODE | +--|NODE |-----+-----+ | 134 | | | +---| 6 |--+->| 7 |--+--| 8 |---+ | | | 135 | | | | +-----+ | +-----+ | +-----+ | | | | 136 | | | | | | | | | | 137 +-----+ +-----+ | +-----+ | +-----+ +-----+ 138 |NODE | |NODE | Tier-3 +->|NODE |--+ Tier-3 |NODE | |NODE | 139 | 1 | | 2 | | 3 | | 4 | | 5 | 140 +-----+ +-----+ +-----+ +-----+ +-----+ 141 | | | | | | | | 142 A O B O <- Servers -> Z O O O 144 Figure 1: Illustration of the basic CLOS 146 5. Justification for BGP SPF Extension 148 In order to simplify layer-3 routing and operations [RFC7938], many 149 data centers use BGP as a routing protocol to create both an underlay 150 and overlay network for their CLOS Topologies. However, BGP is a 151 path-vector routing protocol. Since it does not create a fabric 152 topology, it uses hop-by-hop EBGP peering to facilitate hop-by-hop 153 routing to create the underlay network and to resolve any overlay 154 next hops. The hop-by-hop BGP peering paradigm imposes several 155 restrictions within a CLOS. It severely prohibits a deployment of 156 Route Reflectors/Route Controllers as the EBGP sessions are congruent 157 with the data path. The BGP best-path algorithm is prefix-based and 158 it prevents announcements of prefixes to other BGP speakers until the 159 best-path decision process has been performed for the prefix at each 160 intermediate hop. These restrictions significantly delay the overall 161 convergence of the underlay network within a CLOS network. 163 The LSVR SPF modifications allow BGP to overcome these limitations. 164 Furthermore, using the BGP-LS NLRI format [RFC7752] allows the LSVR 165 data to be advertised for nodes, links, and prefixes in the BGP 166 routing domain and used for SPF computations. 168 6. LSVR Applicability to CLOS Networks 170 With the BGP SPF extensions [I-D.ietf-lsvr-bgp-spf], the BGP best- 171 path computation and route computation are replaced with OSPF-like 172 algorithms [RFC2328] both to determine whether an BGP-LS SPF NLRI has 173 changed and needs to be re-advertised and to compute the BGP routes. 174 These modifications will significantly improve convergence of the 175 underlay while affording the operational benefits of a single routing 176 protocol [RFC7938]. 178 Data center controllers typically require visibility to the BGP 179 topology to compute traffic-engineered paths. These controllers 180 learn the topology and other relevant information via the BGP-LS 181 address family [RFC7752] which is totally independent of the underlay 182 address families (usually IPv4/IPv6 unicast). Furthermore, in 183 traditional BGP underlays, all the BGP routers will need to advertise 184 their BGP-LS information independently. With the BGP SPF extensions, 185 controllers can learn the topology using the same BGP advertisements 186 used to compute the underlay routes. Furthermore, these data center 187 controllers can avail the convergence advantages of the BGP SPF 188 extensions. The placement of controllers can be outside of the 189 forwarding path or within the forwarding path. 191 Alternatively, as each and every router in the BGP SPF domain will 192 have a complete view of the topology, the operator can also choose to 193 configure BGP sessions in hop-by-hop peering model described in 194 [RFC7938] along with BFD [RFC5580]. In doing so, while the hop-by- 195 hop peering model lacks the inherent benefits of the controller-based 196 model, BGP updates need not be serialized by BGP best-path algorithm 197 in either of these models. This helps overall network convergence. 199 6.1. Usage of BGP-LS SPF SAFI 201 The BGP SPF extensions [I-D.ietf-lsvr-bgp-spf] define a new BGP-LS 202 SPF SAFI for announcement of BGP SPF link-state. The NLRI format and 203 its associated attributes follow the format of BGP-LS for node, link, 204 and prefix announcements. Whether the peering model within a CLOS 205 follows hop-by-hop peering described in [RFC7938] or any controller- 206 based or route-reflector peering, an operator can exchange BGP SPF 207 SAFI routes over the BGP peering by simply configuring BGP SPF SAFI 208 between the necessary BGP speakers. 