idnits 2.17.1 draft-ietf-mif-problem-statement-12.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 (April 12, 2011) is 4763 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Unused Reference: 'RFC4477' is defined on line 848, but no explicit reference was found in the text == Outdated reference: A later version (-12) exists of draft-ietf-mif-current-practices-09 == Outdated reference: A later version (-05) exists of draft-ietf-mif-dhcpv6-route-option-01 == Outdated reference: A later version (-12) exists of draft-ietf-mif-dns-server-selection-01 == Outdated reference: A later version (-14) exists of draft-ietf-netext-logical-interface-support-02 -- Obsolete informational reference (is this intentional?): RFC 3315 (Obsoleted by RFC 8415) -- Obsolete informational reference (is this intentional?): RFC 3484 (Obsoleted by RFC 6724) -- Obsolete informational reference (is this intentional?): RFC 3775 (Obsoleted by RFC 6275) -- Obsolete informational reference (is this intentional?): RFC 4294 (Obsoleted by RFC 6434) -- Obsolete informational reference (is this intentional?): RFC 4960 (Obsoleted by RFC 9260) -- Obsolete informational reference (is this intentional?): RFC 5206 (Obsoleted by RFC 8046) -- Obsolete informational reference (is this intentional?): RFC 5245 (Obsoleted by RFC 8445, RFC 8839) Summary: 0 errors (**), 0 flaws (~~), 6 warnings (==), 8 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group M. Blanchet 3 Internet-Draft Viagenie 4 Intended status: Informational P. Seite 5 Expires: October 14, 2011 France Telecom - Orange 6 April 12, 2011 8 Multiple Interfaces and Provisioning Domains Problem Statement 9 draft-ietf-mif-problem-statement-12.txt 11 Abstract 13 This document describes issues encountered by a node attached to 14 multiple provisioning domains. This node receives configuration 15 information from each of its provisioning domains where some 16 configuration objects are global to the node, others are local to the 17 interface. Issues such as selecting the wrong interface to send 18 trafic happen when conflicting node-scoped configuration objects are 19 received and inappropriately used. Moreover, other issues are the 20 result of simulatenous attachment to multiple networks, such as 21 domain selection or addressing and naming space overlaps, regardless 22 of the provisioning mechanism. While multiple provisioning domains 23 are typically seen on nodes with multiple interfaces, this document 24 also discusses single interface nodes situation. 26 Status of this Memo 28 This Internet-Draft is submitted in full conformance with the 29 provisions of BCP 78 and BCP 79. 31 Internet-Drafts are working documents of the Internet Engineering 32 Task Force (IETF). Note that other groups may also distribute 33 working documents as Internet-Drafts. The list of current Internet- 34 Drafts is at http://datatracker.ietf.org/drafts/current/. 36 Internet-Drafts are draft documents valid for a maximum of six months 37 and may be updated, replaced, or obsoleted by other documents at any 38 time. It is inappropriate to use Internet-Drafts as reference 39 material or to cite them other than as "work in progress." 41 This Internet-Draft will expire on October 14, 2011. 43 Copyright Notice 45 Copyright (c) 2011 IETF Trust and the persons identified as the 46 document authors. All rights reserved. 48 This document is subject to BCP 78 and the IETF Trust's Legal 49 Provisions Relating to IETF Documents 50 (http://trustee.ietf.org/license-info) in effect on the date of 51 publication of this document. Please review these documents 52 carefully, as they describe your rights and restrictions with respect 53 to this document. Code Components extracted from this document must 54 include Simplified BSD License text as described in Section 4.e of 55 the Trust Legal Provisions and are provided without warranty as 56 described in the Simplified BSD License. 58 Table of Contents 60 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 61 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3 62 3. Scope and Existing Work . . . . . . . . . . . . . . . . . . . 4 63 3.1. Below IP Interaction . . . . . . . . . . . . . . . . . . . 4 64 3.2. MIF node Characterization . . . . . . . . . . . . . . . . 4 65 3.3. Hosts Requirements . . . . . . . . . . . . . . . . . . . . 5 66 3.4. Mobility and other IP protocols . . . . . . . . . . . . . 6 67 3.5. Address Selection . . . . . . . . . . . . . . . . . . . . 6 68 3.6. Finding and Sharing IP Addresses with Peers . . . . . . . 6 69 3.7. Interface and Provisioning domain selection . . . . . . . 7 70 3.8. Session management . . . . . . . . . . . . . . . . . . . . 7 71 3.9. Socket API . . . . . . . . . . . . . . . . . . . . . . . . 8 72 4. MIF Issues . . . . . . . . . . . . . . . . . . . . . . . . . . 9 73 4.1. DNS resolution issues . . . . . . . . . . . . . . . . . . 9 74 4.2. Node Routing . . . . . . . . . . . . . . . . . . . . . . . 11 75 4.3. Policies conflict . . . . . . . . . . . . . . . . . . . . 12 76 4.4. Session management . . . . . . . . . . . . . . . . . . . . 12 77 4.5. Single Interface on Multiple Provisioning Domains . . . . 13 78 5. Underlying problems and causes . . . . . . . . . . . . . . . . 13 79 6. Security Considerations . . . . . . . . . . . . . . . . . . . 15 80 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 16 81 8. Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 82 9. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 16 83 10. Informative References . . . . . . . . . . . . . . . . . . . . 