idnits 2.17.1 draft-templin-aerolink-45.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 (October 20, 2014) is 3469 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 914, but not defined == Unused Reference: 'RFC0768' is defined on line 2623, but no explicit reference was found in the text == Unused Reference: 'RFC4862' is defined on line 2662, but no explicit reference was found in the text == Unused Reference: 'RFC0879' is defined on line 2694, but no explicit reference was found in the text == Unused Reference: 'RFC4994' is defined on line 2763, but no explicit reference was found in the text == Unused Reference: 'RFC5720' is defined on line 2795, but no explicit reference was found in the text == Unused Reference: 'RFC6146' is defined on line 2810, but no explicit reference was found in the text == Unused Reference: 'RFC6204' is defined on line 2817, but no explicit reference was found in the text == Unused Reference: 'RFC6355' is defined on line 2824, but no explicit reference was found in the text == Unused Reference: 'RFC6691' is defined on line 2832, but no explicit reference was found in the text == Unused Reference: 'RFC6980' is defined on line 2851, but no explicit reference was found in the text == Unused Reference: 'RFC7078' is defined on line 2854, 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-sedhcpv6-04 == Outdated reference: A later version (-09) exists of draft-vandevelde-idr-remote-next-hop-08 -- 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 (~~), 16 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, October 20, 2014 5 rfc6179, rfc6706 (if 6 approved) 7 Intended status: Standards Track 8 Expires: April 23, 2015 10 Transmission of IP Packets over AERO Links 11 draft-templin-aerolink-45.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 April 23, 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 3.5. AERO Link Initialization . . . . . . . . . . . . . . . . 11 71 3.6. AERO Interface Initialization . . . . . . . . . . . . . . 11 72 3.6.1. AERO Relay Behavior . . . . . . . . . . . . . . . . . 11 73 3.6.2. AERO Server Behavior . . . . . . . . . . . . . . . . 11 74 3.6.3. AERO Client Behavior . . . . . . . . . . . . . . . . 12 75 3.7. AERO Interface Routing System . . . . . . . . . . . . . . 13 76 3.8. AERO Interface Neighbor Cache Maintenace . . . . . . . . 13 77 3.9. AERO Interface Sending Algorithm . . . . . . . . . . . . 15 78 3.10. AERO Interface Encapsulation, Re-encapsulation and 79 Decapsulation . . . . . . . . . . . . . . . . . . . . . . 17 80 3.11. AERO Interface Data Origin Authentication . . . . . . . . 19 81 3.12. AERO Interface MTU and Fragmentation . . . . . . . . . . 19 82 3.12.1. Accommodating Large IPv6 ND and DHCPv6 Messages . . 22 83 3.12.2. Integrity . . . . . . . . . . . . . . . . . . . . . 23 84 3.13. AERO Interface Error Handling . . . . . . . . . . . . . . 24 85 3.14. AERO Router Discovery, Prefix Delegation and Address 86 Configuration . . . . . . . . . . . . . . . . . . . . . . 28 87 3.14.1. AERO DHCPv6 Service Model . . . . . . . . . . . . . 28 88 3.14.2. AERO Client Behavior . . . . . . . . . . . . . . . . 28 89 3.14.3. AERO Server Behavior . . . . . . . . . . . . . . . . 31 90 3.15. AERO Intradomain Route Optimization . . . . . . . . . . . 33 91 3.15.1. Reference Operational Scenario . . . . . . . . . . . 33 92 3.15.2. Concept of Operations . . . . . . . . . . . . . . . 35 93 3.15.3. Message Format . . . . . . . . . . . . . . . . . . . 35 94 3.15.4. Sending Predirects . . . . . . . . . . . . . . . . . 36 95 3.15.5. Re-encapsulating and Relaying Predirects . . . . . . 37 96 3.15.6. Processing Predirects and Sending Redirects . . . . 38 97 3.15.7. Re-encapsulating and Relaying Redirects . . . . . . 40 98 3.15.8. Processing Redirects . . . . . . . . . . . . . . . . 41 99 3.15.9. Server-Oriented Redirection . . . . . . . . . . . . 41 100 3.16. Neighbor Unreachability Detection (NUD) . . . . . . . . . 41 101 3.17. Mobility Management . . . . . . . . . . . . . . . . . . . 43 102 3.17.1. Announcing Link-Layer Address Changes . . . . . . . 43 103 3.17.2. Bringing New Links Into Service . . . . . . . . . . 44 104 3.17.3. Removing Existing Links from Service . . . . . . . . 