idnits 2.17.1 draft-ietf-hip-multihoming-00.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == The document seems to contain a disclaimer for pre-RFC5378 work, but was first submitted on or after 10 November 2008. The disclaimer is usually necessary only for documents that revise or obsolete older RFCs, and that take significant amounts of text from those RFCs. If you can contact all authors of the source material and they are willing to grant the BCP78 rights to the IETF Trust, you can and should remove the disclaimer. Otherwise, the disclaimer is needed and you can ignore this comment. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (October 18, 2010) is 4933 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-14) exists of draft-ietf-hip-rfc5206-bis-00 ** Obsolete normative reference: RFC 3484 (Obsoleted by RFC 6724) ** Obsolete normative reference: RFC 4423 (Obsoleted by RFC 9063) ** Obsolete normative reference: RFC 5201 (Obsoleted by RFC 7401) ** Obsolete normative reference: RFC 5202 (Obsoleted by RFC 7402) -- Obsolete informational reference (is this intentional?): RFC 5204 (Obsoleted by RFC 8004) Summary: 4 errors (**), 0 flaws (~~), 3 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group P. Nikander 3 Internet-Draft Ericsson Research NomadicLab 4 Intended status: Standards Track T. Henderson, Ed. 5 Expires: April 21, 2011 The Boeing Company 6 C. Vogt 7 J. Arkko 8 Ericsson Research NomadicLab 9 October 18, 2010 11 Host Multihoming with the Host Identity Protocol 12 draft-ietf-hip-multihoming-00 14 Abstract 16 This document defines host multihoming extensions to the Host 17 Identity Protocol (HIP), by leveraging protocol components defined 18 for host mobility. 20 Status of This Memo 22 This Internet-Draft is submitted in full conformance with the 23 provisions of BCP 78 and BCP 79. 25 Internet-Drafts are working documents of the Internet Engineering 26 Task Force (IETF). Note that other groups may also distribute 27 working documents as Internet-Drafts. The list of current Internet- 28 Drafts is at http://datatracker.ietf.org/drafts/current/. 30 Internet-Drafts are draft documents valid for a maximum of six months 31 and may be updated, replaced, or obsoleted by other documents at any 32 time. It is inappropriate to use Internet-Drafts as reference 33 material or to cite them other than as "work in progress." 35 This Internet-Draft will expire on April 21, 2011. 37 Copyright Notice 39 Copyright (c) 2010 IETF Trust and the persons identified as the 40 document authors. All rights reserved. 42 This document is subject to BCP 78 and the IETF Trust's Legal 43 Provisions Relating to IETF Documents 44 (http://trustee.ietf.org/license-info) in effect on the date of 45 publication of this document. Please review these documents 46 carefully, as they describe your rights and restrictions with respect 47 to this document. Code Components extracted from this document must 48 include Simplified BSD License text as described in Section 4.e of 49 the Trust Legal Provisions and are provided without warranty as 50 described in the Simplified BSD License. 52 This document may contain material from IETF Documents or IETF 53 Contributions published or made publicly available before November 54 10, 2008. The person(s) controlling the copyright in some of this 55 material may not have granted the IETF Trust the right to allow 56 modifications of such material outside the IETF Standards Process. 57 Without obtaining an adequate license from the person(s) controlling 58 the copyright in such materials, this document may not be modified 59 outside the IETF Standards Process, and derivative works of it may 60 not be created outside the IETF Standards Process, except to format 61 it for publication as an RFC or to translate it into languages other 62 than English. 64 Table of Contents 66 1. Introduction and Scope . . . . . . . . . . . . . . . . . . . . 3 67 2. Terminology and Conventions . . . . . . . . . . . . . . . . . 4 68 3. Protocol Model . . . . . . . . . . . . . . . . . . . . . . . . 4 69 3.1. Operating Environment . . . . . . . . . . . . . . . . . . 5 70 3.2. Multihoming Overview . . . . . . . . . . . . . . . . . . . 7 71 4. Protocol Overview . . . . . . . . . . . . . . . . . . . . . . 7 72 4.1. Host Multihoming . . . . . . . . . . . . . . . . . . . . . 8 73 4.2. Site Multihoming . . . . . . . . . . . . . . . . . . . . . 9 74 4.3. Dual host multihoming . . . . . . . . . . . . . . . . . . 10 75 4.4. Combined Mobility and Multihoming . . . . . . . . . . . . 10 76 4.5. Initiating the Protocol in R1 or I2 . . . . . . . . . . . 11 77 5. Other Considerations . . . . . . . . . . . . . . . . . . . . . 12 78 5.1. Address Verification . . . . . . . . . . . . . . . . . . . 12 79 5.2. Preferred Locator . . . . . . . . . . . . . . . . . . . . 12 80 5.3. Interaction with Security Associations . . . . . . . . . . 13 81 6. Processing Rules . . . . . . . . . . . . . . . . . . . . . . . 15 82 6.1. Sending LOCATORs . . . . . . . . . . . . . . . . . . . . . 15 83 6.2. Handling Received LOCATORs . . . . . . . . . . . . . . . . 17 84 6.3. Verifying Address Reachability . . . . . . . . . . . . . . 19 85 6.4. Changing the Preferred Locator . . . . . . . . . . . . . . 19 86 7. Security Considerations . . . . . . . . . . . . . . . . . . . 20 87 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 20 88 9. Authors and Acknowledgments . . . . . . . . . . . . . . . . . 20 89 10. References . . . . . . . . . . . . . . . . . . . . . . . . . . 20 90 10.1. Normative references . . . . . . . . . . . . . . . . . . . 20 91 10.2. Informative references . . . . . . . . . . . . . . . . . . 21 92 Appendix A. Document Revision History . . . . . . . . . . . . . . 21 94 1. Introduction and Scope 96 The Host Identity Protocol [RFC4423] (HIP) supports an architecture 97 that decouples the transport layer (TCP, UDP, etc.) from the 98 internetworking layer (IPv4 and IPv6) by using public/private key 99 pairs, instead of IP addresses, as host identities. When a host uses 100 HIP, the overlying protocol sublayers (e.g., transport layer sockets 101 and Encapsulating Security Payload (ESP) Security Associations (SAs)) 102 are instead bound to representations of these host identities, and 103 the IP addresses are only used for packet forwarding. However, each 104 host must also know at least one IP address at which its peers are 105 reachable. Initially, these IP addresses are the ones used during 106 the HIP base exchange [RFC5201]. 