16 84 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 20 86 1. Introduction 88 A multihomed node may have multiple provisioning domains (via 89 physical and/or virtual interfaces). For example, a node may be 90 simultaneously connected to a wired Ethernet LAN, a 802.11 LAN, a 3G 91 cell network, one or multiple VPN connections or one or multiple 92 tunnels(automatic or manual). Current laptops and smartphones 93 typically have multiple access network interfaces and, thus, are 94 often connected to different provisioning domains. 96 A multihomed node receives configuration information from each of its 97 attached networks, through various mechanisms such as DHCPv4 98 [RFC2131], DHCPv6 [RFC3315], PPP [RFC1661] and IPv6 Router 99 Advertisements [RFC4861]. Some received configuration objects are 100 specific to an interface such as the IP address and the link prefix. 101 Others are typically considered by implementations as being global to 102 the node, such as the routing information (e.g. default gateway), DNS 103 servers IP addresses, and address selection policies, herein named 104 "node-scoped". 106 When the received node-scoped configuration objects have different 107 values from each provisioning domains, such as different DNS servers 108 IP addresses, different default gateways or different address 109 selection policies, the node has to decide which one to use or how it 110 will merge them. 112 Other issues are the result of simulatenous attachment to multiple 113 networks, such as addressing and naming space overlaps, regardless of 114 the provisioning mechanism. 116 The following sections define the multiple interfaces (MIF) node, the 117 scope of this work, describe related work, list issues and then 118 summarize the underlying problems. 120 A companion document [I-D.ietf-mif-current-practices] discusses some 121 current practices of various implementations dealing with MIF. 123 2. Terminology 125 Administrative domain 127 A group of hosts, routers, and networks operated and managed by a 128 single organization [RFC1136]. 130 Provisioning domain 131 A set of consistent configuration information (e.g. Default 132 router, Network prefixes, DNS,...). One administrative domain may 133 have multiple provisioning domains. 135 Reference to IP version 137 When a protocol keyword such as IP, PPP, DHCP is used in this 138 document without any reference to a specific IP version, then it 139 implies both IPv4 and IPv6. A specific IP version keyword such as 140 DHCPv4 or DHCPv6 is meant to be specific to that IP version. 142 3. Scope and Existing Work 144 This section describes existing related work and defines the scope of 145 the problem. 147 3.1. Below IP Interaction 149 Some types of interfaces have link layer characteristics which may be 150 used in determining how multiple provisioning domain issues will be 151 dealt with. For instance, link layers may have authentication and 152 encryption characteristics which could be used as criteria for 153 interface selection. However, network discovery and selection on 154 lower layers as defined by [RFC5113] is out of scope of this 155 document. Moreover, interoperability with lower layer mechanisms 156 such as services defined in IEEE 802.21, which aims at facilitating 157 handover between heterogeneous networks [MIH], is also out of scope. 159 Some mechanisms (e.g., based on a virtual IP interface) 160 allow sharing a single IP address over multiple 161 interfaces to networks with disparate access technologies. From the 162 IP stack view on the node, there is only a single interface and 163 single IP address. Therefore, this situation is out of scope of this 164 current problem statement. Furthermore, link aggregation done under 165 IP where a single interface is shown to the IP stack is also out of 166 scope. 168 3.2. MIF node Characterization 170 A MIF node has the following characteristics: 172 o A [RFC1122] IPv4 and/or [RFC4294] IPv6 compliant node 173 o A MIF node is configured with more than one IP addresses 174 (excluding loopback and link-local) 175 o A MIF node can attach to more than one provisioning domains, as 176 presented to the IP stack. 178 o The interfaces may be virtual or physical. 179 o Configuration objects come from one or more administrative 180 domains. 181 o The IP addresses may be from the same or from different address 182 families, such as IPv4 and IPv6. 183 o Communications using these IP addresses may happen simultaneously 184 and independently. 185 o Some communications using these IP addresses are possible on all 186 the provisioning domains, while some are only possible on a 187 smaller set of the provisioning domains. 188 o While the MIF node may forward packets between its interfaces, 189 forwarding packets is not taken into account in this definition 190 and is out of scope for this document. 192 3.3. Hosts Requirements 194 The requirements for Internet Hosts [RFC1122] describe the multihomed 195 node as if it has multiple IP addresses, which may be associated with 196 one or more physical interfaces connected to the same or different 197 networks. 199 The requirements states that The node maintains a route cache table 200 where each entry contains the local IP address, the destination IP 201 address, Differentiated Services Code Point and Next-hop gateway IP 202 address. The route cache entry would have data about the properties 203 of the path, such as the average round-trip delay measured by a 204 transport protocol. Nowadays, implementations are not caching these 205 informations. 207 [RFC1122] defines two host models: 208 o The "Strong" host model defines a multihomed host as a set of 209 logical hosts within the same physical host. In this model a 210 packet must be sent on an interface that corresponds to the source 211 address of that packet. 212 o The "Weak" host model describes a host that has some embedded 213 gateway functionality. In the weak host model, the host can send 214 and receive packets on any interface. 