idnits 2.17.1 draft-ietf-mif-mpvd-arch-07.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (October 01, 2014) is 3488 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Outdated reference: A later version (-05) exists of draft-korhonen-dmm-prefix-properties-03 -- Obsolete informational reference (is this intentional?): RFC 3315 (Obsoleted by RFC 8415) -- Obsolete informational reference (is this intentional?): RFC 5996 (Obsoleted by RFC 7296) -- Obsolete informational reference (is this intentional?): RFC 6555 (Obsoleted by RFC 8305) Summary: 0 errors (**), 0 flaws (~~), 2 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 MIF Working Group D. Anipko, Ed. 3 Internet-Draft Unaffiliated 4 Intended status: Informational October 01, 2014 5 Expires: April 02, 2015 7 Multiple Provisioning Domain Architecture 8 draft-ietf-mif-mpvd-arch-07 10 Abstract 12 This document is a product of the work of the MIF Architecture Design 13 team. It outlines a solution framework for some of the issues 14 experienced by nodes that can be attached to multiple networks 15 simultaneously. The framework defines the concept of a Provisioning 16 Domain (PvD) which is a a consistent set of network configuration 17 information. PvD aware nodes learn PvD specific information from the 18 networks they are attached to and / or other sources. PvDs are used 19 to enable separation and configuration consistency in presence of 20 multiple concurrent connections. 22 Status of this Memo 24 This Internet-Draft is submitted in full conformance with the 25 provisions of BCP 78 and BCP 79. 27 Internet-Drafts are working documents of the Internet Engineering 28 Task Force (IETF). Note that other groups may also distribute 29 working documents as Internet-Drafts. The list of current Internet- 30 Drafts is at http://datatracker.ietf.org/drafts/current/. 32 Internet-Drafts are draft documents valid for a maximum of six months 33 and may be updated, replaced, or obsoleted by other documents at any 34 time. It is inappropriate to use Internet-Drafts as reference 35 material or to cite them other than as "work in progress." 37 This Internet-Draft will expire on April 02, 2015. 39 Copyright Notice 41 Copyright (c) 2014 IETF Trust and the persons identified as the 42 document authors. All rights reserved. 44 This document is subject to BCP 78 and the IETF Trust's Legal 45 Provisions Relating to IETF Documents (http://trustee.ietf.org/ 46 license-info) in effect on the date of publication of this document. 47 Please review these documents carefully, as they describe your rights 48 and restrictions with respect to this document. Code Components 49 extracted from this document must include Simplified BSD License text 50 as described in Section 4.e of the Trust Legal Provisions and are 51 provided without warranty as described in the Simplified BSD License. 53 Table of Contents 55 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 56 1.1. Requirements Language . . . . . . . . . . . . . . . . . . 3 57 2. Definitions and Types of PvDs . . . . . . . . . . . . . . . . 4 58 2.1. Explicit PvDs . . . . . . . . . . . . . . . . . . . . . . 4 59 2.2. Implicit PvDs and Incremental Adoption of Explicit PvDs . 5 60 2.3. Relationship Between PvDs and Interfaces . . . . . . . . . 6 61 2.4. PvD Identity / Naming . . . . . . . . . . . . . . . . . . 6 62 2.5. The Relationship to Dual-Stack Networks . . . . . . . . . 7 63 3. Conveying PvD information using DHCPv6 and Router Advertisement 7 64 3.1. Separate Messages or One Message? . . . . . . . . . . . . 8 65 3.2. Securing PvD Information . . . . . . . . . . . . . . . . . 8 66 3.3. Backward Compatibility . . . . . . . . . . . . . . . . . . 8 67 3.4. Selective Propagation . . . . . . . . . . . . . . . . . . 8 68 3.5. Retracting / Updating PvD Information . . . . . . . . . . 9 69 3.6. Conveying Configuration Information using IKEv2 . . . . . 9 70 4. Example Network Configurations . . . . . . . . . . . . . . . . 10 71 4.1. A Mobile Node . . . . . . . . . . . . . . . . . . . . . . 10 72 4.2. A Node with a VPN Connection . . . . . . . . . . . . . . . 11 73 4.3. A Home Network and a Network Operator with Multiple PvDs . 12 74 5. Reference Model for the PvD-aware Node . . . . . . . . . . . . 12 75 5.1. Constructions and Maintenance of Separate PvDs . . . . . . 13 76 5.2. Consistent use of PvDs for Network Connections . . . . . . 13 77 5.2.1. Name Resolution . . . . . . . . . . . . . . . . . . . 13 78 5.2.2. Next-hop and Source Address Selection . . . . . . . . 14 79 5.2.3. Listening Applications . . . . . . . . . . . . . . . . 15 80 5.2.3.1. Processing of Incoming Traffic . . . . . . . . . . 15 81 5.2.3.1.1. Connection-oriented APIs . . . . . . . . . . . 15 82 5.2.3.1.2. Connectionless APIs . . . . . . . . . . . . . 16 83 5.2.4. Enforcement of Security Policies . . . . . . . . . . . 16 84 5.3. Connectivity Tests . . . . . . . . . . . . . . . . . . . . 16 85 5.4. Relationship to Interface Management and Connection Manage 17 86 6. PvD support in APIs . . . . . . . . . . . . . . . . . . . . . 17 87 6.1. Basic . . . . . . . . . . . . . . . . . . . . . . . . . . 17 88 6.2. Intermediate . . . . . . . . . . . . . . . . . . . . . . . 18 89 6.3. Advanced . . . . . . . . . . . . . . . . . . . . . . . . . 18 90 7. PvD Trust for PvD-Aware Node . . . . . . . . . . . . . . . . . 18 91 7.1. Untrusted PvDs . . . . . . . . . . . . . . . . . . . . . . 18 92 7.2. Trusted PvDs . . . . . . . . . . . . . . . . . . . . . . . 19 93 7.2.1. Authenticated PvDs . . . . . . . . . . . . . . . . . . 19 94 7.2.2. PvDs Trusted by Attachment . . . . . . . . . . . . . . 20 95 8. Contributors . . . . . . . . . . . . . . . . . . . . . . . . . 20 96 9. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 20 97 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 20 98 11. Security Considerations . . . . . . . . . . . . . . . . . . . 20 99 12. References . . . . . . . . . . . . . . . . . . . . . . . . . . 21 100 12.1. Normative References . . . . . . . . . . . . . . . . . . 21 101 12.2. Informative References . . . . . . . . . . . . . . . . . 21 102 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 22 104 1. Introduction 106 Nodes attached to multiple networks may encounter problems from 107 conflicting configuration between the networks, or attempts to 108 simultaneously use more than one network. While various techniques 109 are currently used to tackle these problems ([RFC6419]), in many 110 cases issues may still appear. The MIF problem statement document 111 [RFC6418] describes the general landscape and discusses many of the 112 specific issues and scenario details. 114 Problems, enumerated in [RFC6418], can be grouped into 3 categories: 116 1. Lack of consistent and distinctive management of configuration 117 elements associated with different networks. 