45 105 3.17.4. Moving to a New Server . . . . . . . . . . . . . . . 45 106 3.18. Encapsulation Protocol Version Considerations . . . . . . 46 107 3.19. Multicast Considerations . . . . . . . . . . . . . . . . 46 108 3.20. Operation on AERO Links Without DHCPv6 Services . . . . . 46 109 3.21. Operation on Server-less AERO Links . . . . . . . . . . . 46 110 3.22. Proxy AERO . . . . . . . . . . . . . . . . . . . . . . . 47 111 3.23. Extending AERO Links Through Security Gateways . . . . . 49 112 3.24. Extending IPv6 AERO Links to the Internet . . . . . . . . 51 113 4. Implementation Status . . . . . . . . . . . . . . . . . . . . 54 114 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 54 115 6. Security Considerations . . . . . . . . . . . . . . . . . . . 55 116 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 55 117 8. References . . . . . . . . . . . . . . . . . . . . . . . . . 56 118 8.1. Normative References . . . . . . . . . . . . . . . . . . 56 119 8.2. Informative References . . . . . . . . . . . . . . . . . 57 120 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 61 122 1. Introduction 124 This document specifies the operation of IP over tunnel virtual links 125 using Asymmetric Extended Route Optimization (AERO). The AERO link 126 can be used for tunneling to neighboring nodes over either IPv6 or 127 IPv4 networks, i.e., AERO views the IPv6 and IPv4 networks as 128 equivalent links for tunneling. Nodes attached to AERO links can 129 exchange packets via trusted intermediate routers that provide 130 forwarding services to reach off-link destinations and redirection 131 services for route optimization that addresses the requirements 132 outlined in [RFC5522]. 134 AERO provides an IPv6 link-local address format known as the AERO 135 address that supports operation of the IPv6 Neighbor Discovery (ND) 136 [RFC4861] protocol and links IPv6 ND to IP forwarding. Admission 137 control and provisioning are supported by the Dynamic Host 138 Configuration Protocol for IPv6 (DHCPv6) [RFC3315], and node mobility 139 is naturally supported through dynamic neighbor cache updates. 140 Although DHCPv6 and IPv6 ND message signalling is used in the control 141 plane, both IPv4 and IPv6 can be used in the data plane. The 142 remainder of this document presents the AERO specification. 144 2. Terminology 146 The terminology in the normative references applies; the following 147 terms are defined within the scope of this document: 149 AERO link 150 a Non-Broadcast, Multiple Access (NBMA) tunnel virtual overlay 151 configured over a node's attached IPv6 and/or IPv4 networks. All 152 nodes on the AERO link appear as single-hop neighbors from the 153 perspective of the virtual overlay. 155 AERO interface 156 a node's attachment to an AERO link. 158 AERO address 159 an IPv6 link-local address constructed as specified in Section 3.3 160 and assigned to a Client's AERO interface. 162 AERO node 163 a node that is connected to an AERO link and that participates in 164 IPv6 ND and DHCPv6 messaging over the link. 166 AERO Client ("Client") 167 a node that assigns an AERO address to an AERO interface and 168 receives an IP prefix via a DHCPv6 Prefix Delegation (PD) exchange 169 with one or more AERO Servers. 171 AERO Server ("Server") 172 a node that configures an AERO interface to provide default 173 forwarding and DHCPv6 services for AERO Clients. The Server 174 assigns the IPv6 link-local subnet router anycast address (fe80::) 175 to the AERO interface and also assigns an administratively 176 assigned IPv6 link-local unicast address used for operation of 177 DHCPv6 and the IPv6 ND protocol. 179 AERO Relay ("Relay") 180 a node that configures an AERO interface to relay IP packets 181 between nodes on the same AERO link and/or forward IP packets 182 between the AERO link and the native Internetwork. The Relay 183 assigns an administratively assigned IPv6 link-local unicast 184 address to the AERO interface the same as for a Server. 186 ingress tunnel endpoint (ITE) 187 an AERO interface endpoint that injects tunneled packets into an 188 AERO link. 190 egress tunnel endpoint (ETE) 191 an AERO interface endpoint that receives tunneled packets from an 192 AERO link. 194 underlying network 195 a connected IPv6 or IPv4 network routing region over which the 196 tunnel virtual overlay is configured. A typical example is an 197 enterprise network. 