210 The BGP-LS SPF SAFI can also co-exist with BGP IP Unicast SAFI which 211 could exchange overlapping IP routes. The routes received by these 212 SAFIs are evaluated, stored, and announced independently according to 213 the rules of [RFC4760]. The tie-breaking of route installation is a 214 matter of the local policies and preferences of the network operator. 216 Finally, as the BGP SPF peering is done following the procedures 217 described in [RFC4271], all the existing transport security 218 mechanisms including [RFC5925] are available for the BGP-LS SPF SAFI. 220 6.1.1. Relationship to Other BGP AFI/SAFI Tuples 222 Normally, the BGP-LS AFI/SAFI is used solely to compute the underlay 223 and is given preference over other AFI/SAFIs. Other BGP SAFIs, e.g., 224 IPv6/IPv6 Unicast VPN would use the BGP-SPF computed routes for next 225 hop resolution. However, if BGP-LS NLRI is also being advertised for 226 controller consumption, there is no need to replicate the Node, Link, 227 and Prefix NLRI in BGP-NLRI. Rather, additional NLRI attributes can 228 be advertised in the BGP-LS SPF AFI/SAFI as required. 230 6.2. Peering Models 232 As previously stated, BGP SPF can be deployed using the existing 233 peering model where there is a single-hop BGP session on each and 234 every link in the data center fabric [RFC7938]. This provides for 235 both the advertisement of routes and the determination of link and 236 neighboring switch availability. With BGP SPF, the underlay will 237 converge faster due to changes to the decision process that will 238 allow NLRI changes to be advertised faster after detecting a change. 240 6.2.1. Sparse Peering Model 242 Alternately, BFD [RFC5580] can be used to swiftly determine the 243 availability of links and the BGP peering model can be significantly 244 sparser than the data center fabric. BGP SPF sessions only need to 245 be established with enough peers to provide a bi-connected graph. If 246 IEBGP is used, then the BGP routers at tier N-1 will act as route- 247 reflectors for the routers at tier N. 249 The obvious usage of sparse peering is to avoid parallel sessions on 250 links between the same two BGP speakers in the data center fabric. 251 However, this use case is not very useful since parallel layer-3 252 links between the same two BGP routers are rare in CLOS or Fat-Tree 253 topologies. Two more interesting scenarios are described below. 255 In current data center topologies, there is often a very dense mesh 256 of links between levels, e.g., leaf and spine, providing 32-way, 257 64-way, or more Equal-Cost Multi-Path (ECMP) paths. In these 258 topologies, it is desirable not to have a BGP session on every link 259 and techniques such as the one described in Section 6.2.2 can be used 260 establish sessions on some subset of northbound links. 262 Alternately, controller-based data center topologies are envisioned 263 where BGP speakers within the data center only establish BGP sessions 264 with two or more controllers. In these topologies, fabric nodes 265 below the first tier (using [RFC7938] hierarchy) will establish BGP 266 multi-hop sessions with the controllers. For the multi-hop sessions, 267 determining the route to the controllers without depending on BGP 268 would need to be through some other means beyond the scope of this 269 document. However, the BGP discovery mechanisms described in 270 Section 6.5 would be one possibility. 