idnits 2.17.00 (12 Aug 2021) /tmp/idnits24184/draft-ietf-6lo-backbone-router-20.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 == Line 344 has weird spacing: '...ss side o ...' -- The document date (23 March 2020) is 782 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: draft-ietf-6lo-ap-nd has been published as RFC 8928 == Outdated reference: draft-ietf-6tisch-architecture has been published as RFC 9030 == Outdated reference: draft-ietf-mboned-ieee802-mcast-problems has been published as RFC 9119 == Outdated reference: A later version (-23) exists of draft-bi-savi-wlan-18 == Outdated reference: A later version (-02) exists of draft-thubert-6lo-unicast-lookup-00 Summary: 0 errors (**), 0 flaws (~~), 7 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 6lo P. Thubert, Ed. 3 Internet-Draft Cisco Systems 4 Updates: 6775, 8505 (if approved) C.E. Perkins 5 Intended status: Standards Track Blue Meadow Networking 6 Expires: 24 September 2020 E. Levy-Abegnoli 7 Cisco Systems 8 23 March 2020 10 IPv6 Backbone Router 11 draft-ietf-6lo-backbone-router-20 13 Abstract 15 This document updates RFC 6775 and RFC 8505 in order to enable proxy 16 services for IPv6 Neighbor Discovery by Routing Registrars called 17 Backbone Routers. Backbone Routers are placed along the wireless 18 edge of a Backbone, and federate multiple wireless links to form a 19 single Multi-Link Subnet. 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 https://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 24 September 2020. 38 Copyright Notice 40 Copyright (c) 2020 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 (https://trustee.ietf.org/ 45 license-info) in effect on the date of publication of this document. 46 Please review these documents carefully, as they describe your rights 47 and restrictions with respect to this document. Code Components 48 extracted from this document must include Simplified BSD License text 49 as described in Section 4.e of the Trust Legal Provisions and are 50 provided without warranty as described in the Simplified BSD License. 52 Table of Contents 54 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 55 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 5 56 2.1. BCP 14 . . . . . . . . . . . . . . . . . . . . . . . . . 5 57 2.2. New Terms . . . . . . . . . . . . . . . . . . . . . . . . 5 58 2.3. Abbreviations . . . . . . . . . . . . . . . . . . . . . . 6 59 2.4. References . . . . . . . . . . . . . . . . . . . . . . . 7 60 3. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . 7 61 3.1. Updating RFC 6775 and RFC 8505 . . . . . . . . . . . . . 10 62 3.2. Access Link . . . . . . . . . . . . . . . . . . . . . . . 11 63 3.3. Route-Over Mesh . . . . . . . . . . . . . . . . . . . . . 13 64 3.4. The Binding Table . . . . . . . . . . . . . . . . . . . . 14 65 3.5. Primary and Secondary 6BBRs . . . . . . . . . . . . . . . 15 66 3.6. Using Optimistic DAD . . . . . . . . . . . . . . . . . . 16 67 4. Multi-Link Subnet Considerations . . . . . . . . . . . . . . 17 68 5. Optional 6LBR serving the Multi-Link Subnet . . . . . . . . . 17 69 6. Using IPv6 ND Over the Backbone Link . . . . . . . . . . . . 18 70 7. Routing Proxy Operations . . . . . . . . . . . . . . . . . . 20 71 8. Bridging Proxy Operations . . . . . . . . . . . . . . . . . . 21 72 9. Creating and Maintaining a Binding . . . . . . . . . . . . . 22 73 9.1. Operations on a Binding in Tentative State . . . . . . . 23 74 9.2. Operations on a Binding in Reachable State . . . . . . . 24 75 9.3. Operations on a Binding in Stale State . . . . . . . . . 25 76 10. Registering Node Considerations . . . . . . . . . . . . . . . 26 77 11. Security Considerations . . . . . . . . . . . . . . . . . . . 27 78 12. Protocol Constants . . . . . . . . . . . . . . . . . . . . . 30 79 13. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 30 80 14. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 30 81 15. Normative References . . . . . . . . . . . . . . . . . . . . 30 82 16. Informative References . . . . . . . . . . . . . . . . . . . 32 83 Appendix A. Possible Future Extensions . . . . . . . . . . . . . 34 84 Appendix B. Applicability and Requirements Served . . . . . . . 35 85 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 37 87 1. Introduction 89 IEEE STD. 802.1 [IEEEstd8021] Ethernet Bridging provides an efficient 90 and reliable broadcast service for wired networks; applications and 91 protocols have been built that heavily depend on that feature for 92 their core operation. Unfortunately, Low-Power Lossy Networks (LLNs) 93 and local wireless networks generally do not provide the broadcast 94 capabilities of Ethernet Bridging in an economical fashion. 96 As a result, protocols designed for bridged networks that rely on 97 multicast and broadcast often exhibit disappointing behaviours when 98 employed unmodified on a local wireless medium (see 99 [I-D.ietf-mboned-ieee802-mcast-problems]). 101 Wi-Fi [IEEEstd80211] Access Points (APs) deployed in an Extended 102 Service Set (ESS) act as Ethernet Bridges [IEEEstd8021], with the 103 property that the bridging state is established at the time of 104 association. This ensures connectivity to the end node (the Wi-Fi 105 STA) and protects the wireless medium against broadcast-intensive 106 Transparent Bridging reactive Lookups. In other words, the 107 association process is used to register the MAC Address of the STA to 108 the AP. The AP subsequently proxies the bridging operation and does 109 not need to forward the broadcast Lookups over the radio. 111 In the same way as Transparent Bridging, IPv6 [RFC8200] Neighbor 112 Discovery [RFC4861] [RFC4862] Protocol (IPv6 ND) is a reactive 113 protocol, based on multicast transmissions to locate an on-link 114 correspondent and ensure the uniqueness of an IPv6 address. The 115 mechanism for Duplicate Address Detection (DAD) [RFC4862] was 116 designed for the efficient broadcast operation of Ethernet Bridging. 117 Since broadcast can be unreliable over wireless media, DAD often 118 fails to discover duplications [I-D.yourtchenko-6man-dad-issues]. In 119 practice, the fact that IPv6 addresses very rarely conflict is mostly 120 attributable to the entropy of the 64-bit Interface IDs as opposed to 121 the succesful operation of the IPv6 ND duplicate address detection 122 and resolution mechanisms. 124 The IPv6 ND Neighbor Solicitation (NS) [RFC4861] message is used for 125 DAD and address Lookup when a node moves, or wakes up and reconnects 126 to the wireless network. The NS message is targeted to a Solicited- 127 Node Multicast Address (SNMA) [RFC4291] and should in theory only 128 reach a very small group of nodes. But in reality, IPv6 multicast 129 messages are typically broadcast on the wireless medium, and so they 130 are processed by most of the wireless nodes over the subnet (e.g., 131 the ESS fabric) regardless of how few of the nodes are subscribed to 132 the SNMA. As a result, IPv6 ND address Lookups and DADs over a large 133 wireless and/or a LowPower Lossy Network (LLN) can consume enough 134 bandwidth to cause a substantial degradation to the unicast traffic 135 service. 137 Because IPv6 ND messages sent to the SNMA group are broadcast at the 138 radio MAC Layer, wireless nodes that do not belong to the SNMA group 139 still have to keep their radio turned on to listen to multicast NS 140 messages, which is a waste of energy for them. In order to reduce 141 their power consumption, certain battery-operated devices such as IoT 142 sensors and smartphones ignore some of the broadcasts, making IPv6 ND 143 operations even less reliable. 145 These problems can be alleviated by reducing the IPv6 ND broadcasts 146 over wireless access links. This has been done by splitting the 147 broadcast domains and routing between subnets, at the extreme by 148 assigning a /64 prefix to each wireless node (see [RFC8273]). But 149 deploying a single large subnet can still be attractive to avoid 150 renumbering in situations that involve large numbers of devices and 151 mobility within a bounded area. 153 A way to reduce the propagation of IPv6 ND broadcast in the wireless 154 domain while preserving a large single subnet is to form a Multi-Link 155 Subnet (MLSN). Each Link in the MLSN, including the backbone, is its 156 own broadcast domain. A key property of MLSNs is that Link-Local 157 unicast traffic, link-scope multicast, and traffic with a hop limit 158 of 1 will not transit to nodes in the same subnet on a different 159 link, something that may produce unexpected behavior in software that 160 expects a subnet to be entirely contained within a single link. 162 This specification considers a special type of MLSN with a central 163 backbone that federates edge (LLN) links, each Link providing its own 164 protection against rogue access and tempering or replaying packets. 165 In particular, the use of classical IPv6 ND on the backbone requires 166 that the all nodes are trusted and that rogue access to the backbone 167 is prevented at all times (see Section 11). 169 In that particular topology, ND proxies can be placed at the boundary 170 of the edge links and the backbone to handle IPv6 ND on behalf of 171 Registered Nodes and forward IPv6 packets back and forth. The ND 172 proxy enables the continuity of IPv6 ND operations beyond the 173 backbone, and enables communication using Global or Unique Local 174 Addresses between any pair of nodes in the MLSN. 176 The 6LoWPAN Backbone Router (6BBR) is a Routing Registrar [RFC8505] 177 that provides proxy-ND services. A 6BBR acting as a Bridging Proxy 178 provides a proxy-ND function with Layer-2 continuity and can be 179 collocated with a Wi-Fi Access Point (AP) as prescribed by IEEE Std 180 802.11 [IEEEstd80211]. A 6BBR acting as a Routing Proxy is 181 applicable to any type of LLN, including LLNs that cannot be bridged 182 onto the backbone, such as IEEE Std 802.15.4 [IEEEstd802154]. 184 Knowledge of which address to proxy for can be obtained by snooping 185 the IPV6 ND protocol (see [I-D.bi-savi-wlan]), but it has been found 186 to be unreliable. An IPv6 address may not be discovered immediately 187 due to a packet loss, or if a "silent" node is not currently using 188 one of its addresses. A change of state (e.g., due to movement) may 189 be missed or misordered, leading to unreliable connectivity and 190 incomplete knowledge of the state of the network. 