108 One consequence of such a decoupling is that new solutions to 109 network-layer mobility and host multihoming are possible. Host 110 mobility is defined in [I-D.ietf-hip-rfc5206-bis] and covers the case 111 in which a host has a single address and changes its network point- 112 of-attachment while desiring to preserve the HIP-enabled security 113 association. Host multihoming is somewhat of a dual case to host 114 mobility, in that a host may simultaneously have more than one 115 network point-of-attachment. There are potentially many variations 116 of host multihoming possible. The scope of this document encompasses 117 messaging and elements of procedure for some basic host multihoming 118 scenarios of interest. 120 Another variation of multihoming that has been heavily studied site 121 multihoming. Solutions for site multihoming in IPv6 networks have 122 been specified by the IETF shim6 working group. The shim6 protocol 123 [RFC5533] bears many architectural similarities to HIP but there are 124 differences in the security model and in the protocol. Future 125 versions of this draft will summarize the differences more 126 completely. 128 While HIP can potentially be used with transports other than the ESP 129 transport format [RFC5202], this document largely assumes the use of 130 ESP and leaves other transport formats for further study. 132 There are a number of situations where the simple end-to-end 133 readdressing functionality defined herein is not sufficient. These 134 include the initial reachability of a multihomed host, location 135 privacy, simultaneous mobility of both hosts, and some modes of NAT 136 traversal. In these situations, there is a need for some helper 137 functionality in the network, such as a HIP rendezvous server 138 [RFC5204]. Such functionality is out of the scope of this document. 139 Finally, making underlying IP multihoming transparent to the 140 transport layer has implications on the proper response of transport 141 congestion control, path MTU selection, and Quality of Service (QoS). 143 Transport-layer mobility triggers, and the proper transport response 144 to a HIP multihoming address change, are outside the scope of this 145 document. 147 2. Terminology and Conventions 149 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 150 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 151 document are to be interpreted as described in RFC 2119 [RFC2119]. 153 Terminology is copied from [I-D.ietf-hip-rfc5206-bis]. 155 LOCATOR. The name of a HIP parameter containing zero or more Locator 156 fields. This parameter's name is distinguished from the Locator 157 fields embedded within it by the use of all capital letters. 159 Locator. A name that controls how the packet is routed through the 160 network and demultiplexed by the end host. It may include a 161 concatenation of traditional network addresses such as an IPv6 162 address and end-to-end identifiers such as an ESP SPI. It may 163 also include transport port numbers or IPv6 Flow Labels as 164 demultiplexing context, or it may simply be a network address. 166 Address. A name that denotes a point-of-attachment to the network. 167 The two most common examples are an IPv4 address and an IPv6 168 address. The set of possible addresses is a subset of the set of 169 possible locators. 171 Preferred locator. A locator on which a host prefers to receive 172 data. With respect to a given peer, a host always has one active 173 Preferred locator, unless there are no active locators. By 174 default, the locators used in the HIP base exchange are the 175 Preferred locators. 177 Credit Based Authorization. A host must verify a mobile or 178 multihomed peer's reachability at a new locator. Credit-Based 179 Authorization authorizes the peer to receive a certain amount of 180 data at the new locator before the result of such verification is 181 known. 183 3. Protocol Model 185 This section is an overview; more detailed specification follows this 186 section. 188 The overall protocol model is the same as in Section 3 of 189 [I-D.ietf-hip-rfc5206-bis]; this section only highlights the 190 differences. 192 3.1. Operating Environment 194 The Host Identity Protocol (HIP) [RFC5201] is a key establishment and 195 parameter negotiation protocol. Its primary applications are for 196 authenticating host messages based on host identities, and 197 establishing security associations (SAs) for the ESP transport format 198 [RFC5202] and possibly other protocols in the future. 200 +--------------------+ +--------------------+ 201 | | | | 202 | +------------+ | | +------------+ | 203 | | Key | | HIP | | Key | | 204 | | Management | <-+-----------------------+-> | Management | | 205 | | Process | | | | Process | | 206 | +------------+ | | +------------+ | 207 | ^ | | ^ | 208 | | | | | | 209 | v | | v | 210 | +------------+ | | +------------+ | 211 | | IPsec | | ESP | | IPsec | | 212 | | Stack | <-+-----------------------+-> | Stack | | 213 | | | | | | | | 214 | +------------+ | | +------------+ | 215 | | | | 216 | | | | 217 | Initiator | | Responder | 218 +--------------------+ +--------------------+ 220 Figure 1: HIP Deployment Model 222 The general deployment model for HIP is shown above, assuming 223 operation in an end-to-end fashion. This document specifies 224 extensions to the HIP protocol to enable end-host mobility and basic 225 multihoming. In summary, these extensions to the HIP base protocol 226 enable the signaling of new addressing information to the peer in HIP 227 messages. The messages are authenticated via a signature or keyed 228 hash message authentication code (HMAC) based on its Host Identity. 