199 underlying interface 200 an AERO node's interface point of attachment to an underlying 201 network. 203 link-layer address 204 an IP address assigned to an AERO node's underlying interface. 205 When UDP encapsulation is used, the UDP port number is also 206 considered as part of the link-layer address. Link-layer 207 addresses are used as the encapsulation header source and 208 destination addresses. 210 network layer address 211 the source or destination address of the encapsulated IP packet. 213 end user network (EUN) 214 an internal virtual or external edge IP network that an AERO 215 Client connects to the rest of the network via the AERO interface. 217 AERO Service Prefix (ASP) 218 an IP prefix associated with the AERO link and from which AERO 219 Client Prefixes (ACPs) are derived (for example, the IPv6 ACP 220 2001:db8:1:2::/64 is derived from the IPv6 ASP 2001:db8::/32). 222 AERO Client Prefix (ACP) 223 a more-specific IP prefix taken from an ASP and delegated to a 224 Client. 226 Throughout the document, the simple terms "Client", "Server" and 227 "Relay" refer to "AERO Client", "AERO Server" and "AERO Relay", 228 respectively. Capitalization is used to distinguish these terms from 229 DHCPv6 client/server/relay [RFC3315]. 231 The terminology of [RFC4861] (including the names of node variables 232 and protocol constants) applies to this document. Also throughout 233 the document, the term "IP" is used to generically refer to either 234 Internet Protocol version (i.e., IPv4 or IPv6). 236 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 237 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 238 document are to be interpreted as described in [RFC2119]. 240 3. Asymmetric Extended Route Optimization (AERO) 242 The following sections specify the operation of IP over Asymmetric 243 Extended Route Optimization (AERO) links: 245 3.1. AERO Link Reference Model 247 .-(::::::::) 248 .-(:::: IP ::::)-. 249 (:: Internetwork ::) 250 `-(::::::::::::)-' 251 `-(::::::)-' 252 | 253 +--------------+ +--------+-------+ +--------------+ 254 |AERO Server S1| | AERO Relay R1 | |AERO Server S2| 255 | Nbr: C1; R1 | | Nbr: S1; S2 | | Nbr: C2; R1 | 256 | default->R1 | |(H1->S1; H2->S2)| | default->R1 | 257 | H1->C1 | +--------+-------+ | H2->C2 | 258 +-------+------+ | +------+-------+ 259 | | | 260 X---+---+-------------------+------------------+---+---X 261 | AERO Link | 262 +-----+--------+ +--------+-----+ 263 |AERO Client C1| |AERO Client C2| 264 | Nbr: S1 | | Nbr: S2 | 265 | default->S1 | | default->S2 | 266 +--------------+ +--------------+ 267 .-. .-. 268 ,-( _)-. ,-( _)-. 269 .-(_ IP )-. .-(_ IP )-. 270 (__ EUN ) (__ EUN ) 271 `-(______)-' `-(______)-' 272 | | 273 +--------+ +--------+ 274 | Host H1| | Host H2| 275 +--------+ +--------+ 277 Figure 1: AERO Link Reference Model 279 Figure 1 above presents the AERO link reference model. In this 280 model: 282 o Relay R1 acts as a default router for its associated Servers S1 283 and S2, and connects the AERO link to the rest of the IP 284 Internetwork 286 o Servers S1 and S2 associate with Relay R1 and also act as default 287 routers for their associated Clients C1 and C2. 289 o Clients C1 and C2 associate with Servers S1 and S2, respectively 290 and also act as default routers for their associated EUNs 292 o Hosts H1 and H2 attach to the EUNs served by Clients C1 and C2, 293 respectively 295 In common operational practice, there may be many additional Relays, 296 Servers and Clients. 298 3.2. AERO Node Types 300 AERO Relays provide default forwarding services to AERO Servers. 301 Relays forward packets between Servers connected to the same AERO 302 link and also forward packets between the AERO link and the native IP 303 Internetwork. Relays present the AERO link to the native 304 Internetwork as a set of one or more AERO Service Prefixes (ASPs) and 305 serve as a gateway between the AERO link and the Internetwork. AERO 306 Relays maintain an AERO interface neighbor cache entry for each AERO 307 Server, and maintain an IP forwarding table entry for each AERO 308 Client Prefix (ACP). 