192 With this specification, the address to be proxied is signaled 193 explicitly through a registration process. A 6LoWPAN node (6LN) 194 registers all its IPv6 Addresses using NS messages with an Extended 195 Address Registration Option (EARO) as specified in [RFC8505] to a 196 6LoWPAN Router (6LR) to which it is directly attached. If the 6LR is 197 a 6BBR then the 6LN is both the Registered Node and the Registering 198 Node. If not, then the 6LoWPAN Border Router (6LBR) that serves the 199 LLN proxies the registration to the 6BBR. In that case, the 6LN is 200 the Registered Node and the 6LBR is the Registering Node. The 6BBR 201 performs IPv6 Neighbor Discovery (IPv6 ND) operations on its Backbone 202 interface on behalf of the 6LNs that have registered addresses on its 203 LLN interfaces without the need of a broadcast over the wireless 204 medium. 206 A Registering Node that resides on the backbone does not register to 207 the SNMA groups associated to its Registered Addresses and defers to 208 the 6BBR to answer or preferably forward to it as unicast the 209 corresponding multicast packets. 211 2. Terminology 213 2.1. BCP 14 215 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 216 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 217 "OPTIONAL" in this document are to be interpreted as described in BCP 218 14 [RFC2119] [RFC8174] when, and only when, they appear in all 219 capitals, as shown here. 221 2.2. New Terms 223 This document introduces the following terminology: 225 Federated: A subnet that comprises a Backbone and one or more 226 (wireless) access links, is said to be federated into one Multi- 227 Link Subnet. The proxy-ND operation of 6BBRs over the Backbone 228 extends IPv6 ND operation over the access links. 230 Sleeping Proxy: A 6BBR acts as a Sleeping Proxy if it answers IPv6 231 ND Neighbor Solicitations over the Backbone on behalf of the 232 Registering Node that is in a sleep state and cannot answer in due 233 time. 235 Routing Proxy: A Routing Proxy provides IPv6 ND proxy functions and 236 enables the MLSN operation over federated links that may not be 237 compatible for bridging. The Routing Proxy advertises its own MAC 238 Address as the Target Link Layer Address (TLLA) in the proxied NAs 239 over the Backbone, and routes at the Network Layer between the 240 federated links. 242 Bridging Proxy: A Bridging Proxy provides IPv6 ND proxy functions 243 while preserving forwarding continuity at the MAC Layer. In that 244 case, the MAC Address and the mobility of the Registering Node is 245 visible across the bridged Backbone. The Bridging Proxy 246 advertises the MAC Address of the Registering Node as the TLLA in 247 the proxied NAs over the Backbone, and proxies ND for all unicast 248 addresses including Link-Local Addresses. Instead of replying on 249 behalf of the Registering Node, a Bridging Proxy will preferably 250 forward the NS Lookup and NUD messages that target the Registered 251 Address to the Registering Node as unicast frames and let it 252 respond in its own. 254 Binding Table: The Binding Table is an abstract database that is 255 maintained by the 6BBR to store the state associated with its 256 registrations. 258 Binding: A Binding is an abstract state associated to one 259 registration, in other words one entry in the Binding Table. 261 2.3. Abbreviations 263 This document uses the following abbreviations: 265 6BBR: 6LoWPAN Backbone Router 266 6LBR: 6LoWPAN Border Router 267 6LN: 6LoWPAN Node 268 6LR: 6LoWPAN Router 269 ARO: Address Registration Option 270 DAC: Duplicate Address Confirmation 271 DAD: Duplicate Address Detection 272 DAR: Duplicate Address Request 273 EARO: Extended Address Registration Option 274 EDAC: Extended Duplicate Address Confirmation 275 EDAR: Extended Duplicate Address Request 276 DODAG: Destination-Oriented Directed Acyclic Graph 277 ID: Identifier 278 LLN: Low-Power and Lossy Network 279 NA: Neighbor Advertisement 280 MAC: Medium Access Control 281 NCE: Neighbor Cache Entry 282 ND: Neighbor Discovery 283 NDP: Neighbor Discovery Protocol 284 NS: Neighbor Solicitation 285 NS(DAD): NDP NS message used for the purpose of duplication 286 avoidance (multicast) 287 NS(Lookup): NDP NS message used for the purpose of address 288 resolution (multicast) 289 NS(NUD): NDP NS message used for the purpose of unreachability 290 detection (unicast) 291 NUD: Neighbor Unreachability Detection 292 ROVR: Registration Ownership Verifier 293 RPL: IPv6 Routing Protocol for LLNs 294 RA: Router Advertisement 295 RS: Router Solicitation 296 SNMA: Solicited-Node Multicast Address 297 LLA: Link Layer Address (aka MAC address) 298 SLLA: Source Link Layer Address 299 TLLA: Target Link Layer Address 300 TID: Transaction ID 302 2.4. References 304 In this document, readers will encounter terms and concepts that are 305 discussed in the following documents: 307 Classical IPv6 ND: "Neighbor Discovery for IP version 6" [RFC4861], 308 "IPv6 Stateless Address Autoconfiguration" [RFC4862] and 309 "Optimistic Duplicate Address Detection" [RFC4429], 311 IPv6 ND over multiple links: "Neighbor Discovery Proxies (proxy-ND)" 312 [RFC4389] and "Multi-Link Subnet Issues" [RFC4903], 314 6LoWPAN: "Problem Statement and Requirements for IPv6 over Low-Power 315 Wireless Personal Area Network (6LoWPAN) Routing" [RFC6606], and 317 6LoWPAN ND: Neighbor Discovery Optimization for Low-Power and Lossy 318 Networks [RFC6775], "Registration Extensions for 6LoWPAN Neighbor 319 Discovery" [RFC8505], and " Address Protected Neighbor Discovery 320 for Low-power and Lossy Networks" [I-D.ietf-6lo-ap-nd]. 322 3. Overview 324 This section and its subsections present a non-normative high level 325 view of the operation of the 6BBR. The following sections cover the 326 normative part. 328 Figure 1 illustrates a backbone link that federates a collection of 329 LLNs as a single IPv6 Subnet, with a number of 6BBRs providing proxy- 330 ND services to their attached LLNs. 332 | 333 +-----+ +-----+ +-----+ IPv6 334 (default) | | (Optional) | | | | Node 335 Router | | 6LBR | | | | or 336 +-----+ +-----+ +-----+ 6LN 337 | Backbone side | | 338 ----+-------+-----------------+---+-------------+----+----- 339 | | | 340 +------+ +------+ +------+ 341 | 6BBR | | 6BBR | | 6BBR | 342 | | | | | | 343 +------+ +------+ +------+ 344 o Wireless side o o o o o o 345 o o o o o o o o o o o o o o o o o o o o 346 o o o o o o o o o o o o o o o o o o o 347 o o o o o o o o o LLN o o o o o o o o o 348 o o o o o o o o o o o o o o 349 o o o 351 Figure 1: Backbone Link and Backbone Routers 353 The LLN may be a hub-and-spoke access link such as (Low-Power) IEEE 354 STD. 802.11 (Wi-Fi) [IEEEstd80211] and IEEE STD. 802.15.1 (Bluetooth) 355 [IEEEstd802151], or a Mesh-Under or a Route-Over network [RFC8505]. 356 The proxy state can be distributed across multiple 6BBRs attached to 357 the same Backbone. 359 The main features of a 6BBR are as follows: 361 * Multi-Link-subnet functions (provided by the 6BBR on the backbone) 362 performed on behalf of Registered Nodes, and 364 * Routing registrar services that reduce multicast within the LLN: 366 - Binding Table management 367 - failover, e.g., due to mobility 369 Each Backbone Router (6BBR) maintains a data structure for its 370 Registered Addresses called a Binding Table. The abstract data that 371 is stored in the Binding Table includes the Registered Address, 372 anchor information on the Registering Node such as connecting 373 interface, Link-Local Address and Link-Layer Address of the 374 Registering Node on that interface, the EARO including ROVR and TID, 375 a state that can be either Reachable, Tentative, or Stale, and other 376 information such as a trust level that may be configured, e.g., to 377 protect a server. The combined Binding Tables of all the 6BBRs on a 378 backbone form a distributed database of Registered Nodes that reside 379 in the LLNs or on the IPv6 Backbone. 381 Unless otherwise configured, a 6BBR does the following: 383 * Create a new entry in a Binding Table for a new Registered Address 384 and ensure that the Address is not duplicated over the Backbone. 386 * Advertise a Registered Address over the Backbone using an NA 387 message, either unsolicited or as a response to a NS message. 388 This includes joining the multicast group associated to the SNMA 389 derived from the Registered Address as specified in section 7.2.1. 390 of [RFC4861] over the Backbone. 392 * The 6BBR MAY respond immediately as a Proxy in lieu of the 393 Registering Node, e.g., if the Registering Node has a sleeping 394 cycle that the 6BBR does not want to interrupt, or if the 6BBR has 395 a recent state that is deemed fresh enough to permit the proxied 396 response. It is preferred, though, that the 6BBR checks whether 397 the Registering Node is still responsive on the Registered 398 Address. To that effect: 400 - as a Bridging Proxy: 401 the 6BBR forwards the multicast DAD and Address Lookup messages 402 as a unicast MAC-Layer frames to the MAC address of the 403 Registering Node that matches the Target in the ND message, and 404 forwards as is the unicast Neighbor Unreachability Detection 405 (NUD) messages, so as to let the Registering Node answer with 406 the ND Message and options that it sees fit; 407 - as a Routing Proxy: 408 the 6BBR checks the liveliness of the Registering Node, e.g., 409 using a NUD verification, before answering on its behalf. 411 * Deliver packets arriving from the LLN, using Neighbor Solicitation 412 messages to look up the destination over the Backbone. 414 * Forward or bridge packets between the LLN and the Backbone. 416 * Verify liveness for a registration, when needed. 418 The first of these functions enables the 6BBR to fulfill its role as 419 a Routing Registrar for each of its attached LLNs. The remaining 420 functions fulfill the role of the 6BBRs as the border routers that 421 federate the Multi-link IPv6 subnet. 