230 --------- 231 | TCP | (sockets bound to HITs) 232 --------- 233 | 234 --------- 235 ----> | ESP | {HIT_s, HIT_d} <-> SPI 236 | --------- 237 | | 238 ---- --------- 239 | MH |-> | HIP | {HIT_s, HIT_d, SPI} <-> {IP_s, IP_d, SPI} 240 ---- --------- 241 | 242 --------- 243 | IP | 244 --------- 246 Figure 2: Architecture for HIP Multihoming (MH) 248 Figure 2 depicts a layered architectural view of a HIP-enabled stack 249 using the ESP transport format. In HIP, upper-layer protocols 250 (including TCP and ESP in this figure) are bound to Host Identity 251 Tags (HITs) and not IP addresses. The HIP sublayer is responsible 252 for maintaining the binding between HITs and IP addresses. The SPI 253 is used to associate an incoming packet with the right HITs. The 254 block labeled "MH" is introduced below. 256 Consider the case when a host is multihomed (has more than one 257 globally routable address) and has multiple addresses available at 258 the HIP layer as alternative locators for fault tolerance. Examples 259 include the use of (possibly multiple) IPv4 and IPv6 addresses on the 260 same interface, or the use of multiple interfaces attached to 261 different service providers. Such host multihoming generally 262 necessitates that a separate ESP SA is maintained for each interface 263 in order to prevent packets that arrive over different paths from 264 falling outside of the ESP anti-replay window [RFC4303]. Multihoming 265 thus makes it possible that the bindings shown on the right side of 266 Figure 2 are one to many (in the outbound direction, one HIT pair to 267 multiple SPIs, and possibly then to multiple IP addresses). However, 268 only one SPI and address pair can be used for any given packet, so 269 the job of the "MH" block depicted above is to dynamically manipulate 270 these bindings. Beyond locally managing such multiple bindings, the 271 peer-to-peer HIP signaling protocol needs to be flexible enough to 272 define the desired mappings between HITs, SPIs, and addresses, and 273 needs to ensure that UPDATE messages are sent along the right network 274 paths so that any HIP-aware middleboxes can observe the SPIs. This 275 document does not specify the "MH" block, nor does it specify 276 detailed elements of procedure for how to handle various multihoming 277 (perhaps combined with mobility) scenarios. The "MH" block may apply 278 to more general problems outside of HIP. However, this document does 279 describe a basic multihoming case (one host adds one address to its 280 initial address and notifies the peer) and leave more complicated 281 scenarios for experimentation and future documents. 283 3.2. Multihoming Overview 285 In host multihoming, a host has multiple locators simultaneously 286 rather than sequentially, as in the case of mobility. By using the 287 LOCATOR parameter defined in [I-D.ietf-hip-rfc5206-bis], a host can 288 inform its peers of additional (multiple) locators at which it can be 289 reached, and can declare a particular locator as a "preferred" 290 locator. Although this document defines a basic mechanism for 291 multihoming, it does not define detailed policies and procedures, 292 such as which locators to choose when more than one pair is 293 available, the operation of simultaneous mobility and multihoming, 294 source address selection policies (beyond those specified in 295 [RFC3484]), and the implications of multihoming on transport 296 protocols and ESP anti-replay windows. 298 4. Protocol Overview 300 In this section, we briefly introduce a number of usage scenarios for 301 HIP multihoming. These scenarios assume that HIP is being used with 302 the ESP transform [RFC5202], although other scenarios may be defined 303 in the future. To understand these usage scenarios, the reader 304 should be at least minimally familiar with the HIP protocol 305 specification [RFC5201]. However, for the (relatively) uninitiated 306 reader, it is most important to keep in mind that in HIP the actual 307 payload traffic is protected with ESP, and that the ESP SPI acts as 308 an index to the right host-to-host context. 310 The scenarios below assume that the two hosts have completed a single 311 HIP base exchange with each other. Both of the hosts therefore have 312 one incoming and one outgoing SA. Further, each SA uses the same 313 pair of IP addresses, which are the ones used in the base exchange. 315 The readdressing protocol is an asymmetric protocol where a mobile or 316 multihomed host informs a peer host about changes of IP addresses on 317 affected SPIs. The readdressing exchange is designed to be 318 piggybacked on existing HIP exchanges. The majority of the packets 319 on which the LOCATOR parameters are expected to be carried are UPDATE 320 packets. However, some implementations may want to experiment with 321 sending LOCATOR parameters also on other packets, such as R1, I2, and 322 NOTIFY. 324 The scenarios below at times describe addresses as being in either an 325 ACTIVE, VERIFIED, or DEPRECATED state. From the perspective of a 326 host, newly-learned addresses of the peer must be verified before put 327 into active service, and addresses removed by the peer are put into a 328 deprecated state. Under limited conditions described in 329 [I-D.ietf-hip-rfc5206-bis], an UNVERIFIED address may be used. 331 Hosts that use link-local addresses as source addresses in their HIP 332 handshakes may not be reachable by a mobile peer. Such hosts SHOULD 333 provide a globally routable address either in the initial handshake 334 or via the LOCATOR parameter. 336 4.1. Host Multihoming 338 A (mobile or stationary) host may sometimes have more than one 339 interface or global address. The host may notify the peer host of 340 the additional interface or address by using the LOCATOR parameter. 341 To avoid problems with the ESP anti-replay window, a host SHOULD use 342 a different SA for each interface or address used to receive packets 343 from the peer host when multiple locator pairs are being used 344 simultaneously rather than sequentially. 346 When more than one locator is provided to the peer host, the host 347 SHOULD indicate which locator is preferred (the locator on which the 348 host prefers to receive traffic). By default, the addresses used in 349 the base exchange are preferred until indicated otherwise. 