idnits 2.17.1 draft-templin-aerolink-40.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 : ---------------------------------------------------------------------------- == The 'Obsoletes: ' line in the draft header should list only the _numbers_ of the RFCs which will be obsoleted by this document (if approved); it should not include the word 'RFC' in the list. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (September 25, 2014) is 3498 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Missing Reference: 'I-D.taylor-v6ops-fragdrop' is mentioned on line 800, but not defined == Unused Reference: 'RFC0768' is defined on line 2446, but no explicit reference was found in the text == Unused Reference: 'RFC4862' is defined on line 2485, but no explicit reference was found in the text == Unused Reference: 'RFC0879' is defined on line 2498, but no explicit reference was found in the text == Unused Reference: 'RFC4994' is defined on line 2553, but no explicit reference was found in the text == Unused Reference: 'RFC6146' is defined on line 2586, but no explicit reference was found in the text == Unused Reference: 'RFC6204' is defined on line 2590, but no explicit reference was found in the text == Unused Reference: 'RFC6355' is defined on line 2597, but no explicit reference was found in the text == Unused Reference: 'RFC6691' is defined on line 2605, but no explicit reference was found in the text == Unused Reference: 'RFC6706' is defined on line 2608, but no explicit reference was found in the text == Unused Reference: 'RFC6980' is defined on line 2624, but no explicit reference was found in the text == Unused Reference: 'RFC7078' is defined on line 2627, but no explicit reference was found in the text ** Obsolete normative reference: RFC 2460 (Obsoleted by RFC 8200) ** Obsolete normative reference: RFC 3315 (Obsoleted by RFC 8415) ** Obsolete normative reference: RFC 3633 (Obsoleted by RFC 8415) ** Obsolete normative reference: RFC 6434 (Obsoleted by RFC 8504) == Outdated reference: A later version (-21) exists of draft-ietf-dhc-sedhcpv6-03 -- Obsolete informational reference (is this intentional?): RFC 879 (Obsoleted by RFC 7805, RFC 9293) -- Obsolete informational reference (is this intentional?): RFC 1981 (Obsoleted by RFC 8201) -- Obsolete informational reference (is this intentional?): RFC 5246 (Obsoleted by RFC 8446) -- Obsolete informational reference (is this intentional?): RFC 5996 (Obsoleted by RFC 7296) -- Obsolete informational reference (is this intentional?): RFC 6204 (Obsoleted by RFC 7084) -- Obsolete informational reference (is this intentional?): RFC 6691 (Obsoleted by RFC 9293) Summary: 4 errors (**), 0 flaws (~~), 15 warnings (==), 7 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group F. Templin, Ed. 3 Internet-Draft Boeing Research & Technology 4 Obsoletes: rfc5320, rfc5558, rfc5720, September 25, 2014 5 rfc6179, rfc6706 (if 6 approved) 7 Intended status: Standards Track 8 Expires: March 29, 2015 10 Transmission of IP Packets over AERO Links 11 draft-templin-aerolink-40.txt 13 Abstract 15 This document specifies the operation of IP over tunnel virtual links 16 using Asymmetric Extended Route Optimization (AERO). Nodes attached 17 to AERO links can exchange packets via trusted intermediate routers 18 that provide forwarding services to reach off-link destinations and 19 redirection services for route optimization. AERO provides an IPv6 20 link-local address format known as the AERO address that supports 21 operation of the IPv6 Neighbor Discovery (ND) protocol and links IPv6 22 ND to IP forwarding. Admission control and provisioning are 23 supported by the Dynamic Host Configuration Protocol for IPv6 24 (DHCPv6), and node mobility is naturally supported through dynamic 25 neighbor cache updates. Although DHCPv6 and IPv6 ND messaging is 26 used in the control plane, both IPv4 and IPv6 are supported in the 27 data plane. 29 Status of This Memo 31 This Internet-Draft is submitted in full conformance with the 32 provisions of BCP 78 and BCP 79. 34 Internet-Drafts are working documents of the Internet Engineering 35 Task Force (IETF). Note that other groups may also distribute 36 working documents as Internet-Drafts. The list of current Internet- 37 Drafts is at http://datatracker.ietf.org/drafts/current/. 39 Internet-Drafts are draft documents valid for a maximum of six months 40 and may be updated, replaced, or obsoleted by other documents at any 41 time. It is inappropriate to use Internet-Drafts as reference 42 material or to cite them other than as "work in progress." 44 This Internet-Draft will expire on March 29, 2015. 46 Copyright Notice 48 Copyright (c) 2014 IETF Trust and the persons identified as the 49 document authors. All rights reserved. 51 This document is subject to BCP 78 and the IETF Trust's Legal 52 Provisions Relating to IETF Documents 53 (http://trustee.ietf.org/license-info) in effect on the date of 54 publication of this document. Please review these documents 55 carefully, as they describe your rights and restrictions with respect 56 to this document. Code Components extracted from this document must 57 include Simplified BSD License text as described in Section 4.e of 58 the Trust Legal Provisions and are provided without warranty as 59 described in the Simplified BSD License. 61 Table of Contents 63 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 64 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 4 65 3. Asymmetric Extended Route Optimization (AERO) . . . . . . . . 6 66 3.1. AERO Link Reference Model . . . . . . . . . . . . . . . . 6 67 3.2. AERO Node Types . . . . . . . . . . . . . . . . . . . . . 7 68 3.3. AERO Addresses . . . . . . . . . . . . . . . . . . . . . 8 69 3.4. AERO Interface Characteristics . . . . . . . . . . . . . 9 70 3.4.1. Coordination of Multiple Underlying Interfaces . . . 11 71 3.5. AERO Interface Neighbor Cache Maintenace . . . . . . . . 11 72 3.6. AERO Interface Sending Algorithm . . . . . . . . . . . . 13 73 3.7. AERO Interface Encapsulation, Re-encapsulation and 74 Decapsulation . . . . . . . . . . . . . . . . . . . . . . 15 75 3.8. AERO Interface Data Origin Authentication . . . . . . . . 16 76 3.9. AERO Interface MTU and Fragmentation . . . . . . . . . . 17 77 3.9.1. Accommodating Large IPv6 ND and DHCPv6 Messages . . . 19 78 3.9.2. Integrity . . . . . . . . . . . . . . . . . . . . . . 20 79 3.10. AERO Interface Error Handling . . . . . . . . . . . . . . 21 80 3.11. AERO Router Discovery, Prefix Delegation and Address 81 Configuration . . . . . . . . . . . . . . . . . . . . . . 25 82 3.11.1. AERO DHCPv6 Service Model . . . . . . . . . . . . . 25 83 3.11.2. AERO Client Behavior . . . . . . . . . . . . . . . . 26 84 3.11.3. AERO Server Behavior . . . . . . . . . . . . . . . . 28 85 3.12. AERO Relay/Server Routing System . . . . . . . . . . . . 31 86 3.13. AERO Redirection . . . . . . . . . . . . . . . . . . . . 31 87 3.13.1. Reference Operational Scenario . . . . . . . . . . . 31 88 3.13.2. Concept of Operations . . . . . . . . . . . . . . . 33 89 3.13.3. Message Format . . . . . . . . . . . . . . . . . . . 33 90 3.13.4. Sending Predirects . . . . . . . . . . . . . . . . . 34 91 3.13.5. Re-encapsulating and Relaying Predirects . . . . . . 35 92 3.13.6. Processing Predirects and Sending Redirects . . . . 36 93 3.13.7. Re-encapsulating and Relaying Redirects . . . . . . 38 94 3.13.8. Processing Redirects . . . . . . . . . . . . . . . . 39 95 3.13.9. Server-Oriented Redirection . . . . . . . . . . . . 39 96 3.14. Neighbor Unreachability Detection (NUD) . . . . . . . . . 39 97 3.15. Mobility Management . . . . . . . . . . . . . . . . . . . 41 98 3.15.1. Announcing Link-Layer Address Changes . . . . . . . 41 99 3.15.2. Bringing New Links Into Service . . . . . . . . . . 42 100 3.15.3. Removing Existing Links from Service . . . . . . . . 42 101 3.15.4. Moving to a New Server . . . . . . . . . . . . . . . 43 102 3.16. Encapsulation Protocol Version Considerations . . . . . . 43 103 3.17. Multicast Considerations . . . . . . . . . . . . . . . . 44 104 3.18. Operation on AERO Links Without DHCPv6 Services . . . . . 44 105 3.19. Operation on Server-less AERO Links . . . . . . . . . . . 44 106 3.20. Proxy AERO . . . . . . . . . . . . . . . . . . . . . . . 45 107 3.21. Extending AERO Links Through Security Gateways . . . . . 46 108 3.22. Extending IPv6 AERO Links to the Internet . . . . . . . . 47 109 4. Implementation Status . . . . . . . . . . . . . . . . . . . . 51 110 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 51 111 6. Security Considerations . . . . . . . . . . . . . . . . . . . 51 112 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 52 113 8. References . . . . . . . . . . . . . . . . . . . . . . . . . 52 114 8.1. Normative References . . . . . . . . . . . . . . . . . . 52 115 8.2. Informative References . . . . . . . . . . . . . . . . . 53 116 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 56 118 1. Introduction 120 This document specifies the operation of IP over tunnel virtual links 121 using Asymmetric Extended Route Optimization (AERO). The AERO link 122 can be used for tunneling to neighboring nodes over either IPv6 or 123 IPv4 networks, i.e., AERO views the IPv6 and IPv4 networks as 124 equivalent links for tunneling. Nodes attached to AERO links can 125 exchange packets via trusted intermediate routers that provide 126 forwarding services to reach off-link destinations and redirection 127 services for route optimization that addresses the requirements 128 outlined in [RFC5522]. 130 AERO provides an IPv6 link-local address format known as the AERO 131 address that supports operation of the IPv6 Neighbor Discovery (ND) 132 [RFC4861] protocol and links IPv6 ND to IP forwarding. Admission 133 control and provisioning are supported by the Dynamic Host 134 Configuration Protocol for IPv6 (DHCPv6) [RFC3315], and node mobility 135 is naturally supported through dynamic neighbor cache updates. 136 Although DHCPv6 and IPv6 ND message signalling is used in the control 137 plane, both IPv4 and IPv6 can be used in the data plane. The 138 remainder of this document presents the AERO specification. 140 2. Terminology 142 The terminology in the normative references applies; the following 143 terms are defined within the scope of this document: 145 AERO link 146 a Non-Broadcast, Multiple Access (NBMA) tunnel virtual overlay 147 configured over a node's attached IPv6 and/or IPv4 networks. All 148 nodes on the AERO link appear as single-hop neighbors from the 149 perspective of the virtual overlay. 151 AERO interface 152 a node's attachment to an AERO link. 154 AERO address 155 an IPv6 link-local address constructed as specified in Section 3.2 156 and assigned to a Client's AERO interface. 158 AERO node 159 a node that is connected to an AERO link and that participates in 160 IPv6 ND and DHCPv6 messaging over the link. 162 AERO Client ("Client") 163 a node that applies an AERO address to an AERO interface and 164 receives an IP prefix via a DHCPv6 Prefix Delegation (PD) exchange 165 with one or more AERO Servers. 167 AERO Server ("Server") 168 a node that configures an AERO interface to provide default 169 forwarding and DHCPv6 services for AERO Clients. The Server 170 applies the IPv6 link-local subnet router anycast address (fe80::) 171 to the AERO interface and also applies an administratively 172 assigned IPv6 link-local unicast address used for operation of 173 DHCPv6 and the IPv6 ND protocol. 175 AERO Relay ("Relay") 176 a node that configures an AERO interface to relay IP packets 177 between nodes on the same AERO link and/or forward IP packets 178 between the AERO link and the native Internetwork. The Relay 179 applies an administratively assigned IPv6 link-local unicast 180 address to the AERO interface the same as for a Server. 182 ingress tunnel endpoint (ITE) 183 an AERO interface endpoint that injects tunneled packets into an 184 AERO link. 186 egress tunnel endpoint (ETE) 187 an AERO interface endpoint that receives tunneled packets from an 188 AERO link. 190 underlying network 191 a connected IPv6 or IPv4 network routing region over which the 192 tunnel virtual overlay is configured. A typical example is an 193 enterprise network. 195 underlying interface 196 an AERO node's interface point of attachment to an underlying 197 network. 199 link-layer address 200 an IP address assigned to an AERO node's underlying interface. 201 When UDP encapsulation is used, the UDP port number is also 202 considered as part of the link-layer address. Link-layer 203 addresses are used as the encapsulation header source and 204 destination addresses. 206 network layer address 207 the source or destination address of the encapsulated IP packet. 209 end user network (EUN) 210 an internal virtual or external edge IP network that an AERO 211 Client connects to the rest of the network via the AERO interface. 213 AERO Service Prefix (ASP) 214 an IP prefix associated with the AERO link and from which AERO 215 Client Prefixes (ACPs) are derived (for example, the IPv6 ACP 216 2001:db8:1:2::/64 is derived from the IPv6 ASP 2001:db8::/32). 218 AERO Client Prefix (ACP) 219 a more-specific IP prefix taken from an ASP and delegated to a 220 Client. 222 Throughout the document, the simple terms "Client", "Server" and 223 "Relay" refer to "AERO Client", "AERO Server" and "AERO Relay", 224 respectively. Capitalization is used to distinguish these terms from 225 DHCPv6 client/server/relay. 227 The terminology of [RFC4861] (including the names of node variables 228 and protocol constants) applies to this document. Also throughout 229 the document, the term "IP" is used to generically refer to either 230 Internet Protocol version (i.e., IPv4 or IPv6). 232 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 233 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 234 document are to be interpreted as described in [RFC2119]. 236 3. Asymmetric Extended Route Optimization (AERO) 238 The following sections specify the operation of IP over Asymmetric 239 Extended Route Optimization (AERO) links: 241 3.1. AERO Link Reference Model 243 .-(::::::::) 244 .-(:::: IP ::::)-. 245 (:: Internetwork ::) 246 `-(::::::::::::)-' 247 `-(::::::)-' 248 | 249 +--------------+ +--------+-------+ +--------------+ 250 |AERO Server S1| | AERO Relay R1 | |AERO Server S2| 251 | Nbr: C1; R1 | | Nbr: S1; S2 | | Nbr: C2; R1 | 252 | default->R1 | |(H1->S1; H2->S2)| | default->R1 | 253 | H1->C1 | +--------+-------+ | H2->C2 | 254 +-------+------+ | +------+-------+ 255 | | | 256 X---+---+-------------------+------------------+---+---X 257 | AERO Link | 258 +-----+--------+ +--------+-----+ 259 |AERO Client C1| |AERO Client C2| 260 | Nbr: S1 | | Nbr: S2 | 261 | default->S1 | | default->S2 | 262 +--------------+ +--------------+ 263 .-. .-. 264 ,-( _)-. ,-( _)-. 265 .-(_ IP )-. .-(_ IP )-. 266 (__ EUN ) (__ EUN ) 267 `-(______)-' `-(______)-' 268 | | 269 +--------+ +--------+ 270 | Host H1| | Host H2| 271 +--------+ +--------+ 273 Figure 1: AERO Link Reference Model 275 Figure 1 above presents the AERO link reference model. In this 276 model: 278 o Relay R1 acts as a default router for its associated Servers S1 279 and S2, and connects the AERO link to the rest of the IP 280 Internetwork 282 o Servers S1 and S2 associate with Relay R1 and also act as default 283 routers for their associated Clients C1 and C2. 285 o Clients C1 and C2 associate with Servers S1 and S2, respectively 286 and also act as default routers for their associated EUNs 288 o Hosts H1 and H2 attach to the EUNs served by Clients C1 and C2, 289 respectively 291 In common operational practice, there may be many additional Relays, 292 Servers and Clients. 294 3.2. AERO Node Types 296 AERO Relays provide default forwarding services to AERO Servers. 297 Relays forward packets between Servers connected to the same AERO 298 link and also forward packets between the AERO link and the native 299 Internetwork. Relays present the AERO link to the native 300 Internetwork as a set of one or more AERO Service Prefixes (ASPs). 301 Each Relay advertises the ASPs for the AERO link into the native IP 302 Internetwork and serves as a gateway between the AERO link and the 303 Internetwork. AERO Relays maintain an AERO interface neighbor cache 304 entry for each AERO Server, and maintain an IP forwarding table entry 305 for each AERO Client Prefix (ACP). 307 AERO Servers provide default forwarding services to AERO Clients. 308 Each Server also peers with each Relay in a dynamic routing protocol 309 instance to advertise its list of associated ACPs. Servers configure 310 a DHCPv6 server function to facilitate Prefix Delegation (PD) 311 exchanges with Clients. Each delegated prefix becomes an ACP taken 312 from an ASP. Servers forward packets between Clients and Relays, as 313 well as between Clients and other Clients associated with the same 314 Server. AERO Servers maintain an AERO interface neighbor cache entry 315 for each AERO Relay. They also maintain both a neighbor cache entry 316 and an IP forwarding table entry for each of their associated 317 Clients. 