423 The operation of IPv6 ND and of proxy-ND are not mutually exclusive 424 on the Backbone, meaning that nodes attached to the Backbone and 425 using IPv6 ND can transparently interact with 6LNs that rely on a 426 6BBR to proxy ND for them, whether the 6LNs are reachable over an LLN 427 or directly attached to the Backbone. 429 The [RFC8505] registration mechanism used to learn addresses to be 430 proxied may co-exist in a 6BBR with a proprietary snooping or the 431 traditional bridging functionality of an Access Point, in order to 432 support legacy LLN nodes that do not support this specification. 434 The registration to a proxy service uses an NS/NA exchange with EARO. 435 The 6BBR operation resembles that of a Mobile IPv6 (MIPv6) [RFC6275] 436 Home Agent (HA). The combination of a 6BBR and a MIPv6 HA enables 437 full mobility support for 6LNs, inside and outside the links that 438 form the subnet. 440 The 6BBRs performs IPv6 ND functions over the backbone as follows: 442 * The EARO [RFC8505] is used in the IPv6 ND exchanges over the 443 Backbone between the 6BBRs to help distinguish duplication from 444 movement. Extended Duplicate Address Messages (EDAR and EDAC) may 445 also be used to communicate with a 6LBR, if one is present. 446 Address duplication is detected using the ROVR field. Conflicting 447 registrations to different 6BBRs for the same Registered Address 448 are resolved using the TID field which forms an order of 449 registrations. 451 * The Link Layer Address (LLA) that the 6BBR advertises for the 452 Registered Address on behalf of the Registered Node over the 453 Backbone can belong to the Registering Node; in that case, the 454 6BBR (acting as a Bridging Proxy (see Section 8)) bridges the 455 unicast packets. Alternatively, the LLA can be that of the 6BBR 456 on the Backbone interface, in which case the 6BBR (acting as a 457 Routing Proxy (see Section 7)) receives the unicast packets at 458 Layer 3 and routes over. 460 3.1. Updating RFC 6775 and RFC 8505 462 This specification adds the EARO as a possible option in RS, NS(DAD) 463 and NA messages over the backbone. This document specifies the use 464 of those ND messages by 6BBRs over the backbone, at a high level in 465 Section 6 and in more detail in Section 9. 467 Note: [RFC8505] requires that the registration NS(EARO) contains an 468 Source Link Layer Address Option (SLLAO). [RFC4862] requires that 469 the NS(DAD) is sent from the unspecified address for which there 470 cannot be a SLLAO. Consequently, an NS(DAD) cannot be confused with 471 a registration. 473 This specification allows to deploy a 6LBR on the backbone where EDAR 474 and EDAC messages coexist with classical ND. It also adds the 475 capability to insert IPv6 ND options in the EDAR and EDAC messages. 476 A 6BBR acting as a 6LR for the Registered Address can insert an SLLAO 477 in the EDAR to the 6LBR in order to avoid a Lookup back. This 478 enables the 6LBR to store the MAC address associated to the 479 Registered Address on a Link and to serve as a mapping server as 480 described in [I-D.thubert-6lo-unicast-lookup]. 482 This specification allows for an address to be registered to more 483 than one 6BBR. Consequently a 6LBR that is deployed on the backbone 484 MUST be capable of maintaining state for each of the 6BBR having 485 registered with the same TID and same ROVR. 487 3.2. Access Link 489 The simplest Multi-Link Subnet topology from the Layer 3 perspective 490 occurs when the wireless network appears as a single hop hub-and- 491 spoke network as shown in Figure 2. The Layer 2 operation may 492 effectively be hub-and-spoke (e.g., Wi-Fi) or Mesh-Under, with a 493 Layer 2 protocol handling the complex topology. 495 | 496 +-----+ +-----+ +-----+ IPv6 497 (default) | | (Optional) | | | | Node 498 Router | | 6LBR | | | | or 499 +-----+ +-----+ +-----+ 6LN 500 | Backbone side | | 501 ----+-------+-----------------+---+-------------+----+----- 502 | | | 503 +------+ +------+ +------+ 504 | 6BBR | | 6BBR | | 6BBR | 505 | 6LR | | 6LR | | 6LR | 506 +------+ +------+ +------+ 507 (6LN) (6LN) (6LN) (6LN) (6LN) (6LN) (6LN) (6LN) 509 Figure 2: Access Link Use case 511 Figure 3 illustrates a flow where 6LN forms an IPv6 Address and 512 registers it to a 6BBR acting as a 6LR [RFC8505]. The 6BBR applies 513 ODAD (see Section 3.6) to the registered address to enable 514 connectivity while the message flow is still in progress. 516 6LN(STA) 6BBR(AP) 6LBR default GW 517 | | | | 518 | LLN Access Link | IPv6 Backbone (e.g., Ethernet) | 519 | | | | 520 | RS(multicast) | | | 521 |---------------->| | | 522 | RA(PIO, Unicast)| | | 523 |<----------------| | | 524 | NS(EARO) | | | 525 |---------------->| | | 526 | | Extended DAR | | 527 | |--------------->| | 528 | | Extended DAC | | 529 | |<---------------| | 530 | | | 531 | | NS-DAD(EARO, multicast) | 532 | |--------> | 533 | |----------------------------------->| 534 | | | 535 | | RS(no SLLAO, for ODAD) | 536 | |----------------------------------->| 537 | | if (no fresher Binding) NS(Lookup) | 538 | | <----------------| 539 | |<-----------------------------------| 540 | | NA(SLLAO, not(O), EARO) | 541 | |----------------------------------->| 542 | | RA(unicast) | 543 | |<-----------------------------------| 544 | | | 545 | IPv6 Packets in optimistic mode | 546 |<---------------------------------------------------->| 547 | | | 548 | | 549 | NA(EARO) | 550 |<----------------| 551 | | 553 Figure 3: Initial Registration Flow to a 6BBR acting as Routing Proxy 555 In this example, a 6LBR is deployed on the backbone link to serve the 556 whole subnet, and EDAR / EDAC messages are used in combination with 557 DAD to enable coexistence with IPv6 ND over the backbone. 559 The RS sent initially by the 6LN (e.g., a Wi-Fi STA) is transmitted 560 as a multicast but since it is intercepted by the 6BBR, it is never 561 effectively broadcast. The multiple arrows associated to the ND 562 messages on the Backbone denote a real Layer 2 broadcast. 564 3.3. Route-Over Mesh 566 A more complex Multi-Link Subnet topology occurs when the wireless 567 network appears as a Layer 3 Mesh network as shown in Figure 4. A 568 so-called Route-Over routing protocol exposes routes between 6LRs 569 towards both 6LRs and 6LNs, and a 6LBR acts as Root of the Layer 3 570 Mesh network and proxy-registers the LLN addresses to the 6BBR. 572 | 573 +-----+ +-----+ +-----+ IPv6 574 (default) | | (Optional) | | | | Node 575 Router | | 6LBR | | | | or 576 +-----+ +-----+ +-----+ 6LN 577 | Backbone side | | 578 ----+-------+-----------------+---+-------------+----+----- 579 | | | 580 +------+ +------+ +------+ 581 | 6BBR | | 6BBR | | 6BBR | 582 +------+ +------+ +------+ 583 | | | 584 +------+ +------+ +------+ 585 | 6LBR | | 6LBR | | 6LBR | 586 +------+ +------+ +------+ 587 (6LN) (6LR) (6LN) (6LR) (6LN) (6LR) (6LR) (6LR)(6LN) 588 (6LN)(6LR) (6LR) (6LN) (6LN) (6LR)(6LN) (6LR) (6LR) (6LR) (6LN) 589 (6LR)(6LR) (6LR) (6LR) (6LR)(6LN) (6LR) (6LR)(6LR) 590 (6LR) (6LR) (6LR) (6LR) (6LN)(6LR) (6LR) (6LR) (6LR) (6LR) 591 (6LN) (6LN)(6LN) (6LN) (6LN) (6LN) (6LN) (6LN) (6LN) (6LN) 593 Figure 4: Route-Over Mesh Use case 595 Figure 5 illustrates IPv6 signaling that enables a 6LN (the 596 Registered Node) to form a Global or a Unique-Local Address and 597 register it to the 6LBR that serves its LLN using [RFC8505] using a 598 neighboring 6LR as relay. The 6LBR (the Registering Node) then 599 proxies the [RFC8505] registration to the 6BBR to obtain proxy-ND 600 services from the 6BBR. 602 The RS sent initially by the 6LN is a transmitted as a multicast and 603 contained within 1-hop broadcast range where hopefully a 6LR is 604 found. The 6LR is expected to be already connected to the LLN and 605 capable to reach the 6LBR, possibly multiple hops away, using unicast 606 messages. 608 6LoWPAN Node 6LR 6LBR 6BBR 609 (mesh leaf) (mesh router) (mesh root) 610 | | | | 611 | 6LoWPAN ND |6LoWPAN ND | 6LoWPAN ND | IPv6 ND 612 | LLN link |Route-Over mesh|Ethernet/serial| Backbone 613 | | |/Internal call | 614 | IPv6 ND RS | | | 615 |-------------->| | | 616 |-----------> | | | 617 |------------------> | | 618 | IPv6 ND RA | | | 619 |<--------------| | | 620 | | | | 621 | NS(EARO) | | | 622 |-------------->| | | 623 | 6LoWPAN ND | Extended DAR | | 624 | |-------------->| | 625 | | | NS(EARO) | 626 | | |-------------->| 627 | | | (proxied) | NS-DAD 628 | | | |------> 629 | | | | (EARO) 630 | | | | 631 | | | NA(EARO) | 632 | | |<--------------| 633 | | Extended DAC | | 634 | |<--------------| | 635 | NA(EARO) | | | 636 |<--------------| | | 637 | | | | 639 Figure 5: Initial Registration Flow over Route-Over Mesh 641 As a non-normative example of a Route-Over Mesh, the 6TiSCH 642 architecture [I-D.ietf-6tisch-architecture] suggests using the RPL 643 [RFC6550] routing protocol and collocating the RPL root with a 6LBR 644 that serves the LLN. The 6LBR is also either collocated with or 645 directly connected to the 6BBR over an IPv6 Link. 647 3.4. The Binding Table 649 Addresses in an LLN that are reachable from the Backbone by way of 650 the 6BBR function must be registered to that 6BBR, using an NS(EARO) 651 with the R flag set [RFC8505]. The 6BBR answers with an NA(EARO) and 652 maintains a state for the registration in an abstract Binding Table. 654 An entry in the Binding Table is called a "Binding". A Binding may 655 be in Tentative, Reachable or Stale state. 657 The 6BBR uses a combination of [RFC8505] and IPv6 ND over the 658 Backbone to advertise the registration and avoid a duplication. 659 Conflicting registrations are solved by the 6BBRs, transparently to 660 the Registering Nodes. 662 Only one 6LN may register a given Address, but the Address may be 663 registered to Multiple 6BBRs for higher availability. 665 Over the LLN, Binding Table management is as follows: 667 * De-registrations (newer TID, same ROVR, null Lifetime) are 668 accepted with a status of 4 ("Removed"); the entry is deleted; 670 * Newer registrations (newer TID, same ROVR, non-null Lifetime) are 671 accepted with a status of 0 (Success); the Binding is updated with 672 the new TID, the Registration Lifetime and the Registering Node; 673 in Tentative state the EDAC response is held and may be 674 overwritten; in other states the Registration Lifetime timer is 675 restarted and the entry is placed in Reachable state. 