351 In the multihoming case, the sender may also have multiple valid 352 locators from which to source traffic. In practice, a HIP 353 association in a multihoming configuration may have both a preferred 354 peer locator and a preferred local locator, although rules for source 355 address selection should ultimately govern the selection of the 356 source locator based on the destination locator. 358 Although the protocol may allow for configurations in which there is 359 an asymmetric number of SAs between the hosts (e.g., one host has two 360 interfaces and two inbound SAs, while the peer has one interface and 361 one inbound SA), it is RECOMMENDED that inbound and outbound SAs be 362 created pairwise between hosts. When an ESP_INFO arrives to rekey a 363 particular outbound SA, the corresponding inbound SA should be also 364 rekeyed at that time. Although asymmetric SA configurations might be 365 experimented with, their usage may constrain interoperability at this 366 time. However, it is recommended that implementations attempt to 367 support peers that prefer to use non-paired SAs. It is expected that 368 this section and behavior will be modified in future revisions of 369 this protocol, once the issue and its implications are better 370 understood. 372 Consider the case between two hosts, one single-homed and one 373 multihomed. The multihomed host may decide to inform the single- 374 homed host about its other address. It is RECOMMENDED that the 375 multihomed host set up a new SA pair for use on this new address. To 376 do this, the multihomed host sends a LOCATOR with an ESP_INFO, 377 indicating the request for a new SA by setting the OLD SPI value to 378 zero, and the NEW SPI value to the newly created incoming SPI. A 379 Locator Type of "1" is used to associate the new address with the new 380 SPI. The LOCATOR parameter also contains a second Type "1" locator, 381 that of the original address and SPI. To simplify parameter 382 processing and avoid explicit protocol extensions to remove locators, 383 each LOCATOR parameter MUST list all locators in use on a connection 384 (a complete listing of inbound locators and SPIs for the host). The 385 multihomed host waits for an ESP_INFO (new outbound SA) from the peer 386 and an ACK of its own UPDATE. As in the mobility case, the peer host 387 must perform an address verification before actively using the new 388 address. Figure 3 illustrates this scenario. 390 Multi-homed Host Peer Host 392 UPDATE(ESP_INFO, LOCATOR, SEQ, [DIFFIE_HELLMAN]) 393 -----------------------------------> 394 UPDATE(ESP_INFO, SEQ, ACK, [DIFFIE_HELLMAN,] ECHO_REQUEST) 395 <----------------------------------- 396 UPDATE(ACK, ECHO_RESPONSE) 397 -----------------------------------> 399 Figure 3: Basic Multihoming Scenario 401 In multihoming scenarios, it is important that hosts receiving 402 UPDATEs associate them correctly with the destination address used in 403 the packet carrying the UPDATE. When processing inbound LOCATORs 404 that establish new security associations on an interface with 405 multiple addresses, a host uses the destination address of the UPDATE 406 containing the LOCATOR as the local address to which the LOCATOR plus 407 ESP_INFO is targeted. This is because hosts may send UPDATEs with 408 the same (locator) IP address to different peer addresses -- this has 409 the effect of creating multiple inbound SAs implicitly affiliated 410 with different peer source addresses. 412 4.2. Site Multihoming 414 A host may have an interface that has multiple globally routable IP 415 addresses. Such a situation may be a result of the site having 416 multiple upper Internet Service Providers, or just because the site 417 provides all hosts with both IPv4 and IPv6 addresses. The host 418 should stay reachable at all or any subset of the currently available 419 global routable addresses, independent of how they are provided. 421 This case is handled the same as if there were different IP 422 addresses, described above in Section 4.1. Note that a single 423 interface may experience site multihoming while the host itself may 424 have multiple interfaces. 426 Note that a host may be multihomed and mobile simultaneously, and 427 that a multihomed host may want to protect the location of some of 428 its interfaces while revealing the real IP address of some others. 430 This document does not presently specify additional site multihoming 431 extensions to HIP; further alignment with the IETF shim6 working 432 group may be considered in the future. 434 4.3. Dual host multihoming 436 Consider the case in which both hosts would like to add an additional 437 address after the base exchange completes. In Figure 4, consider 438 that host1, which used address addr1a in the base exchange to set up 439 SPI1a and SPI2a, wants to add address addr1b. It would send an 440 UPDATE with LOCATOR (containing the address addr1b) to host2, using 441 destination address addr2a, and a new set of SPIs would be added 442 between hosts 1 and 2 (call them SPI1b and SPI2b -- not shown in the 443 figure). Next, consider host2 deciding to add addr2b to the 444 relationship. Host2 must select one of host1's addresses towards 445 which to initiate an UPDATE. It may choose to initiate an UPDATE to 446 addr1a, addr1b, or both. If it chooses to send to both, then a full 447 mesh (four SA pairs) of SAs would exist between the two hosts. This 448 is the most general case; it often may be the case that hosts 449 primarily establish new SAs only with the peer's Preferred locator. 450 The readdressing protocol is flexible enough to accommodate this 451 choice. 453 -<- SPI1a -- -- SPI2a ->- 454 host1 < > addr1a <---> addr2a < > host2 455 ->- SPI2a -- -- SPI1a -<- 457 addr1b <---> addr2a (second SA pair) 458 addr1a <---> addr2b (third SA pair) 459 addr1b <---> addr2b (fourth SA pair) 461 Figure 4: Dual Multihoming Case in Which Each Host Uses LOCATOR to 462 Add a Second Address 464 4.4. Combined Mobility and Multihoming 466 It looks likely that in the future, many mobile hosts will be 467 simultaneously mobile and multihomed, i.e., have multiple mobile 468 interfaces. Furthermore, if the interfaces use different access 469 technologies, it is fairly likely that one of the interfaces may 470 appear stable (retain its current IP address) while some other(s) may 471 experience mobility (undergo IP address change). 