idnits 2.17.1 draft-ietf-shim6-applicability-06.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, and may have content which was first submitted before 10 November 2008. The disclaimer is necessary when there are original authors that you have been unable to contact, or if some do not wish to grant the BCP78 rights to the IETF Trust. If you are able to get all authors (current and original) to grant those rights, 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 (September 2, 2010) is 4982 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- -- Looks like a reference, but probably isn't: '1' on line 174 -- Looks like a reference, but probably isn't: '2' on line 175 ** Obsolete normative reference: RFC 3315 (Obsoleted by RFC 8415) ** Obsolete normative reference: RFC 3484 (Obsoleted by RFC 6724) ** Obsolete normative reference: RFC 3775 (Obsoleted by RFC 6275) ** Obsolete normative reference: RFC 4423 (Obsoleted by RFC 9063) ** Obsolete normative reference: RFC 4960 (Obsoleted by RFC 9260) == Outdated reference: A later version (-09) exists of draft-ietf-csi-dhcpv6-cga-ps-03 == Outdated reference: A later version (-17) exists of draft-ietf-shim6-multihome-shim-api-14 Summary: 5 errors (**), 0 flaws (~~), 3 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group J. Abley 3 Internet-Draft Afilias Canada 4 Intended status: Informational M. Bagnulo 5 Expires: March 6, 2011 A. Garcia-Martinez 6 UC3M 7 September 2, 2010 9 Applicability Statement for the Level 3 Multihoming Shim Protocol 10 (Shim6) 11 draft-ietf-shim6-applicability-06 13 Abstract 15 This document discusses the applicability of the Shim6 IPv6 protocol 16 and associated support protocols and mechanisms to provide site 17 multihoming capabilities in IPv6. 19 Status of this Memo 21 This Internet-Draft is submitted in full conformance with the 22 provisions of BCP 78 and BCP 79. 24 Internet-Drafts are working documents of the Internet Engineering 25 Task Force (IETF). Note that other groups may also distribute 26 working documents as Internet-Drafts. The list of current Internet- 27 Drafts is at http://datatracker.ietf.org/drafts/current/. 29 Internet-Drafts are draft documents valid for a maximum of six months 30 and may be updated, replaced, or obsoleted by other documents at any 31 time. It is inappropriate to use Internet-Drafts as reference 32 material or to cite them other than as "work in progress." 34 This Internet-Draft will expire on March 6, 2011. 36 Copyright Notice 38 Copyright (c) 2010 IETF Trust and the persons identified as the 39 document authors. All rights reserved. 41 This document is subject to BCP 78 and the IETF Trust's Legal 42 Provisions Relating to IETF Documents 43 (http://trustee.ietf.org/license-info) in effect on the date of 44 publication of this document. Please review these documents 45 carefully, as they describe your rights and restrictions with respect 46 to this document. Code Components extracted from this document must 47 include Simplified BSD License text as described in Section 4.e of 48 the Trust Legal Provisions and are provided without warranty as 49 described in the Simplified BSD License. 51 This document may contain material from IETF Documents or IETF 52 Contributions published or made publicly available before November 53 10, 2008. The person(s) controlling the copyright in some of this 54 material may not have granted the IETF Trust the right to allow 55 modifications of such material outside the IETF Standards Process. 56 Without obtaining an adequate license from the person(s) controlling 57 the copyright in such materials, this document may not be modified 58 outside the IETF Standards Process, and derivative works of it may 59 not be created outside the IETF Standards Process, except to format 60 it for publication as an RFC or to translate it into languages other 61 than English. 63 Table of Contents 65 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 66 2. Deployment Scenarios . . . . . . . . . . . . . . . . . . . . . 4 67 3. Address Configuration . . . . . . . . . . . . . . . . . . . . 6 68 3.1. Protocol Version (IPv4 vs. IPv6) . . . . . . . . . . . . . 6 69 3.2. Prefix Lengths . . . . . . . . . . . . . . . . . . . . . . 7 70 3.3. Address Generation . . . . . . . . . . . . . . . . . . . . 7 71 3.4. Use of CGA vs. HBA . . . . . . . . . . . . . . . . . . . . 7 72 4. Shim6 and Ingress Filtering . . . . . . . . . . . . . . . . . 8 73 5. Shim6 Capabilities . . . . . . . . . . . . . . . . . . . . . . 10 74 5.1. Fault Tolerance . . . . . . . . . . . . . . . . . . . . . 10 75 5.1.1. Establishing Communications After an Outage . . . . . 10 76 5.1.2. Short-Lived Communications . . . . . . . . . . . . . . 10 77 5.1.3. Long-Lived Communications . . . . . . . . . . . . . . 11 78 5.2. Load Balancing . . . . . . . . . . . . . . . . . . . . . . 11 79 5.3. Traffic Engineering . . . . . . . . . . . . . . . . . . . 11 80 6. Application Considerations . . . . . . . . . . . . . . . . . . 12 81 7. Interaction with Other Protocols . . . . . . . . . . . . . . . 13 82 7.1. Shim6 and Mobile IPv6 . . . . . . . . . . . . . . . . . . 13 83 7.1.1. Multihomed Home Network . . . . . . . . . . . . . . . 13 84 7.1.2. Shim6 Between the HA and the MN . . . . . . . . . . . 16 85 7.2. Shim6 and SeND . . . . . . . . . . . . . . . . . . . . . . 16 86 7.3. Shim6 and SCTP . . . . . . . . . . . . . . . . . . . . . . 17 87 7.4. Shim6 and NEMO . . . . . . . . . . . . . . . . . . . . . . 17 88 7.5. Shim6 and HIP . . . . . . . . . . . . . . . . . . . . . . 18 89 8. Security Considerations . . . . . . . . . . . . . . . . . . . 19 90 8.1. Privacy Considerations . . . . . . . . . . . . . . . . . . 20 91 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 20 92 10. Contributors . . . . . . . . . . . . . . . . . . . . . . . . . 20 93 11. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 20 94 12. References . . . . . . . . . . . . . . . . . . . . . . . . . . 21 95 12.1. Normative References . . . . . . . . . . . . . . . . . . . 21 96 12.2. Informative References . . . . . . . . . . . . . . . . . . 