119 2. Inappropriate mixed use of configuration elements associated with 120 different networks during a particular network activity or 121 connection. 123 3. Usage of a particular network that is not consistent with the 124 intent of the scenario or involved parties leading to 125 connectivity failure and / or other undesired consequences. 127 An example of (1) is a single, node-scoped list of DNS server IP 128 addresses learned from different networks leading to failures or 129 delays in resolution of names from particular namespaces; an example 130 of (2) is an attempt to resolve the name of an HTTP proxy server 131 learned from network A using a DNS server learned from network B; an 132 example of (3) is the use of an employer-provided VPN connection for 133 peer-to-peer connectivity unrelated to employment activities. 135 This architecture provides solutions to these categories of problems, 136 respectively, by: 138 1. Introducing the formal notion of PvDs, including identity for 139 PvDs, and describing mechanisms for nodes to learn the intended 140 associations between acquired network configuration information 141 elements. 143 2. Introducing a reference model for PvD-aware nodes that prevents 144 the inadvertent mixed use of configuration information which may 145 belong to different PvDs. 147 3. Providing recommendations on PvD selection based on PvD identity 148 and connectivity tests for common scenarios. 150 1.1. Requirements Language 152 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 153 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 154 document are to be interpreted as described in RFC 2119 [RFC2119]. 156 2. Definitions and Types of PvDs 158 Provisioning Domain: 159 A consistent set of network configuration information. 160 Classically, all of the configuration information available on a 161 single interface is provided by a single source (such as a network 162 administrator) and can therefore be treated as a single 163 provisioning domain. In modern IPv6 networks, multihoming can 164 result in more than one provisioning domain being present on a 165 single link. In some scenarios, it is also possible for elements 166 of the same PvD to be present on multiple links. 168 Typical examples of information in a provisioning domain learned 169 from the network are: 171 * Source address prefixes for use by connections within the 172 provisioning domain 174 * IP address(es) of DNS server(s) 176 * Name of HTTP proxy server (if available) 178 * DNS suffixes associated with the network 180 * Default gateway address 182 PvD-aware node: 183 A node that supports the association of network configuration 184 information into PvDs and the use of these PvDs to serve requests 185 for network connections in ways consistent with the 186 recommendations of this architecture. 188 2.1. Explicit PvDs 190 A node may receive explicit information from the network and / or 191 other sources conveying the presence of PvDs and the association of 192 particular network information with a particular PvD. PvDs that are 193 constructed based on such information are referred to as "explicit" 194 in this document. 196 Protocol changes or extensions will likely be required to support 197 explicit PvDs through IETF-defined mechanisms. As an example, one 198 could think of one or more DHCP options carrying PvD identity and / 199 or its elements. 201 A different approach could be the introduction of a DHCP option which 202 only carries the identity of a PvD. Here, the associations between 203 network information elements with the identity is implemented by the 204 respective protocols, for example with a Router Discovery [RFC4861] 205 option associating an address range with a PvD. 207 Another example of a delivery mechanism for PvDs are key exchange or 208 tunneling protocols, such as IKEv2 [RFC5996] that allow the 209 transport of host configuration information. 211 Specific, existing or new features of networking protocols that 212 enable the delivery of PvD identity and association with various 213 network information elements will be defined in companion design 214 documents. 216 Link-specific and / or vendor-proprietary mechanisms for the 217 discovery of PvD information (differing from IETF-defined mechanisms) 218 can be used by nodes either separate from, or in conjunction with, 219 IETF-defined mechanisms; providing they allow the discovery of the 220 necessary elements of the PvD(s). 222 In all cases, nodes must by default ensure that the lifetime of all 223 dynamically discovered PvD configuration is appropriately limited by 224 relevant events. For example, if an interface media state change is 225 indicated, previously discovered information relevant to that 226 interface may no longer be valid and so need to be confirmed or re- 227 discovered. 229 It is expected that the way a node makes use of PvD information is 230 generally independent of the specific mechanism / protocol that the 231 information was received by. 233 In some network topologies, network infrastructure elements may need 234 to advertise multiple PvDs. Generally, the details of how this is 235 performed will be defined in companion design documents. However, 236 where different design choices are possible, the choice that requires 237 a smaller number of packets shall be preferred for efficiency. 239 2.2. Implicit PvDs and Incremental Adoption of Explicit PvDs 241 For some time it is likely that there will be networks which do not 242 advertise explicit PvD information as the deployment of new features 243 in networking protocols is a relatively slow process. 245 When connected to networks which don't advertise explicit PvD 246 information, a PvD-aware node shall automatically create separate 247 PvDs for received configuration. Such PvDs are referred to in this 248 document as "implicit". 250 Through the use of implicit PvDs, PvD-aware nodes may still provide 251 benefits to their users (when compared to non-PvD aware nodes) by 252 following the best practices described in Section 5, using the 253 network information from different interfaces separately to 254 consistently serve network connection request. 256 In mixed mode, i.e., where of multiple networks are available on an 257 attached link only some of which advertise PvD information, the PvD- 258 aware node shall create explicit PvDs from explicitly learned PvD 259 information and associate other learned configuration (without an 260 explicit PvD) with implicit PvD(s) created for that interface. 