677 * Identical registrations (same TID, same ROVR) from the same 678 Registering Node are accepted with a status of 0 (Success). In 679 Tentative state, the response is held and may be overwritten, but 680 the response is eventually produced, carrying the result of the 681 DAD process; 683 * Older registrations (older TID, same ROVR) from the same 684 Registering Node are discarded; 686 * Identical and older registrations (not-newer TID, same ROVR) from 687 a different Registering Node are rejected with a status of 3 688 (Moved); this may be rate limited to avoid undue interference; 690 * Any registration for the same address but with a different ROVR is 691 rejected with a status of 1 (Duplicate). 693 The operation of the Binding Table is specified in detail in 694 Section 9. 696 3.5. Primary and Secondary 6BBRs 698 A Registering Node MAY register the same address to more than one 699 6BBR, in which case the Registering Node uses the same EARO in all 700 the parallel registrations. On the other hand, there is no provision 701 in 6LoWPAN ND for a 6LN (acting as Registered Node) to select its 702 6LBR (acting as Registering Node), so it cannot select more than one 703 either. To allow for this, NS(DAD) and NA messages with an EARO 704 received over the backbone that indicate an identical Binding in 705 another 6BBR (same Registered address, same TID, same ROVR) are 706 silently ignored but for the purpose of selecting the primary 6BBR 707 for that registration. 709 A 6BBR may be either primary or secondary. The primary is the 6BBR 710 that has the highest EUI-64 Address of all the 6BBRs that share a 711 registration for the same Registered Address, with the same ROVR and 712 same Transaction ID, the EUI-64 Address being considered as an 713 unsigned 64bit integer. A given 6BBR can be primary for a given 714 Address and secondary for another Address, regardless of whether or 715 not the Addresses belong to the same 6LN. 717 In the following sections, it is expected that an NA is sent over the 718 backbone only if the node is primary or does not support the concept 719 of primary. More than one 6BBR claiming or defending an address 720 generates unwanted traffic but no reachability issue since all 6BBRs 721 provide reachability from the Backbone to the 6LN. 723 If a Registering Node loses connectivity to its or one of the 6BBRs 724 to which it registered an address, it retries the registration to the 725 (one or more) available 6BBR(s). When doing that, the Registering 726 Node MUST increment the TID in order to force the migration of the 727 state to the new 6BBR, and the reselection of the primary 6BBR if it 728 is the node that was lost. 730 3.6. Using Optimistic DAD 732 Optimistic Duplicate Address Detection [RFC4429] (ODAD) specifies how 733 an IPv6 Address can be used before completion of Duplicate Address 734 Detection (DAD). ODAD guarantees that this behavior will not cause 735 harm if the new Address is a duplicate. 737 Support for ODAD avoids delays in installing the Neighbor Cache Entry 738 (NCE) in the 6BBRs and the default router, enabling immediate 739 connectivity to the registered node. As shown in Figure 3, if the 740 6BBR is aware of the Link-Layer Address (LLA) of a router, then the 741 6BBR sends a Router Solicitation (RS), using the Registered Address 742 as the IP Source Address, to the known router(s). The RS is sent 743 without a Source LLA Option (SLLAO), to avoid invalidating a 744 preexisting NCE in the router. 746 Following ODAD, the router may then send a unicast RA to the 747 Registered Address, and it may resolve that Address using an 748 NS(Lookup) message. In response, the 6BBR sends an NA with an EARO 749 and the Override flag [RFC4861] that is not set. The router can then 750 determine the freshest EARO in case of conflicting NA(EARO) messages, 751 using the method described in section 5.2.1 of [RFC8505]. If the 752 NA(EARO) is the freshest answer, the default router creates a Binding 753 with the SLLAO of the 6BBR (in Routing Proxy mode) or that of the 754 Registering Node (in Bridging Proxy mode) so that traffic from/to the 755 Registered Address can flow immediately. 757 4. Multi-Link Subnet Considerations 759 The Backbone and the federated LLN Links are considered as different 760 links in the Multi-Link Subnet, even if multiple LLNs are attached to 761 the same 6BBR. ND messages are link-scoped and are not forwarded by 762 the 6BBR between the backbone and the LLNs though some packets may be 763 reinjected in Bridging Proxy mode (see Section 8). 765 Legacy nodes located on the backbone expect that the subnet is 766 deployed within a single link and that there is a common Maximum 767 Transmission Unit (MTU) for intra-subnet communication, the Link MTU. 768 They will not perform the IPv6 Path MTU Discovery [RFC8201] for a 769 destination within the subnet. For that reason, the MTU MUST have 770 the same value on the Backbone and all federated LLNs in the MLSN. 771 As a consequence, the 6BBR MUST use the same MTU value in RAs over 772 the Backbone and in the RAs that it transmits towards the LLN links. 774 5. Optional 6LBR serving the Multi-Link Subnet 776 A 6LBR can be deployed to serve the whole MLSN. It may be attached 777 to the backbone, in which case it can be discovered by its capability 778 advertisement (see section 4.3. of [RFC8505]) in RA messages. 780 When a 6LBR is present, the 6BBR uses an EDAR/EDAC message exchange 781 with the 6LBR to check if the new registration corresponds to a 782 duplication or a movement. This is done prior to the NS(DAD) 783 process, which may be avoided if the 6LBR already maintains a 784 conflicting state for the Registered Address. 786 If this registration is duplicate or not the freshest, then the 6LBR 787 replies with an EDAC message with a status code of 1 ("Duplicate 788 Address") or 3 ("Moved"), respectively. If this registration is the 789 freshest, then the 6LBR replies with a status code of 0. In that 790 case, if this registration is fresher than an existing registration 791 for another 6BBR, then the 6LBR also sends an asynchronous EDAC with 792 a status of 4 ("Removed") to that other 6BBR. 794 The EDAR message SHOULD carry the SLLAO used in NS messages by the 795 6BBR for that Binding, and the EDAC message SHOULD carry the Target 796 Link Layer Address Option (TLLAO) associated with the currently 797 accepted registration. This enables a 6BBR to locate the new 798 position of a mobile 6LN in the case of a Routing Proxy operation, 799 and opens the capability for the 6LBR to serve as a mapping server in 800 the future. 802 Note that if Link-Local Addresses are registered, then the scope of 803 uniqueness on which the address duplication is checked is the total 804 collection of links that the 6LBR serves as opposed to the sole link 805 on which the Link-Local Address is assigned. 807 6. Using IPv6 ND Over the Backbone Link 809 On the Backbone side, the 6BBR MUST join the SNMA group corresponding 810 to a Registered Address as soon as it creates a Binding for that 811 Address, and maintain that SNMA membership as long as it maintains 812 the registration. The 6BBR uses either the SNMA or plain unicast to 813 defend the Registered Addresses in its Binding Table over the 814 Backbone (as specified in [RFC4862]). The 6BBR advertises and 815 defends the Registered Addresses over the Backbone Link using RS, 816 NS(DAD) and NA messages with the Registered Address as the Source or 817 Target address. 819 The 6BBR MUST place an EARO in the IPv6 ND messages that it generates 820 on behalf of the Registered Node. Note that an NS(DAD) does not 821 contain an SLLAO and cannot be confused with a proxy registration 822 such as performed by a 6LBR. 824 IPv6 ND operates as follows on the backbone: 826 * Section 7.2.8 of [RFC4861] specifies that an NA message generated 827 as a proxy does not have the Override flag set in order to ensure 828 that if the real owner is present on the link, its own NA will 829 take precedence, and that this NA does not update the NCE for the 830 real owner if one exists. 832 * A node that receives multiple NA messages updates an existing NCE 833 only if the Override flag is set; otherwise the node will probe 834 the cached address. 836 * When an NS(DAD) is received for a tentative address, which means 837 that two nodes form the same address at nearly the same time, 838 section 5.4.3 of [RFC4862] cannot detect which node first claimed 839 the address and the address is abandoned. 841 * In any case, [RFC4862] indicates that a node never responds to a 842 Neighbor Solicitation for a tentative address. 844 This specification adds information about proxied addresses that 845 helps sort out a duplication (different ROVR) from a movement (same 846 ROVR, different TID), and in the latter case the older registration 847 from the fresher one (by comparing TIDs). 849 When a Registering Node moves from one 6BBR to the next, the new 6BBR 850 sends NA messages over the backbone to update existing NCEs. A node 851 that supports this specification and that receives multiple NA 852 messages with an EARO option and the same ROVR MUST favor the NA with 853 the freshest EARO over the others. 855 The 6BBR MAY set the Override flag in the NA messages if it does not 856 compete with the Registering Node for the NCE in backbone nodes. 857 This is assured if the Registering Node is attached via an interface 858 that cannot be bridged onto the backbone, making it impossible for 859 the Registering Node to defend its own addresses there. This may 860 also be signaled by the Registering Node through a protocol extension 861 that is not in scope for this specification. 863 When the Binding is in Tentative state, the 6BBR acts as follows: 865 * an NS(DAD) that indicates a duplication can still not be asserted 866 for first come, but the situation can be avoided using a 6LBR on 867 the backbone that will serialize the order of appearance of the 868 address and ensure first-come/first-serve. 