310 AERO Servers provide default forwarding services to AERO Clients. 311 Each Server also peers with each Relay in a dynamic routing protocol 312 instance to advertise its list of associated ACPs. Servers configure 313 a DHCPv6 server function to facilitate Prefix Delegation (PD) 314 exchanges with Clients. Each delegated prefix becomes an ACP taken 315 from an ASP. Servers forward packets between Clients and Relays, as 316 well as between Clients and other Clients associated with the same 317 Server. (Server-to-Server forwarding is also enabled in some cases). 318 AERO Servers maintain an AERO interface neighbor cache entry for each 319 AERO Relay. They also maintain both a neighbor cache entry and an IP 320 forwarding table entry for each of their associated Clients. 322 AERO Clients act as requesting routers to receive ACPs through DHCPv6 323 PD exchanges with AERO Servers over the AERO link and sub-delegate 324 portions of their ACPs to EUN interfaces. (Each Client MAY associate 325 with a single Server or with multiple Servers, e.g., for fault 326 tolerance and/or load balancing.) Each IPv6 Client receives at least 327 a /64 IPv6 ACP, and may receive even shorter prefixes. Similarly, 328 each IPv4 Client receives at least a /32 IPv4 ACP (i.e., a singleton 329 IPv4 address), and may receive even shorter prefixes. AERO Clients 330 maintain an AERO interface neighbor cache entry for each of their 331 associated Servers as well as for each of their correspondent 332 Clients. 334 AERO Clients that act as hosts typically configure a TUN/TAP 335 interface [TUNTAP] as a point-to-point linkage between the IP layer 336 and the AERO interface. The IP layer therefore sees only the TUN/TAP 337 interface, while the AERO interface provides an intermediate conduit 338 between the TUN/TAP interface and the underlying interfaces. AERO 339 Clients that act as hosts assign one or more IP addresses from their 340 ACPs to the TUN/TAP interface, i.e., and not to the AERO interface. 342 3.3. AERO Addresses 344 An AERO address is an IPv6 link-local address with an embedded ACP 345 and assigned to a Client's AERO interface. The AERO address is 346 formed as follows: 348 fe80::[ACP] 350 For IPv6, the AERO address begins with the prefix fe80::/64 and 351 includes in its interface identifier the base prefix taken from the 352 Client's IPv6 ACP. The base prefix is determined by masking the ACP 353 with the prefix length. For example, if the AERO Client receives the 354 IPv6 ACP: 356 2001:db8:1000:2000::/56 358 it constructs its AERO address as: 360 fe80::2001:db8:1000:2000 362 For IPv4, the AERO address is formed from the lower 64 bits of an 363 IPv4-mapped IPv6 address [RFC4291] that includes the base prefix 364 taken from the Client's IPv4 ACP. For example, if the AERO Client 365 receives the IPv4 ACP: 367 192.0.2.32/28 369 it constructs its AERO address as: 371 fe80::FFFF:192.0.2.32 373 The AERO address remains stable as the Client moves between 374 topological locations, i.e., even if its link-layer addresses change. 376 NOTE: In some cases, prospective neighbors may not have advanced 377 knowledge of the Client's ACP length and may therefore send initial 378 IPv6 ND messages with an AERO destination address that matches the 379 ACP but does not correspond to the base prefix. In that case, the 380 Client MUST accept the address as equivalent to the base address, but 381 then use the base address as the source address of any IPv6 ND 382 message replies. For example, if the Client receives the IPv6 ACP 383 2001:db8:1000:2000::/56 then subsequently receives an IPv6 ND message 384 with destination address fe80::2001:db8:1000:2001, it accepts the 385 message but uses fe80::2001:db8:1000:2000 as the source address of 386 any IPv6 ND replies. 388 3.4. AERO Interface Characteristics 390 AERO interfaces use IP-in-IPv6 encapsulation [RFC2473] to exchange 391 tunneled packets with AERO neighbors attached to an underlying IPv6 392 network, and use IP-in-IPv4 encapsulation [RFC2003][RFC4213] to 393 exchange tunneled packets with AERO neighbors attached to an 394 underlying IPv4 network. AERO interfaces can also coordinate secured 395 tunnel types such as IPsec [RFC4301] or TLS [RFC5246]. When Network 396 Address Translator (NAT) traversal and/or filtering middlebox 397 traversal may be necessary, a UDP header is further inserted 398 immediately above the IP encapsulation header. 