319 AERO Clients act as requesting routers to receive ACPs through DHCPv6 320 PD exchanges with AERO Servers over the AERO link and sub-delegate 321 portions of their ACPs to EUN interfaces. (Each Client MAY associate 322 with a single Server or with multiple Servers, e.g., for fault 323 tolerance and/or load balancing.) Each IPv6 Client receives at least 324 a /64 IPv6 ACP, and may receive even shorter prefixes. Similarly, 325 each IPv4 Client receives at least a /32 IPv4 ACP (i.e., a singleton 326 IPv4 address), and may receive even shorter prefixes. AERO Clients 327 maintain an AERO interface neighbor cache entry for each of their 328 associated Servers as well as for each of their correspondent 329 Clients. 331 AERO Clients that act as hosts typically configure a TUN/TAP 332 interface as a point-to-point linkage between the IP layer and the 333 AERO interface. The IP layer therefore sees only the TUN/TAP 334 interface, while the AERO interface provides an intermediate conduit 335 between the TUN/TAP interface and the underlying interfaces. AERO 336 Clients that act as hosts assign one or more IP addresses from their 337 ACPs to the TUN/TAP interface. 339 3.3. AERO Addresses 341 An AERO address is an IPv6 link-local address with an embedded ACP 342 and assigned to a Client's AERO interface. The AERO address is 343 formed as follows: 345 fe80::[ACP] 347 For IPv6, the AERO address begins with the prefix fe80::/64 and 348 includes in its interface identifier the base prefix taken from the 349 Client's IPv6 ACP. The base prefix is determined by masking the ACP 350 with the prefix length. For example, if the AERO Client receives the 351 IPv6 ACP: 353 2001:db8:1000:2000::/56 355 it constructs its AERO address as: 357 fe80::2001:db8:1000:2000 359 For IPv4, the AERO address is formed from the lower 64 bits of an 360 IPv4-mapped IPv6 address [RFC4291] that includes the base prefix 361 taken from the Client's IPv4 ACP. For example, if the AERO Client 362 receives the IPv4 ACP: 364 192.0.2.32/28 366 it constructs its AERO address as: 368 fe80::FFFF:192.0.2.32 370 The AERO address remains stable as the Client moves between 371 topological locations, i.e., even if its link-layer addresses change. 373 NOTE: In some cases, prospective neighbors may not have advanced 374 knowledge of the Client's ACP length and may therefore send initial 375 IPv6 ND messages with an AERO destination address that matches the 376 ACP but does not correspond to the base prefix. In that case, the 377 Client MUST accept the address as equivalent to the base address, but 378 then use the base address as the source address of any IPv6 ND 379 message replies. For example, if the Client receives the IPv6 ACP 380 2001:db8:1000:2000::/56 then subsequently receives an IPv6 ND message 381 with destination address fe80::2001:db8:1000:2001, it accepts the 382 message but uses fe80::2001:db8:1000:2000 as the source address of 383 any IPv6 ND replies. 385 3.4. AERO Interface Characteristics 387 AERO interfaces use IP-in-IPv6 encapsulation [RFC2473] to exchange 388 tunneled packets with AERO neighbors attached to an underlying IPv6 389 network, and use IP-in-IPv4 encapsulation [RFC2003][RFC4213] to 390 exchange tunneled packets with AERO neighbors attached to an 391 underlying IPv4 network. AERO interfaces can also coordinate secured 392 tunnel types such as IPsec [RFC4301] or TLS [RFC5246]. When Network 393 Address Translator (NAT) traversal and/or filtering middlebox 394 traversal may be necessary, a UDP header is further inserted 395 immediately above the IP encapsulation header. 397 AERO interfaces maintain a neighbor cache, and AERO Clients and 398 Servers use an adaptation of standard unicast IPv6 ND messaging. 399 AERO interfaces use unicast Neighbor Solicitation (NS), Neighbor 400 Advertisement (NA), Router Solicitation (RS) and Router Advertisement 401 (RA) messages the same as for any IPv6 link. AERO interfaces use two 402 redirection message types -- the first known as a Predirect message 403 and the second being the standard Redirect message (see 404 Section 3.12). AERO links further use link-local-only addressing; 405 hence, AERO nodes ignore any Prefix Information Options (PIOs) they 406 may receive in RA messages over an AERO interface. 408 AERO interface ND messages include one or more Target Link-Layer 409 Address Options (TLLAOs) formatted as shown in Figure 2: 411 0 1 2 3 412 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 413 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 414 | Type = 2 | Length = 3 | Reserved | 415 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 416 | Link ID | Preference | UDP Port Number | 417 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 418 | | 419 +-- --+ 420 | | 421 +-- IP Address --+ 422 | | 423 +-- --+ 424 | | 425 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 427 Figure 2: AERO Target Link-Layer Address Option (TLLAO) Format 429 In this format, Link ID is an integer value between 0 and 255 430 corresponding to an underlying interface of the target node, and 431 Preference is an integer value between 0 and 255 indicating the 432 node's preference for this underlying interface (with 255 being the 433 highest preference, 1 being the lowest, and 0 meaning "link 434 disabled"). UDP Port Number and IP Address are set to the addresses 435 used by the target node when it sends encapsulated packets over the 436 underlying interface. When the encapsulation IP address family is 437 IPv4, IP Address is formed as an IPv4-mapped IPv6 address [RFC4291]. 439 When a Relay enables an AERO interface, it assigns an 440 administratively assigned link-local address fe80::ID to the 441 interface. Each fe80::ID address MUST be unique among all Relays and 442 Servers on the link, and MUST NOT collide with any potential AERO 443 addresses. The addresses are typically taken from the range 444 fe80::/96, e.g., as fe80::1, fe80::2, fe80::3, etc. The Relay also 445 maintains an IP forwarding table entry for each Client-Server 446 association and maintains a neighbor cache entry for each Server on 447 the link. Relays do not require the use of IPv6 ND messaging for 448 reachability determination since Relays and Servers engage in a 449 dynamic routing protocol over the AERO interface. At a minimum, 450 however, Relays respond to NS messages by returning an NA. 452 When a Server enables an AERO interface, it assigns the address 453 fe80:: to the interface as a link-local Subnet Router Anycast 454 address, and also assigns an administratively assigned link-local 455 address fe80::ID the same as for Relays. (The Server then accepts 456 DHCPv6 and IPv6 ND solicitation messages destined to either the 457 fe80:: or fe80::ID addresses, but always uses fe80::ID as the source 458 address in the replies it generates.) The Server further configures 459 a DHCPv6 server function to facilitate DHCPv6 PD exchanges with AERO 460 Clients. The Server maintains a neighbor cache entry for each Relay 461 on the link, and manages per-Client neighbor cache entries and IP 462 forwarding table entries based on DHCPv6 exchanges. When the Server 463 receives an NS/RS message on the AERO interface it returns an NA/RA 464 message but does not update the neighbor cache. Each Server also 465 engages in a dynamic routing protocol with all Relays on the link. 466 Finally, the Server provides a simple conduit between Clients and 467 Relays, or between Clients and other Clients. Therefore, packets 468 enter the Server's AERO interface from the link layer and are 469 forwarded back out the link layer without ever leaving the AERO 470 interface and therefore without ever disturbing the network layer. 472 When a Client enables an AERO interface, it invokes DHCPv6 PD to 473 receive an ACP from an AERO Server. Next, it assigns the 474 corresponding AERO address to the AERO interface and creates a 475 neighbor cache entry for the Server, i.e., the PD exchange bootstraps 476 the provisioning of a unique link-local address. The Client 477 maintains a neighbor cache entry for each of its Servers and each of 478 its active correspondent Clients. When the Client receives Redirect/ 479 Predirect messages on the AERO interface it updates or creates 480 neighbor cache entries, including link-layer address information. 481 Unsolicited NA messages update the cached link-layer addresses for 482 correspondent Clients (e.g., following a link-layer address change 483 due to node mobility) but do not create new neighbor cache entries. 484 NS/NA messages used for Neighbor Unreachability Detection (NUD) 485 update timers in existing neighbor cache entires but do not update 486 link-layer addresses nor create new neighbor cache entries. Finally, 487 the Client need not maintain any IP forwarding table entries for its 488 Servers or correspondent Clients. Instead, it can set a single 489 "route-to-interface" default route in the IP forwarding table 490 pointing to the AERO interface, and all forwarding decisions can be 491 made within the AERO interface based on neighbor cache entries. (On 492 systems in which adding a default route would violate security 493 policy, the default route could instead be installed via a 494 "synthesized RA", e.g., as discussed in Section 3.11.2.) 496 3.4.1. Coordination of Multiple Underlying Interfaces 498 AERO interfaces may be configured over multiple underlying 499 interfaces. For example, common mobile handheld devices have both 500 wireless local area network ("WLAN") and cellular wireless links. 501 These links are typically used "one at a time" with low-cost WLAN 502 preferred and highly-available cellular wireless as a standby. In a 503 more complex example, aircraft frequently have many wireless data 504 link types (e.g. satellite-based, terrestrial, air-to-air 505 directional, etc.) with diverse performance and cost properties. 507 If a Client's multiple underlying interfaces are used "one at a time" 508 (i.e., all other interfaces are in standby mode while one interface 509 is active), then Redirect, Predirect and unsolicited NA messages 510 include only a single TLLAO with Link ID set to a constant value. 512 If the Client has multiple active underlying interfaces, then from 513 the perspective of IPv6 ND it would appear to have a single link- 514 local address with multiple link-layer addresses. In that case, 515 Redirect, Predirect and unsolicited NA messages MAY include multiple 516 TLLAOs -- each with a different Link ID that corresponds to a 517 specific underlying interface of the Client. 519 3.5. AERO Interface Neighbor Cache Maintenace 521 Each AERO interface maintains a conceptual neighbor cache that 522 includes an entry for each neighbor it communicates with on the AERO 523 link, the same as for any IPv6 interface [RFC4861]. AERO interface 524 neighbor cache entires are said to be one of "permanent", "static" or 525 "dynamic". 527 Permanent neighbor cache entries are created through explicit 528 administrative action; they have no timeout values and remain in 529 place until explicitly deleted. AERO Relays maintain a permanent 530 neighbor cache entry for each Server on the link, and AERO Servers 531 maintain a permanent neighbor cache entry for each Relay on the link. 533 Static neighbor cache entries are created though DHCPv6 PD exchanges 534 and remain in place for durations bounded by prefix lifetimes. AERO 535 Servers maintain a static neighbor cache entry for each of their 536 associated Clients, and AERO Clients maintain a static neighbor cache 537 for each of their associated Servers. When an AERO Server sends a 538 DHCPv6 Reply message response to a Client's DHCPv6 Solicit/Request or 539 Renew message, it creates or updates a static neighbor cache entry 540 based on the Client's AERO address as the network-layer address, the 541 prefix lifetime as the neighbor cache entry lifetime, the Client's 542 encapsulation IP address and UDP port number as the link-layer 543 address and the prefix length as the length to apply to the AERO 544 address. When an AERO Client receives a DHCPv6 Reply message from a 545 Server, it creates or updates a static neighbor cache entry based on 546 the Reply message link-local source address as the network-layer 547 address, the prefix lifetime as the neighbor cache entry lifetime, 548 and the encapsulation IP source address and UDP source port number as 549 the link-layer address. 551 Dynamic neighbor cache entries are created based on receipt of an 552 IPv6 ND message, and are garbage-collected if not used within a short 553 timescale. AERO Clients maintain dynamic neighbor cache entries for 554 each of their active correspondent Clients with lifetimes based on 555 IPv6 ND messaging constants. When an AERO Client receives a valid 556 Predirect message it creates or updates a dynamic neighbor cache 557 entry for the Predirect target network-layer and link-layer addresses 558 plus prefix length. The node then sets an "AcceptTime" variable in 559 the neighbor cache entry and uses this value to determine whether 560 packets received from the correspondent can be accepted. When an 561 AERO Client receives a valid Redirect message it creates or updates a 562 dynamic neighbor cache entry for the Redirect target network-layer 563 and link-layer addresses plus prefix length. The Client then sets a 564 "ForwardTime" variable in the neighbor cache entry and uses this 565 value to determine whether packets can be sent directly to the 566 correspondent. The Client also maintains a "MaxRetry" variable to 567 limit the number of keepalives sent when a correspondent may have 568 gone unreachable. 570 For dynamic neighbor cache entries, when an AERO Client receives a 571 valid NS message it (re)sets AcceptTime for the neighbor to 572 ACCEPT_TIME. When an AERO Client receives a valid solicited NA 573 message, it (re)sets ForwardTime for the neighbor to FORWARD_TIME and 574 sets MaxRetry to MAX_RETRY. When an AERO Client receives a valid 575 unsolicited NA message, it updates the correspondent's link-layer 576 addresses but DOES NOT reset AcceptTime, ForwardTime or MaxRetry. 578 It is RECOMMENDED that FORWARD_TIME be set to the default constant 579 value 30 seconds to match the default REACHABLE_TIME value specified 580 for IPv6 ND [RFC4861]. 582 It is RECOMMENDED that ACCEPT_TIME be set to the default constant 583 value 40 seconds to allow a 10 second window so that the AERO 584 redirection procedure can converge before AcceptTime decrements below 585 FORWARD_TIME. 587 It is RECOMMENDED that MAX_RETRY be set to 3 the same as described 588 for IPv6 ND address resolution in Section 7.3.3 of [RFC4861]. 590 Different values for FORWARD_TIME, ACCEPT_TIME, and MAX_RETRY MAY be 591 administratively set, if necessary, to better match the AERO link's 592 performance characteristics; however, if different values are chosen, 593 all nodes on the link MUST consistently configure the same values. 594 Most importantly, ACCEPT_TIME SHOULD be set to a value that is 595 sufficiently longer than FORWARD_TIME to allow the AERO redirection 596 procedure to converge. 598 3.6. AERO Interface Sending Algorithm 600 IP packets enter a node's AERO interface either from the network 601 layer (i.e., from a local application or the IP forwarding system), 602 or from the link layer (i.e., from the AERO tunnel virtual link). 603 Packets that enter the AERO interface from the network layer are 604 encapsulated and admitted into the AERO link, i.e., they are 605 tunnelled to an AERO interface neighbor. Packets that enter the AERO 606 interface from the link layer are either re-admitted into the AERO 607 link or delivered to the network layer where they are subject to 608 either local delivery or IP forwarding. Since each AERO node has 609 only partial information about neighbors on the link, AERO interfaces 610 may forward packets with link-local destination addresses at a layer 611 below the network layer. This means that AERO nodes act as both IP 612 routers and sub-IP layer forwarding agents. AERO interface sending 613 considerations for Clients, Servers and Relays are given below. 615 When an IP packet enters a Client's AERO interface from the network 616 layer, if the destination is covered by an ASP the Client searches 617 for a dynamic neighbor cache entry with a non-zero ForwardTime and an 618 AERO address that matches the packet's destination address. (The 619 destination address may be either an address covered by the 620 neighbor's ACP or the (link-local) AERO address itself.) If there is 621 a match, the Client uses a link-layer address in the entry as the 622 link-layer address for encapsulation then admits the packet into the 623 AERO link. If there is no match, the Client instead uses the link- 624 layer address of a neighboring Server as the link-layer address for 625 encapsulation. 627 When an IP packet enters a Server's AERO interface from the link 628 layer, if the destination is covered by an ASP the Server searches 629 for a static neighbor cache entry with an AERO address that matches 630 the packet's destination address. (The destination address may be 631 either an address covered by the neighbor's ACP or the AERO address 632 itself.) If there is a match, the Server uses a link-layer address 633 in the entry as the link-layer address for encapsulation and re- 634 admits the packet into the AERO link. If there is no match, the 635 Server instead uses the link-layer address in any permanent neighbor 636 cache entry as the link-layer address for encapsulation. 