216 The multihomed node computes routes for outgoing datagrams 217 differently depending on the model. Under the strong model, the 218 route is computed based on the source IP address, the destination IP 219 address and the Differentiated Services Code Point. Under the weak 220 model, the source IP address is not used, but only the destination IP 221 address and the Differentiated Services Code Point. 223 3.4. Mobility and other IP protocols 225 The scope of this document is only about nodes implementing [RFC1122] 226 for IPv4 and [RFC4294] for IPv6 without additional features or 227 special-purpose support for transport layers, mobility, multi-homing, 228 or identifier-locator split mechanisms. Dealing with multiple 229 interfaces with such mechanisms is related but considered as a 230 separate problem and is under active study elsewhere in the IETF 231 [RFC4960], [RFC5206], [RFC5533], [RFC5648], 232 [I-D.ietf-mptcp-architecture]. 234 When an application is using one interface while another interface 235 with better characteristics becomes available, the ongoing 236 application session could be transferred to the newly enabled 237 interface. However, in some cases, the ongoing session shall be kept 238 on the current interface while initiating the new sessions on the new 239 interface. The problem of the interface selection is within the MIF 240 scope and may leverage specific node functions (Section 3.8). 241 However, if transfer of IP session is required, IP mobility 242 mechanisms, such as [RFC3775], shall be used. 244 3.5. Address Selection 246 The Default Address Selection specification [RFC3484] defines 247 algorithms for source and destination IP address selections. It is 248 mandatory to be implemented in IPv6 nodes, which also means dual- 249 stack nodes. A node-scoped policy table managed by the IP stack is 250 defined. Mechanisms to update the policy table are being defined 251 [I-D.ietf-6man-addr-select-sol] to update the policy table. 253 Issues on using the Default Address Selection were found in [RFC5220] 254 and [RFC5221] in the context of multiple prefixes on the same link. 256 3.6. Finding and Sharing IP Addresses with Peers 258 Interactive Connectivity Establishment (ICE [RFC5245]) is a technique 259 for NAT traversal for UDP-based (and TCP) media streams established 260 by the offer/answer model. The multiplicity of IP addresses, ports 261 and transport in SDP offers are tested for connectivity by peer-to- 262 peer connectivity checks. The result is candidate IP addresses and 263 ports for establishing a connection with the other peer. However, 264 ICE does not solve issues when incompatible configuration objects are 265 received on different interfaces. 267 Some application protocols do referrals of IP addresses, port numbers 268 and transport for further exchanges. For instance, applications can 269 provide reachability information to itself or to a third party. The 270 general problem of referrals is related to the multiple interface 271 problem, since, in this context, referrals must provide consistent 272 information depending on which provisioning domain is used. 273 Referrals are discussed in [I-D.carpenter-referral-ps] and 274 [I-D.ietf-shim6-app-refer]. 276 3.7. Interface and Provisioning domain selection 278 In a MIF context, the node may handle simultaneously multiple domains 279 with disparate characteristics, especially when supporting multiple 280 access technologies. Selection is simple if the application is 281 restricted to one specific provisioning domain: the application must 282 start on the default provisioning domain if available, otherwise the 283 application does not start. However, if the application can be run 284 on several provisioning domains, the selection problem can be 285 difficult. 287 For example, the interface selection mechanism defined in [TS23.234] 288 uses the following information (non exhaustive list): 290 o preferences provided by the user, 291 o policies provided by network operator, 292 o quality of the radio link, 293 o network resource considerations (e.g. available QoS, IP 294 connectivity check,...), 295 o the application QoS requirements in order to map applications to 296 the best interface 298 However, [TS23.234] is designed for a specific multiple-interfaces 299 use-case. A generic way to handle these characteristics is yet to be 300 defined. 302 3.8. Session management 304 Some implementations, specially in the mobile world, rely on higher- 305 level session manager, also named connection manager, to deal with 306 issues brought by simultaneous attachment to multiple provisioning 307 domains. Typically, the session manager may deal with the selection 308 of the interface, and/or the provisioning domain, on behalf to the 309 applications, or tackle with complex issues such as policies conflict 310 resolution (Section 4.3). As discussed previously in Section 3.7, 311 the session manager may encounter difficulties because of multiple 312 and diverse criteria. 314 Session managers usually leverage the link-layer interface to gather 315 information (e.g lower layer authentication and encryption methods, 316 see Section 3.1) and/or for control purpose. Such link-layer 317 interface may not provide all required services to make a proper 318 decision (e.g. interface selection). Some OS, or terminals, already 319 implement session managers [I-D.ietf-mif-current-practices] and 320 vendor-specific platforms sometimes provides specific socket API 321 (Section 3.9) a session manager can use. However, the generic 322 architecture of a session manager and its associated API are not 323 currently standardized, so session management behavior may differ 324 between OS and platforms. 