22 97 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 23 99 1. Introduction 101 Site multihoming is an arrangement by which a site may use multiple 102 paths to the rest of the Internet to provide better reliability for 103 traffic passing in and out of the site than would be possible with a 104 single path. Some of the motivations for operators to multi-home 105 their network are described in [RFC3582]. 107 In IPv4, site multihoming is achieved by injecting into the global 108 Internet routing system (sometimes referred to as the Default-Free 109 Zone, or DFZ) the additional state required to allow session 110 resilience over re-homing events [RFC4116]. There is concern that 111 this approach will not scale [RFC3221], [RFC4984]. 113 In IPv6, site multihoming in the style of IPv4 is not generally 114 available to end sites due to a strict policy of route aggregation in 115 the DFZ. Site multihoming for sites without provider-independent 116 (PI) addresses is achieved by assigning multiple addresses to each 117 host, one or more from each provider. This multihoming approach 118 provides no transport-layer stability across re-homing events. 120 Shim6 provides layer-3 support for making re-homing events 121 transparent to the transport layer by means of a shim approach. 122 State information relating to the multihoming of two endpoints 123 exchanging unicast traffic is retained on the endpoints themselves, 124 rather than in the network. Communications between Shim6-capable 125 hosts and Shim6-incapable hosts proceed as normal, but without the 126 benefit of transport-layer stability. The Shim6 approach is thought 127 to have better scaling properties with respect to the state held in 128 the DFZ than the IPv4 approach. 130 This note describes the applicability of the Level 3 multihoming 131 (hereafter Shim6) protocol defined in [RFC5533] and the failure 132 detection mechanisms defined in [RFC5534]. 134 The terminology used in this document, including terms like locator, 135 and ULID, is defined in [RFC5533]. 137 2. Deployment Scenarios 139 The goal of the Shim6 protocol is to support locator agility in 140 established communications: different layer-3 endpoint addresses may 141 be used to exchange packets belonging to the same transport-layer 142 session, all the time presenting a consistent identifier pair to 143 upper-layer protocols. 145 In order to be useful, the Shim6 protocol requires that at least one 146 of the peers has more than one address which could be used on the 147 wire (as locators). In the event of communications failure between 148 an active pair of addresses, the Shim6 protocol will attempt to 149 reestablish communication by trying different combinations of 150 locators. 152 While other multi-addressing scenarios are not precluded, the 153 scenario in which the Shim6 protocol is expected to operate is that 154 of a multihomed site which is connected to multiple transit 155 providers, and which receives an IPv6 prefix from each of them. This 156 configuration is intended to provide protection for the end-site in 157 the event of a failure in some subset of the available transit 158 providers, without requiring the end-site to acquire PI address space 159 or requiring any particular cooperation between the transit 160 providers. 162 ,------------------------------------. ,----------------. 163 | Rest of the Internet +-------+ Remote Host R | 164 `--+-----------+------------------+--' `----------------' 165 | | | LR[1] ... LR[m] 166 ,---+----. ,---+----. ,----+---. 167 | ISP[1] | | ISP[2] | ...... | ISP[n] | 168 `---+----' `---+----' `----+---' 169 | | | 170 ,---+-----------+------------------+---. 171 | Multi-Homed Site S assigned | 172 | prefixes P[1], P[2], ..., P[n] | 173 | | 174 | ,--------. L[1] = P[1]:iid[1], | 175 | | Host H | L[2] = P[2]:iid[2], ... | 176 | `--------' L[n] = P[n]:iid[n] | 177 `--------------------------------------' 179 Figure 1 181 In the scenario illustrated in Figure 1 host H communicates with some 182 remote host R. Each of the addresses L[i] configured on host H in the 183 multihomed site S can be reached through provider ISP[i] only, since 184 ISP[i] is solely responsible for advertising a covering prefix for 185 P[i] to the rest of the Internet. 187 The use of locator L[i] on H hence causes inbound traffic towards H 188 to be routed through ISP[i]. Changing the locator from L[i] to L[j] 189 will have the effect of re-routing inbound traffic to H from ISP[i] 190 to ISP[j]. This is the central mechanism by which the Shim6 protocol 191 aims to provide multihoming functionality: by changing locators, host 192 H can change the upstream ISP used to route inbound packets towards 193 itself. Regarding to the outbound traffic to H, the path taken in 194 this case depends on both the actual locator LR[j] chosen by R, and 195 the administrative exit selection policy of site S. 197 The Shim6 protocol has other potential applications beyond site 198 multihoming. For example, since Shim6 is a host-based protocol, it 199 can also be used to support host multihoming. In this case, a 200 failure in communication between a multihomed host and some other 201 remote host might be repaired by selecting a locator associated with 202 a different interface. 204 3. Address Configuration 206 3.1. Protocol Version (IPv4 vs. IPv6) 208 The Shim6 protocol is defined only for IPv6. However, there is no 209 fundamental reason why a Shim6-like approach could not support IPv4 210 addresses as locators, either to provide multihoming support to IPv4- 211 numbered sites, or as part of an IPv4/IPv6 transition strategy. Some 212 extensions to the Shim6 protocol for supporting IPv4 locators have 213 been proposed in [I-D.nordmark-shim6-esd]. 215 The Shim6 protocol, as specified for IPv6, incorporates cryptographic 216 elements in the construction of locators (see [RFC3972], [RFC5535]). 217 Since IPv4 addresses are insufficiently large to contain addresses 218 constructed in this fashion, direct implementation of Shim6 as 219 specified for IPv6 for use with IPv4 addresses might require protocol 220 modifications. 