473 The use of LOCATOR plus ESP_INFO should be flexible enough to handle 474 most such scenarios, although more complicated scenarios have not 475 been studied so far. 477 4.5. Initiating the Protocol in R1 or I2 479 A Responder host MAY include a LOCATOR parameter in the R1 packet 480 that it sends to the Initiator. This parameter MUST be protected by 481 the R1 signature. If the R1 packet contains LOCATOR parameters with 482 a new Preferred locator, the Initiator SHOULD directly set the new 483 Preferred locator to status ACTIVE without performing address 484 verification first, and MUST send the I2 packet to the new Preferred 485 locator. The I1 destination address and the new Preferred locator 486 may be identical. All new non-preferred locators must still undergo 487 address verification once the base exchange completes. 489 Initiator Responder 491 R1 with LOCATOR 492 <----------------------------------- 493 record additional addresses 494 change responder address 495 I2 sent to newly indicated preferred address 496 -----------------------------------> 497 (process normally) 498 R2 499 <----------------------------------- 500 (process normally, later verification of non-preferred locators) 502 Figure 5: LOCATOR Inclusion in R1 504 An Initiator MAY include one or more LOCATOR parameters in the I2 505 packet, independent of whether or not there was a LOCATOR parameter 506 in the R1. These parameters MUST be protected by the I2 signature. 507 Even if the I2 packet contains LOCATOR parameters, the Responder MUST 508 still send the R2 packet to the source address of the I2. The new 509 Preferred locator SHOULD be identical to the I2 source address. If 510 the I2 packet contains LOCATOR parameters, all new locators must 511 undergo address verification as usual, and the ESP traffic that 512 subsequently follows should use the Preferred locator. 514 Initiator Responder 516 I2 with LOCATOR 517 -----------------------------------> 518 (process normally) 519 record additional addresses 520 R2 sent to source address of I2 521 <----------------------------------- 522 (process normally) 524 Figure 6: LOCATOR Inclusion in I2 526 The I1 and I2 may be arriving from different source addresses if the 527 LOCATOR parameter is present in R1. In this case, implementations 528 simultaneously using multiple pre-created R1s, indexed by Initiator 529 IP addresses, may inadvertently fail the puzzle solution of I2 530 packets due to a perceived puzzle mismatch. See, for instance, the 531 example in Appendix A of [RFC5201]. As a solution, the Responder's 532 puzzle indexing mechanism must be flexible enough to accommodate the 533 situation when R1 includes a LOCATOR parameter. 535 5. Other Considerations 537 5.1. Address Verification 539 An address verification method is specified in 540 [I-D.ietf-hip-rfc5206-bis]. It is expected that addresses learned in 541 multihoming scenarios also are subject to the same verification 542 rules. 544 5.2. Preferred Locator 546 When a host has multiple locators, the peer host must decide which to 547 use for outbound packets. It may be that a host would prefer to 548 receive data on a particular inbound interface. HIP allows a 549 particular locator to be designated as a Preferred locator and 550 communicated to the peer. 552 In general, when multiple locators are used for a session, there is 553 the question of using multiple locators for failover only or for 554 load-balancing. Due to the implications of load-balancing on the 555 transport layer that still need to be worked out, this document 556 assumes that multiple locators are used primarily for failover. An 557 implementation may use ICMP interactions, reachability checks, or 558 other means to detect the failure of a locator. 560 5.3. Interaction with Security Associations 562 This document uses the HIP LOCATOR protocol parameter, specified in 563 [I-D.ietf-hip-rfc5206-bis]), that allows the hosts to exchange 564 information about their locator(s) and any changes in their 565 locator(s). The logical structure created with LOCATOR parameters 566 has three levels: hosts, Security Associations (SAs) indexed by 567 Security Parameter Indices (SPIs), and addresses. 569 The relation between these levels for an association constructed as 570 defined in the base specification [RFC5201] and ESP transform 571 [RFC5202] is illustrated in Figure 7. 573 -<- SPI1a -- -- SPI2a ->- 574 host1 < > addr1a <---> addr2a < > host2 575 ->- SPI2a -- -- SPI1a -<- 577 Figure 7: Relation between Hosts, SPIs, and Addresses (Base 578 Specification) 580 In Figure 7, host1 and host2 negotiate two unidirectional SAs, and 581 each host selects the SPI value for its inbound SA. The addresses 582 addr1a and addr2a are the source addresses that the hosts use in the 583 base HIP exchange. These are the "preferred" (and only) addresses 584 conveyed to the peer for use on each SA. That is, although packets 585 sent to any of the hosts' interfaces may be accepted on the inbound 586 SA, the peer host in general knows of only the single destination 587 address learned in the base exchange (e.g., for host1, it sends a 588 packet on SPI2a to addr2a to reach host2), unless other mechanisms 589 exist to learn of new addresses. 591 In general, the bindings that exist in an implementation 592 corresponding to this document can be depicted as shown in Figure 8. 