326 Multiple interfaces management sometimes relies on a virtual 327 interface. For instance, virtual interface allows to support multi- 328 homing, inter-technology handovers and IP flow mobility in a Proxy 329 Mobile IPv6 network [I-D.ietf-netext-logical-interface-support]. 330 This virtual interface allows a multiple-interfaces node sharing a 331 set of IP addresses on multiple physical interfaces and can also add 332 benefits to multi-access scenarios such as 3GPP Multi Access PDN 333 Connectivity [TS23.402]. In most cases, the virtual interface will 334 map several physical network interfaces and the session manager 335 should control both, the configuration of each one of these virtual 336 and physical interfaces, as well as the mapping between the virtual 337 and the sub-interfaces. 339 In multiple interfaces situation, active application sessions should 340 survive to path failures. Here, the session manager may come into 341 play but only relying on existing mechanisms to manage multipath 342 (MPTCP [I-D.ietf-mptcp-architecture]) or failover (MIP6 [RFC3775], 343 SHIM6 [RFC5533]). Description of interaction between these 344 mechanisms and the session manager is out of the scope of this 345 document. 347 3.9. Socket API 349 An Application Programming Interface (API) may expose objects that 350 user applications, or session managers, use for dealing with multiple 351 interfaces. For example, [RFC3542] defines how an application using 352 the Advanced sockets API specifies the interface or the source IP 353 address, through a simple bind() operation or with the IPV6_PKTINFO 354 socket option. 356 Other APIs have been defined to solve similar issues to MIF. For 357 instance, [RFC5014] defines an API to influence the default address 358 selection mechanism by specifying attributes of the source addresses 359 it prefers. [I-D.ietf-shim6-multihome-shim-api] gives another 360 example, in a multihoming context, by defining a socket API enabling 361 interactions between applications and the multihoming shim layer for 362 advanced locator management, and access to information about failure 363 detection and path exploration. 365 4. MIF Issues 367 This section describes the various issues when using a MIF node that 368 has already received configuration objects from its various 369 provisioning domains or when multiple interfaces are used and results 370 in wrong domain selection, addressing or naming space overlaps. They 371 occur, for example, when: 373 1. one interface is on the Internet and one is on a corporate 374 private network. The latter may be through VPN. 375 2. one interface is on one access network (i.e. wifi) and the other 376 one is on another access network (3G) with specific services. 378 4.1. DNS resolution issues 380 A MIF node (M1) has an active interface(I1) connected to a network 381 (N1) which has its DNS server (S1) and another active interface (I2) 382 connected to a network (N2) which has its DNS server (S2). S1 serves 383 with some private namespace "private.example.com". The user or the 384 application uses a name "a.private.example.com" which is within the 385 private namespace of S1 and only resolvable by S1. Any of the 386 following situations may occur: 388 1. M1 stack, based on its routing table, uses I2 to reach S1 to 389 resolve "a.private.example.com". M1 never reaches S1. The name 390 is not resolved. 391 2. M1 keeps only one set of DNS server addresses from the received 392 configuration objects and kept S2 address. M1 sends the forward 393 DNS query for a.private.example.com to S2. S2 responds with an 394 error for an non-existent domain (NXDOMAIN). The name is not 395 resolved. This issue also arises when performing reverse DNS 396 lookup. In the same situation, the reverse DNS query fails. 397 3. M1 keeps only one set of DNS server addresses from the received 398 configuration objects and kept S2 address. M1 sends the DNS 399 query for a.private.example.com to S2. S2 asks its upstream DNS 400 and gets an IP address for a.private.example.com. However, the 401 IP address is not the same one S1 would have given. Therefore, 402 the application tries to connect to the wrong destination node, 403 or to the wrong interface of the latter, which may imply security 404 issues or result in lack of service. 405 4. S1 or S2 has been used to resolve "a.private.example.com" to an 406 [RFC1918] address. Both N1 and N2 are [RFC1918] addressed 407 networks. If addresses overlap, traffic may be sent using the 408 wrong interface. This issue is not related to receiving multiple 409 configuration objects, but to an address overlap between 410 interfaces or attaching networks. 412 5. M1 has resolved an FQDN to locally valid IP address when 413 connected to N1. If the node looses connection to N1, the node 414 may try to connect, via N2, to the same IP address as earlier, 415 but as the address was only locally valid, connection setup 416 fails. Similarly, M1 may have received NXDOMAIN for an FQDN when 417 connected to N1. After detachment from N1, the node should not 418 assume the FQDN continues to be nonexistent on N2. 419 6. M1 requests AAAA record from a DNS server on a network that uses 420 protocol translators and DNS64 [I-D.ietf-behave-dns64]. If the 421 M1 receives synthesized AAAA record, it is guaranteed to be valid 422 only on the network it was learned from. If the M1 uses 423 synthesized AAAA on any other network interface, traffic may be 424 lost, dropped or forwarded to the wrong network. 