222 In addition, there are other factors to take into account when 223 considering the support of IPv4 addresses, in particular IPv4 224 locators. Using multiple IPv4 addresses in a single host in order to 225 support Shim6 style of multihoming would result in an increased IPv4 226 address consumption, which with the current rate of IPv4 addresses 227 would be problematic. Besides, Shim6 may suffer additional problems 228 if locators become translated on the wire. Address translation is 229 more likely to involve IPv4 addresses. IPv4 addressed can be 230 translated to other IPv4 addresses (for example, private IPv4 address 231 into public IPv4 address and vice versa) or to/from IPv6 addresses 232 (for example, as defined by NAT64 233 [I-D.ietf-behave-v6v4-xlate-stateful]). When address translation 234 occurs, a locator exchanged by Shim6 could be different to the 235 address needed to reach the corresponding host, either because the 236 translated version of the locator exchanged by Shim6 is not known or 237 because the translation state does not exist any more in the 238 translator device. Supporting these scenarios would require NAT 239 traversal mechanisms which are not defined yet and which would imply 240 additional complexity (as any other NAT traversal mechanism). 242 3.2. Prefix Lengths 244 The Shim6 protocol does not assume that all the prefixes assigned to 245 the multihomed site have the same prefix length. 247 However, the use of CGA [RFC3972] and HBA [RFC5535] involve encoding 248 information in the lower 64 bits of the locators. This imposes the 249 requirement on address assignment to Shim6-capable hosts that all 250 interface addresses should be able to accommodate 64-bit interface 251 identifiers. It should be noted that this is imposed by RFC4291 252 [RFC4291] 254 3.3. Address Generation 256 The security of the Shim6 protocol is based on the use of CGA and HBA 257 addresses. 259 CGA and HBA generation process can use the information provided by 260 the stateless auto-configuration mechanism defined in [RFC4862] with 261 the additional considerations presented in [RFC3972] and [RFC5535]. 263 Stateful address auto-configuration using DHCP [RFC3315] is not 264 currently supported, because there is no defined mechanism to convey 265 the CGA Parameter Data Structure and other relevant information from 266 the DHCP server to the host. The definition of such mechanism seems 267 to be quite straightforward in the case of the HBA, since only the 268 CGA Parameter Data Structure needs to be delivered from the DHCP 269 server to the Shim6 host, and this data structure does not contain 270 any secret information. In the case of CGAs, the difficulty is 271 increased, since private key information should be exchanged as well 272 as the CGA Parameter Data Structure. However, with appropriate 273 extensions a DHCP server could inform to a host about the SEC value 274 to use when generating an address, or DHCP could even be used by the 275 host to delegate to the server the CPU-intensive task of computing a 276 Modifier for a given combination 277 [I-D.ietf-csi-dhcpv6-cga-ps]. 279 3.4. Use of CGA vs. HBA 281 The choice between CGA and HBA is a trade-off between flexibility and 282 performance. 284 The use of HBA is more efficient in the sense that addresses require 285 less computation than CGA, involving only hash operations for both 286 the generation and the verification of locator sets. However, the 287 locators of an HBA set are determined during the generation process, 288 and cannot be subsequently changed; the addition of new locators to 289 that initial set is not supported, except by re-generation of the 290 entire set which will in turn cause all addresses to change. 292 The use of CGA is more computationally expensive, involving public 293 key cryptography in the verification of locator sets. However, CGAs 294 are more flexible in the sense that they support the dynamic 295 modification of locator sets. 297 Therefore, CGAs are well suited to support dynamic environments such 298 as mobile hosts, where the locator set must be changed frequently. 299 HBAs are better suited for sites where the prefix set remains 300 relatively stable. 302 It should be noted that, since HBAs are defined as a CGA extension, 303 it is possible to generate hybrid HBA/CGA structures that incorporate 304 the strengths of both: i.e. that a single address can be used as an 305 HBA, enabling computationally-cheap validation amongst a fixed set of 306 addresses, and also as a CGA, enabling dynamic manipulation of the 307 locator set. For additional details, see [RFC5535]. 309 4. Shim6 and Ingress Filtering 311 Ingress filtering [RFC2827] prevents address spoofing by dropping 312 packets which come from customer networks with source addresses not 313 belonging to the prefix assigned to them. The problem of deploying 314 ingress filters with multihomed customers is discussed in [RFC3704], 315 in particular considering the case in which non-PI addresses are used 316 by customer networks. This is the case for IPv6 hosts in multihomed 317 networks with PA, and also for a Shim6 host in a multihomed network. 318 Note that this is also the case for other solutions supporting 319 multihoming, such as SCTP [RFC4960], HIP [RFC4423], etc. 321 One solution to this problem is to make the providers aware of the 322 alternative prefixes that can be used by a multihomed site, so that 323 ingress filtering would not be applied to packets with source 324 addresses belonging to these prefixes. This may be possible in some 325 cases, but it cannot be assumed as the general case. 327 [RFC3704] proposes that non-PI addresses should ensure that each 328 packet is delivered to the provider related with the prefix of its 329 source address. To deliver packets to the appropriate outgoing ISP, 330 some routers of the site must consider source addresses in their 331 forwarding decisions, in addition to the usual destination-based 332 forwarding. These routers maintain as many parallel routing tables 333 as valid source prefixes are, and choose a route that is a function 334 of both the source and the destination address. The way these 335 routing tables are populated is out of the scope of this document. 