400 AERO interfaces maintain a neighbor cache, and AERO Clients and 401 Servers use an adaptation of standard unicast IPv6 ND messaging. 402 AERO interfaces use unicast Neighbor Solicitation (NS), Neighbor 403 Advertisement (NA), Router Solicitation (RS) and Router Advertisement 404 (RA) messages the same as for any IPv6 link. AERO interfaces use two 405 redirection message types -- the first known as a Predirect message 406 and the second being the standard Redirect message (see 407 Section 3.15). AERO links further use link-local-only addressing; 408 hence, AERO nodes ignore any Prefix Information Options (PIOs) they 409 may receive in RA messages over an AERO interface. 411 AERO interface ND messages include one or more Target Link-Layer 412 Address Options (TLLAOs) formatted as shown in Figure 2: 414 0 1 2 3 415 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 1 416 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 417 | Type = 2 | Length = 3 | Reserved | 418 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 419 | Link ID | Preference | UDP Port Number | 420 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 421 | | 422 +-- --+ 423 | | 424 +-- IP Address --+ 425 | | 426 +-- --+ 427 | | 428 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 430 Figure 2: AERO Target Link-Layer Address Option (TLLAO) Format 432 In this format, Link ID is an integer value between 0 and 255 433 corresponding to an underlying interface of the target node, and 434 Preference is an integer value between 0 and 255 indicating the 435 node's preference for this underlying interface (with 255 being the 436 highest preference, 1 being the lowest, and 0 meaning "link 437 disabled"). UDP Port Number and IP Address are set to the addresses 438 used by the target node when it sends encapsulated packets over the 439 underlying interface. When the encapsulation IP address family is 440 IPv4, IP Address is formed as an IPv4-mapped IPv6 address [RFC4291]. 442 AERO interfaces may be configured over multiple underlying 443 interfaces. For example, common mobile handheld devices have both 444 wireless local area network ("WLAN") and cellular wireless links. 445 These links are typically used "one at a time" with low-cost WLAN 446 preferred and highly-available cellular wireless as a standby. In a 447 more complex example, aircraft frequently have many wireless data 448 link types (e.g. satellite-based, terrestrial, air-to-air 449 directional, etc.) with diverse performance and cost properties. 451 If a Client's multiple underlying interfaces are used "one at a time" 452 (i.e., all other interfaces are in standby mode while one interface 453 is active), then Redirect, Predirect and unsolicited NA messages 454 include only a single TLLAO with Link ID set to a constant value. 456 If the Client has multiple active underlying interfaces, then from 457 the perspective of IPv6 ND it would appear to have a single link- 458 local address with multiple link-layer addresses. In that case, 459 Redirect, Predirect and unsolicited NA messages MAY include multiple 460 TLLAOs -- each with a different Link ID that corresponds to a 461 specific underlying interface of the Client. 463 3.5. AERO Link Initialization 465 When an administrative authority first deploys a set of AERO Relays 466 and Servers that comprise an AERO link, they assign a unique domain 467 name for the link, e.g., "example.com". Next, if the administrative 468 policy permits Clients within the domain to serve as correspondent 469 nodes for Internet mobile nodes, the administrative authority adds a 470 Fully Qualified Domain Name (FQDN) for each of the AERO link's ASPs 471 to the Domain Name System (DNS) [RFC1035]. The FQDN is based on the 472 suffix "aero.linkupnetworks.net" with a wildcard-terminated reverse 473 mapping of the ASP [RFC3596][RFC4592], and resolves to a DNS PTR 474 resource record. For example, for the ASP '2001:db8:1::/48' within 475 the domain name "example.com", the DNS database contains: 477 '*.1.0.0.0.8.b.d.0.1.0.0.2.aero.linkupnetworks.net. PTR example.com' 479 This mapping advertises the AERO link's ASPs to prospective mobile 480 nodes. 