272 6.2.2. Bi-Connected Graph Heuristic 274 With this heuristic, discovery of BGP peers is assumed, e.g., as 275 described in Section 6.5. Additionally, it assumed that the 276 direction of the peering can be ascertained. In the context of a 277 data center fabric, direction is either northbound (toward the 278 spine), southbound (toward the Top-Of-Rack (TOR) switches) or east- 279 west (same level in hierarchy. The determination of the direction is 280 beyond the scope of this document. However, it would be reasonable 281 to assume a technique where the TOR switches can be identified and 282 the number of hops to the TOR is used to determine the direction. 284 In this heuristic, BGP speakers allow passive session establishment 285 for southbound BGP sessions. For northbound sessions, BGP speakers 286 will attempt to maintain two northbound BGP sessions with different 287 switches (in data center fabrics there is normally a single layer-3 288 connection anyway). For east-west sessions, passive BGP session 289 establishment is allowed. However, BGP speaker will never actively 290 establish an east-west BGP session unless it can't establish two 291 northbound BGP sessions. 293 6.3. BGP Spine/Leaf Topology Policy 295 One of the advantages of using BGP SPF as the underlay protocol is 296 that BGP policy can be applied at any level. In Spine/Leaf 297 topologies, it is not necessary to advertise BGP-LS NLRI received by 298 leaves northbound to the spine nodes at the level above. If a common 299 AS is used for the spine nodes, This can easily be accomplished with 300 EBGP and a simple policy to filter advertisements from the leaves to 301 the spine if the first AS in the AS path is the spine AS. 303 In the figure below, the leaves would not advertise any NLRI with AS 304 64512 as the first AS in the AS path. 306 +--------+ +--------+ +--------+ 307 AS 64512 | | | | | | 308 for Spine | Spine1 +----+ Spine2 +- ......... -+ SpineN | 309 Nodes at | | | | | | 310 this Level +-+-+-+-++ ++-+-+-+-+ +-+-+-+-++ 311 +------+ | | | | | | | | | | | 312 | +-----|-|-|------+ | | | | | | | 313 | | +--|-|-|--------+-|-|-----------------+ | | | 314 | | | | | | +---+ | | | | | 315 | | | | | | | +--|-|-------------------+ | | 316 | | | | | | | | | | +------+ +----+ 317 | | | | | | | | | +--------------|----------+ | 318 | | | | | | | | +-------------+ | | | 319 | | | | | +----|--|----------------|--|--------+ | | 320 | | | | +------|--|--------------+ | | | | | 321 | | | +------+ | | | | | | | | 322 ++--+--++ +-+-+--++ ++-+--+-+ ++-+--+-+ 323 | Leaf1 |~~~~~~| Leaf2 | ........ | LeafX | | LeafY | 324 +-------+ +-------+ +-------+ +-------+ 326 Figure 2: Spine/Leaf Topology Policy 328 6.4. BGP Peer Discovery Requirements 330 The most basic requirement is to be able to discover the address of a 331 single-hop peer without pre-configuration. This is being 332 accomplished today with using IPv6 Router Advertisements (RA) 333 [RFC4861] and assuming that a BGP sessions is desired with any 334 discovered peer. Beyond the basic requirement, it is useful to have 335 to following information relating to the BGP session: 337 o Autonomous System (AS) and BGP Identifier of a potential peer. 338 The latter can be used for debugging and to decrease the 339 likelihood of BGP session establishment collisions. 341 o Security capabilities supported and for cryptographic 342 authentication, the security capabilities and possibly a key-chain 343 [RFC8177] to be used. 345 o Session Policy Identifier - A group number or name used to 346 associate common session parameters with the peer. For example, 347 in a data center, BGP sessions with a Top of Rack (ToR) device 348 could have parameters than BGP sessions between leaf and spine. 