638 When an IP packet enters a Relay's AERO interface from the network 639 layer, the Relay searches its IP forwarding table for an entry that 640 is covered by an ASP and also matches the destination. If there is a 641 match, the Relay uses the link-layer address in the neighbor cache 642 entry for the next-hop Server as the link-layer address for 643 encapsulation and admits the packet into the AERO link. When an IP 644 packet enters a Relay's AERO interface from the link-layer, if the 645 destination is not a link-local address and is does not match an ASP 646 the Relay removes the packet from the AERO interface and uses IP 647 forwarding to forward the packet to the Internetwork. If the 648 destination address is a link-local or non-link-local address that 649 matches an ASP, and there is a more-specific ACP entry in the IP 650 forwarding table, the Relay uses the link-layer address in the 651 corresponding neighbor cache entry for the next-hop Server as the 652 link-layer address for encapsulation and re-admits the packet into 653 the AERO link. When an IP packet enters a Relay's AERO interface 654 from either the network layer or link-layer, and the packet's 655 destination address matches an ASP but there is no more-specific ACP 656 entry, the Relay drops the packet and returns an ICMP Destination 657 Unreachable message (see: Section 3.10). 659 When an AERO Server receives a packet from a Relay via the AERO 660 interface, the Server MUST NOT forward the packet back to the same or 661 a different Relay. 663 When an AERO Relay receives a packet from a Server via the AERO 664 interface, the Relay MUST NOT forward the packet back to the same 665 Server. 667 When an AERO node re-admits a packet into the AERO link without 668 involving the network layer, the node MUST NOT decrement the network 669 layer TTL/Hop-count. 671 Note that in the above that the link-layer address for encapsulation 672 may be determined through consulting either the neighbor cache or the 673 IP forwarding table. IP forwarding is therefore linked to IPv6 ND 674 via the AERO address. 676 3.7. AERO Interface Encapsulation, Re-encapsulation and Decapsulation 678 AERO interfaces encapsulate IP packets according to whether they are 679 entering the AERO interface from the network layer or if they are 680 being re-admitted into the same AERO link they arrived on. This 681 latter form of encapsulation is known as "re-encapsulation". 683 AERO interfaces encapsulate packets per the base tunneling 684 specifications (e.g., [RFC2003][RFC2473][RFC4213][RFC4301][RFC5246], 685 etc.) except that the interface copies the "TTL/Hop Limit", "Type of 686 Service/Traffic Class" and "Congestion Experienced" values in the 687 packet's IP header into the corresponding fields in the encapsulation 688 IP header. For packets undergoing re-encapsulation, the AERO 689 interface instead copies the "TTL/Hop Limit", "Type of Service/ 690 Traffic Class" and "Congestion Experienced" values in the original 691 encapsulation IP header into the corresponding fields in the new 692 encapsulation IP header (i.e., the values are transferred between 693 encapsulation headers and *not* copied from the encapsulated packet's 694 network-layer header). 696 The AERO interface encapsulates the packet per the base tunneling 697 specification except that it inserts a UDP header following the 698 encapsulation IP header and before any other encapsulation headers. 699 The AERO interface sets the UDP source port to a constant value that 700 it will use in each successive packet it sends and sets the UDP 701 length field to the length of the encapsulated packet plus 8 bytes 702 for the UDP header itself. For packets sent via a Server, the AERO 703 interface sets the UDP destination port to 8060, i.e., the IANA- 704 registered port number for AERO. For packets sent to a correspondent 705 Client, the AERO interface sets the UDP destination port to the port 706 value stored in the neighbor cache entry for this correspondent. The 707 AERO interface also sets the UDP checksum field to zero (see: 708 [RFC6935][RFC6936]) unless an integrity check is required (see: 709 Section 3.9.2). 711 The AERO interface next sets the IP protocol number in the 712 encapsulation header to 17 (i.e., the IP protocol number for UDP). 713 When IPv6 is used as the encapsulation protocol, the interface then 714 sets the flow label value in the encapsulation header the same as 715 described in [RFC6438]. When IPv4 is used as the encapsulation 716 protocol, the AERO interface sets the DF bit as discussed in 717 Section 3.9. 719 AERO interfaces decapsulate packets destined either to the node 720 itself or to a destination reached via an interface other than the 721 AERO interface the packet was received on. When the AERO interface 722 receives a UDP packet, it examines the first octet of the 723 encapsulated packet. The packet is accepted if the most significant 724 four bits of the first octet encode the value '0110' (i.e., the 725 version number value for IPv6) or the value '0100' (i.e., the version 726 number value for IPv4). Otherwise, the packet is accepted if the 727 first octet encodes a valid IP protocol number per the IANA 728 "protocol-numbers" registry that matches a supported encapsulation 729 type. Otherwise, the packet is discarded. 731 Further decapsulation then proceeds according to the appropriate base 732 tunneling specification. 734 3.8. AERO Interface Data Origin Authentication 736 AERO nodes employ simple data origin authentication procedures for 737 encapsulated packets they receive from other nodes on the AERO link. 738 In particular: 740 o AERO Relays and Servers accept encapsulated packets with a link- 741 layer source address that matches a permanent neighbor cache 742 entry. 744 o AERO Servers accept authentic encapsulated DHCPv6 messages, and 745 create or update a static neighbor cache entry for the source 746 based on the specific message type. 748 o AERO Servers accept encapsulated packets if there is a static 749 neighbor cache entry with an AERO address that matches the 750 packet's network-layer source address and with a link-layer 751 address that matches the packet's link-layer source address. 753 o AERO Clients accept encapsulated packets if there is a static 754 neighbor cache entry with a link-layer source address that matches 755 the packet's link-layer source address. 757 o AERO Clients and Servers accept encapsulated packets if there is a 758 dynamic neighbor cache entry with an AERO address that matches the 759 packet's network-layer source address, with a link-layer address 760 that matches the packet's link-layer source address, and with a 761 non-zero AcceptTime. 763 Note that this simple data origin authentication only applies to 764 environments in which link-layer addresses cannot be spoofed. 765 Additional security mitigations may be necessary in other 766 environments. 768 3.9. AERO Interface MTU and Fragmentation 770 The AERO interface is the node's point of attachment to the AERO 771 link. AERO links over IP networks have a maximum link MTU of 64KB 772 minus the encapsulation overhead (termed here "ENCAPS"), since the 773 maximum packet size in the base IP specifications is 64KB 774 [RFC0791][RFC2460] (while IPv6 jumbograms can be up to 4GB, they are 775 considered optional for IPv6 nodes [RFC2675][RFC6434]). 777 IPv6 specifies a minimum link MTU of 1280 bytes [RFC2460]. This is 778 the minimum packet size the AERO interface MUST admit without 779 returning an ICMP Packet Too Big (PTB) message. Although IPv4 780 specifies a smaller minimum link MTU of 68 bytes [RFC0791], AERO 781 interfaces also observe a 1280 byte minimum for IPv4. Additionally, 782 the vast majority of links in the Internet configure an MTU of at 783 least 1500 bytes. Original source hosts have therefore become 784 conditioned to expect that IP packets up to 1500 bytes in length will 785 either be delivered to the final destination or a suitable PTB 786 message returned. However, PTB messages may be lost in the network 787 [RFC2923] resulting in failure of the IP MTU discovery mechanisms 788 [RFC1191][RFC1981]. 790 For these reasons, AERO interfaces MUST admit packets up to 1500 791 bytes in length even if some fragmentation is necessary. AERO 792 interfaces MAY admit even larger packets as long as they can be 793 accommodated without fragmentation. 795 For AERO links over IPv4, the IP ID field is only 16 bits in length, 796 meaning that fragmentation at high data rates could result in data 797 corruption due to reassembly misassociations [RFC6864][RFC4963] (see: 798 Section 3.9.2). For AERO links over both IPv4 and IPv6, studies have 799 also shown that IP fragments are dropped unconditionally over some 800 network paths [I-D.taylor-v6ops-fragdrop]. For these reasons, when 801 fragmentation is needed it is performed within the AERO interface 802 (i.e., instead of at the encapsulating IP layer) through the 803 insertion of an IPv6 Fragment Header [RFC2460]. Since the Fragment 804 Header reduces the room available for packet data, but the original 805 source has no way to control its insertion, the Fragment Header 806 length MUST be included in the ENCAPS length even for packets in 807 which the header does not appear. 809 The source AERO interface (i.e., the tunnel ingress) therefore sends 810 encapsulated packets to the destination AERO interface (i.e., the 811 tunnel egress) according to the following algorithm: 813 o For IP packets that are no larger than (1280-ENCAPS) bytes, the 814 tunnel ingress encapsulates the packet and admits it into the 815 tunnel without fragmentation. For IPv4 AERO links, tunnel ingress 816 sets the Don't Fragment (DF) bit to 0 so that these packets will 817 be delivered to the tunnel egress even if there is a restricting 818 link in the path, i.e., unless lost due to congestion or routing 819 errors. 821 o For IP packets that are larger than (1280-ENCAPS) bytes but no 822 larger than 1500 bytes, the tunnel ingress encapsulates the packet 823 and inserts a Fragment Header above the UDP/IP encapsulation 824 headers. Next, the tunnel ingress uses the fragmentation 825 algorithm in [RFC2460] to break the packet into two non- 826 overlapping fragments where the first fragment (including ENCAPS) 827 is no larger than 1024 bytes and the second is no larger than the 828 first. Each fragment consists of identical UDP/IP encapsulation 829 headers, followed by the Fragment Header followed by the fragment 830 of the encapsulated packet itself. The tunnel ingress then admits 831 both fragments into the tunnel, and for IPv4 sets the DF bit to 0 832 in the IP encapsulation header. These fragmented encapsulated 833 packets will be delivered to the tunnel egress. 835 o For IPv4 packets that are larger than 1500 bytes and with the DF 836 bit set to 0, the tunnel ingress uses ordinary IP fragmentation to 837 break the unencapsulated packet into a minimum number of non- 838 overlapping fragments where the first fragment is no larger than 839 1024-ENCAPS and all other fragments are no larger than the first 840 fragment. The tunnel ingress then encapsulates each fragment (and 841 for IPv4 sets the DF bit to 0) then admits them into the tunnel. 842 These encapsulated fragments will be delivered to the final 843 destination via the tunnel egress. 845 o For all other IP packets, if the packet is too large to enter the 846 underlying interface following encapsulation, the tunnel ingress 847 drops the packet and returns a network-layer (L3) PTB message to 848 the original source with MTU set to the larger of 1500 bytes or 849 the underlying interface MTU minus ENCAPS. Otherwise, the tunnel 850 ingress encapsulates the packet and admits it into the tunnel 851 without fragmentation (and for IPv4 sets the DF bit to 1) and 852 translates any link-layer (L2) PTB messages it may receive from 853 the network into corresponding L3 PTB messages to send to the 854 original source as specified in Section 3.10. Since both L2 and 855 L3 PTB messages may be either lost or contain insufficient 856 information, however, it is RECOMMENDED that original sources that 857 send unfragmentable IP packets larger than 1500 bytes use 858 Packetization Layer Path MTU Discovery (PLPMTUD) [RFC4821]. 860 While sending packets according to the above algorithm, the tunnel 861 ingress MAY also send 1500 byte probe packets to determine whether 862 they can reach the tunnel egress without fragmentation. If the 863 probes succeed, the tunnel ingress can begin sending packets that are 864 no larger than 1500 bytes without fragmentation (and for IPv4 with DF 865 set to 1). Since the path MTU within the tunnel may fluctuate due to 866 routing changes, the tunnel ingress SHOULD continue to send 867 additional probes subject to rate limiting and SHOULD process any L2 868 PTB messages as an indication that the path MTU may have decreased. 869 If the path MTU within the tunnel becomes insufficient, the source 870 MUST resume fragmentation. 872 To construct a probe, the tunnel ingress prepares an NS message with 873 a Nonce option plus trailing NULL padding octets added to a length of 874 1500 bytes without including the length of the padding in the IPv6 875 Payload Length field, but with the length included in the 876 encapsulating IP header. The tunnel ingress then encapsulates the 877 padded NS message in the encapsulation headers (and for IPv4 sets DF 878 to 1) then sends the message to the tunnel egress. If the tunnel 879 egress returns a solicited NA message with a matching Nonce option, 880 the tunnel ingress deems the probe successful. 882 When the tunnel egress receives the fragments of a fragmented packet, 883 it reassembles them into a whole packet per the reassembly algorithm 884 in [RFC2460] then discards the Fragment Header. The tunnel egress 885 therefore MUST be capable of reassembling packets up to 1500+ENCAPS 886 bytes in length; hence, it is RECOMMENDED that the tunnel egress be 887 capable of reassembling at least 2KB. 889 3.9.1. Accommodating Large IPv6 ND and DHCPv6 Messages 891 IPv6 ND and DHCPv6 messages MUST be accommodated even if some 892 fragmentation is necessary. These packets are therefore accommodated 893 through a modification of the second rule in the above algorithm as 894 follows: 896 o For IPv6 ND and DHCPv6 messages that are larger than (1280-ENCAPS) 897 bytes, the tunnel ingress encapsulates the packet and inserts a 898 Fragment Header above the UDP/IP encapsulation headers. Next, the 899 tunnel ingress uses the fragmentation algorithm in [RFC2460] to 900 break the packet into a minimum number of non-overlapping 901 fragments where the first fragment (including ENCAPS) is no larger 902 than 1024 bytes and the remaining fragments are no larger than the 903 first. The tunnel ingress then encapsulates each fragment (and 904 for IPv4 sets the DF bit to 0) then admits them into the tunnel. 906 IPv6 ND and DHCPv6 messages that exceed the minimum reassembly size 907 listed above rarely occur in the modern era, however the tunnel 908 egress SHOULD be able to reassemble them if they do. This means that 909 the tunnel egress SHOULD include a configuration knob allowing the 910 operator to set a larger reassembly buffer size if large IPv6ND and 911 DHCPv6 messages become more common in the future. 913 The tunnel ingress can send large IPv6 ND and DHCPv6 messages without 914 fragmentation if there is assurance that large packets can traverse 915 the tunnel without fragmentation. The tunnel ingress MAY send probe 916 packets of 1500 bytes or larger as specified above to determine a 917 size for which fragmentation can be avoided. 919 3.9.2. Integrity 921 When fragmentation is needed, there must be assurance that reassembly 922 can be safely conducted without incurring data corruption. Sources 923 of corruption can include implementation errors, memory errors and 924 misassociation of fragments from a first datagram with fragments of 925 another datagram. The first two conditions (implementation and 926 memory errors) are mitigated by modern systems and implementations 927 that have demonstrated integrity through decades of operational 928 practice. The third condition (reassembly misassociations) must be 929 accounted for by AERO. 931 The AERO fragmentation procedure described in the above algorithms 932 uses the IPv6 Fragment Header and reuses standard IPv6 fragmentation 933 and reassembly code. Since the Fragment Header includes a 32-bit ID 934 field, there would need to be 2^32 packets alive in the network 935 before a second packet with a duplicate ID enters the system with the 936 (remote) possibility for a reassembly misassociation. For 1280 byte 937 packets, and for a maximum network lifetime value of 60 938 seconds[RFC2460], this means that the tunnel ingress would need to 939 produce ~(7 *10^12) bits/sec in order for a duplication event to be 940 possible. This exceeds the bandwidth of data link technologies of 941 the modern era, but not necessarily so going forward into the future. 