870 * an NS or an NA that denotes an older registration for the same 871 Registered Node is not interpreted as a duplication as specified 872 in section 5.4.3 and 5.4.4 of [RFC4862], respectively. 874 When the Binding is no longer in Tentative state, the 6BBR acts as 875 follows: 877 * an NS or an NA with an EARO that denotes a duplicate registration 878 (different ROVR) is answered with an NA message that carries an 879 EARO with a status of 1 (Duplicate), unless the received message 880 is an NA that carries an EARO with a status of 1. 882 In any state, the 6BBR acts as follows: 884 * an NS or an NA with an EARO that denotes an older registration 885 (same ROVR) is answered with an NA message that carries an EARO 886 with a status of 3 (Moved) to ensure that the stale state is 887 removed rapidly. 889 This behavior is specified in more detail in Section 9. 891 This specification enables proxy operation for the IPv6 ND resolution 892 of LLN devices and a prefix that is used across a Multi-Link Subnet 893 MAY be advertised as on-link over the Backbone. This is done for 894 backward compatibility with existing IPv6 hosts by setting the L flag 895 in the Prefix Information Option (PIO) of RA messages [RFC4861]. 897 For movement involving a slow reattachment, the NUD procedure defined 898 in [RFC4861] may time out too quickly. Nodes on the backbone SHOULD 899 support [RFC7048] whenever possible. 901 7. Routing Proxy Operations 903 A Routing Proxy provides IPv6 ND proxy functions for Global and 904 Unique Local addresses between the LLN and the backbone, but not for 905 Link-Local addresses. It operates as an IPv6 border router and 906 provides a full Link-Layer isolation. 908 In this mode, it is not required that the MAC addresses of the 6LNs 909 are visible at Layer 2 over the Backbone. It is thus useful when the 910 messaging over the Backbone that is associated to wireless mobility 911 becomes expensive, e.g., when the Layer 2 topology is virtualized 912 over a wide area IP underlay. 914 This mode is definitely required when the LLN uses a MAC address 915 format that is different from that on the Backbone (e.g., EUI-64 vs. 916 EUI-48). Since a 6LN may not be able to resolve an arbitrary 917 destination in the MLSN directly, a prefix that is used across a MLSN 918 MUST NOT be advertised as on-link in RA messages sent towards the 919 LLN. 921 In order to maintain IP connectivity, the 6BBR installs a connected 922 Host route to the Registered Address on the LLN interface, via the 923 Registering Node as identified by the Source Address and the SLLA 924 option in the NS(EARO) messages. 926 When operating as a Routing Proxy, the 6BBR MUST use its Layer 2 927 Address on its Backbone Interface in the SLLAO of the RS messages and 928 the TLLAO of the NA messages that it generates to advertise the 929 Registered Addresses. 931 For each Registered Address, multiple peers on the Backbone may have 932 resolved the Address with the 6BBR MAC Address, maintaining that 933 mapping in their Neighbor Cache. The 6BBR SHOULD maintain a list of 934 the peers on the Backbone which have associated its MAC Address with 935 the Registered Address. If that Registered Address moves to another 936 6BBR, the previous 6BBR SHOULD unicast a gratuitous NA to each such 937 peer, to supply the LLA of the new 6BBR in the TLLA option for the 938 Address. A 6BBR that does not maintain this list MAY multicast a 939 gratuitous NA message; this NA will possibly hit all the nodes on the 940 Backbone, whether or not they maintain an NCE for the Registered 941 Address. In either case, the 6BBR MAY set the Override flag if it is 942 known that the Registered Node cannot attach to the backbone, so as 943 to avoid interruptions and save probing flows in the future. 945 If a correspondent fails to receive the gratuitous NA, it will keep 946 sending traffic to a 6BBR to which the node was previously 947 registered. Since the previous 6BBR removed its Host route to the 948 Registered Address, it will look up the address over the backbone, 949 resolve the address with the LLA of the new 6BBR, and forward the 950 packet to the correct 6BBR. The previous 6BBR SHOULD also issue a 951 redirect message [RFC4861] to update the cache of the correspondent. 953 8. Bridging Proxy Operations 955 A Bridging Proxy provides IPv6 ND proxy functions between the LLN and 956 the backbone while preserving the forwarding continuity at the MAC 957 Layer. It acts as a Layer 2 Bridge for all types of unicast packets 958 including link-scoped, and appears as an IPv6 Host on the Backbone. 960 The Bridging Proxy registers any Binding including for a Link-Local 961 address to the 6LBR (if present) and defends it over the backbone in 962 IPv6 ND procedures. 964 To achieve this, the Bridging Proxy intercepts the IPv6 ND messages 965 and may reinject them on the other side, respond directly or drop 966 them. For instance, an ND(Lookup) from the backbone that matches a 967 Binding can be responded directly, or turned into a unicast on the 968 LLN side to let the 6LN respond. 970 As a Bridging Proxy, the 6BBR MUST use the Registering Node's Layer 2 971 Address in the SLLAO of the NS/RS messages and the TLLAO of the NA 972 messages that it generates to advertise the Registered Addresses. 973 The Registering Node's Layer 2 address is found in the SLLA of the 974 registration NS(EARO), and maintained in the Binding Table. 976 The Multi-Link Subnet prefix SHOULD NOT be advertised as on-link in 977 RA messages sent towards the LLN. If a destination address is seen 978 as on-link, then a 6LN may use NS(Lookup) messages to resolve that 979 address. In that case, the 6BBR MUST either answer the NS(Lookup) 980 message directly or reinject the message on the backbone, either as a 981 Layer 2 unicast or a multicast. 983 If the Registering Node owns the Registered Address, meaning that the 984 Registering Node is the Registered Node, then its mobility does not 985 impact existing NCEs over the Backbone. In a network where proxy 986 registrations are used, meaning that the Registering Node acts on 987 behalf of the Registered Node, if the Registered Node selects a new 988 Registering Node then the existing NCEs across the Backbone pointing 989 at the old Registering Node must be updated. In that case, the 6BBR 990 SHOULD attempt to fix the existing NCEs across the Backbone pointing 991 at other 6BBRs using NA messages as described in Section 7. 993 This method can fail if the multicast message is not received; one or 994 more correspondent nodes on the Backbone might maintain an stale NCE, 995 and packets to the Registered Address may be lost. When this 996 condition happens, it is eventually discovered and resolved using NUD 997 as defined in [RFC4861]. 999 9. Creating and Maintaining a Binding 1001 Upon receiving a registration for a new Address (i.e., an NS(EARO) 1002 with the R flag set), the 6BBR creates a Binding and operates as a 1003 6LR according to [RFC8505], interacting with the 6LBR if one is 1004 present. 1006 An implementation of a Routing Proxy that creates a Binding MUST also 1007 create an associated Host route pointing to the registering node in 1008 the LLN interface from which the registration was received. 1010 Acting as a 6BBR, the 6LR operation is modified as follows: 1012 * Acting as Bridging Proxy the 6LR MUST proxy ND over the backbone 1013 for registered Link-Local Addresses. 1015 * EDAR and EDAC messages SHOULD carry a SLLAO and a TLLAO, 1016 respectively. 1018 * An EDAC message with a status of 9 (6LBR Registry Saturated) is 1019 assimilated as a status of 0 if a following DAD process protects 1020 the address against duplication. 1022 This specification enables nodes on a Backbone Link to co-exist along 1023 with nodes implementing IPv6 ND [RFC4861] as well as other non- 1024 normative specifications such as [I-D.bi-savi-wlan]. It is possible 1025 that not all IPv6 addresses on the Backbone are registered and known 1026 to the 6LBR, and an EDAR/EDAC echange with the 6LBR might succeed 1027 even for a duplicate address. Consequently the 6BBR still needs to 1028 perform IPv6 ND DAD over the backbone after an EDAC with a status 1029 code of 0 or 9. 1031 For the DAD operation, the Binding is placed in Tentative state for a 1032 duration of TENTATIVE_DURATION (Section 12), and an NS(DAD) message 1033 is sent as a multicast message over the Backbone to the SNMA 1034 associated with the registered Address [RFC4862]. The EARO from the 1035 registration MUST be placed unchanged in the NS(DAD) message. 1037 If a registration is received for an existing Binding with a non-null 1038 Registration Lifetime and the registration is fresher (same ROVR, 1039 fresher TID), then the Binding is updated, with the new Registration 1040 Lifetime, TID, and possibly Registering Node. In Tentative state 1041 (see Section 9.1), the current DAD operation continues unaltered. In 1042 other states (see Section 9.2 and Section 9.3 ), the Binding is 1043 placed in Reachable state for the Registration Lifetime, and the 6BBR 1044 returns an NA(EARO) to the Registering Node with a status of 0 1045 (Success). 1047 Upon a registration that is identical (same ROVR, TID, and 1048 Registering Node), the 6BBR does not alter its current state. In 1049 Reachable State it returns an NA(EARO) back to the Registering Node 1050 with a status of 0 (Success). A registration that is not as fresh 1051 (same ROVR, older TID) is ignored. 1053 If a registration is received for an existing Binding and a 1054 registration Lifetime of zero, then the Binding is removed, and the 1055 6BBR returns an NA(EARO) back to the Registering Node with a status 1056 of 0 (Success). An implementation of a Routing Proxy that removes a 1057 binding MUST remove the associated Host route pointing on the 1058 registering node. 1060 The old 6BBR removes its Binding Table entry and notifies the 1061 Registering Node with a status of 3 (Moved) if a new 6BBR claims a 1062 fresher registration (same ROVR, fresher TID) for the same address. 