593 In this figure, a host can have multiple inbound SPIs (and, not 594 shown, multiple outbound SPIs) associated with another host. 595 Furthermore, each SPI may have multiple addresses associated with it. 596 These addresses that are bound to an SPI are not used to lookup the 597 incoming SA. Rather, the addresses are those that are provided to 598 the peer host, as hints for which addresses to use to reach the host 599 on that SPI. The LOCATOR parameter is used to change the set of 600 addresses that a peer associates with a particular SPI. 602 address11 603 / 604 SPI1 - address12 605 / 606 / address21 607 host -- SPI2 < 608 \ address22 609 \ 610 SPI3 - address31 611 \ 612 address32 614 Figure 8: Relation between Hosts, SPIs, and Addresses (General Case) 616 A host may establish any number of security associations (or SPIs) 617 with a peer. The main purpose of having multiple SPIs with a peer is 618 to group the addresses into collections that are likely to experience 619 fate sharing. For example, if the host needs to change its addresses 620 on SPI2, it is likely that both address21 and address22 will 621 simultaneously become obsolete. In a typical case, such SPIs may 622 correspond with physical interfaces; see below. Note, however, that 623 especially in the case of site multihoming, one of the addresses may 624 become unreachable while the other one still works. In the typical 625 case, however, this does not require the host to inform its peers 626 about the situation, since even the non-working address still 627 logically exists. 629 A basic property of HIP SAs is that the inbound IP address is not 630 used to lookup the incoming SA. Therefore, in Figure 8, it may seem 631 unnecessary for address31, for example, to be associated only with 632 SPI3 -- in practice, a packet may arrive to SPI1 via destination 633 address address31 as well. However, the use of different source and 634 destination addresses typically leads to different paths, with 635 different latencies in the network, and if packets were to arrive via 636 an arbitrary destination IP address (or path) for a given SPI, the 637 reordering due to different latencies may cause some packets to fall 638 outside of the ESP anti-replay window. For this reason, HIP provides 639 a mechanism to affiliate destination addresses with inbound SPIs, 640 when there is a concern that anti-replay windows might be violated. 641 In this sense, we can say that a given inbound SPI has an "affinity" 642 for certain inbound IP addresses, and this affinity is communicated 643 to the peer host. Each physical interface SHOULD have a separate SA, 644 unless the ESP anti-replay window is loose. 646 Moreover, even when the destination addresses used for a particular 647 SPI are held constant, the use of different source interfaces may 648 also cause packets to fall outside of the ESP anti-replay window, 649 since the path traversed is often affected by the source address or 650 interface used. A host has no way to influence the source interface 651 on which a peer sends its packets on a given SPI. A host SHOULD 652 consistently use the same source interface and address when sending 653 to a particular destination IP address and SPI. For this reason, a 654 host may find it useful to change its SPI or at least reset its ESP 655 anti-replay window when the peer host readdresses. 657 An address may appear on more than one SPI. This creates no 658 ambiguity since the receiver will ignore the IP addresses during SA 659 lookup anyway. However, this document does not specify such cases. 661 When the LOCATOR parameter is sent in an UPDATE packet, then the 662 receiver will respond with an UPDATE acknowledgment. When the 663 LOCATOR parameter is sent in an R1 or I2 packet, the base exchange 664 retransmission mechanism will confirm its successful delivery. 665 LOCATORs may experimentally be used in NOTIFY packets; in this case, 666 the recipient MUST consider the LOCATOR as informational and not 667 immediately change the current preferred address, but can test the 668 additional locators when the need arises. The use of the LOCATOR in 669 a NOTIFY message may not be compatible with middleboxes. 671 6. Processing Rules 673 Processing rules are specified in [I-D.ietf-hip-rfc5206-bis]. Future 674 versions of this document will specify multihoming-specific 675 processing rules here. 677 6.1. Sending LOCATORs 679 The decision of when to send LOCATORs is basically a local policy 680 issue. However, it is RECOMMENDED that a host send a LOCATOR 681 whenever it recognizes a change of its IP addresses in use on an 682 active HIP association, and assumes that the change is going to last 683 at least for a few seconds. Rapidly sending LOCATORs that force the 684 peer to change the preferred address SHOULD be avoided. 686 When a host decides to inform its peers about changes in its IP 687 addresses, it has to decide how to group the various addresses with 688 SPIs. The grouping should consider also whether middlebox 689 interaction requires sending the same LOCATOR in separate UPDATEs on 690 different paths. Since each SPI is associated with a different 691 Security Association, the grouping policy may also be based on ESP 692 anti-replay protection considerations. In the typical case, simply 693 basing the grouping on actual kernel level physical and logical 694 interfaces may be the best policy. Grouping policy is outside of the 695 scope of this document. 697 Note that the purpose of announcing IP addresses in a LOCATOR is to 698 provide connectivity between the communicating hosts. In most cases, 699 tunnels or virtual interfaces such as IPsec tunnel interfaces or 700 Mobile IP home addresses provide sub-optimal connectivity. 701 Furthermore, it should be possible to replace most tunnels with HIP 702 based "non-tunneling", therefore making most virtual interfaces 703 fairly unnecessary in the future. Therefore, virtual interfaces 704 SHOULD NOT be announced in general. On the other hand, there are 705 clearly situations where tunnels are used for diagnostic and/or 706 testing purposes. In such and other similar cases announcing the IP 707 addresses of virtual interfaces may be appropriate. 709 Hosts MUST NOT announce broadcast or multicast addresses in LOCATORs. 710 Link-local addresses MAY be announced to peers that are known to be 711 neighbors on the same link, such as when the IP destination address 712 of a peer is also link-local. The announcement of link-local 713 addresses in this case is a policy decision; link-local addresses 714 used as Preferred locators will create reachability problems when the 715 host moves to another link. In any case, link-local addresses MUST 716 NOT be announced to a peer unless that peer is known to be on the 717 same link. 719 Once the host has decided on the groups and assignment of addresses 720 to the SPIs, it creates a LOCATOR parameter that serves as a complete 721 representation of the addresses and affiliated SPIs intended for 722 active use. We now describe a few cases introduced in Section 4. We 723 assume that the Traffic Type for each locator is set to "0" (other 724 values for Traffic Type may be specified in documents that separate 725 the HIP control plane from data plane traffic). Other mobility and 726 multihoming cases are possible but are left for further 727 experimentation. 