942 Although typical wireless data links used by AERO Clients support 943 vastly lower data rates, the aggregate data rates between AERO 944 Servers and Relays may be substantial. However, high speed data 945 links in the network core are expected to configure larger MTUs, 946 e.g., 4KB, 8KB or even larger. Hence, no integrity check is included 947 to cover the AERO fragmentation and reassembly procedures. 949 When the tunnel ingress sends an IPv4-encapsulated packet with the DF 950 bit set to 0 in the above algorithms, there is a chance that the 951 packet may be fragmented by an IPv4 router somewhere within the 952 tunnel. Since the largest such packet is only 1280 bytes, however, 953 it is very likely that the packet will traverse the tunnel without 954 incurring a restricting link. Even when a link within the tunnel 955 configures an MTU smaller than 1280 bytes, it is very likely that it 956 does so due to limited performance characteristics [RFC3819]. This 957 means that the tunnel would not be able to convey fragmented 958 IPv4-encapsulated packets fast enough to produce reassembly 959 misassociations, as discussed above. However, AERO must also account 960 for the possibility of tunnel paths that include "poorly managed" 961 IPv4 link MTUs. 963 Since the IPv4 header includes only a 16-bit ID field, there would 964 only need to be 2^16 packets alive in the network before a second 965 packet with a duplicate ID enters the system. For 1280 byte packets, 966 and for a maximum network lifetime value of 120 seconds[RFC0791], 967 this means that the tunnel ingress would only need to produce ~(5 968 *10^6) bits/sec in order for a duplication event to be possible - a 969 value that is well within range for many modern wired and wireless 970 data link technologies. 972 Therefore, if there is strong operational assurance that no IPv4 973 links capable of supporting data rates of 5Mbps or more configure an 974 MTU smaller than 1280 the tunnel ingress MAY omit an integrity check 975 for the IPv4 fragmentation and reassembly procedures; otherwise, the 976 tunnel ingress SHOULD include an integrity check. When an upper 977 layer encapsulation (e.g., IPsec) already includes an integrity 978 check, the tunnel ingress need not include an additional check. 979 Otherwise, the tunnel ingress calculates the UDP checksum over the 980 encapsulated packet and writes the value into the UDP encapsulation 981 header, i.e., instead of writing the value 0. The tunnel egress will 982 then verify the UDP checksum and discard the packet if the checksum 983 is incorrect. 985 3.10. AERO Interface Error Handling 987 When an AERO node admits encapsulated packets into the AERO 988 interface, it may receive link-layer (L2) or network-layer (L3) error 989 indications. 991 An L2 error indication is an ICMP error message generated by a router 992 on the path to the neighbor or by the neighbor itself. The message 993 includes an IP header with the address of the node that generated the 994 error as the source address and with the link-layer address of the 995 AERO node as the destination address. 997 The IP header is followed by an ICMP header that includes an error 998 Type, Code and Checksum. For ICMPv6 [RFC4443], the error Types 999 include "Destination Unreachable", "Packet Too Big (PTB)", "Time 1000 Exceeded" and "Parameter Problem". For ICMPv4 [RFC0792], the error 1001 Types include "Destination Unreachable", "Fragmentation Needed" (a 1002 Destination Unreachable Code that is analogous to the ICMPv6 PTB), 1003 "Time Exceeded" and "Parameter Problem". 1005 The ICMP header is followed by the leading portion of the packet that 1006 generated the error, also known as the "packet-in-error". For 1007 ICMPv6, [RFC4443] specifies that the packet-in-error includes: "As 1008 much of invoking packet as possible without the ICMPv6 packet 1009 exceeding the minimum IPv6 MTU" (i.e., no more than 1280 bytes). For 1010 ICMPv4, [RFC0792] specifies that the packet-in-error includes: 1011 "Internet Header + 64 bits of Original Data Datagram", however 1012 [RFC1812] Section 4.3.2.3 updates this specification by stating: "the 1013 ICMP datagram SHOULD contain as much of the original datagram as 1014 possible without the length of the ICMP datagram exceeding 576 1015 bytes". 1017 The L2 error message format is shown in Figure 3: 1019 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1020 ~ ~ 1021 | L2 IP Header of | 1022 | error message | 1023 ~ ~ 1024 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1025 | L2 ICMP Header | 1026 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1027 ~ ~ P 1028 | IP and other encapsulation | a 1029 | headers of original L3 packet | c 1030 ~ ~ k 1031 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ e 1032 ~ ~ t 1033 | IP header of | 1034 | original L3 packet | i 1035 ~ ~ n 1036 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1037 ~ ~ e 1038 | Upper layer headers and | r 1039 | leading portion of body | r 1040 | of the original L3 packet | o 1041 ~ ~ r 1042 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1044 Figure 3: AERO Interface L2 Error Message Format 1046 The AERO node rules for processing these L2 error messages is as 1047 follows: 1049 o When an AERO node receives an L2 Parameter Problem message, it 1050 processes the message the same as described as for ordinary ICMP 1051 errors in the normative references [RFC0792][RFC4443]. 1053 o When an AERO node receives persistent L2 IPv4 Time Exceeded 1054 messages, the IP ID field may be wrapping before earlier fragments 1055 have been processed. In that case, the node SHOULD begin 1056 including IPv4 integrity checks (see: Section 3.9.2). 1058 o When an AERO Client receives persistent L2 Destination Unreachable 1059 messages in response to tunneled packets that it sends to one of 1060 its dynamic neighbor correspondents, the Client SHOULD test the 1061 path to the correspondent using Neighbor Unreachability Detection 1062 (NUD) (see Section 3.14). If NUD fails, the Client SHOULD set 1063 ForwardTime for the corresponding dynamic neighbor cache entry to 1064 0 and allow future packets destined to the correspondent to flow 1065 through a Server. 1067 o When an AERO Client receives persistent L2 Destination Unreachable 1068 messages in response to tunneled packets that it sends to one of 1069 its static neighbor Servers, the Client SHOULD test the path to 1070 the Server using NUD. If NUD fails, the Client SHOULD delete the 1071 neighbor cache entry and attempt to associate with a new Server. 1073 o When an AERO Server receives persistent L2 Destination Unreachable 1074 messages in response to tunneled packets that it sends to one of 1075 its static neighbor Clients, the Server SHOULD test the path to 1076 the Client using NUD. If NUD fails, the Server SHOULD cancel the 1077 DHCPv6 PD lease for the Client's ACP, withdraw its route for the 1078 ACP from the AERO routing system and delete the neighbor cache 1079 entry (see Sections 3.11 and 3.12). 1081 o When an AERO Relay or Server receives an L2 Destination 1082 Unreachable message in response to a tunneled packet that it sends 1083 to one of its permanent neighbors, it discards the message since 1084 the routing system is likely in a temporary transitional state 1085 that will soon re-converge. 1087 o When an AERO node receives an L2 PTB message, it translates the 1088 message into an L3 PTB message if possible (*) and forwards the 1089 message toward the original source as described below. 1091 To translate an L2 PTB message to an L3 PTB message, the AERO node 1092 first caches the MTU field value of the L2 ICMP header. The node 1093 next discards the L2 IP and ICMP headers, and also discards the 1094 encapsulation headers of the original L3 packet. Next the node 1095 encapsulates the included segment of the original L3 packet in an L3 1096 IP and ICMP header, and sets the ICMP header Type and Code values to 1097 appropriate values for the L3 IP protocol. In the process, the node 1098 writes the maximum of 1500 bytes and (L2 MTU - ENCAPS) into the MTU 1099 field of the L3 ICMP header. 1101 The node next writes the IP source address of the original L3 packet 1102 as the destination address of the L3 PTB message and determines the 1103 next hop to the destination. If the next hop is reached via the AERO 1104 interface, the node uses the IPv6 address "::" or the IPv4 address 1105 "0.0.0.0" as the IP source address of the L3 PTB message. Otherwise, 1106 the node uses one of its non link-local addresses as the source 1107 address of the L3 PTB message. The node finally calculates the ICMP 1108 checksum over the L3 PTB message and writes the Checksum in the 1109 corresponding field of the L3 ICMP header. The L3 PTB message 1110 therefore is formatted as follows: 1112 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1113 ~ ~ 1114 | L3 IP Header of | 1115 | error message | 1116 ~ ~ 1117 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1118 | L3 ICMP Header | 1119 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1120 ~ ~ p 1121 | IP header of | k 1122 | original L3 packet | t 1123 ~ ~ 1124 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ i 1125 ~ ~ n 1126 | Upper layer headers and | 1127 | leading portion of body | e 1128 | of the original L3 packet | r 1129 ~ ~ r 1130 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1132 Figure 4: AERO Interface L3 Error Message Format 1134 After the node has prepared the L3 PTB message, it either forwards 1135 the message via a link outside of the AERO interface without 1136 encapsulation, or encapsulates and forwards the message to the next 1137 hop via the AERO interface. 1139 When an AERO Relay receives an L3 packet for which the destination 1140 address is covered by an ASP, if there is no more-specific routing 1141 information for the destination the Relay drops the packet and 1142 returns an L3 Destination Unreachable message. The Relay first 1143 writes the IP source address of the original L3 packet as the 1144 destination address of the L3 Destination Unreachable message and 1145 determines the next hop to the destination. If the next hop is 1146 reached via the AERO interface, the Relay uses the IPv6 address "::" 1147 or the IPv4 address "0.0.0.0" as the IP source address of the L3 1148 Destination Unreachable message and forwards the message to the next 1149 hop within the AERO interface. Otherwise, the Relay uses one of its 1150 non link-local addresses as the source address of the L3 Destination 1151 Unreachable message and forwards the message via a link outside the 1152 AERO interface. 1154 When an AERO node receives any L3 error message via the AERO 1155 interface, it examines the destination address in the L3 IP header of 1156 the message. If the next hop toward the destination address of the 1157 error message is via the AERO interface, the node re-encapsulates and 1158 forwards the message to the next hop within the AERO interface. 1159 Otherwise, if the source address in the L3 IP header of the message 1160 is the IPv6 address "::" or the IPv4 address "0.0.0.0", the node 1161 writes one of its non link-local addresses as the source address of 1162 the L3 message and recalculates the IP and/or ICMP checksums. The 1163 node finally forwards the message via a link outside of the AERO 1164 interface. 1166 (*) Note that in some instances the packet-in-error field of an L2 1167 PTB message may not include enough information for translation to an 1168 L3 PTB message. In that case, the AERO interface simply discards the 1169 L2 PTB message. It can therefore be said that translation of L2 PTB 1170 messages to L3 PTB messages can provide a useful optimization when 1171 possible, but is not critical for sources that correctly use PLPMTUD. 1173 3.11. AERO Router Discovery, Prefix Delegation and Address 1174 Configuration 1176 3.11.1. AERO DHCPv6 Service Model 1178 Each AERO Server configures a DHCPv6 server function to facilitate PD 1179 requests from Clients. Each Server is pre-configured with an 1180 identical list of ACP-to-Client ID mappings for all Clients enrolled 1181 in the AERO system, as well as any information necessary to 1182 authenticate Clients. The configuration information is maintained by 1183 a central administrative authority for the AERO link and securely 1184 propagated to all Servers whenever a new Client is enrolled or an 1185 existing Client is withdrawn. 1187 With these identical configurations, each Server can function 1188 independently of all other Servers, including the maintenance of 1189 active leases. Therefore, no Server-to-Server DHCPv6 state 1190 synchronization is necessary, and Clients can optionally hold 1191 separate leases for the same ACP from multiple Servers. 1193 In this way, Clients can easily associate with multiple Servers, and 1194 can receive new leases from new Servers before deprecating leases 1195 held through old Servers. This enables a graceful "make-before- 1196 break" capability. 1198 3.11.2. AERO Client Behavior 1200 AERO Clients discover the link-layer addresses of AERO Servers via 1201 static configuration, or through an automated means such as DNS name 1202 resolution. In the absence of other information, the Client resolves 1203 the Fully-Qualified Domain Name (FQDN) "linkupnetworks.[domainname]" 1204 where "linkupnetworks" is a constant text string and "[domainname]" 1205 is the connection-specific DNS suffix for the Client's underlying 1206 network connection (e.g., "example.com"). After discovering the 1207 link-layer addresses, the Client associates with one or more of the 1208 corresponding Servers. 1210 To associate with a Server, the Client acts as a requesting router to 1211 request an ACP through a DHCPv6 PD exchange[RFC3315][RFC3633] in 1212 which the Client's Solicit/Request messages use the IPv6 1213 "unspecified" address (i.e., "::") as the IPv6 source address, 1214 'All_DHCP_Relay_Agents_and_Servers' as the IPv6 destination address 1215 and the link-layer address of the Server as the link-layer 1216 destination address. The Client also includes a Client Identifier 1217 option with a DHCP Unique Identifier (DUID) plus any necessary 1218 authentication options to identify itself to the DHCPv6 server, and 1219 includes a Client Link Layer Address Option (CLLAO) [RFC6939] with 1220 the format shown in Figure 5: 1222 0 1 2 3 1223 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1224 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1225 | OPTION_CLIENT_LINKLAYER_ADDR | option-length | 1226 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1227 | link-layer type (16 bits) | Link ID | Preference | 1228 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1230 Figure 5: AERO Client Link-Layer Address Option (CLLAO) Format 1232 The Client sets the CLLAO 'option-length' field to 4 and sets the 1233 'link-layer type' field to TBD1 (see: IANA Considerations), then 1234 includes appropriate Link ID and Preference values for the underlying 1235 interface over which the Solicit/Request will be issued (note that 1236 these are the same values that would be included in a TLLAO as shown 1237 in Figure 2). If the Client is pre-provisioned with an ACP 1238 associated with the AERO service, it MAY also include the ACP in the 1239 Solicit/Request message Identity Association (IA) option to indicate 1240 its preferred ACP to the DHCPv6 server. The Client then sends the 1241 encapsulated DHCPv6 request via the underlying interface. 1243 When the Client receives its ACP and the set of ASPs via a DHCPv6 1244 Reply from the AERO Server, it creates a static neighbor cache entry 1245 with the Server's link-local address as the network-layer address and 1246 the Server's encapsulation address as the link-layer address. The 1247 Client then records the lifetime for the ACP in the neighbor cache 1248 entry and marks the neighbor cache entry as "default", i.e., the 1249 Client considers the Server as a default router. If the Reply 1250 message contains a Vendor-Specific Information Option (see: 1251 Section 3.11.3) the Client also caches each ASP in the option. 1253 The Client then applies the AERO address to the AERO interface and 1254 sub-delegates the ACP to nodes and links within its attached EUNs 1255 (the AERO address thereafter remains stable as the Client moves). 1256 The Client also assigns a default IP route to the AERO interface as a 1257 route-to-interface, i.e., with no explicit next-hop. The next hop 1258 will then be determined after a packet has been submitted to the AERO 1259 interface by inspecting the neighbor cache (see above). 1261 On some platforms (e.g., popular cell phone operating systems), the 1262 act of assigning a default IPv6 route to the AERO interface may not 1263 be permitted from a user application due to security policy. 