1063 The old 6BBR MAY preserve a temporary state in order to forward 1064 packets in flight. The state may for instance be a NCE formed based 1065 on a received NA message. It may also be a Binding Table entry in 1066 Stale state and pointing at the new 6BBR on the backbone, or any 1067 other abstract cache entry that can be used to resolve the Link-Layer 1068 Address of the new 6BBR. The old 6BBR SHOULD also use REDIRECT 1069 messages as specified in [RFC4861] to update the correspondents for 1070 the Registered Address, pointing to the new 6BBR. 1072 9.1. Operations on a Binding in Tentative State 1074 The Tentative state covers a DAD period over the backbone during 1075 which an address being registered is checked for duplication using 1076 procedures defined in [RFC4862]. 1078 For a Binding in Tentative state: 1080 * The Binding MUST be removed if an NA message is received over the 1081 Backbone for the Registered Address with no EARO, or containing an 1082 EARO that indicates an existing registration owned by a different 1083 Registering Node (different ROVR). In that case, an NA is sent 1084 back to the Registering Node with a status of 1 (Duplicate) to 1085 indicate that the binding has been rejected. This behavior might 1086 be overridden by policy, in particular if the registration is 1087 trusted, e.g., based on the validation of the ROVR field (see 1088 [I-D.ietf-6lo-ap-nd]). 1090 * The Binding MUST be removed if an NS(DAD) message is received over 1091 the Backbone for the Registered Address with no EARO, or 1092 containing an EARO with a different ROVR that indicates a 1093 tentative registration by a different Registering Node. In that 1094 case, an NA is sent back to the Registering Node with a status of 1095 1 (Duplicate). This behavior might be overridden by policy, in 1096 particular if the registration is trusted, e.g., based on the 1097 validation of the ROVR field (see [I-D.ietf-6lo-ap-nd]). 1099 * The Binding MUST be removed if an NA or an NS(DAD) message is 1100 received over the Backbone for the Registered Address containing 1101 an EARO with a that indicates a fresher registration ([RFC8505]) 1102 for the same Registering Node (same ROVR). In that case, an NA 1103 MUST be sent back to the Registering Node with a status of 3 1104 (Moved). 1106 * The Binding MUST be kept unchanged if an NA or an NS(DAD) message 1107 is received over the Backbone for the Registered Address 1108 containing an EARO with a that indicates an older registration 1109 ([RFC8505]) for the same Registering Node (same ROVR). The 1110 message is answered with an NA that carries an EARO with a status 1111 of 3 (Moved) and the Override flag not set. This behavior might 1112 be overridden by policy, in particular if the registration is not 1113 trusted. 1115 * Other NS(DAD) and NA messages from the Backbone are ignored. 1117 * NS(Lookup) and NS(NUD) messages SHOULD be optimistically answered 1118 with an NA message containing an EARO with a status of 0 and the 1119 Override flag not set (see Section 3.6). If optimistic DAD is 1120 disabled, then they SHOULD be queued to be answered when the 1121 Binding goes to Reachable state. 1123 When the TENTATIVE_DURATION (Section 12) timer elapses, the Binding 1124 is placed in Reachable state for the Registration Lifetime, and the 1125 6BBR returns an NA(EARO) to the Registering Node with a status of 0 1126 (Success). 1128 The 6BBR also attempts to take over any existing Binding from other 1129 6BBRs and to update existing NCEs in backbone nodes. This is done by 1130 sending an NA message with an EARO and the Override flag not set over 1131 the backbone (see Section 7 and Section 8). 1133 9.2. Operations on a Binding in Reachable State 1135 The Reachable state covers an active registration after a successful 1136 DAD process. 1138 If the Registration Lifetime is of a long duration, an implementation 1139 might be configured to reassess the availability of the Registering 1140 Node at a lower period, using a NUD procedure as specified in 1141 [RFC7048]. If the NUD procedure fails, the Binding SHOULD be placed 1142 in Stale state immediately. 1144 For a Binding in Reachable state: 1146 * The Binding MUST be removed if an NA or an NS(DAD) message is 1147 received over the Backbone for the Registered Address containing 1148 an EARO that indicates a fresher registration ([RFC8505]) for the 1149 same Registered Node (i.e., same ROVR but fresher TID). A status 1150 of 4 (Removed) is returned in an asynchronous NA(EARO) to the 1151 Registering Node. Based on configuration, an implementation may 1152 delay this operation by a timer with a short setting, e.g., a few 1153 seconds to a minute, in order to a allow for a parallel 1154 registration to reach this node, in which case the NA might be 1155 ignored. 1157 * NS(DAD) and NA messages containing an EARO that indicates a 1158 registration for the same Registered Node that is not as fresh as 1159 this binding MUST be answered with an NA message containing an 1160 EARO with a status of 3 (Moved). 1162 * An NS(DAD) with no EARO or with an EARO that indicates a duplicate 1163 registration (i.e., different ROVR) MUST be answered with an NA 1164 message containing an EARO with a status of 1 (Duplicate) and the 1165 Override flag not set, unless the received message is an NA that 1166 carries an EARO with a status of 1, in which case the node 1167 refrains from answering. 1169 * Other NS(DAD) and NA messages from the Backbone are ignored. 1171 * NS(Lookup) and NS(NUD) messages SHOULD be answered with an NA 1172 message containing an EARO with a status of 0 and the Override 1173 flag not set. The 6BBR MAY check whether the Registering Node is 1174 still available using a NUD procedure over the LLN prior to 1175 answering; this behaviour depends on the use case and is subject 1176 to configuration. 1178 When the Registration Lifetime timer elapses, the Binding is placed 1179 in Stale state for a duration of STALE_DURATION (Section 12). 1181 9.3. Operations on a Binding in Stale State 1183 The Stale state enables tracking of the Backbone peers that have a 1184 NCE pointing to this 6BBR in case the Registered Address shows up 1185 later. 1187 If the Registered Address is claimed by another 6LN on the Backbone, 1188 with an NS(DAD) or an NA, the 6BBR does not defend the Address. 1190 For a Binding in Stale state: 1192 * The Binding MUST be removed if an NA or an NS(DAD) message is 1193 received over the Backbone for the Registered Address containing 1194 no EARO or an EARO that indicates either a fresher registration 1195 for the same Registered Node or a duplicate registration. A 1196 status of 4 (Removed) MAY be returned in an asynchronous NA(EARO) 1197 to the Registering Node. 1199 * NS(DAD) and NA messages containing an EARO that indicates a 1200 registration for the same Registered Node that is not as fresh as 1201 this MUST be answered with an NA message containing an EARO with a 1202 status of 3 (Moved). 1204 * If the 6BBR receives an NS(Lookup) or an NS(NUD) message for the 1205 Registered Address, the 6BBR MUST attempt a NUD procedure as 1206 specified in [RFC7048] to the Registering Node, targeting the 1207 Registered Address, prior to answering. If the NUD procedure 1208 succeeds, the operation in Reachable state applies. If the NUD 1209 fails, the 6BBR refrains from answering. 1211 * Other NS(DAD) and NA messages from the Backbone are ignored. 1213 When the STALE_DURATION (Section 12) timer elapses, the Binding MUST 1214 be removed. 1216 10. Registering Node Considerations 1218 A Registering Node MUST implement [RFC8505] in order to interact with 1219 a 6BBR (which acts as a routing registrar). Following [RFC8505], the 1220 Registering Node signals that it requires IPv6 proxy-ND services from 1221 a 6BBR by registering the corresponding IPv6 Address using an 1222 NS(EARO) message with the R flag set. 1224 The Registering Node may be the 6LN owning the IPv6 Address, or a 1225 6LBR that performs the registration on its behalf in a Route-Over 1226 mesh. 1228 A 6LN MUST register all of its IPv6 Addresses to its 6LR, which is 1229 the 6BBR when they are connected at Layer 2. Failure to register an 1230 address may result in the address being unreachable by other parties. 1231 This would happen for instance if the 6BBR propagates the NS(Lookup) 1232 from the backbone only to the LLN nodes that do not register their 1233 addresses. 1235 The Registering Node MUST refrain from using multicast NS(Lookup) 1236 when the destination is not known as on-link, e.g., if the prefix is 1237 advertised in a PIO with the L flag that is not set. In that case, 1238 the Registering Node sends its packets directly to its 6LR. 1240 The Registering Node SHOULD also follow BCP 202 [RFC7772] in order to 1241 limit the use of multicast RAs. It SHOULD also implement Simple 1242 Procedures for Detecting Network Attachment in IPv6 [RFC6059] (DNA 1243 procedures) to detect movements, and support Packet-Loss Resiliency 1244 for Router Solicitations [RFC7559] in order to improve reliability 1245 for the unicast RS messages. 1247 11. Security Considerations 1249 The procedures in this document modify the mechanisms used for IPv6 1250 ND and DAD and should not affect other aspects of IPv6 or higher- 1251 level-protocol operation. As such, the main classes of attacks that 1252 are in play are those which week to block neighbor discovery or to 1253 forcibly claim an address that another node is attempting to use. In 1254 the absence of cryptographic protection at higher layers, the latter 1255 class of attacks can have significant consequences, with the attacker 1256 being able to read all the "stolen" traffic that was directed to the 1257 target of the attack. 1259 This specification applies to LLNs and a backbone in which the 1260 individual links are protected against rogue access, on the LLN by 1261 authenticating a node that attaches to the network and encrypting at 1262 the MAC layer the transmissions, and on the backbone side using the 1263 physical security and access control measures that are typically 1264 applied there, so packets may neither be forged or nor overheard. 