262 2.3. Relationship Between PvDs and Interfaces 264 By default, implicit PvDs are limited to the network configuration 265 information received on a single interface and by default one such 266 PvD is formed for each interface. If additional information is 267 available to the host (through mechanisms out of scope of this 268 document), the host may form implicit PvDs with different 269 granularity. For example, PvDs spanning multiple interfaces such a 270 home network with a router that has multiple internal interfaces, or 271 multiple PvDs on a single interface such as a network that has 272 multiple uplink connections. 274 Explicit PvDs, in practice will often also be scoped only for 275 configuration related to a particular interface. However, there are 276 no such requirements or limitations defined in this architecture. 277 Explicit PvDs may include information related to more than one 278 interface if the node learns the presence of the same PvD on those 279 interfaces and the authentication of the PvD ID meets the level 280 required by the node policy (generally, authentication of a PvD ID 281 may be also required in scenarios involving only one connected 282 interface and / or PvD). 284 This architecture intends to support such scenarios, among others. 285 Hence, it shall be noted that no hierarchical relationship exists 286 between interfaces and PvDs: it is possible for multiple PvDs to be 287 simultaneously accessible over one interface, as well as a single PvD 288 to be simultaneously accessible over multiple interfaces. 290 2.4. PvD Identity / Naming 292 For explicit PvDs, the PvD ID is a value that is, or has a high 293 probability of being globally unique, and is received as part of PvD 294 information. It shall be possible to generate a human-readable form 295 of the PvD ID to present to the end-user, either based on the PvD ID 296 itself, or using meta-data associated with the ID. For implicit PvDs, 297 the node assigns a locally generated ID with a high probability of 298 being globally unique to each implicit PvD. 300 A PvD-aware node may use these IDs to select a PvD with a matching ID 301 for special-purpose connection requests in accordance with node 302 policy, as chosen by advanced applications, or to present a human- 303 readable representation of the IDs to the end-user for selection of 304 PvDs. 306 A single network provider may operate multiple networks, including 307 networks at different locations. In such cases, the provider may 308 chose whether to advertise single or multiple PvD identities at all 309 or some of those networks as it suits their business needs. This 310 architecture does not impose any specific requirements in this 311 regard. 313 When multiple nodes are connected to the same link with one or more 314 explicit PvDs available, this architecture assumes that the 315 information about all available PvDs is made available by the 316 networks to all the connected nodes. At the same time, connected 317 nodes may have different heuristics, policies and / or other 318 settings, including their configured sets of trusted PvDs. This may 319 lead to different PvDs actually being used by different nodes for 320 their connections. 322 Possible extensions, whereby networks advertize different sets of 323 PvDs to different connected nodes are out of scope of this document. 325 2.5. The Relationship to Dual-Stack Networks 327 When applied to dual-stack networks, the PvD definition allows for 328 multiple PvDs to be created whereby each PvD contains information 329 relevant to only one address family, or for a single PvD containing 330 information for multiple address families. This architecture 331 requires that accompanying design documents describing PvD-related 332 protocol changes must support PvDs containing information from 333 multiple address families. PvD-aware nodes must be capable of 334 creating and using both single-family and multi-family PvDs. 336 For explicit PvDs, the choice of either of these approaches is a 337 policy decision for the network administrator and / or the node user/ 338 administrator. Since some of the IP configuration information that 339 can be learned from the network can be applicable to multiple address 340 families (for instance DHCP Address Selection Policy Opt [RFC7078]), 341 it is likely that dual-stack networks will deploy single PvDs for 342 both address families. 344 By default for implicit PvDs, PvD-aware nodes shall include multiple 345 IP families into a single implicit PvD created for an interface. At 346 the time of writing, in dual-stack networks it appears to be common 347 practice for the configuration of both address families to be 348 provided by a single source. 350 A PvD-aware node that provides an API to use, enumerate and inspect 351 PvDs and / or their properties shall provide the ability to filter 352 PvDs and / or their properties by address family. 354 3. Conveying PvD information using DHCPv6 and Router Advertisements 356 DHCPv6 and Router Advertisements are the two most common methods of 357 configuring hosts. To support the architecture described in this 358 document, these protocols would need to be extended to convey 359 explicit PvD information. The following sections describe topic 360 which must be considered before finalizing a mechanism to augment 361 DHCPv6 and RAs with PvD information. 363 3.1. Separate Messages or One Message? 365 When information related to several PvDs is available from the same 366 configuration source, there are two possible ways of distributing 367 this information: One way is to send information from each different 368 provisioning domain in separate messages. The second method is 369 combining the information from multiple PvDs into a single message. 370 The latter method has the advantage of being more efficient but could 371 have problems with to authentication and authorization, as well as 372 potential issues with accommodating information not tagged with any 373 PvD information. 375 3.2. Securing PvD Information 377 DHCPv6 and RAs both provide some form of authentication to ensure the 378 identity of the source as well as the integrity of the secured 379 message content. While this is useful, determining authenticity does 380 tell a node whether the configuration source is actually allowed to 381 provide information from a given PvD. To resolve this, there must be 382 a mechanism for the PvD owner to attach some form of authorization 383 token to the configuration information that is delivered. 385 3.3. Backward Compatibility 387 The extensions to RAs and DHCPv6 should be defined in such a manner 388 than unmodified hosts (i.e. hosts not aware of PvDs) will continue 389 to function as well as they did prior to PvD information being added. 