1264 Typically, those platforms include a TUN/TAP interface that acts as a 1265 point-to-point conduit between user applications and the AERO 1266 interface. In that case, the Client can instead generate a 1267 "synthesized RA" message. The message conforms to [RFC4861] and is 1268 prepared as follows: 1270 o the IPv6 source address is fe80:: 1272 o the IPv6 destination address is all-nodes multicast 1274 o the Router Lifetime is set to a time that is no longer than the 1275 ACP DHCPv6 lifetime 1277 o the message does not include a Source Link Layer Address Option 1278 (SLLAO) 1280 o the message includes a Prefix Information Option (PIO) with a /64 1281 prefix taken from the ACP as the prefix for autoconfiguration 1283 The Client then sends the synthesized RA message via the TUN/TAP 1284 interface, where the operating system kernel will interpret it as 1285 though it were generated by an actual router. The operating system 1286 will then install a default route and use StateLess Address 1287 AutoConfiguration (SLAAC) to configure an IPv6 address on the TUN/TAP 1288 interface. Methods for similarly installing an IPv4 default route 1289 and IPv4 address on the TUN/TAP interface are based on synthesized 1290 DHCPv4 messages [RFC2131]. Note that in this method, the Client 1291 appears as a mobility proxy for applications that bind to the (point- 1292 to-point) TUN/TAP interface. The arrangement can be likened to a 1293 Proxy AERO scenario in which the mobile node and Client are located 1294 within the same physical platform (see Section 3.20 for further 1295 details on Proxy AERO). 1297 The Client subsequently renews its ACP delegation through each of its 1298 Servers by performing DHCPv6 Renew/Reply exchanges with its AERO 1299 address as the IPv6 source address, 1300 'All_DHCP_Relay_Agents_and_Servers' as the IPv6 destination address, 1301 the link-layer address of a Server as the link-layer destination 1302 address and the same Client identifier, authentication options and 1303 CLLAO option as was used in the initial PD request. Note that if the 1304 Client does not issue a DHCPv6 Renew before the Server has terminated 1305 the lease (e.g., if the Client has been out of touch with the Server 1306 for a considerable amount of time), the Server's Reply will report 1307 NoBinding and the Client must re-initiate the DHCPv6 PD procedure. 1308 If the Client sends synthesized RA and/or DHCPv4 messages (see 1309 above), it also sends a new synthesized message when issuing a DHCPv6 1310 Renew or when re-initiating the DHCPv6 PD procedure. 1312 Since the Client's AERO address is configured from the unique ACP 1313 delegation it receives, there is no need for Duplicate Address 1314 Detection (DAD) on AERO links. Other nodes maliciously attempting to 1315 hijack an authorized Client's AERO address will be denied access to 1316 the network by the DHCPv6 server due to an unacceptable link-layer 1317 address and/or security parameters (see: Security Considerations). 1319 AERO Clients ignore the IP address and UDP port number in any S/TLLAO 1320 options in ND messages they receive directly from another AERO 1321 Client, but examine the Link ID and Preference values to match the 1322 message with the correct link-layer address information. 1324 When a source Client forwards a packet to a prospective destination 1325 Client (i.e., one for which the packet's destination address is 1326 covered by an ASP), the source Client initiates an AERO route 1327 optimization procedure as specified in Section 3.13. 1329 3.11.3. AERO Server Behavior 1331 AERO Servers configure a DHCPv6 server function on their AERO links. 1332 AERO Servers arrange to add their encapsulation layer IP addresses 1333 (i.e., their link-layer addresses) to the DNS resource records for 1334 the FQDN "linkupnetworks.[domainname]" before entering service. 1336 When an AERO Server receives a prospective Client's DHCPv6 PD 1337 Solicit/Request message, it first authenticates the message. If 1338 authentication succeeds, the Server determines the correct ACP to 1339 delegate to the Client by matching the Client's DUID within an online 1340 directory service (e.g., LDAP). The Server then delegates the ACP 1341 and creates a static neighbor cache entry for the Client's AERO 1342 address with lifetime set to no more than the lease lifetime and the 1343 Client's link-layer address as the link-layer address for the Link ID 1344 specified in the CLLAO option. The Server then creates an IP 1345 forwarding table entry so that the AERO routing system will propagate 1346 the ACP to all Relays (see: Section 3.12). Finally, the Server sends 1347 a DHCPv6 Reply message to the Client while using fe80::ID as the IPv6 1348 source address, the Client's AERO address as the IPv6 destination 1349 address, and the Client's link-layer address as the destination link- 1350 layer address. The Server also includes a Server Unicast option with 1351 server-address set to fe80::ID so that all future Client/Server 1352 transactions will be link-local-only unicast over the AERO link. 1354 When the Server sends the DHCPv6 Reply message, it also includes a 1355 DHCPv6 Vendor-Specific Information Option with 'enterprise-number' 1356 set to "TBD2" (see: IANA Considerations). The option is formatted as 1357 shown in[RFC3315] and with the AERO enterprise-specific format shown 1358 in Figure 6: 1360 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1361 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1362 | OPTION_VENDOR_OPTS | option-len | 1363 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1364 | enterprise-number ("TBD2") | 1365 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1366 | Reserved | Prefix Length | 1367 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1368 | | 1369 + ASP (1) + 1370 | | 1371 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1372 | Reserved | Prefix Length | 1373 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1374 | | 1375 + ASP (2) + 1376 | | 1377 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1378 | Reserved | Prefix Length | 1379 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1380 | | 1381 + ASP (3) + 1382 | | 1383 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1384 . (etc.) . 1385 . . 1386 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1388 Figure 6: AERO Vendor-Specific Information Option 1390 Per Figure 6, the option includes one or more ASP. The ASP field 1391 contains the IP prefix as it would appear in the interface identifier 1392 portion of the corresponding AERO address (see: Section 3.3). For 1393 IPv6, valid values for the Prefix Length field are 0 through 64; for 1394 IPv4, valid values are 0 through 32. 1396 After the initial DHCPv6 PD exchange, the AERO Server maintains the 1397 neighbor cache entry for the Client as long as the lease lifetime 1398 remains current. If the Client issues a Renew/Reply exchange, the 1399 Server extends the lifetime. If the Client issues a Release/Reply 1400 exchange, or if the Client does not issue a Renew/Reply within the 1401 lease lifetime, the Server deletes the neighbor cache entry for the 1402 Client and withdraws the IP route from the AERO routing system. 1404 3.12. AERO Relay/Server Routing System 1406 Relays require full topology information of all Client/Server 1407 associations, while individual Servers only require partial topology 1408 information, i.e., they only need to know the ACPs associated with 1409 their current set of associated Clients. This is accomplished 1410 through the use of an internal instance of the Border Gateway 1411 Protocol (BGP) [RFC4271] coordinated between Servers and Relays. 1412 This internal BGP instance does not interact with the public Internet 1413 BGP instance; therefore, the AERO link is presented to the IP 1414 Internetwork as a small set of ASPs as opposed to the full set of 1415 individual ACPs. 1417 In a reference BGP arrangement, each AERO Server is configured as an 1418 Autonomous System Border Router (ASBR) for a stub Autonomous System 1419 (AS) (possibly using a private AS Number (ASN) [RFC1930]), and each 1420 Server further peers with each Relay but does not peer with other 1421 Servers. Similarly, Relays need not peer with each other, since they 1422 will receive all updates from all Servers and will therefore have a 1423 consistent view of the AERO link ACP delegations. 1425 Each Server maintains a working set of associated Clients, and 1426 dynamically announces new ACPs and withdraws departed ACPs in its BGP 1427 updates to Relays (this is typically accomplished via a "redistribute 1428 static" routing directive). Relays do not send BGP updates to 1429 Servers, however, such that the BGP route reporting is unidirectional 1430 from the Servers to the Relays. 1432 The Relays therefore discover the full topology of the AERO link in 1433 terms of the working set of ACPs associated with each Server, while 1434 the Servers only discover the ACPs of their associated Clients. 1435 Since Clients are expected to remain associated with their current 1436 set of Servers for extended timeframes, the amount of BGP control 1437 messaging between Servers and Relays should be minimal. However, BGP 1438 peers SHOULD dampen any route oscillations caused by impatient 1439 Clients that repeatedly associate and disassociate with Servers. 1441 3.13. AERO Redirection 1443 3.13.1. Reference Operational Scenario 1445 Figure 7 depicts the AERO redirection reference operational scenario, 1446 using IPv6 addressing as the example (while not shown, a 1447 corresponding example for IPv4 addressing can be easily constructed). 1448 The figure shows an AERO Relay ('R1'), two AERO Servers ('S1', 'S2'), 1449 two AERO Clients ('C1', 'C2') and two ordinary IPv6 hosts ('H1', 1450 'H2'): 1452 +--------------+ +--------------+ +--------------+ 1453 | Server S1 | | Relay R1 | | Server S2 | 1454 +--------------+ +--------------+ +--------------+ 1455 fe80::2 fe80::1 fe80::3 1456 L2(S1) L2(R1) L2(S2) 1457 | | | 1458 X-----+-----+------------------+-----------------+----+----X 1459 | AERO Link | 1460 L2(A) L2(B) 1461 fe80::2001:db8:0:0 fe80::2001:db8:1:0 1462 +--------------+ +--------------+ 1463 |AERO Client C1| |AERO Client C2| 1464 +--------------+ +--------------+ 1465 2001:DB8:0::/48 2001:DB8:1::/48 1466 | | 1467 .-. .-. 1468 ,-( _)-. 2001:db8:0::1 2001:db8:1::1 ,-( _)-. 1469 .-(_ IP )-. +---------+ +---------+ .-(_ IP )-. 1470 (__ EUN )--| Host H1 | | Host H2 |--(__ EUN ) 1471 `-(______)-' +---------+ +---------+ `-(______)-' 1473 Figure 7: AERO Reference Operational Scenario 1475 In Figure 7, Relay ('R1') applies the address fe80::1 to its AERO 1476 interface with link-layer address L2(R1), Server ('S1') applies the 1477 address fe80::2 with link-layer address L2(S1),and Server ('S2') 1478 applies the address fe80::3 with link-layer address L2(S2). Servers 1479 ('S1') and ('S2') next arrange to add their link-layer addresses to a 1480 published list of valid Servers for the AERO link. 1482 AERO Client ('C1') receives the ACP 2001:db8:0::/48 in a DHCPv6 PD 1483 exchange via AERO Server ('S1') then applies the address 1484 fe80::2001:db8:0:0 to its AERO interface with link-layer address 1485 L2(C1). Client ('C1') configures a default route and neighbor cache 1486 entry via the AERO interface with next-hop address fe80::2 and link- 1487 layer address L2(S1), then sub-delegates the ACP to its attached 1488 EUNs. IPv6 host ('H1') connects to the EUN, and configures the 1489 address 2001:db8:0::1. 1491 AERO Client ('C2') receives the ACP 2001:db8:1::/48 in a DHCPv6 PD 1492 exchange via AERO Server ('S2') then applies the address 1493 fe80::2001:db8:1:0 to its AERO interface with link-layer address 1494 L2(C2). Client ('C2') configures a default route and neighbor cache 1495 entry via the AERO interface with next-hop address fe80::3 and link- 1496 layer address L2(S2), then sub-delegates the ACP to its attached 1497 EUNs. IPv6 host ('H1') connects to the EUN, and configures the 1498 address 2001:db8:1::1. 1500 3.13.2. Concept of Operations 1502 Again, with reference to Figure 7, when source host ('H1') sends a 1503 packet to destination host ('H2'), the packet is first forwarded over 1504 the source host's attached EUN to Client ('C1'). Client ('C1') then 1505 forwards the packet via its AERO interface to Server ('S1') and also 1506 sends a Predirect message toward Client ('C2') via Server ('S1'). 1507 Server ('S1') then re-encapsulates and forwards both the packet and 1508 the Predirect message out the same AERO interface toward Client 1509 ('C2') via Relay ('R1'). 1511 When Relay ('R1') receives the packet and Predirect message, it 1512 consults its forwarding table to discover Server ('S2') as the next 1513 hop toward Client ('C2'). Relay ('R1') then forwards both the packet 1514 and the Predirect message to Server ('S2'), which then forwards them 1515 to Client ('C2'). 1517 After Client ('C2') receives the Predirect message, it process the 1518 message and returns a Redirect message toward Client ('C1') via 1519 Server ('S2'). During the process, Client ('C2') also creates or 1520 updates a dynamic neighbor cache entry for Client ('C1'). 1522 When Server ('S2') receives the Redirect message, it re-encapsulates 1523 the message and forwards it on to Relay ('R1'), which forwards the 1524 message on to Server ('S1') which forwards the message on to Client 1525 ('C1'). After Client ('C1') receives the Redirect message, it 1526 processes the message and creates or updates a dynamic neighbor cache 1527 entry for Client ('C2'). 1529 Following the above Predirect/Redirect message exchange, forwarding 1530 of packets from Client ('C1') to Client ('C2') without involving any 1531 intermediate nodes is enabled. The mechanisms that support this 1532 exchange are specified in the following sections. 1534 3.13.3. Message Format 1536 AERO Redirect/Predirect messages use the same format as for ICMPv6 1537 Redirect messages depicted in Section 4.5 of [RFC4861], but also 1538 include a new "Prefix Length" field taken from the low-order 8 bits 1539 of the Redirect message Reserved field. For IPv6, valid values for 1540 the Prefix Length field are 0 through 64; for IPv4, valid values are 1541 0 through 32. The Redirect/Predirect messages are formatted as shown 1542 in Figure 8: 1544 0 1 2 3 1545 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1546 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1547 | Type (=137) | Code (=0/1) | Checksum | 1548 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1549 | Reserved | Prefix Length | 1550 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1551 | | 1552 + + 1553 | | 1554 + Target Address + 1555 | | 1556 + + 1557 | | 1558 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1559 | | 1560 + + 1561 | | 1562 + Destination Address + 1563 | | 1564 + + 1565 | | 1566 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1567 | Options ... 1568 +-+-+-+-+-+-+-+-+-+-+-+- 1570 Figure 8: AERO Redirect/Predirect Message Format 1572 3.13.4. Sending Predirects 1574 When a Client forwards a packet with a source address from one of its 1575 ACPs toward a destination address covered by an ASP (i.e., toward 1576 another AERO Client connected to the same AERO link), the source 1577 Client MAY send a Predirect message forward toward the destination 1578 Client via the Server. 1580 In the reference operational scenario, when Client ('C1') forwards a 1581 packet toward Client ('C2'), it MAY also send a Predirect message 1582 forward toward Client ('C2'), subject to rate limiting (see 1583 Section 8.2 of [RFC4861]). Client ('C1') prepares the Predirect 1584 message as follows: 1586 o the link-layer source address is set to 'L2(C1)' (i.e., the link- 1587 layer address of Client ('C1')). 1589 o the link-layer destination address is set to 'L2(S1)' (i.e., the 1590 link-layer address of Server ('S1')). 1592 o the network-layer source address is set to fe80::2001:db8:0:0 1593 (i.e., the AERO address of Client ('C1')). 1595 o the network-layer destination address is set to fe80::2001:db8:1:0 1596 (i.e., the AERO address of Client ('C2')). 1598 o the Type is set to 137. 1600 o the Code is set to 1 to indicate "Predirect". 1602 o the Prefix Length is set to the length of the prefix to be 1603 assigned to the Target Address. 1605 o the Target Address is set to fe80::2001:db8:0:0 (i.e., the AERO 1606 address of Client ('C1')). 1608 o the Destination Address is set to the source address of the 1609 originating packet that triggered the Predirection event. (If the 1610 originating packet is an IPv4 packet, the address is constructed 1611 in IPv4-compatible IPv6 address format). 1613 o the message includes one or more TLLAOs with Link ID and 1614 Preference set to appropriate values for Client ('C1')'s 1615 underlying interfaces, and with UDP Port Number and IP Address set 1616 to 0'. 1618 o the message SHOULD include a Timestamp option and a Nonce option. 1620 o the message includes a Redirected Header Option (RHO) that 1621 contains the originating packet truncated if necessary to ensure 1622 that at least the network-layer header is included but the size of 1623 the message does not exceed 1280 bytes. 