337 As proposed in [I-D.huitema-multi6-hosts], it is required for site 338 exit routers (at least) to be part of a single connected source based 339 routing domain: 341 Multiple site exits 342 | | | | 343 -----+-----+-----+-----+----- 344 ( ) 345 ( Source based routing domain ) 346 ( ) 347 ----+----+----+----+----+---- 348 ( ) 349 ( Generic routing domain ) 350 ( ) 351 ----------------------------- 353 Figure 2 355 In this way, packets arriving to this connected source based routing 356 domain would be delivered to the appropriate exit router. 358 Some particular cases of this generic deployment scenario are: 360 - a single exit router, in which the router chooses the exit provider 361 according to the source address of the packet to be forwarded 363 - a site in which all routers perform source address based forwarding 365 - a site in which only site-exit routers perform source address based 366 forwarding, and these site-exit routers are connected through point- 367 to-point tunnels, so that packets can be tunneled to the appropriate 368 exit router according to its source address 370 For hosts attached directly to networks of different providers, a 371 host solution to ensure that packets are forwarded to the appropriate 372 interface according to its source address must be provided. This 373 problem is under discussion in the Multiple Interfaces (MIF) IETF 374 Working Group. 376 Shim6 has no means to enforce neither host nor network forwarding for 377 a given locator to be used as source address. If any notification is 378 received from the router dropping the packets with legitimate source 379 addresses as a result of ingress filtering, the affected locator 380 could be associated to a low preference (or not being used at all). 381 But even if such notification is not received, or not processed by 382 the Shim6 layer, defective ingress filtering configuration will be 383 treated as a communication failure, and Shim6 re-homing would finally 384 select a working path in which packets are not filtered, if this path 385 exists. Note that this behavior results from the powerful end-to-end 386 resilience properties exhibited by REAP. 388 5. Shim6 Capabilities 390 5.1. Fault Tolerance 392 5.1.1. Establishing Communications After an Outage 394 If a host within a multihomed site attempts to establish a 395 communication with a remote host and selects a locator which 396 corresponds to a failed transit path, bidirectional communication 397 between the two hosts will not succeed. In order to establish a new 398 communication, the initiating host must try different combinations of 399 (source, destination) locator pairs until it finds a pair that works. 400 The mechanism for this default address selection is described in 401 [RFC3484]. A commentary on this mechanism in the context of 402 multihomed environments can be found in 403 [I-D.bagnulo-ipv6-rfc3484-update]. 405 Since a Shim6 context is normally only established between two hosts 406 after initial communication has been set up, there is no opportunity 407 for Shim6 to participate in the discovery of a suitable, initial 408 (source, destination) locator pair. The same consideration holds for 409 referrals, as it is described in Section 6. 411 5.1.2. Short-Lived Communications 413 The Shim6 context establishment operation requires a 4-way packet 414 exchange, and involves some overhead on the participating hosts in 415 memory and CPU. 417 For short-lived communications between two hosts, the benefit of 418 establishing a Shim6 context might not exceed the cost, perhaps 419 because the protocols concerned are fault tolerant and can arrange 420 their own recovery (e.g. DNS) or because the frequency of re-homing 421 events is sufficiently low that the probability of such a failure 422 occurring during a short-lived exchange is not considered 423 significant. 425 It is anticipated that the exchange of Shim6 context will provide 426 most benefit for exchanges between hosts which are long-lived. For 427 this reason the default behaviour of Shim6-capable hosts is expected 428 to employ deferred context-establishment. This default behaviour 429 will be able to be overridden by applications which prefer immediate 430 context establishment regardless of transaction longevity. 432 It must be noted that all the above considerations refer to the 433 lifetime of the interaction between the peers and not about the 434 lifetime of a particular connection (e.g. TCP connection). In other 435 words, the Shim6 context is established between ULID pairs and it 436 affects all the communication between these ULIDs. So, two nodes 437 with multiple short-lived communications using the same ULID pair 438 would benefit as much from the Shim6 features as two nodes having a 439 single long-lived communication. One example of such scenario would 440 be a web client software downloading web contents from a server over 441 multiple TCP connections. Each TCP connection is short-lived, but 442 the communication/contact between the two ULID could be long-lived. 444 5.1.3. Long-Lived Communications 446 As discussed in Section 5.1.2, hosts engaged in long-lived 447 communications will suffer lower proportional overhead, and greater 448 probability of benefit than those performing brief transactions. 450 Deferred context setup ensures that session establishment time will 451 not be increased by the use of Shim6. 453 5.2. Load Balancing 455 The Shim6 protocol does not support load balancing within a single 456 context: all packets associated with a particular context are 457 exchanged using a single locator pair per direction, with the 458 exception of forked contexts, which are created upon explicit 459 requests from the upper-layer protocol. 461 It may be possible to extend the Shim6 protocol to use multiple 462 locator pairs in a single context, but the impact of such an 463 extension on upper-layer protocols (e.g. on TCP congestion control) 464 should be considered carefully. 