390 This could imply that some information may need to be duplicated in 391 order to be conveyed to legacy hosts. Similarly, PvD aware hosts 392 need to be able to correctly utiilize legacy configuration sources 393 which do not provide PvD information. There are also several 394 initiatives that are aimed at adding some form of additional 395 information to prefixes [I-D.bhandari-dhc-class-based-prefix] and 396 [I-D.korhonen-dmm-prefix-properties] and any new mechanism should try 397 to consider co-existence with such deployed mechanisms. 399 3.4. Selective Propagation 400 When a configuration source has information regarding several PvDs, 401 it is currently unclear whether the source should provide information 402 about all PvDs to any host that requests this information. While 403 this may be reasonable in some cases, it might become an unreasonable 404 burden once the number of PvDs starts increasing. One way to 405 restrict the propagation of information which is of no use to a 406 specific host is for the host to indicate the PvD information they 407 require within their configuration request. One way this could be 408 accomplished is by using a DHCPv6 ORO containing the PvDs that are of 409 interest. The configuration source can then respond with only the 410 requested information. 412 By default, a configuration source SHOULD provide information related 413 to all provisioning domains without expecting the client to request 414 the PvD(s) it requires. This is necessary to ensure that hosts that 415 do not support a selective PvD information request mechanism will 416 work. Also, note that IPv6 neighbor discovery does not provide any 417 functionality analogous to the DHCPv6 ORO. 419 In this case, when a host receives superfluous PvD information, it 420 can simply be discarded. Also, in constrained networks such as LLNs, 421 the amount of configuration information needs to be restricted to 422 ensure that the load on the hosts is bearable while keeping the 423 information identical across all the hosts. 425 If selective propagation is required, some form of PvD discovery 426 mechanism needs to be specified so that hosts / applications can be 427 pre-provisioned to request a specific PvD. Alternately, the set of 428 PvDs that the network can provide to the host can be propagated to 429 the host using RAs or stateless DHCPv6. The discovery mechanism may 430 potentially support the discovery of available PvDs on a per-host 431 basis. 433 3.5. Retracting / Updating PvD Information 435 After PvD information is provisioned to a host, it may become 436 outdated or superseded by updated information before the hosts would 437 normally request updates. To resolve this requires that the 438 mechanism be able to update and / or withdraw all (or some subset) of 439 the information related to a given PvD. For efficiency reasons, there 440 should be a way to specify that all information from the PvD needs to 441 be reconfigured instead of individually updating each item associated 442 with the PvD. 444 3.6. Conveying Configuration Information using IKEv2 445 Internet Key Exchange protocol version 2 (IKEv2) [RFC5996] [RFC5739] 446 is another widely used method of configuring host IP information. 447 For IKEv2, the provisioning domain could be implicitly learned from 448 the Identification - Responder (IDr) payloads that the IKEv2 449 initiator and responder inject during their IKEv2 exchange. The IP 450 configuration may depend on the named IDr. Another possibility could 451 be adding a specific provisioning domain identifying payload 452 extensions to IKEv2. All of the considerations for DHCPv6 and RAs 453 listed above potentially apply to IKEv2 as well. 455 4. Example Network Configurations 457 4.1. A Mobile Node 459 Consider a mobile node with two network interfaces: one to the mobile 460 network, the other to the Wi-Fi network. When the mobile node is 461 only connected to the mobile network, it will typically have one PvD, 462 implicit or explicit. When the mobile node discovers and connects to 463 a Wi-Fi network, it will have zero or more (typically one) additional 464 PvD(s). 466 Some existing OS implementations only allow one active network 467 connection. In this case, only the PvD(s) associated with the active 468 interface can be used at any given time. 470 As an example, the mobile network can explicitly deliver PvD 471 information through the PDP context activation process. Then, the 472 PvD aware mobile node will treat the mobile network as an explicit 473 PvD. Conversely, the legacy Wi-Fi network may not explicitly 474 communicate PvD information to the mobile node. The PvD aware mobile 475 node will associate network configuration for the Wi-Fi network with 476 an implicit PvD in this case. 478 The following diagram illustrates the use of different PvDs in this 479 scenario: 481 <----------- Wi-Fi 'Internet' PvD -------> 482 +--------+ 483 | +----+ | +----+ _ __ _ _ 484 | |WiFi| | | | ( ` ) ( ` )_ 485 | |-IF +-|----+ |--------------------------( `) 486 | | | | |WiFi| ( ) (_ Internet _) 487 | +----+ | | AP | ( ) `- __ _) - 488 | | | | ( Service ) 489 | | +----+ ( Provider's ) 490 | | ( Networks - 491 | +----+ | `_ ) _ _ 492 | |CELL| | ( ) ( ` )_ 493 | |-IF +-|-------------------------------------( `) 494 | | | | (_ __) (_ Internet _) 495 | +----+ | `- -- `- __ _) - 496 +--------+ 497 <------- Mobile 'Internet' PvD -----------> 499 An example of PvD use with Wi-Fi and mobile interfaces. 501 4.2. A Node with a VPN Connection 503 If the node has established a VPN connection, zero or more (typically 504 one) additional PvD(s) will be created. These may be implicit or 505 explicit. The routing to IP addresses reachable within this PvD will 506 be set up via the VPN connection, and the routing of packets to 507 addresses outside the scope of this PvD will remain unaffected. If a 508 node already has N connected PvDs, after the VPN session has been 509 established typically there will be N+1 connected PvDs. 511 The following diagram illustrates the use of different PvDs in this 512 scenario: 514 <----------- 'Internet' PvD ------> 515 +--------+ 516 | +----+ | +----+ _ __ _ _ 517 | |Phy | | | | ( ` ) ( ` )_ 518 | |-IF +-|----+ |--------------------( `) 519 | | | | | | ( ) (_ Internet _) 520 | +----+ | | | ( ) `- __ _) - 521 | | |Home| ( Service ) || 522 | | |gate| ( Provider's ) || 523 | | |-way| ( Network - || 524 | +----+ | | | `_ ) +---------+ +------------+ 525 | |VPN | | | | ( ) | VPN | | | 526 | |-IF +-|----+ |---------------------+ Gateway |--+ Private | 527 | | | | | | (_ __) | | | Services | 528 | +----+ | +----+ `- -- +---------+ +------------+ 529 +--------+ 530 <-------------- Explicit 'VPN' PvD -----> 532 An example of PvD use with VPN. 534 4.3. A Home Network and a Network Operator with Multiple PvDs 536 An operator may use separate PvDs for individual services which they 537 offer to their customers. These may be used so that services can be 538 designed and provisioned to be completely independent of each other, 539 allowing for complete flexibility in combinations of services which 540 are offered to customers. 