426 Some networks requires the user to authenticate on a captive web 427 portal before providing Internet connectivity. This may lead to 428 specific DNS resolution issues. Consider a MIF node (M1) with an 429 active interface(I1) connected to a network (N1), which has its DNS 430 server (S1), and another active interface (I2) connected to a network 431 (N2), which has its DNS server (S2). Until the user has not 432 authenticated, S1 is configured to respond to any A or AAAA record 433 query with the IP address of a captive portal, so as to redirect web 434 browsers to an access control portal web page. This captive portal 435 can be reached only via I1. When the user has authenticated to the 436 captive portal, M1 can resolve an FQDN when connected to N1. 437 However, if the address is only locally valid on N1, any of the issue 438 described above may occur. When the user has not authenticated, any 439 of the following situations may occur: 441 1. M1 keeps only one set of DNS server addresses from the received 442 configuration objects and kept S2 address. M1 sends the forward 443 DNS query for a.example.com to S2. S2 responds with the correct 444 answer, R1. M1 attempts to contact R1 by way of I1. The 445 connection fails. Or, the connection succeeds, bypassing the 446 security policy on N1, possibly exposing the owner of M1 to 447 prosecution. 448 2. M1 keeps only one set of DNS server addresses from the received 449 configuration objects and kept S1 address. M1 sends the DNS 450 query for a.example.com to S1. S1 provides the address of its 451 captive portal. M1 attempts to contact this IP address using I1. 452 The application fails to connect, resulting in lack of service. 453 Or, the application succeeds in connecting, but connects to the 454 captive portal rather than the intended destination, resulting in 455 lack of service (i.e. IP connectivity check issue described in 456 Section 4.4). 458 4.2. Node Routing 460 A MIF node (M1) has an active interface(I1) connected to a network 461 (N1) and another active interface (I2) connected to a network (N2). 462 The user or the application is trying to reach an IP address (IP1). 463 Any of the following situations may occur: 465 1. For IP1, M1 has one default route (R1) via network (N1). To 466 reach IP1, M1 stack uses R1 and sends through I1. If IP1 is only 467 reachable by N2, IP1 is never reached or is not the right target. 468 2. For the IP1 address family, M1 has one default route (R1, R2) per 469 network (N1, N2). IP1 is reachable by both networks, but N2 path 470 has better characteristics, such as better round-trip time, least 471 cost, better bandwidth, etc.... These preferences could be 472 defined by user, provisioned by the network operator, or else. 473 M1 stack uses R1 and tries to send through I1. IP1 is reached 474 but the service would be better by I2. 475 3. For the IP1 address family, M1 has a default route (R1), a 476 specific X.0.0.0/8 route R1B (for example but not restricted to 477 RFC1918 prefix) to N1 and a default route (R2) to N2. IP1 is 478 reachable by N2 only, but the prefix (X.0.0.0/8) is used in both 479 networks. Because of the most specific route R1B, M1 stack sends 480 through I2 and never reach the target. 482 A MIF node may have multiple routes to a destination. However, by 483 default, it does not have any hint concerning which interface would 484 be the best to use for that destination. The first-hop selection may 485 leverage on local routing policy, allowing some actors (e.g. network 486 operator or service provider) to influence the routing table, i.e. 487 make decision regarding which interface to use. For instance, a user 488 on such multihomed node might want a local policy to influence which 489 interface will be used based on various conditions. Some SDOs have 490 defined policy-based routing selection mechanisms. For instance, the 491 Access Network Discovery and Selection Function (ANDSF) [TS23.402] 492 provides inter-systems routing policies to terminals with both a 3GPP 493 and non-3GPP interfaces. However, the routing selection may still be 494 difficult, due to disjoint criteria as discussed in Section 3.8. 495 Moreover, information required to make the right decision may not be 496 available. For instance, interfaces to lower layer may not provide 497 all required hints to the selection (e.g. information on interface 498 quality). 500 A node usually has a node-scoped routing table. However, a MIF node 501 is connected to multiple provisioning domains; if each of these 502 domains pushes routing policies to the node, then conflicts between 503 policies may happen and the node has no easy way to merge or 504 reconciliate them. 506 On a MIF node, some source addresses are not valid if used on some 507 interfaces. For example, an RFC1918 source address might be 508 appropriate on the VPN interface but not on the public interface of 509 the MIF node. If the source address is not chosen appropriately, 510 then packets may be filtered in the path if source address filtering 511 is in place ([RFC2827], [RFC3704]) and reply packets may never come 512 back to the source. 514 4.3. Policies conflict 516 The distribution of configuration policies (e.g. address selection, 517 routing, DNS selection...) to end nodes is being discussed (e.g. 518 ANDSF in [TS23.402], [I-D.ietf-mif-dhcpv6-route-option]). If 519 implemented in multiple provisioning domains, such mechanisms may 520 conflict and bring issues to the multihomed node. Considering a MIF 521 node (M1) with an active interface(I1) connected to a network (N1) 522 and another active interface (I2) connected to a network (N2), the 523 following conflicts may occur: 524 1. M1 receives from both networks (N1 and N2) an update of its 525 default address selection policy. However, the policies are 526 specific to each network. The policies are merged by M1 stack. 527 Based on the merged policy, the chosen source address is from N1 528 but packets are sent to N2. The source address is not reachable 529 from N2, therefore the return packet is lost. Merging address 530 selection policies may have important impacts on routing. 