466 When many contexts are considered together in aggregation, e.g. on a 467 single host which participates in many simultaneous contexts or in a 468 site full of hosts, some degree of load sharing should occur 469 naturally due to the selection of different locator pairs in each 470 context. However, there is no mechanism defined to ensure that this 471 natural load sharing is arranged to provide a statistical balance 472 between transit providers. 474 5.3. Traffic Engineering 476 The Shim6 protocol provides some lightweight traffic engineering 477 capabilities in the form of the Locator Preferences option, which 478 allows a host to inform a remote host of local preferences for 479 locator selection. 481 This mechanism is only available after a Shim6 context has been 482 established, and it is a host-based capability rather than a site- 483 based capability. There is no defined mechanism which would allow 484 use of the Locator Preferences option amongst a site full of hosts to 485 be managed centrally. 487 6. Application Considerations 489 Shim6 provides multihoming support without forcing changes in the 490 applications running on the host. The fact that an address has been 491 generated according to the CGA or HBA specification does not require 492 any specific action from the application, e.g. it can obtain remote 493 CGA or HBA addresses as a result of a getaddrinfo() call to trigger a 494 DNS Request. The storage of CGA or HBA addresses in DNS does not 495 require also any modification of this protocol, since they are 496 recorded using AAAA records. Moreover, neither the ULID/locator 497 management [RFC5533] nor the failure detection and recovery [RFC5534] 498 functions require application awareness. 500 However, a specific API [I-D.ietf-shim6-multihome-shim-api] is being 501 developed for those applications which might require additional 502 capabilities in ULID/locator management, such as the locator pair in 503 use for a given context, or the set of local or remote locators 504 available for it. This API can also be used to disable Shim6 505 operation when required. 507 It is worth to note that callbacks can benefit naturally from Shim6 508 support. In a callback, an application in B retrieves IP_A, the IP 509 address of a peer A, and B uses IP_A to establish a new communication 510 with A. As long as the address exchanged, IP_A is the ULID for the 511 initial communication between A and B, and B uses the same address as 512 in the initial communication, and this initial communication is alive 513 (or the context has not been deleted), the new communication could 514 use the locators exchanged by Shim6 for the first communication. In 515 this case, communication could proceed even if the ULID of A is not 516 reachable. 518 However, Shim6 does not provide specific protection to current 519 applications when they use referrals. A referral is the exchange of 520 the IP address IP_A of a party A by party B to party C, so that party 521 C could use IP_A to communicate with party A 522 [I-D.ietf-multi6-app-refer]. In a normal case, the ULID IP_A would 523 be the only information sent by B to C as referral. But if IP_A is 524 no longer valid as locator in A, C could have trouble in establishing 525 a communication with A. Increased failure protection for referrals 526 could be obtained if B exchanged the whole list of alternative 527 locators of A, although in this case the application protocol should 528 be modified. Note that B could send to C the current locator of A, 529 instead of the ULID of A, as a way of using the most recent 530 reachability information about A. While in this case no modification 531 of the application protocol is required, some concerns arise: host A 532 may not accept one of its locator as ULID for initiating a 533 communication, and if CGA are used, the locator may not be a CGA so a 534 Shim6 context among A and C could not be created. 536 7. Interaction with Other Protocols 538 7.1. Shim6 and Mobile IPv6 540 We next consider some scenarios in which the Shim6 protocol and the 541 MIPv6 protocol [RFC3775] might be used simultaneously. 543 7.1.1. Multihomed Home Network 545 In this case, the Home Network of the Mobile Node (MN) is multihomed. 546 This implies the availability of multiple Home Network prefixes, 547 resulting on multiple HoAs for each MN. Since the MN is a node 548 within a multihomed site, it seems reasonable to expect that the MN 549 should be able to benefit from the multihoming capabilities provided 550 by the Shim6 protocol. Moreover, the MN needs to be able to obtain 551 the multihoming benefits even when it is roaming away from the Home 552 Network: if the MN is away from the Home Network while the Home 553 Network suffers a failure in a transit path, the MN should be able to 554 continue communicating using alternate paths to reach the Home 555 Network. 557 The resulting scenario is the following: 559 +------------------------------------+ 560 | Internet | 561 +------------------------------------+ 562 | | 563 +----+ +----+ 564 |ISP1| |ISP2| 565 +----+ +----+ 566 | | 567 +------------------------------------+ 568 | Multihomed Home Network | 569 | Prefixes: P1 and P2 | 570 | | 571 | Home Agent | 572 | // | 573 +------------------//----------------+ 574 // 575 // 576 +-----+ 577 | MN | HoA1, HoA2 578 +-----+ 580 Figure 3 582 So, in this configuration, the Shim6 protocol is used to provide 583 multiple communication paths to all the nodes within the multihomed 584 sites (including the mobile nodes) and the MIPv6 protocol is used to 585 support mobility of the mobile nodes of the multihomed site. 587 The proposed protocol architecture would be the following: 589 +--------------+ 590 | Application | 591 +--------------+ 592 | Transport | 593 +--------------+ 594 | IP | 595 | +----------+ | 596 | | IPSec | | 597 | +----------+<--ULIDs 598 | | Shim6 | | 599 | +----------+<--HoAs 600 | | MIPv6 | | 601 | +----------+<--CoAs 602 | | 603 +--------------+ 605 Figure 4 607 In this architecture, the upper layer protocols and IPSec would use 608 ULIDs of the Shim6 protocol. Only the HoAs will be presented by the 609 upper layers to the Shim6 layer as potential ULIDs. Two Shim6 610 entities will exchange their own available HoAs as locators. 