482 3.6. AERO Interface Initialization 484 3.6.1. AERO Relay Behavior 486 When a Relay enables an AERO interface, it first assigns an 487 administratively-assigned link-local address fe80::ID to the 488 interface. Each fe80::ID address MUST be unique among all Relays and 489 Servers on the link, and MUST NOT collide with any potential AERO 490 addresses. The addresses are typically taken from the range 491 fe80::/96, e.g., as fe80::1, fe80::2, fe80::3, etc. The Relay then 492 engages in a dynamic routing protocol session with all Servers on the 493 link (see: Section 3.7), and advertises the set of ASPs into the 494 native IP Internetwork. 496 Each Relay subsequently maintains an IP forwarding table entry for 497 each Client-Server association, and maintains a neighbor cache entry 498 for each Server on the link. Relays do not require the use of IPv6 499 ND messaging since the dynamic routing protocol already provides 500 reachability information. At a minimum, however, Relays respond to a 501 Server's NS messages by returning an NA. 503 3.6.2. AERO Server Behavior 505 When a Server enables an AERO interface, it assigns the address 506 fe80:: to the interface as a link-local Subnet Router Anycast 507 address, and also assigns an administratively assigned link-local 508 address fe80::ID the same as for Relays. (The Server then accepts 509 DHCPv6 and IPv6 ND solicitation messages destined to either the 510 fe80:: or fe80::ID addresses, but always uses fe80::ID as the source 511 address in the replies it generates.) The Server further configures 512 a DHCPv6 server function to facilitate DHCPv6 PD exchanges with AERO 513 Clients. The Server maintains a neighbor cache entry for each Relay 514 on the link, and manages per-Client neighbor cache entries and IP 515 forwarding table entries based on control message exchanges. Each 516 Server also engages in a dynamic routing protocol with each Relay on 517 the link (see: Section 3.7). 519 When the Server receives an NS/RS message on the AERO interface it 520 returns an NA/RA message but does not update the neighbor cache. The 521 Server further provides a simple conduit between Clients and Relays, 522 between Clients and other Clients and between Clients and other 523 Servers. Therefore, packets enter the Server's AERO interface from 524 the link layer and are forwarded back out the link layer without ever 525 leaving the AERO interface and therefore without ever disturbing the 526 network layer. 528 3.6.3. AERO Client Behavior 530 When a Client enables an AERO interface, it invokes DHCPv6 PD to 531 receive an ACP from an AERO Server. Next, it assigns the 532 corresponding AERO address to the AERO interface and creates a 533 neighbor cache entry for the Server, i.e., the PD exchange bootstraps 534 the provisioning of a unique link-local address. The Client 535 maintains a neighbor cache entry for each of its Servers and each of 536 its active correspondent Clients. When the Client receives Redirect/ 537 Predirect messages on the AERO interface it updates or creates 538 neighbor cache entries, including link-layer address information. 539 Unsolicited NA messages update the cached link-layer addresses for 540 correspondent Clients (e.g., following a link-layer address change 541 due to node mobility) but do not create new neighbor cache entries. 542 NS/NA messages used for Neighbor Unreachability Detection (NUD) 543 update timers in existing neighbor cache entires but do not update 544 link-layer addresses nor create new neighbor cache entries. 546 Finally, the Client need not maintain any IP forwarding table entries 547 for its Servers or correspondent Clients. Instead, it can set a 548 single "route-to-interface" default route in the IP forwarding table, 549 and all forwarding decisions can be made within the AERO interface 550 based on neighbor cache entries. (On systems in which adding a 551 default route would violate security policy, the default route could 552 instead be installed via a "synthesized RA", e.g., as discussed in 553 Section 3.14.2.) 