1266 In particular, the LLN MAC is required to provide secure unicast to/ 1267 from the Backbone Router and secure broadcast from the routers in a 1268 way that prevents tampering with or replaying the ND messages. 1270 For the IPv6 ND operation over the backbone, and unless the classical 1271 ND is disabled (e.g., by configuration), the classical ND messages 1272 are interpreted as emitted by the address owner and have precedence 1273 over the 6BBR that is only a proxy. 1275 It results that the security threats that are detailed in section 1276 11.1 of [RFC4861] fully apply to this specification as well. In very 1277 short: 1279 * Any node that can send a packet on the backbone can take over any 1280 address including addresses of LLN nodes by claiming it with an NA 1281 message and the Override bit set. This means that the real owner 1282 will stop receiving its packets. 1284 * Any node that can send a packet on the backbone can forge traffic 1285 and pretend it is issued from a address that it does not own, even 1286 if it did not claim the address using ND. 1288 * Any node that can send a packet on the backbone can present itself 1289 as a preferred router to intercept all traffic outgoing the 1290 subnet. It may even expose a prefix on the subnet as not-on-link 1291 and intercept all the traffic within the subnet. 1293 * If the rogue can receive a packet from the backbone it can also 1294 snoop all the intercepted traffic, be it by stealing an address or 1295 the role of a router. 1297 This means that any rogue access to the backbone must be prevented at 1298 all times, and that nodes that are attached to the backbone must be 1299 fully trusted / never compromised. 1301 Using address registration as the sole ND mechanism on a link and 1302 coupling it with [I-D.ietf-6lo-ap-nd] guarantees the ownership of a 1303 registered address within that link. 1305 * The protection is based on a proof-of-ownership encoded in the 1306 ROVR field and protects against address theft and impersonation by 1307 a 6LN, because the 6LR can challenge the Registered Node for a 1308 proof-of-ownership. 1310 * The protection extends to the full LLN in the case of an LLN Link, 1311 but does not extend over the backbone since the 6BBR cannot 1312 provide the proof-of-ownership when it defends the address. 1314 A possible attack over the backbone can be done by sending an NS with 1315 an EARO and expecting the NA(EARO) back to contain the TID and ROVR 1316 fields of the existing state. With that information, the attacker 1317 can easily increase the TID and take over the Binding. 1319 If the classical ND is disabled on the backbone and the use of 1320 [I-D.ietf-6lo-ap-nd] and a 6LBR are mandated, the network will 1321 benefit from the following new advantages: 1323 Zero-trust security for ND flows within the whole subnet: the 1324 increased security that [I-D.ietf-6lo-ap-nd] provides on the LLN 1325 will also apply to the backbone; it becomes impossible for an 1326 attached node to claim an address that belongs to another node 1327 using ND, and the network can filter packets that are not 1328 originated by the owner of the source address (SAVI), as long as 1329 that the routers are known and trusted. 1331 Remote ND DoS attack avoidance: the complete list of addresses in 1332 the network will be known to the 6LBR and available to the default 1333 router; with that information the router does not need to send a 1334 multicast NA(Lookup) in case of a Neighbor Cache miss for an 1335 incoming packet, which is a source of remote DoS attack against 1336 the network 1338 Less IPv6 ND-related multicast on the backbone: DAD and NS(Lookup) 1339 become unicast queries to the 6LBR 1341 Better DAD operation on wireless: DAD has been found to fail to 1342 detect duplications on large Wi-Fi infrastructures due to the 1343 unreliable broadcast operation on wireless; using a 6LBR enables a 1344 unicast lookup 1346 Less Layer-2 churn on the backbone: Using the Routing Proxy 1347 approach, the Link-Layer address of the LLN devices and their 1348 mobility are not visible in the backbone; only the Link-Layer 1349 addresses of the 6BBR and backbone nodes are visible at Layer 2 on 1350 the backbone. This is mandatory for LLNs that cannot be bridged 1351 on the backbone, and useful in any case to scale down, stabilize 1352 the forwarding tables at Layer 2 and avoid the gratuitous frames 1353 that are typically broadcasted to fix the transparent bridging 1354 tables when a wireless node roams from an AP to the next. 1356 This specification introduces a 6BBR that is a router on the path of 1357 the LLN traffic and a 6LBR that is used for the lookup. They could 1358 be interesting targets for an attacker. A compromised 6BBR can 1359 accept a registration but block the traffic, or refrain from 1360 proxying. A compromised 6LBR may accept unduly the transfer of 1361 ownership of an address, or block a new comer by faking that its 1362 address is a duplicate. But those attacks are possible in a 1363 classical network from a compromised default router and a DHCP 1364 server, respectively, and can be prevented using the same methods. 1366 A possible attack over the LLN can still be done by compromising a 1367 6LR. A compromised 6LR may modify the ROVR of EDAR messages in 1368 flight and transfer the ownership of the Registered Address to itself 1369 or a tier. It may also claim that a ROVR was validated when it 1370 really wasn't, and reattribute an address to self or to an attached 1371 6LN. This means that 6LRs, as well as 6LBRs and 6BBRS must still be 1372 fully trusted / never compromised. 1374 This specification mandates to check on the 6LBR on the backbone 1375 before doing the classical DAD, in case the address already exists. 1376 This may delay the DAD operation and should be protected by a short 1377 timer, in the order of 100ms or less, which will only represent a 1378 small extra delay versus the 1s wait of the DAD operation. 1380 12. Protocol Constants 1382 This Specification uses the following constants: 1384 TENTATIVE_DURATION: 800 milliseconds 1386 STALE_DURATION: see below 1388 In LLNs with long-lived Addresses such as LPWANs, STALE_DURATION 1389 SHOULD be configured with a relatively long value to cover an 1390 interval when the address may be reused, and before it is safe to 1391 expect that the address was definitively released. A good default 1392 value can be 24 hours. In LLNs where addresses are renewed rapidly, 1393 e.g., for privacy reasons, STALE_DURATION SHOULD be configured with a 1394 relatively shorter value, by default 5 minutes. 1396 13. IANA Considerations 1398 This document has no request to IANA. 1400 14. Acknowledgments 1402 Many thanks to Dorothy Stanley, Thomas Watteyne and Jerome Henry for 1403 their various contributions. Also many thanks to Timothy Winters and 1404 Erik Nordmark for their help, review and support in preparation to 1405 the IESG cycle, and to Kyle Rose, Elwyn Davies, Barry Leiba, Mirja 1406 Kuhlewind, Alvaro Retana, Roman Danyliw and very especially Dominique 1407 Barthel and Benjamin Kaduk for their useful contributions through the 1408 IETF last call and IESG process. 1410 15. Normative References 1412 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1413 Requirement Levels", BCP 14, RFC 2119, 1414 DOI 10.17487/RFC2119, March 1997, 1415 . 1417 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 1418 Architecture", RFC 4291, DOI 10.17487/RFC4291, February 1419 2006, . 1421 [RFC4429] Moore, N., "Optimistic Duplicate Address Detection (DAD) 1422 for IPv6", RFC 4429, DOI 10.17487/RFC4429, April 2006, 1423 . 1425 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 1426 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 1427 DOI 10.17487/RFC4861, September 2007, 1428 . 1430 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 1431 Address Autoconfiguration", RFC 4862, 1432 DOI 10.17487/RFC4862, September 2007, 1433 . 1435 [RFC6059] Krishnan, S. and G. Daley, "Simple Procedures for 1436 Detecting Network Attachment in IPv6", RFC 6059, 1437 DOI 10.17487/RFC6059, November 2010, 1438 . 1440 [RFC6775] Shelby, Z., Ed., Chakrabarti, S., Nordmark, E., and C. 1441 Bormann, "Neighbor Discovery Optimization for IPv6 over 1442 Low-Power Wireless Personal Area Networks (6LoWPANs)", 1443 RFC 6775, DOI 10.17487/RFC6775, November 2012, 1444 . 1446 [RFC7048] Nordmark, E. and I. Gashinsky, "Neighbor Unreachability 1447 Detection Is Too Impatient", RFC 7048, 1448 DOI 10.17487/RFC7048, January 2014, 1449 . 1451 [RFC7559] Krishnan, S., Anipko, D., and D. Thaler, "Packet-Loss 1452 Resiliency for Router Solicitations", RFC 7559, 1453 DOI 10.17487/RFC7559, May 2015, 1454 . 1456 [RFC7772] Yourtchenko, A. and L. Colitti, "Reducing Energy 1457 Consumption of Router Advertisements", BCP 202, RFC 7772, 1458 DOI 10.17487/RFC7772, February 2016, 1459 . 1461 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 1462 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 1463 May 2017, . 1465 [RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6 1466 (IPv6) Specification", STD 86, RFC 8200, 1467 DOI 10.17487/RFC8200, July 2017, 1468 . 1470 [RFC8201] McCann, J., Deering, S., Mogul, J., and R. Hinden, Ed., 1471 "Path MTU Discovery for IP version 6", STD 87, RFC 8201, 1472 DOI 10.17487/RFC8201, July 2017, 1473 . 1475 [RFC8505] Thubert, P., Ed., Nordmark, E., Chakrabarti, S., and C. 1476 Perkins, "Registration Extensions for IPv6 over Low-Power 1477 Wireless Personal Area Network (6LoWPAN) Neighbor 1478 Discovery", RFC 8505, DOI 10.17487/RFC8505, November 2018, 1479 . 1481 16. Informative References 1483 [RFC4389] Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery 1484 Proxies (ND Proxy)", RFC 4389, DOI 10.17487/RFC4389, April 1485 2006, . 1487 [RFC4903] Thaler, D., "Multi-Link Subnet Issues", RFC 4903, 1488 DOI 10.17487/RFC4903, June 2007, 1489 . 1491 [RFC5415] Calhoun, P., Ed., Montemurro, M., Ed., and D. Stanley, 1492 Ed., "Control And Provisioning of Wireless Access Points 1493 (CAPWAP) Protocol Specification", RFC 5415, 1494 DOI 10.17487/RFC5415, March 2009, 1495 . 1497 [RFC5568] Koodli, R., Ed., "Mobile IPv6 Fast Handovers", RFC 5568, 1498 DOI 10.17487/RFC5568, July 2009, 1499 . 