611 Therefore, Shim6 provides failover between different HoAs and allows 612 preserving established communications when an outage affects the path 613 through the ISP that has delegated the HoA used for initiating the 614 communication (similarly to the case of a host within a multihomed 615 site). The CoAs are not presented to the Shim6 layer and are not 616 included in the local locator set in this case. The CoAs are managed 617 by the MIPv6 layer, which binds each HoA to a CoA. 619 So, in this case, the upper layer protocols select a ULID pair for 620 the communication. The Shim6 protocol translates the ULID pair to an 621 alternative locator in case that is needed. Both the ULIDs and the 622 alternative locators are HoAs. Next, the MIPv6 layer maps the 623 selected HoA to the corresponding CoA, which is the actual address 624 included in the wire. 626 The Shim6 context is established between the MN and the CN, and it 627 would allow the communication to use all the available HoAs to 628 provide fault tolerance. The MIPv6 protocol is used between the MN 629 and the HA in the case of the bidirectional tunnel mode, and between 630 the MN and the CN in case of the RO (Route Optimization) mode. 632 7.1.2. Shim6 Between the HA and the MN 634 Another scenario where a Shim6-MIPv6 interaction may be useful is the 635 case where a Shim6 context is established between the MN and the HA 636 in order to provide fault tolerance capabilities to the bidirectional 637 tunnel between them. 639 Consider the case where the HA has multiple addresses (whether 640 because the Home Network is multihomed or because the HA has multiple 641 interfaces) and/or the MN has multiple addresses (whether because the 642 visited network is multihomed or because the MN has multiple 643 interfaces). In this case, if a failure affects the address pair 644 that is being used to run the tunnel between the MN and HA, 645 additional mechanisms need to be used to preserve the communication. 647 One possibility would be to use MIPv6 capabilities, by simply 648 changing the CoA used as the tunnel endpoint. However, MIPv6 lacks 649 of failure detection mechanisms that would allow the MN and/or the HA 650 to detect the failure and trigger the usage of an alternative 651 address. Shim6 provides such failure detection protocol, so one 652 possibility would be re-using the failure detection function from the 653 Shim6 failure detection protocol in MIPv6. In this case, the Shim6 654 protocol wouldn't be used to create Shim6 context and provide fault 655 tolerance, but just its failure detection functionality would be re- 656 used. 658 The other possibility would be to use the Shim6 protocol to create a 659 Shim6 context between the HA and the MN so that the Shim6 detects any 660 failure and re-homes the communication in a transparent fashion to 661 MIPv6. In this case, the Shim6 protocol would be associated to the 662 tunnel interface. 664 7.2. Shim6 and SeND 666 Secure Neighbor Discovery (SeND) [RFC3971] uses CGAs to prove address 667 ownership for Neighbor Discovery [RFC4861]. The Shim6 protocol can 668 use either CGAs or HBAs to protect locator sets included in Shim6 669 contexts. It is expected that some hosts will need to participate in 670 both SeND and Shim6 simultaneously. 672 In the case that both the SeND and Shim6 protocols are using the CGA 673 technique to generate addresses, then there is no conflict: the host 674 will generate addresses for both purposes as CGAs, and since it will 675 be in control of the associated private key, the same CGA can be used 676 for the different protocols. 678 In the case that a Shim6-capable host is using HBAs to protect its 679 locator sets, the host will need to generate hybrid HBA/CGA addresses 680 as defined in [RFC5535] and discussed briefly in Section 3.4. In 681 this case, the CGA Parameter Data Structure containing a valid public 682 key and the Multi-Prefix extension are included as inputs to the hash 683 function. 685 7.3. Shim6 and SCTP 687 The SCTP [RFC4960] protocol provides a reliable, stream-based 688 communications channel between two hosts which provides a superset of 689 the capabilities of TCP. One of the notable features of SCTP is that 690 it allows the exchange of endpoint addresses between hosts, and is 691 able to recover from the failure of a particular endpoint pair in a 692 manner which is conceptually similar to locator selection in Shim6. 694 SCTP is a transport-layer protocol, higher in the protocol stack than 695 Shim6, and hence there is no fundamental incompatibility which would 696 prevent a Shim6-capable host from communicating using SCTP. 698 However, since SCTP and Shim6 both aim to exchange addressing 699 information between hosts in order to meet the same generic goal, it 700 is possible that their simultaneous use might result in unexpected 701 behaviour, e.g. lead to race conditions. 703 The capabilities of SCTP with respect to path maintenance of a 704 reliable, connection-oriented stream protocol are more extensive than 705 the more general layer-3 locator agility provided by Shim6. 706 Therefore, It is recommended that Shim6 is not used for SCTP 707 sessions, and that path maintenance is provided solely by SCTP. 708 There are at least two ways to enforce this behaviour. One option 709 would be to make the stack, and in particular the Shim6 sublayer, 710 aware of SCTP sockets and in this case refrain from creating a Shim6 711 context. The other option is that the upper layer, SCTP in this 712 case, informs using a Shim6 capable API like the one proposed in 713 [I-D.ietf-shim6-multihome-shim-api] that no Shim6 context must be 714 created for this particular communication. 716 Note that the issues described here for SCTP may also arise for a 717 multipath TCP solution. 719 7.4. Shim6 and NEMO 721 The NEMO [RFC3963] protocol extensions to MIPv6 allow a Mobile 722 Network to communicate through a bidirectional tunnel via a Mobile 723 Router (MR) to a NEMO-compliant Home Agent (HA) located in a Home 724 Network. 