729 1. Host multihoming (addition of an address). We only describe the 730 simple case of adding an additional address to a (previously) 731 single-homed, non-mobile host. The host SHOULD set up a new SA 732 pair between this new address and the preferred address of the 733 peer host. To do this, the multihomed host creates a new inbound 734 SA and creates a new SPI. For the outgoing UPDATE message, it 735 inserts an ESP_INFO parameter with an OLD SPI field of "0", a NEW 736 SPI field corresponding to the new SPI, and a KEYMAT Index as 737 selected by local policy. The host adds to the UPDATE message a 738 LOCATOR with two Type "1" Locators: the original address and SPI 739 active on the association, and the new address and new SPI being 740 added (with the SPI matching the NEW SPI contained in the 741 ESP_INFO). The Preferred bit SHOULD be set depending on the 742 policy to tell the peer host which of the two locators is 743 preferred. The UPDATE also contains a SEQ parameter and 744 optionally a DIFFIE_HELLMAN parameter, and follows rekeying 745 procedures with respect to this new address. The UPDATE message 746 SHOULD be sent to the peer's Preferred address with a source 747 address corresponding to the new locator. 749 The sending of multiple LOCATORs, locators with Locator Type "0", and 750 multiple ESP_INFO parameters is for further study. Note that the 751 inclusion of LOCATOR in an R1 packet requires the use of Type "0" 752 locators since no SAs are set up at that point. 754 6.2. Handling Received LOCATORs 756 A host SHOULD be prepared to receive a LOCATOR parameter in the 757 following HIP packets: R1, I2, UPDATE, and NOTIFY. 759 This document describes sending both ESP_INFO and LOCATOR parameters 760 in an UPDATE. The ESP_INFO parameter is included when there is a 761 need to rekey or key a new SPI, and is otherwise included for the 762 possible benefit of HIP-aware middleboxes. The LOCATOR parameter 763 contains a complete map of the locators that the host wishes to make 764 or keep active for the HIP association. 766 In general, the processing of a LOCATOR depends upon the packet type 767 in which it is included. Here, we describe only the case in which 768 ESP_INFO is present and a single LOCATOR and ESP_INFO are sent in an 769 UPDATE message; other cases are for further study. The steps below 770 cover each of the cases described in Section 6.1. 772 The processing of ESP_INFO and LOCATOR parameters is intended to be 773 modular and support future generalization to the inclusion of 774 multiple ESP_INFO and/or multiple LOCATOR parameters. A host SHOULD 775 first process the ESP_INFO before the LOCATOR, since the ESP_INFO may 776 contain a new SPI value mapped to an existing SPI, while a Type "1" 777 locator will only contain a reference to the new SPI. 779 When a host receives a validated HIP UPDATE with a LOCATOR and 780 ESP_INFO parameter, it processes the ESP_INFO as follows. The 781 ESP_INFO parameter indicates whether an SA is being rekeyed, created, 782 deprecated, or just identified for the benefit of middleboxes. The 783 host examines the OLD SPI and NEW SPI values in the ESP_INFO 784 parameter: 786 1. (no rekeying) If the OLD SPI is equal to the NEW SPI and both 787 correspond to an existing SPI, the ESP_INFO is gratuitous 788 (provided for middleboxes) and no rekeying is necessary. 790 2. (rekeying) If the OLD SPI indicates an existing SPI and the NEW 791 SPI is a different non-zero value, the existing SA is being 792 rekeyed and the host follows HIP ESP rekeying procedures by 793 creating a new outbound SA with an SPI corresponding to the NEW 794 SPI, with no addresses bound to this SPI. Note that locators in 795 the LOCATOR parameter will reference this new SPI instead of the 796 old SPI. 798 3. (new SA) If the OLD SPI value is zero and the NEW SPI is a new 799 non-zero value, then a new SA is being requested by the peer. 800 This case is also treated like a rekeying event; the receiving 801 host must create a new SA and respond with an UPDATE ACK. 803 4. (deprecating the SA) If the OLD SPI indicates an existing SPI and 804 the NEW SPI is zero, the SA is being deprecated and all locators 805 uniquely bound to the SPI are put into the DEPRECATED state. 807 If none of the above cases apply, a protocol error has occurred and 808 the processing of the UPDATE is stopped. 810 Next, the locators in the LOCATOR parameter are processed. For each 811 locator listed in the LOCATOR parameter, check that the address 812 therein is a legal unicast or anycast address. That is, the address 813 MUST NOT be a broadcast or multicast address. Note that some 814 implementations MAY accept addresses that indicate the local host, 815 since it may be allowed that the host runs HIP with itself. 817 The below assumes that all locators are of Type "1" with a Traffic 818 Type of "0"; other cases are for further study. 820 For each Type "1" address listed in the LOCATOR parameter, the host 821 checks whether the address is already bound to the SPI indicated. If 822 the address is already bound, its lifetime is updated. If the status 823 of the address is DEPRECATED, the status is changed to UNVERIFIED. 824 If the address is not already bound, the address is added, and its 825 status is set to UNVERIFIED. Mark all addresses corresponding to the 826 SPI that were NOT listed in the LOCATOR parameter as DEPRECATED. 828 As a result, at the end of processing, the addresses listed in the 829 LOCATOR parameter have either a state of UNVERIFIED or ACTIVE, and 830 any old addresses on the old SA not listed in the LOCATOR parameter 831 have a state of DEPRECATED. 