531 2. A node usually has a node-scoped routing table. However, each of 532 the connected provisioning domains (N1 and N2) may push routing 533 policies to the node, then conflicts between policies may happen 534 and the node has no easy way to merge or reconciliate them. 535 3. M1 receives from one of the network an update of its access 536 selection policy, e.g. via the 3GPP/ANDSF [TS23.402]. However, 537 the policy is in conflict with the local policy (e.g. user 538 defined, or default OS policy). Assuming that the network 539 provides list of overloaded access network, if the policy sent by 540 the network is ignored, packet may be sent to an access network 541 with poor quality of communication. 543 4.4. Session management 545 Consider that a node has selected an interface and managed to 546 configure it (i.e. the node obtained a valid IP address from the 547 network). However, the Internet connectivity is not available. The 548 problem could be due to the following reasons: 549 1. The network requires a web-based authentication (e.g. the access 550 network is a WiFi Hot Spot). 551 2. IP interface is configured active but layer 2 is so poor (e.g. 552 poor radio condition) that no layer 3 traffic can succeed. 554 In this situation, the session management should be able to perform 555 IP connectivity checks before selecting an interface. 557 Session issues may also arise when the node discovers a new 558 provisioning domain. Consider a MIF node (M1) has an active 559 interface(I1) connected to a network (N1) where an application is 560 running a TCP session. A new network (N2) becomes available. If N2 561 is selected (e.g. because of better quality of communication), M1 562 gets IP connectivity to N2 and updates the routing table priority. 563 So, if no specific route to the correspondent node and if the node 564 implements the weak host model [RFC1122], the TCP connection breaks 565 as next hop changes. In order to continue communicating with the 566 correspondent node, M1 should try to re-connect the server via N2. 567 In some situation, it could be preferable to maintain current 568 sessions on N1 while new sessions start on N2. 570 4.5. Single Interface on Multiple Provisioning Domains 572 When a node using a single interface is connected to multiple 573 networks, such as different default routers, similar issues as 574 described above happen. Even with a single interface, a node may 575 wish to connect to more than one provisioning domain: that node may 576 use more than one IP source address and may have more than one 577 default router. The node may want to access services that can only 578 be reached using one of the provisioning domain. In this case, it 579 needs to use the right outgoing source address and default gateway to 580 reach that service. In this situation, that node may also need to 581 use different DNS servers to get domain names in those different 582 provisioning domains. 584 5. Underlying problems and causes 586 This section lists the underlying problems, and their causes, which 587 lead to the issues discussed in the previous section. The problems 588 can be divided into five categories: 1) Configuration 2) DNS 589 resolution 3) Routing 4) Address selection and 5) session management 590 and API. They are shown as below: 592 1. Configuration. In a MIF context, configuration information 593 specific to a provisioning domain may be ignored because: 594 1. Configuration objects (e.g. DNS servers, NTP servers, ...) 595 are node-scoped. So the IP stack is not able to maintain the 596 mapping between information and corresponding provisioning 597 domain. 598 2. Same configuration objects (e.g. DNS server addresses, NTP 599 server addresses, ..) received from multiple provisioning 600 domains may be overwritten. 602 3. Host implementations usually do not keep separate network 603 configuration (such as DNS server addresses) per provisioning 604 domain. 605 2. DNS resolution 606 1. Some FQDN can be resolvable only by sending queries to the 607 right server (e.g. intranet services). However, DNS query 608 could be sent to the wrong interface because DNS server 609 addresses may be node-scoped. 610 2. A DNS answer may be only valid on a specific provisioning 611 domain but applications may not be aware of that mapping 612 because DNS answers may not be kept with the provisioning 613 from which the answer comes from. 614 3. Routing 615 1. In the MIF context, routing information could be specific to 616 each interface. This could lead to routing issue because, in 617 current node implementations, routing tables are node-scoped. 618 2. Current node implementations do not take into account the 619 Differentiated Services Code Point or path characteristics in 620 the routing table. 621 3. Even if implementations take into account path 622 characteristics, the node has no way to properly merge or 623 reconciliate the provisioning domain preferences. 624 4. a node attached to multiple provisioning domain could be 625 provided with incompatible selection policies. If the 626 different actors (e.g. user and network operator) are allowed 627 to provide their own policies, the node has no way to 628 properly merge or reconciliate multiple selection policies. 629 5. The problem of first hop selection could not be solved via 630 configuration (Section 3.7), and may leverage on 631 sophisticated and specific mechanisms (Section 3.8). 632 4. Address selection 633 1. Default Address Selection policies may be specific to their 634 corresponding provisioning domain. However, a MIF node may 635 not be able to manage per-provisioning domain address 636 selection policies because default Address Selection policy 637 is node-scoped. 