726 If either or both of the MR or HA are multihomed, then a Shim6 727 context established preserves the integrity of the bidirectional 728 tunnel between them in the event that a transit failure occurs in the 729 connecting path. 731 Once the tunnel between MR and HA is established, hosts within the 732 Mobile Network which are Shim6-capable can establish contexts with 733 remote hosts in order to receive the same multihoming benefits as any 734 host located within the Home Network. 736 7.5. Shim6 and HIP 738 Shim6 and the Host Identity Protocol ( HIP [RFC4423]) are 739 architecturally similar in the sense that both solutions allow two 740 hosts to use different locators to support communications between 741 stable ULIDs. The signaling exchange to establish the demultiplexing 742 context on the hosts is very similar for both protocols. However, 743 there are a few key differences. First, Shim6 avoids defining a new 744 namespace for ULIDs, preferring instead to use a routable locator as 745 a ULID, while HIP uses public keys and hashes thereof as ULIDs. The 746 use of a routable locator as ULID better supports deferred context 747 establishment, application callbacks, and application referrals, and 748 avoids management and resolution costs of a new namespace, but 749 requires additional security mechanisms to securely bind the ULID 750 with the locators. Second, Shim6 uses an explicit context header on 751 data packets for which the ULIDs differ from the locators in use 752 (this header is only needed after a failure/rehoming event occurs), 753 while HIP compresses this context tag into the ESP SPI field of a 754 BEET-mode security association BEET [I-D.nikander-esp-beet-mode]. 755 Third, HIP as presently defined requires the use of public-key 756 operations in its signaling exchange and ESP encryption in the data 757 plane, while the use of Shim6 requires neither (if only HBA addresses 758 are used). HIP by default provides data protection, while this is a 759 non goal for Shim6. 761 The Shim6 working group was chartered to provide a solution to a 762 specific problem, multihoming, which minimizes deployment disruption, 763 while HIP is considered more of an experimental approach intended to 764 solve several more general problems (mobility, multihoming and loss 765 of end-to-end addressing transparency) through an explicit 766 identifier/locator split. Communicating hosts that are willing and 767 interested to run HIP (perhaps extended with Shim6's failure 768 detection protocol) likely have no reason to also run Shim6. In this 769 sense, HIP may be viewed as a possible long-term evolution or 770 extension of the Shim6 architecture, or one possible implementation 771 of the extended Shim6 design ESD [I-D.nordmark-shim6-esd]. 773 8. Security Considerations 775 This section considers the applicability of the Shim6 protocol from a 776 security perspective, i.e. which security features can expect 777 applications and users of the Shim6 protocol. 779 First of all, it should be noted that the Shim6 protocol is not a 780 security protocol, like for instance HIP. This means that as opposed 781 to HIP, it is an explicit non goal of the Shim6 protocol to provide 782 enhanced security for the communications that use the Shim6 protocol. 783 The goal of the Shim6 protocol design in terms of security is not to 784 introduce new vulnerabilities that were not present in the current 785 non-Shim6 enabled communications. In particular, it is an explicit 786 non goal of the Shim6 protocol security to provide protection from 787 on-path attackers. On-path attackers are able to sniff and spoof 788 packets in the current Internet, and they are able to do the same in 789 Shim6 communications (as long as the communication flows through the 790 path they are located on). So, summarizing, the Shim6 protocol does 791 not provide data packet protection from on-path attackers. 793 However, the Shim6 protocol does use several security techniques. 794 The goal of these security measures is to protect the Shim6 signaling 795 protocol from new attacks resulting from the adoption of the Shim6 796 protocol. In particular, the use of HBA/CGA prevents on-path and 797 off-path attackers to introduce new locators in the locator set of a 798 Shim6 context, preventing redirection attacks [RFC4218]. Moreover, 799 the usage of probes before re-homing to a different locator as a 800 destination address prevents flooding attacks from off-path 801 attackers. 803 In addition, the usage of a 4-way handshake for establishing the 804 Shim6 context protects against DoS attacks, so hosts implementing the 805 Shim6 protocol should not be more vulnerable to DoS attacks than 806 regular IPv6 hosts. 808 Finally, many Shim6 signaling messages contain a Context Tag, meaning 809 that only attackers that know the Context Tag can forge them. As a 810 consequence, only on-path attackers can generate false Shim6 811 signaling packets for an established context. The impact of these 812 attacks would be limited since they would not be able to add 813 additional locators to the locator set (because of the HBA/CGA 814 protection). In general the possible attacks have similar effects to 815 the ones that an on-path attacker can launch on any regular IPv6 816 communication. The residual threats are described in the Security 817 Considerations of the Shim6 protocol specification [RFC5533]. 819 8.1. Privacy Considerations 821 The Shim6 protocol is designed to provide some basic privacy 822 features. In particular, HBAs are generated in such a way, that the 823 different addresses assigned to a host cannot be trivially linked 824 together as belonging to the same host, since there is nothing in 825 common in the addresses themselves. Similar features are provided 826 when the CGA protection is used. This means that it is not trivial 827 to determine that a set of addresses is assigned to a single Shim6 828 host. 830 However, the Shim6 protocol does exchange the locator set in clear 831 text and it also uses a fixed Context Tag when using different 832 locators in a given context. This implies that an attacker observing 833 the Shim6 context establishment exchange or seeing different payload 834 packets exchanged through different locators, but with the same 835 Context Tag, can determine the set of addresses assigned to a host. 836 However, this requires that the attacker is located along the path 837 and that it can capture the Shim6 signaling packets. A more in depth 838 analysis of the privacy of the Shim6 protocol can be found in 839 [I-D.bagnulo-shim6-privacy]. 