542 From the perspective of the home network and the node, this model is 543 functionally very similar to being multihomed to multiple upstream 544 operators: Each of the different services offered by the service 545 provider is its own PvD with associated PvD information. In this 546 case, the operator may provide a generic / default PvD (explicit or 547 implicit), which provides Internet access to the customer. 548 Additional services would then be provisioned as explicit PvDs for 549 subscribing customers. 551 The following diagram illustrates this, using video-on-demand as a 552 service-specific PvD: 554 <------ Implicit 'Internet' PvD ------> 555 +----+ +----+ _ __ _ _ 556 | | | | ( ` ) ( ` )_ 557 | PC +-----+ |--------------------------( `) 558 | | | | ( ) (_ Internet _) 559 +----+ | | ( ) `- __ _) - 560 |Home| ( Service ) 561 |gate| ( Provider's ) 562 |-way| ( Network - 563 +-----+ | | `_ ) +---------+ 564 | Set | | | ( ) |ISP Video| 565 | Top +----+ |---------------------------+on Demand| 566 | Box | | | (_ __) | Service | 567 +-----+ +----+ `- -- +---------+ 568 <-- Explicit 'Video-on-Demand' PvD --> 570 An example of PvD use within a home network. 572 In this case, the number of PvDs that a single operator could 573 provision is based on the number of independently provisioned 574 services which they offer. Some examples may include: 576 o Real-time packet voice 578 o Streaming video 580 o Interactive video (n-way video conferencing) 582 o Interactive gaming 584 o Best effort / Internet access 586 5. Reference Model for the PvD-aware Node 587 5.1. Constructions and Maintenance of Separate PvDs 589 It is assumed that normally, the configuration information contained 590 in a single PvD shall be sufficient for a node to fulfill a network 591 connection request by an application, and hence there should be no 592 need to attempt to merge information across different PvDs. 594 Nevertheless, even when a PvD lacks some necessary configuration 595 information, merging of information associated with different PvD(s) 596 shall not be done automatically as this will typically lead to the 597 issues described in [RFC6418]. 599 A node may use other sources, for example: node local policy, user 600 input or other mechanisms not defined by the IETF for any of the 601 following: 603 o Construction of a PvD in its entirety (analogous to statically 604 configuring IP on an interface) 606 o Supplementing some, or all learned PvDs with particular 607 configuration elements 609 o Merging of information from different PvDs (if this is explicitly 610 allowed by policy) 612 As an example, a node administrator could inject a DNS server which 613 is not ISP-specific into PvDs for use on any of the networks that the 614 node could attach to. Such creation / augmentation of PvD(s) could 615 be static or dynamic. The specific mechanism(s) for implementing 616 this are outside of scope of this document. 618 5.2. Consistent use of PvDs for Network Connections 620 PvDs enable PvD-aware nodes to consistently use the correct set of 621 configuration elements to serve specific network requests from 622 beginning to end. This section provides examples of such use. 624 5.2.1. Name Resolution 626 When a PvD-aware node needs to resolve the name of the destination 627 for use by a connection request, the node could use one, or multiple 628 PvDs for a given name lookup. 630 The node shall chose a single PvD if, for example, the node policy 631 required the use of a particular PvD for a specific purpose (e.g. to 632 download an MMS message using a specific APN over a cellular 633 connection). To make this selection, the node could use a match 634 between the PvD DNS suffix and an FQDN which is being resolved or 635 match of PvD ID, as determined by the node policy. 637 The node may pick multiple PvDs, if for example, the PvDs are for 638 general purpose Internet connectivity, and the node is attempting to 639 maximize the probability of connectivity similar to the Happy 640 Eyeballs [RFC6555] approach. In this case, the node could perform 641 DNS lookups in parallel, or in sequence. Alternatively, the node may 642 use only one PvD for the lookup, based on the PvD connectivity 643 properties, user configuration of preferred Internet PvD, etc. 645 If an application implements an API that provides a way of explicitly 646 specifying the desired interface or PvD, that interface or PvD should 647 be used for name resolution (and the subsequent connection attempt), 648 provided that the host's configuration permits this. 650 In either case, by default a node uses information obtained via a 651 name service lookup to establish connections only within the same PvD 652 as the lookup results were obtained. 654 For clarification, when it is written that the name service lookup 655 results were obtained "from a PvD", it should be understood to mean 656 that the name service query was issued against a name service which 657 is configured for use in a particular PvD. In that sense, the 658 results are "from" that particular PvD. 660 Some nodes may support transports and / or APIs which provide an 661 abstraction of a single connection, aggregating multiple underlying 662 connections. MPTCP [RFC6182] is an example of such a transport 663 protocol. For connections provided by such transports/APIs, a PvD- 664 aware node may use different PvDs for servicing that logical 665 connection, provided that all operations on the underlying 666 connections are performed consistently within their corresponding 667 PvD(s). 669 5.2.2. Next-hop and Source Address Selection 671 For the purpose of this example, let us assume that the preceding 672 name lookup succeeded in a particular PvD. For each obtained 673 destination address, the node shall perform a next-hop lookup among 674 routers associated with that PvD. As an example, the node could 675 determine such associations via matching the source address prefixes/ 676 specific routes advertized by the router against known PvDs, or 677 receiving an explicit PvD affiliation advertized through a new Router 678 Discovery [RFC4861] option. 680 For each destination, once the best next-hop is found, the node 681 selects the best source address according to rules defined in 682 [RFC6724], but with the constraint that the source address must 683 belong to a range associated with the used PvD. If needed, the node 684 would use prefix policy from the same PvD for selecting the best 685 source address from multiple candidates. 