638 2. On a MIF node, some source addresses are not valid if used on 639 some interfaces or even on some default routers on the same 640 interface. In this situation, the source address should be 641 taken into account in the routing table; but current node 642 implementations do not support such a feature. 643 3. Source address or address selection policies could be 644 specified by applications. However, there is no advanced 645 APIs to allow applications realizing such operations. 646 5. Session management and API 647 1. Some implementations, specially in the mobile world, have 648 higher-level API and/or session manager (aka connection 649 manager) to address MIF issues. These mechanisms are not 650 standardized and do not necessarily behave the same way 651 across different OS, and/or platforms, in the presence of the 652 MIF problems. This lack of consistency is an issue for user 653 and operator who could experience different session manager 654 behaviors depending on the terminal. 655 2. Session managers usually leverage on interface to link layer 656 to gather information (e.g lower layer authentication and 657 encryption methods) and/or for control purpose. However, 658 such link layer interface may not provide all required 659 services (e.g. may not provide all information allowing to 660 make a proper interface selection). 661 3. A MIF node can support different session managers, which may 662 have contradictory ways to solve the MIF issues. For 663 instance, because of different selection algorithms, two 664 different session managers could select different domains in 665 a same context. Or, when dealing with different domain 666 selection policies, a session manager may give precedence to 667 user policy while another could favor mobile operator policy. 668 4. When host routing is updated and if weak host model is 669 supported, ongoing TCP sessions may break if routes changes 670 for these sessions. When TCP sessions should be bound to the 671 interface, the strong host model should be used. 672 5. When provided by different actors (e.g. user, network, 673 default-OS), policies may conflict and, thus, need to be 674 reconciliated at the host level. Policy conflict resolution 675 may impact other functions (e.g. naming, routing). 676 6. Even if the node has managed to configure an interface, 677 Internet connectivity could be not available. It could be 678 due to an access control function coming into play above the 679 layer 3, or because of poor layer 2 conditions. IP 680 connectivity check should be performed before selecting an 681 interface. 683 6. Security Considerations 685 The problems discussed in this document have security implications, 686 such as when the packets sent on the wrong interface might be leaking 687 some confidential information. Configuration parameters from one 688 provisioning domain could cause a denial of service on another 689 provisioning domain (e.g. DNS issues). Moreover, the undetermined 690 behavior of IP stacks in the multihomed context bring additional 691 threats where an interface on a multihomed node might be used to 692 conduct attacks targeted to the networks connected by the other 693 interfaces.corrupted provisioning domain selection policy may induce 694 a node to make decisions causing certain traffic to be forwarded to 695 the attacker. 697 Additional security concerns are raised by possible future mechanisms 698 that provide additional information to the node so that it can make a 699 more intelligent decision with regards to the issues discussed in 700 this document. Such future mechanisms may themselves be vulnerable 701 and may not be easy to protect in the general case. 703 7. IANA Considerations 705 This document has no actions for IANA. 707 8. Authors 709 This document is a joint effort with authors of the MIF requirements 710 draft [I-D.yang-mif-req]. The authors of this document, in 711 alphabetical order, include: Marc Blanchet, Jacqni Qin, Pierrick 712 Seite, Carl Williams and Peny Yang. 714 9. Acknowledgements 716 The initial Internet-Drafts prior to the MIF working group and the 717 discussions during the MIF BOF meeting and on the mailing list around 718 the MIF charter scope on the mailing list brought very good input to 719 the problem statement. This draft steals a lot of text from these 720 discussions and initial drafts (e.g. [I-D.yang-mif-req], 721 [I-D.hui-ip-multiple-connections-ps], 722 [I-D.ietf-mif-dns-server-selection]). Therefore, the editor would 723 like to acknowledge the following people (in no specific order), from 724 which some text has been taken from: Jari Arkko, Keith Moore, Sam 725 Hartman, George Tsirtsis, Scott Brim, Ted Lemon, Bernie Volz, Giyeong 726 Son, Gabriel Montenegro, Julien Laganier, Teemu Savolainen, Christian 727 Vogt, Lars Eggert, Margaret Wasserman, Hui Deng, Ralph Droms, Ted 728 Hardie, Christian Huitema, Remi Denis-Courmont, Alexandru Petrescu, 729 Zhen Cao, Gaetan Feige, Telemaco Melia and Juan-Carlos Zuniga. Sorry 730 if some contributors have not been named. 732 10. Informative References 734 [I-D.carpenter-referral-ps] 735 Carpenter, B., Jiang, S., and Z. Cao, "Problem Statement 736 for Referral", draft-carpenter-referral-ps-02 (work in 737 progress), February 2011. 739 [I-D.hui-ip-multiple-connections-ps] 740 Hui, M. and H. Deng, "Problem Statement and Requirement of 741 Simple IP Multi-homing of the Host", 742 draft-hui-ip-multiple-connections-ps-02 (work in 743 progress), March 2009. 745 [I-D.ietf-6man-addr-select-sol] 746 Matsumoto, A., Fujisaki, T., and R. Hiromi, "Solution 747 approaches for address-selection problems", 748 draft-ietf-6man-addr-select-sol-03 (work in progress), 749 March 2010. 