841 9. IANA Considerations 843 This document has no actions for IANA. 845 10. Contributors 847 The analysis on the interaction between the Shim6 protocol and the 848 other protocols presented in this note benefited from the advice of 849 various people including Tom Henderson, Erik Nordmark, Hesham 850 Soliman, Vijay Devarpalli, John Loughney and Dave Thaler. 852 11. Acknowledgements 854 Joe Abley's work was supported in part by the US National Science 855 Foundation (research grant SCI-0427144) and DNS-OARC. 857 Marcelo Bagnulo worked on this document while visiting Ericsson 858 Research Laboratory Nomadiclab. 860 Shinta Sugimoto reviewed this document and provided comments and 861 text. 863 Iljitsch van Beijnum, Brian Carpenter, Sam Xia reviewed this document 864 and provided comments. 866 12. References 868 12.1. Normative References 870 [RFC2827] Ferguson, P. and D. Senie, "Network Ingress Filtering: 871 Defeating Denial of Service Attacks which employ IP Source 872 Address Spoofing", BCP 38, RFC 2827, May 2000. 874 [RFC3315] Droms, R., Bound, J., Volz, B., Lemon, T., Perkins, C., 875 and M. Carney, "Dynamic Host Configuration Protocol for 876 IPv6 (DHCPv6)", RFC 3315, July 2003. 878 [RFC3484] Draves, R., "Default Address Selection for Internet 879 Protocol version 6 (IPv6)", RFC 3484, February 2003. 881 [RFC3704] Baker, F. and P. Savola, "Ingress Filtering for Multihomed 882 Networks", BCP 84, RFC 3704, March 2004. 884 [RFC3775] Johnson, D., Perkins, C., and J. Arkko, "Mobility Support 885 in IPv6", RFC 3775, June 2004. 887 [RFC3963] Devarapalli, V., Wakikawa, R., Petrescu, A., and P. 888 Thubert, "Network Mobility (NEMO) Basic Support Protocol", 889 RFC 3963, January 2005. 891 [RFC3971] Arkko, J., Kempf, J., Zill, B., and P. Nikander, "SEcure 892 Neighbor Discovery (SEND)", RFC 3971, March 2005. 894 [RFC3972] Aura, T., "Cryptographically Generated Addresses (CGA)", 895 RFC 3972, March 2005. 897 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 898 Architecture", RFC 4291, February 2006. 900 [RFC4423] Moskowitz, R. and P. Nikander, "Host Identity Protocol 901 (HIP) Architecture", RFC 4423, May 2006. 903 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 904 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 905 September 2007. 907 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 908 Address Autoconfiguration", RFC 4862, September 2007. 910 [RFC4960] Stewart, R., "Stream Control Transmission Protocol", 911 RFC 4960, September 2007. 913 [RFC4984] Meyer, D., Zhang, L., and K. Fall, "Report from the IAB 914 Workshop on Routing and Addressing", RFC 4984, 915 September 2007. 917 [RFC5533] Nordmark, E. and M. Bagnulo, "Shim6: Level 3 Multihoming 918 Shim Protocol for IPv6", RFC 5533, June 2009. 920 [RFC5534] Arkko, J. and I. van Beijnum, "Failure Detection and 921 Locator Pair Exploration Protocol for IPv6 Multihoming", 922 RFC 5534, June 2009. 924 [RFC5535] Bagnulo, M., "Hash-Based Addresses (HBA)", RFC 5535, 925 June 2009. 927 12.2. Informative References 929 [I-D.bagnulo-ipv6-rfc3484-update] 930 Bagnulo, M., "Updating RFC 3484 for multihoming support", 931 draft-bagnulo-ipv6-rfc3484-update-00 (work in progress), 932 December 2005. 934 [I-D.bagnulo-shim6-privacy] 935 Bagnulo, M., "Privacy Analysis for the SHIM6 protocol", 936 draft-bagnulo-shim6-privacy-01 (work in progress), 937 October 2006. 939 [I-D.huitema-multi6-hosts] 940 Huitema, C. and R. Draves, "Host-Centric IPv6 941 Multihoming", draft-huitema-multi6-hosts-03 (work in 942 progress), February 2004. 944 [I-D.ietf-behave-v6v4-xlate-stateful] 945 Bagnulo, M., Matthews, P., and I. Beijnum, "Stateful 946 NAT64: Network Address and Protocol Translation from IPv6 947 Clients to IPv4 Servers", 948 draft-ietf-behave-v6v4-xlate-stateful-12 (work in 949 progress), July 2010. 951 [I-D.ietf-csi-dhcpv6-cga-ps] 952 Jiang, S., Shen, S., and T. Chown, "DHCPv6 and CGA 953 Interaction: Problem Statement", 954 draft-ietf-csi-dhcpv6-cga-ps-03 (work in progress), 955 June 2010. 957 [I-D.ietf-multi6-app-refer] 958 Nordmark, E., "Multi6 Application Referral Issues", 959 draft-ietf-multi6-app-refer-00 (work in progress), 960 January 2005. 962 [I-D.ietf-shim6-multihome-shim-api] 963 Komu, M., Bagnulo, M., Slavov, K., and S. Sugimoto, 964 "Socket Application Program Interface (API) for 965 Multihoming Shim", draft-ietf-shim6-multihome-shim-api-14 966 (work in progress), August 2010. 968 [I-D.nikander-esp-beet-mode] 969 Nikander, P. and J. Melen, "A Bound End-to-End Tunnel 970 (BEET) mode for ESP", draft-nikander-esp-beet-mode-09 971 (work in progress), August 2008. 973 [I-D.nordmark-shim6-esd] 974 Nordmark, E., "Extended Shim6 Design for ID/loc split and 975 Traffic Engineering", draft-nordmark-shim6-esd-01 (work in 976 progress), February 2008. 978 [RFC3221] Huston, G., "Commentary on Inter-Domain Routing in the 979 Internet", RFC 3221, December 2001. 981 [RFC3582] Abley, J., Black, B., and V. Gill, "Goals for IPv6 Site- 982 Multihoming Architectures", RFC 3582, August 2003. 984 [RFC4116] Abley, J., Lindqvist, K., Davies, E., Black, B., and V. 985 Gill, "IPv4 Multihoming Practices and Limitations", 986 RFC 4116, July 2005. 988 [RFC4218] Nordmark, E. and T. Li, "Threats Relating to IPv6 989 Multihoming Solutions", RFC 4218, October 2005. 991 Authors' Addresses 993 Joe Abley 994 Afilias Canada, Inc. 995 Suite 204 996 4141 Yonge Street 997 Toronto, Ontario M2P 2A8 998 Canada 1000 Phone: +1 416 673 4176 1001 Email: jabley@ca.afilias.info 1002 URI: http://afilias.info/ 1003 Marcelo Bagnulo 1004 U. Carlos III de Madrid 1005 Av. Universidad 30 1006 Leganes, Madrid 28911 1007 Spain 1009 Phone: +34 91 6248814 1010 Email: marcelo@it.uc3m.es 1011 URI: http://www.it.uc3m.es/ 1013 Alberto Garcia Martinez 1014 U. Carlos III de Madrid 1015 Av. Universidad 30 1016 Leganes, Madrid 28911 1017 Spain 1019 Phone: +34 91 6248782 1020 Email: alberto@it.uc3m.es 1021 URI: http://www.it.uc3m.es/