687 When destination / source pairs are identified, they are sorted using 688 the [RFC6724] destination sorting rules and prefix policy table from 689 the used PvD. 691 5.2.3. Listening Applications 693 Consider a host connected to several PvDs, running an application 694 that opens a listening socket / transport API object. The 695 application is authorized by the host policy to use a subset of 696 connected PvDs that may or may not be equal to the complete set of 697 the connected PvDs. As an example, in the case where there are 698 different PvDs on the Wi-Fi and cellular interfaces, for general 699 Internet traffic the host could use only one, preferred PvD at a time 700 (and accordingly, advertise to remote peers the host name and 701 addresses associated with that PvD), or it could use one PvD as the 702 default for outgoing connections, while still allowing use of the 703 other PvDs simultaneously. 705 Another example is a host with an established VPN connection. Here, 706 security policy could be used to permit or deny application's access 707 to the VPN (and other) PvD(s). 709 For non-PvD aware applications, the operating system has policies 710 that determine the authorized set of PvDs and the preferred outgoing 711 PvD. For PvD-aware applications, both the authorized set of PvDs and 712 the default outgoing PvD can be determined as the common subset 713 produced between the OS policies and the set of PvD IDs or 714 characteristics provided by the application. 716 Application input could be provided on per-application, per- 717 transport-API-object or per-transport-API-call basis. The API for 718 application input may have an option for specifying whether the input 719 should be treated as a preference instead of a requirement. 721 5.2.3.1. Processing of Incoming Traffic 723 Unicast IP packets are received on a specific IP address associated 724 with a PvD. For multicast packets, the host can derive the PvD 725 association from other configuration information, such as an explicit 726 PvD property or local policy. 728 The node OS or middleware may apply more advanced techniques for 729 determining the resultant PvD and / or authorization of the incoming 730 traffic. Those techniques are outside of scope of this document. 732 If the determined receiving PvD of a packet is not in the allowed 733 subset of PvDs for the particular application / transport API object, 734 the packet should be handled in the same way as if there were no 735 listener. 737 5.2.3.1.1. Connection-oriented APIs 739 For connection-oriented APIs, when the initial incoming packet is 740 received, the packet PvD is remembered for the established connection 741 and used for handling of outgoing traffic for that connection. While 742 typically, connection-oriented APIs use a connection-oriented 743 transport protocol, such as TCP, it is possible to have a connection- 744 oriented API that uses a generally connectionless transport protocol, 745 such as UDP. 747 For APIs/protocols that support multiple IP traffic flows associated 748 with a single transport API connection object (for example, multi 749 path TCP), the processing rules may be adjusted accordingly. 751 5.2.3.1.2. Connectionless APIs 753 For connectionless APIs, the host should provide an API that PvD- 754 aware applications can use to query the PvD associated with the 755 packet. For outgoing traffic on this transport API object, the OS 756 should use the selected outgoing PvDs, determined as described above. 758 5.2.4. Enforcement of Security Policies 760 By themselves, PvDs do not define, and cannot be used for 761 communication of, security policies. When implemented in a network, 762 this architecture provides the host with information about connected 763 networks. The actual behavior of the host then depends on the host's 764 policies (provisioned through mechanisms out of scope of this 765 document), applied taking received PvD information into account. In 766 some scenarios, e.g. a VPN, such policies could require the host to 767 use only a particular VPN PvD for some / all of the application's 768 traffic (VPN 'disable split tunneling' also known as 'force 769 tunneling' behavior), or apply such restrictions only to selected 770 applications and allow the simultaneous use of the VPN PvD together 771 with the other connected PvDs by the other or all applications (VPN 772 'split tunneling' behavior). 774 5.3. Connectivity Tests 776 Although some PvDs may appear as valid candidates for PvD selection 777 (e.g. good link quality, consistent connection parameters, etc.), 778 they may provide limited or no connectivity to the desired network or 779 the Internet. For example, some PvDs provide limited IP connectivity 780 (e.g., scoped to the link or to the access network), but require the 781 node to authenticate through a web portal to get full access to the 782 Internet. This may be more likely to happen for PvDs which are not 783 trusted by a given PvD-aware node. 785 An attempt to use such a PvD may lead to limited network connectivity 786 or application connection failures. To prevent the latter, a PvD- 787 aware node may perform a connectivity test for the PvD before using 788 it to serve application network connection requests. In current 789 implementations, some nodes already implement this e.g., by trying to 790 reach a dedicated web server (see [RFC6419]). 792 Section 5.2 describes how a PvD-aware node shall maintain and use 793 multiple PvDs separately. The PvD-aware node shall perform a 794 connectivity test and, only after validation of the PvD, consider 795 using it to serve application connections requests. Ongoing 796 connectivity tests are also required, since during the IP session, 797 the end-to-end connectivity could be disrupted for various reasons 798 (e.g. L2 problems, IP QoS issues); hence, a connectivity monitoring 799 function is needed to check the connectivity status and remove the 800 PvD from the set of usable PvDs if necessary. 802 There may be cases where a connectivity test for PvD selection may 803 not be appropriate and should be complemented, or replaced, by PvD 804 selection based on other factors. For example, this could be 805 realized by leveraging some 3GPP and IEEE mechanisms, which would 806 allow the exposure of some PvD characteristics to the node (e.g. 807 3GPP Access Network Discovery and Selection Function (ANDSF) 808 [TS23402], IEEE 802.11u [IEEE802.11u]/ANQP). 810 5.4. Relationship to Interface Management and Connection Managers 812 Current devices, such as mobile handsets make use of proprietary 813 mechanisms and custom applications to manage connectivity in 814 environments with multiple interfaces and multiple sets of network 815 configuration. These mechanisms or applications are commonly known 816 as connection managers [RFC6419]. 