1501 [RFC6606] Kim, E., Kaspar, D., Gomez, C., and C. Bormann, "Problem 1502 Statement and Requirements for IPv6 over Low-Power 1503 Wireless Personal Area Network (6LoWPAN) Routing", 1504 RFC 6606, DOI 10.17487/RFC6606, May 2012, 1505 . 1507 [RFC6275] Perkins, C., Ed., Johnson, D., and J. Arkko, "Mobility 1508 Support in IPv6", RFC 6275, DOI 10.17487/RFC6275, July 1509 2011, . 1511 [RFC6550] Winter, T., Ed., Thubert, P., Ed., Brandt, A., Hui, J., 1512 Kelsey, R., Levis, P., Pister, K., Struik, R., Vasseur, 1513 JP., and R. Alexander, "RPL: IPv6 Routing Protocol for 1514 Low-Power and Lossy Networks", RFC 6550, 1515 DOI 10.17487/RFC6550, March 2012, 1516 . 1518 [RFC6830] Farinacci, D., Fuller, V., Meyer, D., and D. Lewis, "The 1519 Locator/ID Separation Protocol (LISP)", RFC 6830, 1520 DOI 10.17487/RFC6830, January 2013, 1521 . 1523 [RFC8273] Brzozowski, J. and G. Van de Velde, "Unique IPv6 Prefix 1524 per Host", RFC 8273, DOI 10.17487/RFC8273, December 2017, 1525 . 1527 [I-D.yourtchenko-6man-dad-issues] 1528 Yourtchenko, A. and E. Nordmark, "A survey of issues 1529 related to IPv6 Duplicate Address Detection", Work in 1530 Progress, Internet-Draft, draft-yourtchenko-6man-dad- 1531 issues-01, 3 March 2015, . 1534 [I-D.nordmark-6man-dad-approaches] 1535 Nordmark, E., "Possible approaches to make DAD more robust 1536 and/or efficient", Work in Progress, Internet-Draft, 1537 draft-nordmark-6man-dad-approaches-02, 19 October 2015, 1538 . 1541 [I-D.ietf-6man-rs-refresh] 1542 Nordmark, E., Yourtchenko, A., and S. Krishnan, "IPv6 1543 Neighbor Discovery Optional RS/RA Refresh", Work in 1544 Progress, Internet-Draft, draft-ietf-6man-rs-refresh-02, 1545 31 October 2016, . 1548 [I-D.ietf-6lo-ap-nd] 1549 Thubert, P., Sarikaya, B., Sethi, M., and R. Struik, 1550 "Address Protected Neighbor Discovery for Low-power and 1551 Lossy Networks", Work in Progress, Internet-Draft, draft- 1552 ietf-6lo-ap-nd-20, 9 March 2020, 1553 . 1555 [I-D.ietf-6tisch-architecture] 1556 Thubert, P., "An Architecture for IPv6 over the TSCH mode 1557 of IEEE 802.15.4", Work in Progress, Internet-Draft, 1558 draft-ietf-6tisch-architecture-28, 29 October 2019, 1559 . 1562 [I-D.ietf-mboned-ieee802-mcast-problems] 1563 Perkins, C., McBride, M., Stanley, D., Kumari, W., and J. 1564 Zuniga, "Multicast Considerations over IEEE 802 Wireless 1565 Media", Work in Progress, Internet-Draft, draft-ietf- 1566 mboned-ieee802-mcast-problems-11, 11 December 2019, 1567 . 1570 [I-D.bi-savi-wlan] 1571 Bi, J., Wu, J., Wang, Y., and T. Lin, "A SAVI Solution for 1572 WLAN", Work in Progress, Internet-Draft, draft-bi-savi- 1573 wlan-18, 17 November 2019, 1574 . 1576 [I-D.thubert-6lo-unicast-lookup] 1577 Thubert, P. and E. Levy-Abegnoli, "IPv6 Neighbor Discovery 1578 Unicast Lookup", Work in Progress, Internet-Draft, draft- 1579 thubert-6lo-unicast-lookup-00, 25 January 2019, 1580 . 1583 [IEEEstd8021] 1584 IEEE standard for Information Technology, "IEEE Standard 1585 for Information technology -- Telecommunications and 1586 information exchange between systems Local and 1587 metropolitan area networks Part 1: Bridging and 1588 Architecture". 1590 [IEEEstd80211] 1591 IEEE standard for Information Technology, "IEEE Standard 1592 for Information technology -- Telecommunications and 1593 information exchange between systems Local and 1594 metropolitan area networks-- Specific requirements Part 1595 11: Wireless LAN Medium Access Control (MAC) and Physical 1596 Layer (PHY) Specifications". 1598 [IEEEstd802151] 1599 IEEE standard for Information Technology, "IEEE Standard 1600 for Information Technology - Telecommunications and 1601 Information Exchange Between Systems - Local and 1602 Metropolitan Area Networks - Specific Requirements. - Part 1603 15.1: Wireless Medium Access Control (MAC) and Physical 1604 Layer (PHY) Specifications for Wireless Personal Area 1605 Networks (WPANs)". 1607 [IEEEstd802154] 1608 IEEE standard for Information Technology, "IEEE Standard 1609 for Local and metropolitan area networks -- Part 15.4: 1610 Low-Rate Wireless Personal Area Networks (LR-WPANs)". 1612 Appendix A. Possible Future Extensions 1614 With the current specification, the 6LBR is not leveraged to avoid 1615 multicast NS(Lookup) on the Backbone. This could be done by adding a 1616 lookup procedure in the EDAR/EDAC exchange. 1618 By default the specification does not have a fine-grained trust 1619 model: all nodes that can authenticate to the LLN MAC or attach to 1620 the backbone are equally trusted. It would be desirable to provide a 1621 stronger authorization model, e.g., whereby nodes that associate 1622 their address with a proof-of-ownership [I-D.ietf-6lo-ap-nd] should 1623 be more trusted than nodes that do not. Such a trust model and 1624 related signaling could be added in the future to override the 1625 default operation and favor trusted nodes. 1627 Future documents may extend this specification by allowing the 6BBR 1628 to redistribute Host routes in routing protocols that would operate 1629 over the Backbone, or in MIPv6 [RFC6275], or FMIP [RFC5568], or the 1630 Locator/ID Separation Protocol (LISP) [RFC6830] to support mobility 1631 on behalf of the 6LNs, etc... LISP may also be used to provide an 1632 equivalent to the EDAR/EDAC exchange using a Map Server / Map 1633 Resolver as a replacement to the 6LBR. 1635 Appendix B. Applicability and Requirements Served 1637 This document specifies proxy-ND functions that can be used to 1638 federate an IPv6 Backbone Link and multiple IPv6 LLNs into a single 1639 Multi-Link Subnet. The proxy-ND functions enable IPv6 ND services 1640 for Duplicate Address Detection (DAD) and Address Lookup that do not 1641 require broadcasts over the LLNs. 1643 The term LLN is used to cover multiple types of WLANs and WPANs, 1644 including (Low-Power) Wi-Fi, BLUETOOTH(R) Low Energy, IEEE STD 1645 802.11ah and IEEE STD.802.15.4 wireless meshes, covering the types of 1646 networks listed in Appendix B.3 of [RFC8505] "Requirements Related to 1647 Various Low-Power Link Types". 1649 Each LLN in the subnet is attached to an IPv6 Backbone Router (6BBR). 1650 The Backbone Routers interconnect the LLNs and advertise the 1651 Addresses of the 6LNs over the Backbone Link using proxy-ND 1652 operations. 1654 This specification updates IPv6 ND over the Backbone to distinguish 1655 Address movement from duplication and eliminate stale state in the 1656 Backbone routers and Backbone nodes once a 6LN has roamed. This way, 1657 mobile nodes may roam rapidly from one 6BBR to the next and 1658 requirements in Appendix B.1 of [RFC8505] "Requirements Related to 1659 Mobility" are met. 1661 A 6LN can register its IPv6 Addresses and thereby obtain proxy-ND 1662 services over the Backbone, meeting the requirements expressed in 1663 Appendix B.4 of [RFC8505], "Requirements Related to Proxy 1664 Operations". 1666 The negative impact of the IPv6 ND-related broadcasts can be limited 1667 to one of the federated links, enabling the number of 6LNs to grow. 1668 The Routing Proxy operation avoids the need to expose the MAC 1669 addresses of the 6LNs onto the backbone, keeping the Layer 2 topology 1670 simple and stable. This meets the requirements in Appendix B.6 of 1671 [RFC8505] "Requirements Related to Scalability", as long has the 1672 6BBRs are dimensioned for the number of registrations that each needs 1673 to support. 1675 In the case of a Wi-Fi access link, a 6BBR may be collocated with the 1676 Access Point (AP), or with a Fabric Edge (FE) or a CAPWAP [RFC5415] 1677 Wireless LAN Controller (WLC). In those cases, the wireless client 1678 (STA) is the 6LN that makes use of [RFC8505] to register its IPv6 1679 Address(es) to the 6BBR acting as Routing Registrar. The 6LBR can be 1680 centralized and either connected to the Backbone Link or reachable 1681 over IP. The 6BBR proxy-ND operations eliminate the need for 1682 wireless nodes to respond synchronously when a Lookup is performed 1683 for their IPv6 Addresses. This provides the function of a Sleep 1684 Proxy for ND [I-D.nordmark-6man-dad-approaches]. 1686 For the TimeSlotted Channel Hopping (TSCH) mode of [IEEEstd802154], 1687 the 6TiSCH architecture [I-D.ietf-6tisch-architecture] describes how 1688 a 6LoWPAN ND host could connect to the Internet via a RPL mesh 1689 Network, but doing so requires extensions to the 6LOWPAN ND protocol 1690 to support mobility and reachability in a secure and manageable 1691 environment. The extensions detailed in this document also work for 1692 the 6TiSCH architecture, serving the requirements listed in 1693 Appendix B.2 of [RFC8505] "Requirements Related to Routing 1694 Protocols". 1696 The registration mechanism may be seen as a more reliable alternate 1697 to snooping [I-D.bi-savi-wlan]. It can be noted that registration 1698 and snooping are not mutually exclusive. Snooping may be used in 1699 conjunction with the registration for nodes that do not register 1700 their IPv6 Addresses. The 6BBR assumes that if a node registers at 1701 least one IPv6 Address to it, then the node registers all of its 1702 Addresses to the 6BBR. With this assumption, the 6BBR can possibly 1703 cancel all undesirable multicast NS messages that would otherwise 1704 have been delivered to that node. 1706 Scalability of the Multi-Link Subnet [RFC4903] requires avoidance of 1707 multicast/broadcast operations as much as possible even on the 1708 Backbone [I-D.ietf-mboned-ieee802-mcast-problems]. Although hosts 1709 can connect to the Backbone using IPv6 ND operations, multicast RAs 1710 can be saved by using [I-D.ietf-6man-rs-refresh], which also requires 1711 the support of [RFC7559]. 1713 Authors' Addresses 1715 Pascal Thubert (editor) 1716 Cisco Systems, Inc 1717 Building D 1718 45 Allee des Ormes - BP1200 1719 06254 MOUGINS - Sophia Antipolis 1720 France 1722 Phone: +33 497 23 26 34 1723 Email: pthubert@cisco.com 1725 Charles E. Perkins 1726 Blue Meadow Networking 1727 Saratoga, 95070 1728 United States of America 1730 Email: charliep@computer.org 1732 Eric Levy-Abegnoli 1733 Cisco Systems, Inc 1734 Building D 1735 45 Allee des Ormes - BP1200 1736 06254 MOUGINS - Sophia Antipolis 1737 France 1739 Phone: +33 497 23 26 20 1740 Email: elevyabe@cisco.com