1625 Note that the act of sending Predirect messages is cited as "MAY", 1626 since Client ('C1') may have advanced knowledge that the direct path 1627 to Client ('C2') would be unusable or otherwise undesirable. If the 1628 direct path later becomes unusable after the initial route 1629 optimization, Client ('C1') simply allows packets to again flow 1630 through Server ('S1'). 1632 3.13.5. Re-encapsulating and Relaying Predirects 1634 When Server ('S1') receives a Predirect message from Client ('C1'), 1635 it first verifies that the TLLAOs in the Predirect are a proper 1636 subset of the Link IDs in Client ('C1')'s neighbor cache entry. If 1637 the Client's TLLAOs are not acceptable, Server ('S1') discards the 1638 message. Otherwise, Server ('S1') validates the message according to 1639 the ICMPv6 Redirect message validation rules in Section 8.1 of 1641 [RFC4861], except that the Predirect has Code=1. Server ('S1') also 1642 verifies that Client ('C1') is authorized to use the Prefix Length in 1643 the Predirect when applied to the AERO address in the network-layer 1644 source address by searching for the AERO address in the neighbor 1645 cache. If validation fails, Server ('S1') discards the Predirect; 1646 otherwise, it copies the correct UDP Port numbers and IP Addresses 1647 for Client ('C1')'s links into the (previously empty) TLLAOs. 1649 Server ('S1') then examines the network-layer destination address of 1650 the Predirect to determine the next hop toward Client ('C2') by 1651 searching for the AERO address in the neighbor cache. Since Client 1652 ('C2') is not one of its neighbors, Server ('S1') re-encapsulates the 1653 Predirect and relays it via Relay ('R1') by changing the link-layer 1654 source address of the message to 'L2(S1)' and changing the link-layer 1655 destination address to 'L2(R1)'. Server ('S1') finally forwards the 1656 re-encapsulated message to Relay ('R1') without decrementing the 1657 network-layer TTL/Hop Limit field. 1659 When Relay ('R1') receives the Predirect message from Server ('S1') 1660 it determines that Server ('S2') is the next hop toward Client ('C2') 1661 by consulting its forwarding table. Relay ('R1') then re- 1662 encapsulates the Predirect while changing the link-layer source 1663 address to 'L2(R1)' and changing the link-layer destination address 1664 to 'L2(S2)'. Relay ('R1') then relays the Predirect via Server 1665 ('S2'). 1667 When Server ('S2') receives the Predirect message from Relay ('R1') 1668 it determines that Client ('C2') is a neighbor by consulting its 1669 neighbor cache. Server ('S2') then re-encapsulates the Predirect 1670 while changing the link-layer source address to 'L2(S2)' and changing 1671 the link-layer destination address to 'L2(C2)'. Server ('S2') then 1672 forwards the message to Client ('C2'). 1674 3.13.6. Processing Predirects and Sending Redirects 1676 When Client ('C2') receives the Predirect message, it accepts the 1677 Predirect only if the message has a link-layer source address of one 1678 of its Servers (e.g., L2(S2)). Client ('C2') further accepts the 1679 message only if it is willing to serve as a redirection target. 1680 Next, Client ('C2') validates the message according to the ICMPv6 1681 Redirect message validation rules in Section 8.1 of [RFC4861], except 1682 that it accepts the message even though Code=1 and even though the 1683 network-layer source address is not that of it's current first-hop 1684 router. 1686 In the reference operational scenario, when Client ('C2') receives a 1687 valid Predirect message, it either creates or updates a dynamic 1688 neighbor cache entry that stores the Target Address of the message as 1689 the network-layer address of Client ('C1') , stores the link-layer 1690 addresses found in the TLLAOs as the link-layer addresses of Client 1691 ('C1') and stores the Prefix Length as the length to be applied to 1692 the network-layer address for forwarding purposes. Client ('C2') 1693 then sets AcceptTime for the neighbor cache entry to ACCEPT_TIME. 1695 After processing the message, Client ('C2') prepares a Redirect 1696 message response as follows: 1698 o the link-layer source address is set to 'L2(C2)' (i.e., the link- 1699 layer address of Client ('C2')). 1701 o the link-layer destination address is set to 'L2(S2)' (i.e., the 1702 link-layer address of Server ('S2')). 1704 o the network-layer source address is set to fe80::2001:db8:1:0 1705 (i.e., the AERO address of Client ('C2')). 1707 o the network-layer destination address is set to fe80::2001:db8:0:0 1708 (i.e., the AERO address of Client ('C1')). 1710 o the Type is set to 137. 1712 o the Code is set to 0 to indicate "Redirect". 1714 o the Prefix Length is set to the length of the prefix to be applied 1715 to the Target Address. 1717 o the Target Address is set to fe80::2001:db8:1:0 (i.e., the AERO 1718 address of Client ('C2')). 1720 o the Destination Address is set to the destination address of the 1721 originating packet that triggered the Redirection event. (If the 1722 originating packet is an IPv4 packet, the address is constructed 1723 in IPv4-compatible IPv6 address format). 1725 o the message includes one or more TLLAOs with Link ID and 1726 Preference set to appropriate values for Client ('C2')'s 1727 underlying interfaces, and with UDP Port Number and IP Address set 1728 to '0'. 1730 o the message SHOULD include a Timestamp option and MUST echo the 1731 Nonce option received in the Predirect (i.e., if a Nonce option is 1732 included). 1734 o the message includes as much of the RHO copied from the 1735 corresponding AERO Predirect message as possible such that at 1736 least the network-layer header is included but the size of the 1737 message does not exceed 1280 bytes. 1739 After Client ('C2') prepares the Redirect message, it sends the 1740 message to Server ('S2'). 1742 3.13.7. Re-encapsulating and Relaying Redirects 1744 When Server ('S2') receives a Redirect message from Client ('C2'), it 1745 first verifies that the TLLAOs in the Redirect are a proper subset of 1746 the Link IDs in Client ('C2')'s neighbor cache entry. If the 1747 Client's TLLAOs are not acceptable, Server ('S2') discards the 1748 message. Otherwise, Server ('S2') validates the message according to 1749 the ICMPv6 Redirect message validation rules in Section 8.1 of 1750 [RFC4861]. Server ('S2') also verifies that Client ('C2') is 1751 authorized to use the Prefix Length in the Redirect when applied to 1752 the AERO address in the network-layer source address by searching for 1753 the AERO address in the neighbor cache. If validation fails, Server 1754 ('S2') discards the Predirect; otherwise, it copies the correct UDP 1755 Port numbers and IP Addresses for Client ('C2')'s links into the 1756 (previously empty) TLLAOs. 1758 Server ('S2') then examines the network-layer destination address of 1759 the Predirect to determine the next hop toward Client ('C2') by 1760 searching for the AERO address in the neighbor cache. Since Client 1761 ('C2') is not a neighbor, Server ('S2') re-encapsulates the Predirect 1762 and relays it via Relay ('R1') by changing the link-layer source 1763 address of the message to 'L2(S2)' and changing the link-layer 1764 destination address to 'L2(R1)'. Server ('S2') finally forwards the 1765 re-encapsulated message to Relay ('R1') without decrementing the 1766 network-layer TTL/Hop Limit field. 1768 When Relay ('R1') receives the Predirect message from Server ('S2') 1769 it determines that Server ('S1') is the next hop toward Client ('C1') 1770 by consulting its forwarding table. Relay ('R1') then re- 1771 encapsulates the Predirect while changing the link-layer source 1772 address to 'L2(R1)' and changing the link-layer destination address 1773 to 'L2(S1)'. Relay ('R1') then relays the Predirect via Server 1774 ('S1'). 1776 When Server ('S1') receives the Predirect message from Relay ('R1') 1777 it determines that Client ('C1') is a neighbor by consulting its 1778 neighbor cache. Server ('S1') then re-encapsulates the Predirect 1779 while changing the link-layer source address to 'L2(S1)' and changing 1780 the link-layer destination address to 'L2(C1)'. Server ('S1') then 1781 forwards the message to Client ('C1'). 1783 3.13.8. Processing Redirects 1785 When Client ('C1') receives the Redirect message, it accepts the 1786 message only if it has a link-layer source address of one of its 1787 Servers (e.g., ''L2(S1)'). Next, Client ('C1') validates the message 1788 according to the ICMPv6 Redirect message validation rules in 1789 Section 8.1 of [RFC4861], except that it accepts the message even 1790 though the network-layer source address is not that of it's current 1791 first-hop router. Following validation, Client ('C1') then processes 1792 the message as follows. 1794 In the reference operational scenario, when Client ('C1') receives 1795 the Redirect message, it either creates or updates a dynamic neighbor 1796 cache entry that stores the Target Address of the message as the 1797 network-layer address of Client ('C2'), stores the link-layer 1798 addresses found in the TLLAOs as the link-layer addresses of Client 1799 ('C2') and stores the Prefix Length as the length to be applied to 1800 the network-layer address for forwarding purposes. Client ('C1') 1801 then sets ForwardTime for the neighbor cache entry to FORWARD_TIME. 1803 Now, Client ('C1') has a neighbor cache entry with a valid 1804 ForwardTime value, while Client ('C2') has a neighbor cache entry 1805 with a valid AcceptTime value. Thereafter, Client ('C1') may forward 1806 ordinary network-layer data packets directly to Client ('C2') without 1807 involving any intermediate nodes, and Client ('C2') can verify that 1808 the packets came from an acceptable source. (In order for Client 1809 ('C2') to forward packets to Client ('C1'), a corresponding 1810 Predirect/Redirect message exchange is required in the reverse 1811 direction; hence, the mechanism is asymmetric.) 1813 3.13.9. Server-Oriented Redirection 1815 In some environments, the Server nearest the target Client may need 1816 to serve as the redirection target, e.g., if direct Client-to-Client 1817 communications are not possible. In that case, the Server prepares 1818 the Redirect message the same as if it were the destination Client 1819 (see: Section 3.13.6), except that it writes its own link-layer 1820 address in the TLLAO option. The Server must then maintain a 1821 neighbor cache entry for the redirected source Client. 1823 3.14. Neighbor Unreachability Detection (NUD) 1825 AERO nodes perform Neighbor Unreachability Detection (NUD) by sending 1826 unicast NS messages to elicit solicited NA messages from neighbors 1827 the same as described in [RFC4861]. NUD is performed either 1828 reactively in response to persistent L2 errors (see Section 3.10) or 1829 proactively to refresh existing neighbor cache entries. 1831 When an AERO node sends an NS/NA message, it MUST use its link-local 1832 address as the IPv6 source address and the link-local address of the 1833 neighbor as the IPv6 destination address. When an AERO node receives 1834 an NS message or a solicited NA message, it accepts the message if it 1835 has a neighbor cache entry for the neighbor; otherwise, it ignores 1836 the message. 1838 When a source Client is redirected to a target Client it SHOULD 1839 proactively test the direct path by sending an initial NS message to 1840 elicit a solicited NA response. While testing the path, the source 1841 Client can optionally continue sending packets via the Server, 1842 maintain a small queue of packets until target reachability is 1843 confirmed, or (optimistically) allow packets to flow directly to the 1844 target. The source Client SHOULD thereafter continue to proactively 1845 test the direct path to the target Client (see Section 7.3 of 1846 [RFC4861]) periodically in order to keep dynamic neighbor cache 1847 entries alive. 1849 In particular, while the source Client is actively sending packets to 1850 the target Client it SHOULD also send NS messages separated by 1851 RETRANS_TIMER milliseconds in order to receive solicited NA messages. 1852 If the source Client is unable to elicit a solicited NA response from 1853 the target Client after MAX_RETRY attempts, it SHOULD set ForwardTime 1854 to 0 and resume sending packets via one of its Servers. Otherwise, 1855 the source Client considers the path usable and SHOULD thereafter 1856 process any link-layer errors as a hint that the direct path to the 1857 target Client has either failed or has become intermittent. 1859 When a target Client receives an NS message from a source Client, it 1860 resets AcceptTime to ACCEPT_TIME if a neighbor cache entry exists; 1861 otherwise, it discards the NS message. If ForwardTime is non-zero, 1862 the target Client then sends a solicited NA message to the link-layer 1863 address of the source Client; otherwise, it sends the solicited NA 1864 message to the link-layer address of one of its Servers. 1866 When a source Client receives a solicited NA message from a target 1867 Client, it resets ForwardTime to FORWARD_TIME if a neighbor cache 1868 entry exists; otherwise, it discards the NA message. 1870 When ForwardTime for a dynamic neighbor cache entry expires, the 1871 source Client resumes sending any subsequent packets via a Server and 1872 may (eventually) attempt to re-initiate the AERO redirection process. 1873 When AcceptTime for a dynamic neighbor cache entry expires, the 1874 target Client discards any subsequent packets received directly from 1875 the source Client. When both ForwardTime and AcceptTime for a 1876 dynamic neighbor cache entry expire, the Client deletes the neighbor 1877 cache entry. 1879 3.15. Mobility Management 1881 3.15.1. Announcing Link-Layer Address Changes 1883 When a Client needs to change its link-layer address, e.g., due to a 1884 mobility event, it performs an immediate DHCPv6 Rebind/Reply exchange 1885 via each of its Servers using the new link-layer address as the 1886 source and with a CLLAO that includes the correct Link ID and 1887 Preference values. If authentication succeeds, the Server then 1888 update its neighbor cache and sends a DHCPv6 Reply. Note that if the 1889 Client does not issue a DHCPv6 Rebind before the Server has 1890 terminated the lease (e.g., if the Client has been out of touch with 1891 the Server for a considerable amount of time), the Server's Reply 1892 will report NoBinding and the Client must re-initiate the DHCPv6 PD 1893 procedure. 1895 Next, the Client sends unsolicited NA messages to each of its 1896 correspondent Client neighbors using the same procedures as specified 1897 in Section 7.2.6 of [RFC4861], except that it sends the messages as 1898 unicast to each neighbor via a Server instead of multicast. In this 1899 process, the Client should send no more than 1900 MAX_NEIGHBOR_ADVERTISEMENT messages separated by no less than 1901 RETRANS_TIMER seconds to each neighbor. 1903 With reference to Figure 7, when Client ('C2') needs to change its 1904 link-layer address it sends unicast unsolicited NA messages to Client 1905 ('C1') via Server ('S2') as follows: 1907 o the link-layer source address is set to 'L2(C2)' (i.e., the link- 1908 layer address of Client ('C2')). 1910 o the link-layer destination address is set to 'L2(S2)' (i.e., the 1911 link-layer address of Server ('S2')). 1913 o the network-layer source address is set to fe80::2001:db8:1:0 1914 (i.e., the AERO address of Client ('C2')). 1916 o the network-layer destination address is set to fe80::2001:db8:0:0 1917 (i.e., the AERO address of Client ('C1')). 1919 o the Type is set to 136. 1921 o the Code is set to 0. 1923 o the Solicited flag is set to 0. 1925 o the Override flag is set to 1. 1927 o the Target Address is set to fe80::2001:db8:1:0 (i.e., the AERO 1928 address of Client ('C2')). 1930 o the message includes one or more TLLAOs with Link ID and 1931 Preference set to appropriate values for Client ('C2')'s 1932 underlying interfaces, and with UDP Port Number and IP Address set 1933 to '0'. 1935 o the message SHOULD include a Timestamp option. 1937 When Server ('S1') receives the NA message, it relays the message in 1938 the same way as described for relaying Redirect messages in 1939 Section 3.13.7. In particular, Server ('S1') copies the correct UDP 1940 port numbers and IP addresses into the TLLAOs, changes the link-layer 1941 source address to its own address, changes the link-layer destination 1942 address to the address of Relay ('R1'), then forwards the NA message 1943 via the relaying chain the same as for a Redirect. 1945 When Client ('C1') receives the NA message, it accepts the message 1946 only if it already has a neighbor cache entry for Client ('C2') then 1947 updates the link-layer addresses for Client ('C2') based on the 1948 addresses in the TLLAOs. However, Client ('C1') MUST NOT update 1949 ForwardTime since Client ('C2') will not have updated AcceptTime. 1951 Note that these unsolicited NA messages are unacknowledged; hence, 1952 Client ('C2') has no way of knowing whether Client ('C1') has 1953 received them. If the messages are somehow lost, however, Client 1954 ('C1') will soon learn of the mobility event via the NUD procedures 1955 specified in Section 3.14. 