818 Connection managers sometimes rely on policy servers to allow a node 819 that is connected to multiple networks to perform network selection. 820 They can also make use of routing guidance from the network (e.g. 821 3GPP ANDSF [TS23402]). Although connection managers solve some 822 connectivity problems, they rarely address network selection problems 823 in a comprehensive manner. With proprietary solutions, it is 824 challenging to present coherent behavior to the end user of the 825 device, as different platforms present different behaviors even when 826 connected to the same network, with the same type of interface, and 827 for the same purpose. The architecture described in this document 828 should improve the hosts behavior by providing the hosts with tools 829 and guidance to make informed network selection decisions. 831 6. PvD support in APIs 833 For all levels of PvD support in APIs described in this chapter, it 834 is expected that the notifications about changes in the set of 835 available PvDs are exposed as part of the API surface. 837 6.1. Basic 838 Applications are not PvD-aware in any manner and only submit 839 connection requests. The node performs PvD selection implicitly, 840 without any application participation, based purely on node-specific 841 administrative policies and / or choices made by the user from a user 842 interface provided by the operating environment, not by the 843 application. 845 As an example, PvD selection can be done at the name service lookup 846 step by using the relevant configuration elements, such as those 847 described in [RFC6731]. As another example, PvD selection could be 848 made based on application identity or type (i.e., a node could always 849 use a particular PvD for a VOIP application). 851 6.2. Intermediate 853 Applications indirectly participate in PvD selection by specifying 854 hard requirements and soft preferences. As an example, a real time 855 communication application intending to use the connection for the 856 exchange of real time audio / video data may indicate a preference or 857 a requirement for connection quality, which could affect PvD 858 selection (different PvDs could correspond to Internet connections 859 with different loss rates and latencies). 861 Another example is the connection of an infrequently executed 862 background activity, which checks for application updates and 863 performs large downloads when updates are available. For such 864 connections, a cheaper or zero cost PvD may be preferable, even if 865 such a connection has a higher relative loss rate or lower bandwidth. 866 The node performs PvD selection based on applications' inputs and 867 policies and / or user preferences. Some / all properties of the 868 resultant PvD may be exposed to applications. 870 6.3. Advanced 872 PvDs are directly exposed to applications for enumeration and 873 selection. Node polices and / or user choices may still override the 874 applications' preferences and limit which PvD(s) can be enumerated 875 and / or used by the application, irrespective of any preferences 876 which the application may have specified. Depending on the 877 implementation, such restrictions (imposed by node policy and / or 878 user choice) may or may not be visible to the application. 880 7. PvD Trust for PvD-Aware Node 882 7.1. Untrusted PvDs 884 Implicit and explicit PvDs for which no trust relationship exists are 885 considered untrusted. Only PvDs which meet the requirements in 886 Section 7.2 are trusted; any other PvD is untrusted. 888 In order to avoid the various forms of misinformation that could 889 occur when PvDs are untrusted, nodes that implement PvD separation 890 cannot assume that two explicit PvDs with the same identifier are 891 actually the same PvD. A node that makes this assumption will be 892 vulnerable to attacks where, for example, an open Wifi hotspot might 893 assert that it was part of another PvD and thereby attempt to draw 894 traffic intended for that PvD onto its own network. 896 Since implicit PvD identifiers are synthesized by the node, this 897 issue cannot arise with implicit PvDs. 899 Mechanisms exist (for example, [RFC6731]) whereby a PvD can provide 900 configuration information that asserts special knowledge about the 901 reachability of resources through that PvD. Such assertions cannot be 902 validated unless the node has a trust relationship with the PvD; 903 therefore, assertions of this type must be ignored by nodes that 904 receive them from untrusted PvDs. Failure to ignore such assertions 905 could result in traffic being diverted from legitimate destinations 906 to spoofed destinations. 908 7.2. Trusted PvDs 910 Trusted PvDs are PvDs for which two conditions apply: First, a trust 911 relationship must exist between the node that is using the PvD 912 configuration and the source that provided that configuration; this 913 is the authorization portion of the trust relationship. Second, 914 there must be some way to validate the trust relationship. This is 915 the authentication portion of the trust relationship. Two mechanisms 916 for validating the trust relationship are defined. 918 It shall be possible to validate the trust relationship for all 919 advertised elements of a trusted PvD, irrespective of whether the PvD 920 elements are communicated as a whole, e.g., in a single DHCP option, 921 or separately, e.g., in supplementary RA options. The feasibility of 922 mechanisms to implement a trust relationship for all PvD elements 923 will be determined in the respective companion design documents. 925 7.2.1. Authenticated PvDs 927 One way to validate the trust relationship between a node and the 928 source of a PvD is through the combination of cryptographic 929 authentication and an identifier configured on the node. In some 930 cases, the two could be the same; for example, if authentication is 931 by a shared secret, the secret would have to be associated with the 932 PvD identifier. Without a PvD Identifier / shared key tuple, 933 authentication would be impossible, and hence authentication and 934 authorization are combined. 936 However, if authentication is done using a public key mechanism such 937 as a TLS certificate or DANE, authentication by itself is not enough 938 since theoretically any PvD could be authenticated in this way. In 939 addition to authentication, the node would need configuration to 940 trust the identifier being authenticated. Validating the 941 authenticated PvD name against a list of PvD names configured as 942 trusted on the node would constitute the authorization step in this 943 case. 945 7.2.2. PvDs Trusted by Attachment 947 In some cases, a trust relationship may be validated by some means 948 other than those described in Section 7.2.1 simply by virtue of the 949 connection through which the PvD was obtained. For instance, a 950 handset connected to a mobile network may know through the mobile 951 network infrastructure that it is connected to a trusted PvD. 952 Whatever mechanism was used to validate that connection constitutes 953 the authentication portion of the PvD trust relationship. 