350 In a data center fabric, it is often useful to know whether a peer is 351 southbound (towards the servers) or northbound (towards the spine or 352 super-spine), e.g., Section 6.2.2. A potential requirement would be 353 to determine this dynamically. One mechanism, without specifying all 354 the details, might be for the ToRs to be identified when installed 355 and for the others switches in the fabric to determine their level 356 based on the distance from the closest ToR. 358 If there are multiple links between BGP speakers or the links between 359 BGP speakers are unnumbered, it is also useful to be able to 360 establish multi-hop sessions using the loopback addresses. This will 361 often require the discovery protocol to install route(s) toward the 362 potential peer loopback addresses prior to BGP session establishment. 364 Finally, a simple BGP discovery protocol could also be used to 365 establish a multi-hop session with one or more controllers by 366 advertising connectivity to one or more controllers. However, once 367 the multi-hop session actually traverses multiple nodes, it is 368 bordering a distance-vector routing protocol and possibly this is not 369 a good requirement for the discovery protocol. 371 6.5. BGP Peer Discovery 373 6.5.1. BGP Peer Discovery Alternatives 375 While BGP peer discovery is not part of [I-D.ietf-lsvr-bgp-spf], 376 there are, at least, three proposals for BGP peer discovery. At 377 least one of these mechanisms will be adopted and will be applicable 378 to deployments other than the data center. It is strongly 379 RECOMMENDED that the accepted mechanism be used in conjunction with 380 BGP SPF in data centers. The BGP discovery mechanism should 381 discovery both peer addresses and endpoints for BFD discovery. 382 Additionally, it would be great if there were a heuristic for 383 determining whether the peer is at a tier above or below the 384 discovering BGP speaker (refer to Section 6.2.2). 386 The BGP discovery mechanisms under consideration are 387 [I-D.acee-idr-lldp-peer-discovery], 388 [I-D.xu-idr-neighbor-autodiscovery], and [I-D.ietf-lsvr-l3dl]. 390 6.5.2. Data Center Interconnect (DCI) Applicability 392 Since BGP SPF is to be used for the routing underlay and DCI gateway 393 boxes typically have direct or very simple connectivity, BGP external 394 sessions would typically not include the BGP SPF SAFI. 396 6.6. Non-CLOS/FAT Tree Topology Applicability 398 The BGP SPF extensions [I-D.ietf-lsvr-bgp-spf] can be used in other 399 topologies and avail the inherent convergence improvements. 400 Additionally, sparse peering techniques may be utilized Section 6.2. 401 However, determining whether or to establish a BGP session is more 402 complex and the heuristic described in Section 6.2.2 cannot be used. 403 In such topologies, other techniques such as those described in 404 [I-D.ietf-lsr-dynamic-flooding] may be employed. One potential 405 deployment would be the underlay for a Service Provider (SP) backbone 406 where usage of a single protocol, i.e., BGP, is desired. 408 7. BGP Policy Applicability 410 Existing BGP policy including aggregation and prefix filtering may be 411 used in conjunction with the BGP-LS SPF SAFI. When aggregation 412 policy is used, BGP-LS SPF prefix NLRI will be originated for the 413 aggregate prefix and BGP-LS SPF prefix NLRI for components will be 414 filtered. Additionally, link and node NLRI may be filtered and the 415 abstracted by the prefix NLRI. 417 When BGP policy is used with the BGP-LS SPF SAFI, BGP speakers in the 418 BGP-LS SPF routing domain will not all have the same set of NLRI and 419 will compute a different BGP local routing table. Consequently, care 420 must be taken to assure routing is consistent and blackholes or 421 routing loops do not ensue. However, this is no different than if 422 tradition BGP routing using the IPv4 and IPv6 address families were 423 used. 425 8. IANA Considerations 427 No IANA updates are requested by this document. 429 9. Security Considerations 431 This document introduces no new security considerations above and 432 beyond those already specified in the [RFC4271] and 433 [I-D.ietf-lsvr-bgp-spf]. 