833 Once the host has processed the locators, if the LOCATOR parameter 834 contains a new Preferred locator, the host SHOULD initiate a change 835 of the Preferred locator. This requires that the host first verifies 836 reachability of the associated address, and only then changes the 837 Preferred locator; see Section 6.4. 839 If a host receives a locator with an unsupported Locator Type, and 840 when such a locator is also declared to be the Preferred locator for 841 the peer, the host SHOULD send a NOTIFY error with a Notify Message 842 Type of LOCATOR_TYPE_UNSUPPORTED, with the Notification Data field 843 containing the locator(s) that the receiver failed to process. 844 Otherwise, a host MAY send a NOTIFY error if a (non-preferred) 845 locator with an unsupported Locator Type is received in a LOCATOR 846 parameter. 848 6.3. Verifying Address Reachability 850 Address verification is defined in [I-D.ietf-hip-rfc5206-bis]. 852 When address verification is in progress for a new Preferred locator, 853 the host SHOULD select a different locator listed as ACTIVE, if one 854 such locator is available, to continue communications until address 855 verification completes. Alternatively, the host MAY use the new 856 Preferred locator while in UNVERIFIED status to the extent Credit- 857 Based Authorization permits. Credit-Based Authorization is explained 858 in [I-D.ietf-hip-rfc5206-bis]. Once address verification succeeds, 859 the status of the new Preferred locator changes to ACTIVE. 861 6.4. Changing the Preferred Locator 863 A host MAY want to change the Preferred outgoing locator for 864 different reasons, e.g., because traffic information or ICMP error 865 messages indicate that the currently used preferred address may have 866 become unreachable. Another reason may be due to receiving a LOCATOR 867 parameter that has the "P" bit set. 869 To change the Preferred locator, the host initiates the following 870 procedure: 872 1. If the new Preferred locator has ACTIVE status, the Preferred 873 locator is changed and the procedure succeeds. 875 2. If the new Preferred locator has UNVERIFIED status, the host 876 starts to verify its reachability. The host SHOULD use a 877 different locator listed as ACTIVE until address verification 878 completes if one such locator is available. Alternatively, the 879 host MAY use the new Preferred locator, even though in UNVERIFIED 880 status, to the extent Credit-Based Authorization permits. Once 881 address verification succeeds, the status of the new Preferred 882 locator changes to ACTIVE and its use is no longer governed by 883 Credit-Based Authorization. 885 3. If the peer host has not indicated a preference for any address, 886 then the host picks one of the peer's ACTIVE addresses randomly 887 or according to policy. This case may arise if, for example, 888 ICMP error messages that deprecate the Preferred locator arrive, 889 but the peer has not yet indicated a new Preferred locator. 891 4. If the new Preferred locator has DEPRECATED status and there is 892 at least one non-deprecated address, the host selects one of the 893 non-deprecated addresses as a new Preferred locator and 894 continues. If the selected address is UNVERIFIED, the address 895 verification procedure described above will apply. 897 7. Security Considerations 899 Security considerations are addressed in [I-D.ietf-hip-rfc5206-bis]. 901 8. IANA Considerations 903 None. 905 9. Authors and Acknowledgments 907 Pekka Nikander and Jari Arkko originated this document, and Christian 908 Vogt and Thomas Henderson (editor) later joined as co-authors. Greg 909 Perkins contributed the initial draft of the security section. Petri 910 Jokela was a co-author of the initial individual submission. 912 The authors thank Miika Komu, Mika Kousa, Jeff Ahrenholz, and Jan 913 Melen for many improvements to the document. 915 10. References 917 10.1. Normative references 919 [I-D.ietf-hip-rfc5206-bis] Nikander, P., Henderson, T., Vogt, C., 920 and J. Arkko, "End-Host Mobility and 921 Multihoming with the Host Identity 922 Protocol", draft-ietf-hip-rfc5206-bis-00 923 (work in progress), August 2010. 925 [RFC2119] Bradner, S., "Key words for use in RFCs 926 to Indicate Requirement Levels", BCP 14, 927 RFC 2119, March 1997. 929 [RFC3484] Draves, R., "Default Address Selection 930 for Internet Protocol version 6 (IPv6)", 931 RFC 3484, February 2003. 933 [RFC4303] Kent, S., "IP Encapsulating Security 934 Payload (ESP)", RFC 4303, December 2005. 936 [RFC4423] Moskowitz, R. and P. Nikander, "Host 937 Identity Protocol (HIP) Architecture", 938 RFC 4423, May 2006. 940 [RFC5201] Moskowitz, R., Nikander, P., Jokela, P., 941 and T. Henderson, "Host Identity 942 Protocol", RFC 5201, April 2008. 944 [RFC5202] Jokela, P., Moskowitz, R., and P. 945 Nikander, "Using the Encapsulating 946 Security Payload (ESP) Transport Format 947 with the Host Identity Protocol (HIP)", 948 RFC 5202, April 2008. 950 10.2. Informative references 952 [RFC5204] Laganier, J. and L. Eggert, "Host 953 Identity Protocol (HIP) Rendezvous 954 Extension", RFC 5204, April 2008. 956 [RFC5533] Nordmark, E. and M. Bagnulo, "Shim6: 957 Level 3 Multihoming Shim Protocol for 958 IPv6", RFC 5533, June 2009. 960 Appendix A. Document Revision History 962 To be removed upon publication 964 +----------+--------------------------------------------------------+ 965 | Revision | Comments | 966 +----------+--------------------------------------------------------+ 967 | draft-00 | Initial version with multihoming text imported from | 968 | | RFC5206. | 969 +----------+--------------------------------------------------------+ 971 Authors' Addresses 973 Pekka Nikander 974 Ericsson Research NomadicLab 975 JORVAS FIN-02420 976 FINLAND 978 Phone: +358 9 299 1 979 EMail: pekka.nikander@nomadiclab.com 980 Thomas R. Henderson (editor) 981 The Boeing Company 982 P.O. Box 3707 983 Seattle, WA 984 USA 986 EMail: thomas.r.henderson@boeing.com 988 Christian Vogt 989 Ericsson Research NomadicLab 990 Hirsalantie 11 991 JORVAS FIN-02420 992 FINLAND 994 Phone: 995 EMail: christian.vogt@ericsson.com 997 Jari Arkko 998 Ericsson Research NomadicLab 999 JORVAS FIN-02420 1000 FINLAND 1002 Phone: +358 40 5079256 1003 EMail: jari.arkko@ericsson.com