1957 3.15.2. Bringing New Links Into Service 1959 When a Client needs to bring a new underlying interface into service 1960 (e.g., when it activates a new data link), it performs an immediate 1961 Rebind/Reply exchange via each of its Servers using the new link- 1962 layer address as the source address and with a CLLAO that includes 1963 the new Link ID and Preference values. If authentication succeeds, 1964 the Server then updates its neighbor cache and sends a DHCPv6 Reply. 1965 The Client MAY then send unsolicited NA messages to each of its 1966 correspondent Clients to inform them of the new link-layer address as 1967 described in Section 3.15.1. 1969 3.15.3. Removing Existing Links from Service 1971 When a Client needs to remove an existing underlying interface from 1972 service (e.g., when it de-activates an existing data link), it 1973 performs an immediate Rebind/Reply exchange via each of its Servers 1974 over any available link with a CLLAO that includes the deprecated 1975 Link ID and a Preference value of 0. If authentication succeeds, the 1976 Server then updates its neighbor cache and sends a DHCPv6 Reply. The 1977 Client SHOULD then send unsolicited NA messages to each of its 1978 correspondent Clients to inform them of the deprecated link-layer 1979 address as described in Section 3.15.1. 1981 3.15.4. Moving to a New Server 1983 When a Client associates with a new Server, it performs the Client 1984 procedures specified in Section 3.12. 1986 When a Client disassociates with an existing Server, it sends a 1987 DHCPv6 Release message to the unicast link-local network layer 1988 address of the old Server. The Client SHOULD send the message via a 1989 new Server (i.e., by setting the link-layer destination address to 1990 the address of the new Server) in case the old Server is unreachable 1991 at the link layer, e.g., if the old Server is in a different network 1992 partition. The new Server will forward the message to a Relay, which 1993 will in turn forward the message to the old Server. 1995 When the old Server receives the DHCPv6 Release, it first 1996 authenticates the message. If authentication succeeds, the old 1997 Server withdraws the IP route from the AERO routing system and 1998 deletes the neighbor cache entry for the Client. (The old Server MAY 1999 impose a small delay before deleting the neighbor cache entry so that 2000 any packets already in the system can still be delivered to the 2001 Client.) The old Server then returns a DHCPv6 Reply message via a 2002 Relay. The Client can then use the Reply message to verify that the 2003 termination signal has been processed, and can delete both the 2004 default route and the neighbor cache entry for the old Server. (Note 2005 that the Server's Reply to the Client's Release message may be lost, 2006 e.g., if the AERO routing system has not yet converged. Since the 2007 Client is responsible for reliability, however, it will retry until 2008 it gets an indication that the Release was successful.) 2010 Clients SHOULD NOT move rapidly between Servers in order to avoid 2011 causing excessive oscillations in the AERO routing system. Such 2012 oscillations could result in intermittent reachability for the Client 2013 itself, while causing little harm to the network due to routing 2014 protocol dampening. Examples of when a Client might wish to change 2015 to a different Server include a Server that has gone unreachable, 2016 topological movements of significant distance, etc. 2018 3.16. Encapsulation Protocol Version Considerations 2020 A source Client may connect only to an IPvX underlying network, while 2021 the target Client connects only to an IPvY underlying network. In 2022 that case, the target and source Clients have no means for reaching 2023 each other directly (since they connect to underlying networks of 2024 different IP protocol versions) and so must ignore any redirection 2025 messages and continue to send packets via the Server. 2027 3.17. Multicast Considerations 2029 When the underlying network does not support multicast, AERO nodes 2030 map IPv6 link-scoped multicast addresses (including 2031 'All_DHCP_Relay_Agents_and_Servers') to the link-layer address of a 2032 Server. 2034 When the underlying network supports multicast, AERO nodes use the 2035 multicast address mapping specification found in [RFC2529] for IPv4 2036 underlying networks and use a direct multicast mapping for IPv6 2037 underlying networks. (In the latter case, "direct multicast mapping" 2038 means that if the IPv6 multicast destination address of the 2039 encapsulated packet is "M", then the IPv6 multicast destination 2040 address of the encapsulating header is also "M".) 2042 3.18. Operation on AERO Links Without DHCPv6 Services 2044 When Servers on the AERO link do not provide DHCPv6 services, 2045 operation can still be accommodated through administrative 2046 configuration of ACPs on AERO Clients. In that case, administrative 2047 configurations of AERO interface neighbor cache entries on both the 2048 Server and Client are also necessary. However, this may interfere 2049 with the ability for Clients to dynamically change to new Servers, 2050 and can expose the AERO link to misconfigurations unless the 2051 administrative configurations are carefully coordinated. 2053 3.19. Operation on Server-less AERO Links 2055 In some AERO link scenarios, there may be no Servers on the link and/ 2056 or no need for Clients to use a Server as an intermediary trust 2057 anchor. In that case, each Client acts as a Server unto itself to 2058 establish neighbor cache entries by performing direct Client-to- 2059 Client IPv6 ND message exchanges, and some other form of trust basis 2060 must be applied so that each Client can verify that the prospective 2061 neighbor is authorized to use its claimed ACP. 2063 When there is no Server on the link, Clients must arrange to receive 2064 ACPs and publish them via a secure alternate prefix delegation 2065 authority through some means outside the scope of this document. 2067 3.20. Proxy AERO 2069 Proxy Mobile IPv6 (PMIPv6) [RFC5213][RFC5844] presents a localized 2070 mobility management scheme for use within an access network domain. 2071 It is typically used in cellular wireless service provider networks, 2072 and allows mobile nodes to receive and retain a stable IP address 2073 without needing to implement any special mobility protocols. In the 2074 PMIPv6 architecture, access network devices known as Mobility Access 2075 Gateways (MAGs) provide mobile nodes with an access link abstraction 2076 and receive prefixes for the mobile nodes from a Local Mobility 2077 Anchor (LMA). 2079 The AERO Client (acting as a MAG) can similarly provide proxy 2080 services for mobile nodes that do not participate in AERO messaging. 2081 The proxy Client presents an access link abstraction to mobile nodes, 2082 and performs DHCPv6 PD exchanges over the AERO interface with an AERO 2083 Server (acting as an LMA) to receive a prefix for address 2084 provisioning of the mobile node. 2086 When a mobile node comes onto an access link presented by a proxy 2087 Client, the Client authenticates the node and obtains a unique 2088 identifier that it can use as the DUID in its DHCPv6 PD messages to 2089 the Server. When the Server delegates a prefix, the Client creates a 2090 new AERO address for the mobile node and assigns the delegated prefix 2091 to the mobile node's access link. The Client then generates address 2092 autoconfiguration messages (e.g., IPv6 RA, DHCPv6, DHCPv4, etc.) over 2093 the access link and configures itself as a default router for the 2094 mobile node. Since the Client may serve many such mobile nodes 2095 simultaneously, it may configure multiple AERO addresses, i.e., one 2096 for each mobile node. 2098 When two mobile nodes are associated with the same proxy Client, the 2099 Client can forward traffic between the mobiles without involving the 2100 Server since it configures the AERO addresses of each mobile and 2101 therefore also has the necessary routing information. When two 2102 mobiles are associated with different Clients, the first mobile 2103 node's Client can initiate standard AERO route optimization using the 2104 mobile's AERO address as the source for route optimization messaging. 2105 This may result in a route optimization where the first mobile node's 2106 Client discovers a direct path to the second mobile node's Client. 2108 When a mobile node moves to a new proxy Client, the old proxy Client 2109 issues a DHCPv6 Release message and sends unsolicited NA messages to 2110 any of the mobile node's correspondents the same as specified for 2111 announcing link-layer address changes in Section 3.15.1. However, 2112 since the old Client has no way of knowing where the mobile has moved 2113 to, it sets the Code field in the NA message to 1. When the 2114 correspondent receives such an NA message, it deletes the neighbor 2115 cache entry for the departed mobile node and again allows packets to 2116 flow through its Server. 2118 In addition to the use of DHCPv6 PD signaling, the AERO approach 2119 differs from PMIPv6 in its use of the NBMA virtual link model instead 2120 of point-to-point tunnels. This provides a more agile interface for 2121 Client-to-Server coordinations, and also facilitates simple route 2122 optimization. The AERO routing system is also arranged in such a 2123 fashion that Clients get the same service from any Server they happen 2124 to associate with. This provides a natural fault tolerance and load 2125 balancing capability such as desired for distributed mobility 2126 management. All other considerations are the same as specified in 2127 [RFC5213][RFC5844]. 2129 3.21. Extending AERO Links Through Security Gateways 2131 When an enterprise mobile device moves from a campus LAN connection 2132 to a public Internet link, it must re-enter the enterprise via a 2133 security gateway that has both a physical interface connection to the 2134 Internet and a physical interface connection to the enterprise 2135 internetwork. This most often entails the establishment of a Virtual 2136 Private Network (VPN) link over the public Internet from the mobile 2137 device to the security gateway. During this process, the mobile 2138 device supplies the security gateway with its public Internet address 2139 as the link-layer address for the VPN. The mobile device then acts 2140 as an AERO Client to negotiate with the security gateway to obtain 2141 its ACP. 2143 In order to satisfy this need, the security gateway also operates as 2144 an AERO Server with support for AERO Client proxying. In particular, 2145 when a mobile device (i.e., the Client) connects via the security 2146 gateway (i.e., the Server), the Server provides the Client with an 2147 ACP in a DHCPv6 PD exchange the same as if it were attached to an 2148 enterprise campus access link. The Server then replaces the Client's 2149 link-layer source address with the Server's enterprise-facing link- 2150 layer address in all AERO messages the Client sends toward neighbors 2151 on the AERO link. The AERO messages are then delivered to other 2152 devices on the AERO link as if they were originated by the security 2153 gateway instead of by the AERO Client. In the reverse direction, the 2154 AERO messages sourced by devices within the enterprise network can be 2155 forwarded to the security gateway, which then replaces the link-layer 2156 destination address with the Client's link-layer address and replaces 2157 the link-layer source address with its own (Internet-facing) link- 2158 layer address. 2160 After receiving the ACP, the Client can send IP packets that use an 2161 address taken from the ACP as the network layer source address, the 2162 Client's link-layer address as the link-layer source address, and the 2163 Server's Internet-facing link-layer address as the link-layer 2164 destination address. The Server will then rewrite the link-layer 2165 source address with the Server's own enterprise-facing link-layer 2166 address and rewrite the link-layer destination address with the 2167 target AERO node's link-layer address, and the packets will enter the 2168 enterprise network as though they were sourced from a device located 2169 within the enterprise. In the reverse direction, when a packet 2170 sourced by a node within the enterprise network uses a destination 2171 address from the Client's ACP, the packet will be delivered to the 2172 security gateway which then rewrites the link-layer destination 2173 address to the Client's link-layer address and rewrites the link- 2174 layer source address to the Server's Internet-facing link-layer 2175 address. The Server then delivers the packet across the VPN to the 2176 AERO Client. In this way, the AERO virtual link is essentially 2177 extended *through* the security gateway to the point at which the VPN 2178 link and AERO link are effectively grafted together by the link-layer 2179 address rewriting performed by the security gateway. All AERO 2180 messaging services (including route optimization and mobility 2181 signaling) are therefore extended to the Client. 2183 In order to support this virtual link grafting, the security gateway 2184 (acting as an AERO Server) must keep static neighbor cache entries 2185 for all of its associated Clients located on the public Internet. 2186 The neighbor cache entry is keyed by the AERO Client's AERO address 2187 the same as if the Client were located within the enterprise 2188 internetwork. The neighbor cache is then managed in all ways as 2189 though the Client were an ordinary AERO Client. This includes the 2190 AERO IPv6 ND messaging signaling for Route Optimization and Neighbor 2191 Unreachability Detection. 2193 Note that the main difference between a security gateway acting as an 2194 AERO Server and an enterprise-internal AERO Server is that the 2195 security gateway has at least one enterprise-internal physical 2196 interface and at least one public Internet physical interface. 2197 Conversely, the enterprise-internal AERO Server has only enterprise- 2198 internal physical interfaces. For this reason security gateway 2199 proxying is needed to ensure that the public Internet link-layer 2200 addressing space is kept separate from the enterprise-internal link- 2201 layer addressing space. This is afforded through a natural extension 2202 of the security association caching already performed for each VPN 2203 client by the security gateway. 2205 3.22. Extending IPv6 AERO Links to the Internet 2207 When an IPv6 host ('H1') with an address from an ACP owned by AERO 2208 Client ('C1') sends packets to a correspondent IPv6 host ('H2'), the 2209 packets eventually arrive at the IPv6 router that owns ('H2')s 2210 prefix. This IPv6 router may or may not be an AERO Client ('C2') 2211 either within the same home network as ('C1') or in a different home 2212 network. 2214 If Client ('C1') is currently located outside the boundaries of its 2215 home network, it will connect back into the home network via a 2216 security gateway acting as an AERO Server. The packets sent by 2217 ('H1') via ('C1') will then be forwarded through the security gateway 2218 then through the home network and finally to ('C2') where they will 2219 be delivered to ('H2'). This could lead to sub-optimal performance 2220 when ('C2') could instead be reached via a more direct route without 2221 involving the security gateway. 2223 Consider the case when host ('H1') has the IPv6 address 2224 2001:db8:1::1, and Client ('C1') has the ACP 2001:db8:1::/64 with 2225 underlying IPv6 Internet address of 2001:db8:1000::1. Also, host 2226 ('H2') has the IPv6 address 2001:db8:2::1, and Client ('C2') has the 2227 ACP 2001:db8:2::/64 with underlying IPv6 Internet address of 2228 2001:db8:2000::1. While Client ('C1') may not initially know whether 2229 ('C2') is in fact an AERO Client, it can attempt route optimization 2230 using an approach similar to the Return Routability procedure 2231 specified for Mobile IPv6 (MIPv6) [RFC6275]. In order to support 2232 this process, both Clients MUST intercept and decapsulate packets 2233 that have a subnet router anycast address corresponding to any of the 2234 /64 prefixes covered by their respective ACPs. 2236 To initiate the process, Client ('C1') creates a specially-crafted 2237 encapsulated AERO Predirect message that will be routed through its 2238 home network then through ('C2')s home network and finally to ('C2') 2239 itself. Client ('C1') prepares the initial message in the exchange 2240 as follows: 2242 o The encapsulating IPv6 header source address is set to 2243 2001:db8:1:: (i.e., the IPv6 subnet router anycast address for 2244 ('C1')s ACP) 2246 o The encapsulating IPv6 header destination address is set to 2247 2001:db8:2:: (i.e., the presumed IPv6 subnet router anycast 2248 address for ('C2')s ACP) 2250 o The encapsulating IPv6 header is followed by a UDP header with 2251 source and destination port set to 8060 2253 o The encapsulated IPv6 header source address is set to 2254 fe80::2001:db8:1:0 (i.e., the AERO address for ('C1')) 2256 o The encapsulated IPv6 header destination address is set to 2257 fe80::2001:db8:2:0 (i.e., the presumed AERO address for ('C2')) 2259 o The encapsulated AERO Predirect message includes all of the 2260 securing information that would occur in a MIPv6 "Home Test Init" 2261 message (format TBD) 2263 Client ('C1') then further encapsulates the message in the 2264 encapsulating headers necessary to convey the packet to the security 2265 gateway (e.g., through IPsec encapsulation) so that the message now 2266 appears "double-encapsulated". ('C1') then sends the message to the 2267 security gateway, which re-encapsulates and forwards it over the home 2268 network from where it will eventually reach ('C2'). 