435 10. Acknowledgements 437 The authors would like to thank Alvaro Retana and Yan Filyurin for 438 the review and comments. 440 11. References 442 11.1. Normative References 444 [I-D.ietf-lsvr-bgp-spf] 445 Patel, K., Lindem, A., Zandi, S., and W. Henderickx, 446 "Shortest Path Routing Extensions for BGP Protocol", 447 draft-ietf-lsvr-bgp-spf-04 (work in progress), December 448 2018. 450 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 451 Requirement Levels", BCP 14, RFC 2119, 452 DOI 10.17487/RFC2119, March 1997, . 455 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 456 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 457 May 2017, . 459 11.2. Informative References 461 [CLOS] "A Study of Non-Blocking Switching Networks", The Bell 462 System Technical Journal, Vol. 32(2), DOI 463 10.1002/j.1538-7305.1953.tb01433.x, March 1953. 465 [I-D.acee-idr-lldp-peer-discovery] 466 Lindem, A., Patel, K., Zandi, S., Haas, J., and X. Xu, 467 "BGP Logical Link Discovery Protocol (LLDP) Peer 468 Discovery", draft-acee-idr-lldp-peer-discovery-04 (work in 469 progress), December 2018. 471 [I-D.ietf-lsr-dynamic-flooding] 472 Li, T., Psenak, P., Ginsberg, L., Przygienda, T., Cooper, 473 D., Jalil, L., and S. Dontula, "Dynamic Flooding on Dense 474 Graphs", draft-ietf-lsr-dynamic-flooding-00 (work in 475 progress), February 2019. 477 [I-D.ietf-lsvr-l3dl] 478 Bush, R., Austein, R., and K. Patel, "Layer 3 Discovery 479 and Liveness", draft-ietf-lsvr-l3dl-00 (work in progress), 480 April 2019. 482 [I-D.xu-idr-neighbor-autodiscovery] 483 Xu, X., Talaulikar, K., Bi, K., Tantsura, J., and N. 484 Triantafillis, "BGP Neighbor Discovery", draft-xu-idr- 485 neighbor-autodiscovery-11 (work in progress), April 2019. 487 [RFC2328] Moy, J., "OSPF Version 2", STD 54, RFC 2328, 488 DOI 10.17487/RFC2328, April 1998, . 491 [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A 492 Border Gateway Protocol 4 (BGP-4)", RFC 4271, 493 DOI 10.17487/RFC4271, January 2006, . 496 [RFC4760] Bates, T., Chandra, R., Katz, D., and Y. Rekhter, 497 "Multiprotocol Extensions for BGP-4", RFC 4760, 498 DOI 10.17487/RFC4760, January 2007, . 501 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 502 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 503 DOI 10.17487/RFC4861, September 2007, . 506 [RFC4957] Krishnan, S., Ed., Montavont, N., Njedjou, E., Veerepalli, 507 S., and A. Yegin, Ed., "Link-Layer Event Notifications for 508 Detecting Network Attachments", RFC 4957, 509 DOI 10.17487/RFC4957, August 2007, . 512 [RFC5580] Tschofenig, H., Ed., Adrangi, F., Jones, M., Lior, A., and 513 B. Aboba, "Carrying Location Objects in RADIUS and 514 Diameter", RFC 5580, DOI 10.17487/RFC5580, August 2009, 515 . 517 [RFC5925] Touch, J., Mankin, A., and R. Bonica, "The TCP 518 Authentication Option", RFC 5925, DOI 10.17487/RFC5925, 519 June 2010, . 521 [RFC7752] Gredler, H., Ed., Medved, J., Previdi, S., Farrel, A., and 522 S. Ray, "North-Bound Distribution of Link-State and 523 Traffic Engineering (TE) Information Using BGP", RFC 7752, 524 DOI 10.17487/RFC7752, March 2016, . 527 [RFC7938] Lapukhov, P., Premji, A., and J. Mitchell, Ed., "Use of 528 BGP for Routing in Large-Scale Data Centers", RFC 7938, 529 DOI 10.17487/RFC7938, August 2016, . 532 [RFC8177] Lindem, A., Ed., Qu, Y., Yeung, D., Chen, I., and J. 533 Zhang, "YANG Data Model for Key Chains", RFC 8177, 534 DOI 10.17487/RFC8177, June 2017, . 537 Authors' Addresses 539 Keyur Patel 540 Arrcus, Inc. 541 2077 Gateway Pl 542 San Jose, CA 95110 543 USA 545 Email: keyur@arrcus.com 547 Acee Lindem 548 Cisco Systems 549 301 Midenhall Way 550 Cary, NC 95110 551 USA 553 Email: acee@cisco.com 555 Shawn Zandi 556 Linkedin 557 222 2nd Street 558 San Francisco, CA 94105 559 USA 561 Email: szandi@linkedin.com 563 Gaurav Dawra 564 Linkedin 565 222 2nd Street 566 San Francisco, CA 94105 567 USA 569 Email: gdawra@linkedin.com