751 [I-D.ietf-behave-dns64] 752 Bagnulo, M., Sullivan, A., Matthews, P., and I. Beijnum, 753 "DNS64: DNS extensions for Network Address Translation 754 from IPv6 Clients to IPv4 Servers", 755 draft-ietf-behave-dns64-11 (work in progress), 756 October 2010. 758 [I-D.ietf-mif-current-practices] 759 Wasserman, M. and P. Seite, "Current Practices for 760 Multiple Interface Hosts", 761 draft-ietf-mif-current-practices-09 (work in progress), 762 March 2011. 764 [I-D.ietf-mif-dhcpv6-route-option] 765 Dec, W., Mrugalski, T., Sun, T., and B. Sarikaya, "DHCPv6 766 Route Option", draft-ietf-mif-dhcpv6-route-option-01 (work 767 in progress), March 2011. 769 [I-D.ietf-mif-dns-server-selection] 770 Savolainen, T. and J. Kato, "Improved DNS Server Selection 771 for Multi-Homed Nodes", 772 draft-ietf-mif-dns-server-selection-01 (work in progress), 773 March 2011. 775 [I-D.ietf-mptcp-architecture] 776 Ford, A., Raiciu, C., Handley, M., Barre, S., and J. 777 Iyengar, "Architectural Guidelines for Multipath TCP 778 Development", draft-ietf-mptcp-architecture-05 (work in 779 progress), January 2011. 781 [I-D.ietf-netext-logical-interface-support] 782 Melia, T. and S. Gundavelli, "Logical Interface Support 783 for multi-mode IP Hosts", 784 draft-ietf-netext-logical-interface-support-02 (work in 785 progress), March 2011. 787 [I-D.ietf-shim6-app-refer] 788 Nordmark, E., "Shim6 Application Referral Issues", 789 draft-ietf-shim6-app-refer-00 (work in progress), 790 July 2005. 792 [I-D.ietf-shim6-multihome-shim-api] 793 Komu, M., Bagnulo, M., Slavov, K., and S. Sugimoto, 794 "Socket Application Program Interface (API) for 795 Multihoming Shim", draft-ietf-shim6-multihome-shim-api-17 796 (work in progress), April 2011. 798 [I-D.yang-mif-req] 799 Yang, P., Seite, P., Williams, C., and J. Qin, 800 "Requirements on multiple Interface (MIF) of simple IP", 801 draft-yang-mif-req-00 (work in progress), March 2009. 803 [MIH] IEEE, "IEEE Standard for Local and Metropolitan Area 804 Networks - Part 21: Media Independent Handover Services, 805 IEEE LAN/MAN Std 802.21-2008, January 2009.", 2010. 807 [RFC1122] Braden, R., "Requirements for Internet Hosts - 808 Communication Layers", STD 3, RFC 1122, October 1989. 810 [RFC1136] Hares, S. and D. Katz, "Administrative Domains and Routing 811 Domains: A model for routing in the Internet", RFC 1136, 812 December 1989. 814 [RFC1661] Simpson, W., "The Point-to-Point Protocol (PPP)", STD 51, 815 RFC 1661, July 1994. 817 [RFC1918] Rekhter, Y., Moskowitz, R., Karrenberg, D., Groot, G., and 818 E. Lear, "Address Allocation for Private Internets", 819 BCP 5, RFC 1918, February 1996. 821 [RFC2131] Droms, R., "Dynamic Host Configuration Protocol", 822 RFC 2131, March 1997. 824 [RFC2827] Ferguson, P. and D. Senie, "Network Ingress Filtering: 825 Defeating Denial of Service Attacks which employ IP Source 826 Address Spoofing", BCP 38, RFC 2827, May 2000. 828 [RFC3315] Droms, R., Bound, J., Volz, B., Lemon, T., Perkins, C., 829 and M. Carney, "Dynamic Host Configuration Protocol for 830 IPv6 (DHCPv6)", RFC 3315, July 2003. 832 [RFC3484] Draves, R., "Default Address Selection for Internet 833 Protocol version 6 (IPv6)", RFC 3484, February 2003. 835 [RFC3542] Stevens, W., Thomas, M., Nordmark, E., and T. Jinmei, 836 "Advanced Sockets Application Program Interface (API) for 837 IPv6", RFC 3542, May 2003. 839 [RFC3704] Baker, F. and P. Savola, "Ingress Filtering for Multihomed 840 Networks", BCP 84, RFC 3704, March 2004. 842 [RFC3775] Johnson, D., Perkins, C., and J. Arkko, "Mobility Support 843 in IPv6", RFC 3775, June 2004. 845 [RFC4294] Loughney, J., "IPv6 Node Requirements", RFC 4294, 846 April 2006. 848 [RFC4477] Chown, T., Venaas, S., and C. Strauf, "Dynamic Host 849 Configuration Protocol (DHCP): IPv4 and IPv6 Dual-Stack 850 Issues", RFC 4477, May 2006. 852 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 853 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 854 September 2007. 856 [RFC4960] Stewart, R., "Stream Control Transmission Protocol", 857 RFC 4960, September 2007. 859 [RFC5014] Nordmark, E., Chakrabarti, S., and J. Laganier, "IPv6 860 Socket API for Source Address Selection", RFC 5014, 861 September 2007. 863 [RFC5113] Arkko, J., Aboba, B., Korhonen, J., and F. Bari, "Network 864 Discovery and Selection Problem", RFC 5113, January 2008. 866 [RFC5206] Nikander, P., Henderson, T., Vogt, C., and J. Arkko, "End- 867 Host Mobility and Multihoming with the Host Identity 868 Protocol", RFC 5206, April 2008. 870 [RFC5220] Matsumoto, A., Fujisaki, T., Hiromi, R., and K. Kanayama, 871 "Problem Statement for Default Address Selection in Multi- 872 Prefix Environments: Operational Issues of RFC 3484 873 Default Rules", RFC 5220, July 2008. 875 [RFC5221] Matsumoto, A., Fujisaki, T., Hiromi, R., and K. Kanayama, 876 "Requirements for Address Selection Mechanisms", RFC 5221, 877 July 2008. 879 [RFC5245] Rosenberg, J., "Interactive Connectivity Establishment 880 (ICE): A Protocol for Network Address Translator (NAT) 881 Traversal for Offer/Answer Protocols", RFC 5245, 882 April 2010. 884 [RFC5533] Nordmark, E. and M. Bagnulo, "Shim6: Level 3 Multihoming 885 Shim Protocol for IPv6", RFC 5533, June 2009. 887 [RFC5648] Wakikawa, R., Devarapalli, V., Tsirtsis, G., Ernst, T., 888 and K. Nagami, "Multiple Care-of Addresses Registration", 889 RFC 5648, October 2009. 891 [TS23.234] 892 3GPP, "3GPP system to Wireless Local Area Network (WLAN) 893 interworking; TS 23.234", 2009. 895 [TS23.402] 896 3GPP, "Architecture enhancements for non- 3GPP accesses; 897 TS 23.402", 2010. 899 Authors' Addresses 901 Marc Blanchet 902 Viagenie 903 2875 boul. Laurier, suite D2-630 904 Quebec, QC G1V 2M2 905 Canada 907 Email: Marc.Blanchet@viagenie.ca 908 URI: http://viagenie.ca 910 Pierrick Seite 911 France Telecom - Orange 912 4, rue du Clos Courtel, BP 91226 913 Cesson-Sevigne 35512 914 France 916 Email: pierrick.seite@orange-ftgroup.com