2270 At the same time, ('C1') creates and sends a second encapsulated AERO 2271 Predirect message that will be routed through the IPv6 Internet 2272 without involving the security gateway. Client ('C1') prepares the 2273 message as follows: 2275 o The encapsulating IPv6 header source address is set to 2276 2001:db8:1000:1 (i.e., the Internet IPv6 address of ('C1')) 2278 o The encapsulating IPv6 header destination address is set to 2279 2001:db8:2:: (i.e., the presumed IPv6 subnet router anycast 2280 address for ('C2')s ACP) 2282 o The encapsulating IPv6 header is followed by a UDP header with 2283 source and destination port set to 8060 2285 o The encapsulated IPv6 header source address is set to 2286 fe80::2001:db8:1:0 (i.e., the AERO address for ('C1')) 2288 o The encapsulated IPv6 header destination address is set to 2289 fe80::2001:db8:2:0 (i.e., the presumed AERO address for ('C2')) 2291 o The encapsulated AERO Predirect message includes all of the 2292 securing information that would occur in a MIPv6 "Care-of Test 2293 Init" message (format TBD) 2295 If ('C2') is indeed an AERO Client, it will receive both Predirect 2296 messages through its home network. ('C2') then return a 2297 corresponding Redirect for each of the Predirect messages with the 2298 source and destination addresses in the inner and outer headers 2299 reversed. The first message includes all of the securing information 2300 that would occur in a MIPv6 "Home Test" message, while the second 2301 message includes all of the securing information that would occur in 2302 a MIPv6 "Care-of Test" message (formats TBD). 2304 When ('C1') receives the Redirect messages, it performs the necessary 2305 security procedures per the MIPv6 specification. It then prepares an 2306 encapsulated NS message that includes the same source and destination 2307 addresses as for the "Care-of Test Init" Predirect message, and 2308 includes all of the securing information that would occur in a MIPv6 2309 "Binding Update" message (format TBD) and sends the message to 2310 ('C2'). 2312 When ('C2') receives the NS message, if the securing information is 2313 correct it creates or updates a neighbor cache entry for ('C1') with 2314 fe80::2001:db8:1:0 as the network-layer address, 2001:db8:1000::1 as 2315 the link-layer address and with AcceptTime set to ACCEPT_TIME. 2316 ('C2') then sends an encapsulated NA message back to ('C1') that 2317 includes the same source and destination addresses as for the "Care- 2318 of Test" Redirect message, and includes all of the securing 2319 information that would occur in a MIPv6 "Binding Acknowledgement" 2320 message (format TBD) and sends the message to ('C1'). 2322 When ('C1') receives the NA message, it creates or updates a neighbor 2323 cache entry for ('C2') with fe80::2001:db8:2:0 as the network-layer 2324 address and 2001:db8:2:: as the link-layer address and with 2325 ForwardTime set to FORWARD_TIME, thus completing the route 2326 optimization in the forward direction. 2328 ('C1') subsequently forwards encapsulated packets with outer source 2329 address 2001:db8:1000::1, with outer destination address 2330 2001:db8:2::, with inner source address taken from the 2001:db8:1::, 2331 and with inner destination address taken from 2001:db8:2:: due to the 2332 fact that it has a securely-established neighbor cache entry with 2333 non-zero ForwardTime. ('C2') subsequently accepts any such 2334 encapsulated packets due to the fact that it has a securely- 2335 established neighbor cache entry with non-zero AcceptTime.. 2337 In order to keep neighbor cache entries alive, ('C1') periodically 2338 sends additional NS messages to ('C2') and receives any NA responses. 2339 If ('C1') moves to a different point of attachment after the initial 2340 route optimization, it sends a new secured NS message to ('C2') as 2341 above to update ('C2')s neighbor cache. 2343 If ('C2') has packets to send to ('C1'), it performs a corresponding 2344 route optimization in the opposite direction following the same 2345 procedures described above. In the process, the already-established 2346 unidirectional neighbor cache entries within ('C1') and ('C2') are 2347 updated to include the now-bidirectional information. In particular, 2348 the AcceptTime and ForwardTime variables for both neighbor cache 2349 entries are updated to non-zero values, and the link-layer address 2350 for ('C1')s neighbor cache entry for ('C2') is reset to 2351 2001:db8:2000::1. 2353 Note that two AERO Clients can use full security protocol messaging 2354 instead of Return Routability, e.g., if strong authentication and/or 2355 confidentiality are desired. In that case, security protocol key 2356 exchanges such as specified for MOBIKE [RFC4555] would be used to 2357 establish security associations and neighbor cache entries between 2358 the AERO clients. Thereafter, AERO NS/NA messaging can be used to 2359 maintain neighbor cache entries, test reachability, and to announce 2360 mobility events. If reachability testing fails, e.g., if both 2361 Clients move at roughly the same time, the Clients can tear down the 2362 security association and neighbor cache entries and again allow 2363 packets to flow through their home network (which may result in a new 2364 route optimization event). 2366 4. Implementation Status 2368 An application-layer implementation is in progress. 2370 5. IANA Considerations 2372 IANA is instructed to assign a new 2-octet Hardware Type number 2373 "TBD1" for AERO in the "arp-parameters" registry per Section 2 of 2374 [RFC5494]. The number is assigned from the 2-octet Unassigned range 2375 with Hardware Type "AERO" and with this document as the reference. 2377 IANA is instructed to assign a 4-octet Enterprise Number "TBD2" for 2378 AERO in the "enterprise-numbers" registry per [RFC3315]. 2380 6. Security Considerations 2382 AERO link security considerations are the same as for standard IPv6 2383 Neighbor Discovery [RFC4861] except that AERO improves on some 2384 aspects. In particular, AERO uses a trust basis between Clients and 2385 Servers, where the Clients only engage in the AERO mechanism when it 2386 is facilitated by a trust anchor. Unless there is some other means 2387 of authenticating the Client's identity (e.g., link-layer security), 2388 AERO nodes SHOULD also use DHCPv6 securing services (e.g., DHCPv6 2389 authentication, Secure DHCPv6 [I-D.ietf-dhc-sedhcpv6], etc.) for 2390 Client authentication and network admission control. 2392 AERO Redirect, Predirect and unsolicited NA messages SHOULD include a 2393 Timestamp option (see Section 5.3 of [RFC3971]) that other AERO nodes 2394 can use to verify the message time of origin. AERO Predirect, NS and 2395 RS messages SHOULD include a Nonce option (see Section 5.3 of 2396 [RFC3971]) that recipients echo back in corresponding responses. 2398 AERO links must be protected against link-layer address spoofing 2399 attacks in which an attacker on the link pretends to be a trusted 2400 neighbor. Links that provide link-layer securing mechanisms (e.g., 2401 IEEE 802.1X WLANs) and links that provide physical security (e.g., 2402 enterprise network wired LANs) provide a first line of defense that 2403 is often sufficient. In other instances, additional securing 2404 mechanisms such as Secure Neighbor Discovery (SeND) [RFC3971], IPsec 2405 [RFC4301] or TLS [RFC5246] may be necessary. 2407 AERO Clients MUST ensure that their connectivity is not used by 2408 unauthorized nodes on their EUNs to gain access to a protected 2409 network, i.e., AERO Clients that act as routers MUST NOT provide 2410 routing services for unauthorized nodes. (This concern is no 2411 different than for ordinary hosts that receive an IP address 2412 delegation but then "share" the address with unauthorized nodes via a 2413 NAT function.) 2415 On some AERO links, establishment and maintenance of a direct path 2416 between neighbors requires secured coordination such as through the 2417 Internet Key Exchange (IKEv2) protocol [RFC5996] to establish a 2418 security association. 2420 7. Acknowledgements 2422 Discussions both on IETF lists and in private exchanges helped shape 2423 some of the concepts in this work. Individuals who contributed 2424 insights include Mikael Abrahamsson, Mark Andrews, Fred Baker, 2425 Stewart Bryant, Brian Carpenter, Wojciech Dec, Ralph Droms, Sri 2426 Gundavelli, Brian Haberman, Joel Halpern, Sascha Hlusiak, Lee Howard, 2427 Andre Kostur, Ted Lemon, Joe Touch and Bernie Volz. Members of the 2428 IESG also provided valuable input during their review process that 2429 greatly improved the document. Special thanks go to Stewart Bryant, 2430 Joel Halpern and Brian Haberman for their shepherding guidance. 2432 This work has further been encouraged and supported by Boeing 2433 colleagues including Keith Bartley, Dave Bernhardt, Cam Brodie, 2434 Balaguruna Chidambaram, Claudiu Danilov, Wen Fang, Anthony Gregory, 2435 Jeff Holland, Ed King, Gen MacLean, Kent Shuey, Brian Skeen, Mike 2436 Slane, Julie Wulff, Yueli Yang, and other members of the BR&T and BIT 2437 mobile networking teams. 2439 Earlier works on NBMA tunneling approaches are found in 2440 [RFC2529][RFC5214][RFC5569]. 2442 8. References 2444 8.1. Normative References 2446 [RFC0768] Postel, J., "User Datagram Protocol", STD 6, RFC 768, 2447 August 1980. 2449 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, September 2450 1981. 2452 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 2453 RFC 792, September 1981. 2455 [RFC2003] Perkins, C., "IP Encapsulation within IP", RFC 2003, 2456 October 1996. 2458 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 2459 Requirement Levels", BCP 14, RFC 2119, March 1997. 2461 [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 2462 (IPv6) Specification", RFC 2460, December 1998. 2464 [RFC2473] Conta, A. and S. Deering, "Generic Packet Tunneling in 2465 IPv6 Specification", RFC 2473, December 1998. 2467 [RFC3315] Droms, R., Bound, J., Volz, B., Lemon, T., Perkins, C., 2468 and M. Carney, "Dynamic Host Configuration Protocol for 2469 IPv6 (DHCPv6)", RFC 3315, July 2003. 2471 [RFC3633] Troan, O. and R. Droms, "IPv6 Prefix Options for Dynamic 2472 Host Configuration Protocol (DHCP) version 6", RFC 3633, 2473 December 2003. 2475 [RFC3971] Arkko, J., Kempf, J., Zill, B., and P. Nikander, "SEcure 2476 Neighbor Discovery (SEND)", RFC 3971, March 2005. 2478 [RFC4213] Nordmark, E. and R. Gilligan, "Basic Transition Mechanisms 2479 for IPv6 Hosts and Routers", RFC 4213, October 2005. 2481 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 2482 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 2483 September 2007. 2485 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 2486 Address Autoconfiguration", RFC 4862, September 2007. 2488 [RFC6434] Jankiewicz, E., Loughney, J., and T. Narten, "IPv6 Node 2489 Requirements", RFC 6434, December 2011. 2491 8.2. Informative References 2493 [I-D.ietf-dhc-sedhcpv6] 2494 Jiang, S., Shen, S., Zhang, D., and T. Jinmei, "Secure 2495 DHCPv6 with Public Key", draft-ietf-dhc-sedhcpv6-03 (work 2496 in progress), June 2014. 2498 [RFC0879] Postel, J., "TCP maximum segment size and related topics", 2499 RFC 879, November 1983. 2501 [RFC1191] Mogul, J. and S. Deering, "Path MTU discovery", RFC 1191, 2502 November 1990. 2504 [RFC1812] Baker, F., "Requirements for IP Version 4 Routers", RFC 2505 1812, June 1995. 2507 [RFC1930] Hawkinson, J. and T. Bates, "Guidelines for creation, 2508 selection, and registration of an Autonomous System (AS)", 2509 BCP 6, RFC 1930, March 1996. 2511 [RFC1981] McCann, J., Deering, S., and J. Mogul, "Path MTU Discovery 2512 for IP version 6", RFC 1981, August 1996. 2514 [RFC2131] Droms, R., "Dynamic Host Configuration Protocol", RFC 2515 2131, March 1997. 2517 [RFC2529] Carpenter, B. and C. Jung, "Transmission of IPv6 over IPv4 2518 Domains without Explicit Tunnels", RFC 2529, March 1999. 2520 [RFC2675] Borman, D., Deering, S., and R. Hinden, "IPv6 Jumbograms", 2521 RFC 2675, August 1999. 2523 [RFC2923] Lahey, K., "TCP Problems with Path MTU Discovery", RFC 2524 2923, September 2000. 2526 [RFC3819] Karn, P., Bormann, C., Fairhurst, G., Grossman, D., 2527 Ludwig, R., Mahdavi, J., Montenegro, G., Touch, J., and L. 2528 Wood, "Advice for Internet Subnetwork Designers", BCP 89, 2529 RFC 3819, July 2004. 2531 [RFC4271] Rekhter, Y., Li, T., and S. Hares, "A Border Gateway 2532 Protocol 4 (BGP-4)", RFC 4271, January 2006. 2534 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 2535 Architecture", RFC 4291, February 2006. 2537 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 2538 Internet Protocol", RFC 4301, December 2005. 2540 [RFC4443] Conta, A., Deering, S., and M. Gupta, "Internet Control 2541 Message Protocol (ICMPv6) for the Internet Protocol 2542 Version 6 (IPv6) Specification", RFC 4443, March 2006. 2544 [RFC4555] Eronen, P., "IKEv2 Mobility and Multihoming Protocol 2545 (MOBIKE)", RFC 4555, June 2006. 2547 [RFC4821] Mathis, M. and J. Heffner, "Packetization Layer Path MTU 2548 Discovery", RFC 4821, March 2007. 2550 [RFC4963] Heffner, J., Mathis, M., and B. Chandler, "IPv4 Reassembly 2551 Errors at High Data Rates", RFC 4963, July 2007. 2553 [RFC4994] Zeng, S., Volz, B., Kinnear, K., and J. Brzozowski, 2554 "DHCPv6 Relay Agent Echo Request Option", RFC 4994, 2555 September 2007. 2557 [RFC5213] Gundavelli, S., Leung, K., Devarapalli, V., Chowdhury, K., 2558 and B. Patil, "Proxy Mobile IPv6", RFC 5213, August 2008. 2560 [RFC5214] Templin, F., Gleeson, T., and D. Thaler, "Intra-Site 2561 Automatic Tunnel Addressing Protocol (ISATAP)", RFC 5214, 2562 March 2008. 2564 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 2565 (TLS) Protocol Version 1.2", RFC 5246, August 2008. 2567 [RFC5494] Arkko, J. and C. Pignataro, "IANA Allocation Guidelines 2568 for the Address Resolution Protocol (ARP)", RFC 5494, 2569 April 2009. 2571 [RFC5522] Eddy, W., Ivancic, W., and T. Davis, "Network Mobility 2572 Route Optimization Requirements for Operational Use in 2573 Aeronautics and Space Exploration Mobile Networks", RFC 2574 5522, October 2009. 2576 [RFC5569] Despres, R., "IPv6 Rapid Deployment on IPv4 2577 Infrastructures (6rd)", RFC 5569, January 2010. 2579 [RFC5844] Wakikawa, R. and S. Gundavelli, "IPv4 Support for Proxy 2580 Mobile IPv6", RFC 5844, May 2010. 2582 [RFC5996] Kaufman, C., Hoffman, P., Nir, Y., and P. Eronen, 2583 "Internet Key Exchange Protocol Version 2 (IKEv2)", RFC 2584 5996, September 2010. 2586 [RFC6146] Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful 2587 NAT64: Network Address and Protocol Translation from IPv6 2588 Clients to IPv4 Servers", RFC 6146, April 2011. 2590 [RFC6204] Singh, H., Beebee, W., Donley, C., Stark, B., and O. 2591 Troan, "Basic Requirements for IPv6 Customer Edge 2592 Routers", RFC 6204, April 2011. 2594 [RFC6275] Perkins, C., Johnson, D., and J. Arkko, "Mobility Support 2595 in IPv6", RFC 6275, July 2011. 2597 [RFC6355] Narten, T. and J. Johnson, "Definition of the UUID-Based 2598 DHCPv6 Unique Identifier (DUID-UUID)", RFC 6355, August 2599 2011. 2601 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 2602 for Equal Cost Multipath Routing and Link Aggregation in 2603 Tunnels", RFC 6438, November 2011. 2605 [RFC6691] Borman, D., "TCP Options and Maximum Segment Size (MSS)", 2606 RFC 6691, July 2012. 2608 [RFC6706] Templin, F., "Asymmetric Extended Route Optimization 2609 (AERO)", RFC 6706, August 2012. 2611 [RFC6864] Touch, J., "Updated Specification of the IPv4 ID Field", 2612 RFC 6864, February 2013. 2614 [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and 2615 UDP Checksums for Tunneled Packets", RFC 6935, April 2013. 2617 [RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement 2618 for the Use of IPv6 UDP Datagrams with Zero Checksums", 2619 RFC 6936, April 2013. 2621 [RFC6939] Halwasia, G., Bhandari, S., and W. Dec, "Client Link-Layer 2622 Address Option in DHCPv6", RFC 6939, May 2013. 2624 [RFC6980] Gont, F., "Security Implications of IPv6 Fragmentation 2625 with IPv6 Neighbor Discovery", RFC 6980, August 2013. 2627 [RFC7078] Matsumoto, A., Fujisaki, T., and T. Chown, "Distributing 2628 Address Selection Policy Using DHCPv6", RFC 7078, January 2629 2014. 2631 Author's Address 2633 Fred L. Templin (editor) 2634 Boeing Research & Technology 2635 P.O. Box 3707 2636 Seattle, WA 98124 2637 USA 2639 Email: fltemplin@acm.org