954 Presumably, such a handset would be configured from the factory (or 955 else through mobile operator or user preference settings) to trust 956 the PvD, and this would constitute the authorization portion of this 957 type of trust relationship. 959 8. Contributors 961 The following individuals contributed to this document (listed in no 962 specific order): Alper Yegin (alper.yegin@yegin.org), Aaron Yi Ding 963 (yding@cs.helsinki.fi), Zhen Cao (caozhenpku@gmail.com), Dapeng Liu 964 (liudapeng@chinamobile.com), Dave Thaler (dthaler@microsoft.com), 965 Dmitry Anipko (dmitry.anipko@microsoft.com), Hui Deng 966 (denghui@chinamobile.com), Jouni Korhonen (jouni.nospam@gmail.com), 967 Juan Carlos Zuniga (JuanCarlos.Zuniga@InterDigital.com), Konstantinos 968 Pentikousis (k.pentikousis@huawei.com), Marc Blanchet 969 (marc.blanchet@viagenie.ca), Margaret Wasserman 970 (margaretw42@gmail.com), Pierrick Seite (pierrick.seite@orange.com), 971 Suresh Krishnan (suresh.krishnan@ericsson.com), Teemu Savolainen 972 (teemu.savolainen@nokia.com), Ted Lemon (ted.lemon@nominum.com) and 973 Tim Chown (tjc@ecs.soton.ac.uk). 975 9. Acknowledgments 977 The authors would like to thank (in no specific order) Ian Farrer, 978 Marcus Stenberg and Mikael Abrahamsson for their review and comments. 980 10. IANA Considerations 982 This memo does not include any IANA requests. 984 11. Security Considerations 986 There are at least three different forms of attacks that can be 987 performed using configuration sources that support multiple 988 provisioning domains. 990 Tampering with provided configuration information: An attacker may 991 attempt to modify information provided inside the PvD container 992 option. These attacks can easily be prevented by using message 993 integrity features provided by the underlying protocol used to 994 carry the configuration information. E.g. SEND [RFC3971] would 995 detect any form of tampering with the RA contents and the DHCPv6 996 [RFC3315] AUTH option that would detect any form of tampering with 997 the DHCPv6 message contents. This attack can also be performed by 998 a compromised configuration source by modifying information inside 999 a specific PvD, in which case the mitigations proposed in the next 1000 subsection may be helpful. 1002 Rogue configuration source: A compromised configuration source, such 1003 as a router or a DHCPv6 server, may advertise information about 1004 PvDs that it is not authorized to advertise. e.g. A coffee shop 1005 WLAN may advertise configuration information purporting to be from 1006 an enterprise and may try to attract enterprise related traffic. 1007 The only real way to prevent this is is for the PvD related 1008 configuration container to contain embedded authentication and 1009 authorization information from the owner of the PvD. This provides 1010 the client with a way of detecting the attack by verifying the 1011 authentication and authorization information provided inside the 1012 PvD container option, after verifying its trust of the PvD owner 1013 (e.g. a certificate with a well-known / common trust anchor). 1015 Replay attacks: A compromised configuration source or an on-link 1016 attacker may try to capture advertised configuration information 1017 and replay it on a different link, or at a future point in time. 1018 This can be avoided by including a replay protection mechanism 1019 such as a timestamp or a nonce inside the PvD container to ensure 1020 the validity of the provided information. 1022 12. References 1024 12.1. Normative References 1026 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1027 Requirement Levels", BCP 14, RFC 2119, March 1997. 1029 12.2. Informative References 1031 [I-D.bhandari-dhc-class-based-prefix] 1032 Systems, C., Halwasia, G., Gundavelli, S., Deng, H., 1033 Thiebaut, L., Korhonen, J. and I. Farrer, "DHCPv6 class 1034 based prefix", Internet-Draft draft-bhandari-dhc-class- 1035 based-prefix-05, July 2013. 1037 [I-D.korhonen-dmm-prefix-properties] 1038 Korhonen, J., Patil, B., Gundavelli, S., Seite, P. and D. 1039 Liu, "IPv6 Prefix Mobility Management Properties", 1040 Internet-Draft draft-korhonen-dmm-prefix-properties-03, 1041 October 2012. 1043 [IEEE802.11u] 1044 IEEE, "IEEE Standard 802.11u-2011 (Amendment 9: 1045 Interworking with External Networks)", 2011. 1047 [RFC3315] Droms, R., Bound, J., Volz, B., Lemon, T., Perkins, C. and 1048 M. Carney, "Dynamic Host Configuration Protocol for IPv6 1049 (DHCPv6)", RFC 3315, July 2003. 1051 [RFC3971] Arkko, J., Kempf, J., Zill, B. and P. Nikander, "SEcure 1052 Neighbor Discovery (SEND)", RFC 3971, March 2005. 1054 [RFC4861] Narten, T., Nordmark, E., Simpson, W. and H. Soliman, 1055 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 1056 September 2007. 1058 [RFC5739] Eronen, P., Laganier, J. and C. Madson, "IPv6 1059 Configuration in Internet Key Exchange Protocol Version 2 1060 (IKEv2)", RFC 5739, February 2010. 1062 [RFC5996] Kaufman, C., Hoffman, P., Nir, Y. and P. Eronen, "Internet 1063 Key Exchange Protocol Version 2 (IKEv2)", RFC 5996, 1064 September 2010. 1066 [RFC6182] Ford, A., Raiciu, C., Handley, M., Barre, S. and J. 1067 Iyengar, "Architectural Guidelines for Multipath TCP 1068 Development", RFC 6182, March 2011. 1070 [RFC6418] Blanchet, M. and P. Seite, "Multiple Interfaces and 1071 Provisioning Domains Problem Statement", RFC 6418, 1072 November 2011. 1074 [RFC6419] Wasserman, M. and P. Seite, "Current Practices for 1075 Multiple-Interface Hosts", RFC 6419, November 2011. 1077 [RFC6555] Wing, D. and A. Yourtchenko, "Happy Eyeballs: Success with 1078 Dual-Stack Hosts", RFC 6555, April 2012. 1080 [RFC6724] Thaler, D., Draves, R., Matsumoto, A. and T. Chown, 1081 "Default Address Selection for Internet Protocol Version 6 1082 (IPv6)", RFC 6724, September 2012. 1084 [RFC6731] Savolainen, T., Kato, J. and T. Lemon, "Improved Recursive 1085 DNS Server Selection for Multi-Interfaced Nodes", RFC 1086 6731, December 2012. 1088 [RFC7078] Matsumoto, A., Fujisaki, T. and T. Chown, "Distributing 1089 Address Selection Policy Using DHCPv6", RFC 7078, January 1090 2014. 1092 [TS23402] 3GPP, "3GPP TS 23.402; Architecture enhancements for non- 1093 3GPP accesses; release 12", 2014. 1095 Author's Address 1096 Dmitry Anipko, editor 1097 Unaffiliated 1099 Phone: +1 425 442 6356 1100 Email: dmitry.anipko@gmail.com