555 3.7. AERO Interface Routing System 557 Relays require full topology information of all Client/Server 558 associations, while individual Servers only need to know the ACPs 559 associated with their current set of associated Clients. This is 560 accomplished through the use of an internal instance of the Border 561 Gateway Protocol (BGP) [RFC4271] coordinated between Servers and 562 Relays. This internal BGP instance does not interact with the public 563 Internet BGP instance; therefore, the AERO link is presented to the 564 IP Internetwork as a small set of ASPs as opposed to the full set of 565 individual ACPs. 567 In a reference BGP arrangement, each AERO Server is configured as an 568 Autonomous System Border Router (ASBR) for a stub Autonomous System 569 (AS) (possibly using a private AS Number (ASN) [RFC1930]), and each 570 Server further peers with each Relay but does not peer with other 571 Servers. Similarly, Relays do not peer with each other, since they 572 will reliably receive all updates from all Servers and will therefore 573 have a consistent view of the AERO link ACP delegations. 575 Each Server maintains a working set of associated Clients, and 576 dynamically announces new ACPs and withdraws departed ACPs in its BGP 577 updates to Relays. Relays do not send BGP updates to Servers, 578 however, such that the BGP route reporting is unidirectional from 579 Servers to Relays. 581 Relays therefore discover the full topology of the AERO link in terms 582 of the working set of ACPs associated with each Server, while Servers 583 only discover the ACPs of their associated Clients. Since Clients 584 are expected to remain associated with their current set of Servers 585 for extended timeframes, the amount of BGP control messaging between 586 Servers and Relays should be minimal. However, BGP Servers SHOULD 587 dampen any route oscillations caused by impatient Clients that 588 repeatedly associate and disassociate with them. 590 In environments where sustained packet forwarding over AERO Relays 591 would present an undesirable data plane burden, Relays can instead be 592 configured to report all ACPs to all Servers while including a BGP 593 Remote-Next-Hop [I-D.vandevelde-idr-remote-next-hop]. The Server 594 then creates a neighbor cache entry for each ACP with the Remote- 595 Next-Hop as the link-layer address to enable Server-to-Server route 596 optimization. 598 3.8. AERO Interface Neighbor Cache Maintenace 600 Each AERO interface maintains a conceptual neighbor cache that 601 includes an entry for each neighbor it communicates with on the AERO 602 link, the same as for any IPv6 interface [RFC4861]. AERO interface 603 neighbor cache entires are said to be one of "permanent", "static" or 604 "dynamic". 606 Permanent neighbor cache entries are created through explicit 607 administrative action; they have no timeout values and remain in 608 place until explicitly deleted. AERO Relays maintain a permanent 609 neighbor cache entry for each Server on the link, and AERO Servers 610 maintain a permanent neighbor cache entry for each Relay. Each entry 611 maintains the mapping between the neighbor's fe80::ID network-layer 612 address and corresponding link-layer address. 614 Static neighbor cache entries are created though DHCPv6 PD exchanges 615 and remain in place for durations bounded by prefix lifetimes. AERO 616 Servers maintain a static neighbor cache entry for each of their 617 associated Clients, and AERO Clients maintain a static neighbor cache 618 for each of their associated Servers. When an AERO Server sends a 619 DHCPv6 Reply message response to a Client's DHCPv6 Solicit/Request or 620 Renew message, it creates or updates a static neighbor cache entry 621 based on the Client's AERO address as the network-layer address, the 622 prefix lifetime as the neighbor cache entry lifetime, the Client's 623 encapsulation IP address and UDP port number as the link-layer 624 address and the prefix length as the length to apply to the AERO 625 address. When an AERO Client receives a DHCPv6 Reply message from a 626 Server, it creates or updates a static neighbor cache entry based on 627 the Reply message link-local source address as the network-layer 628 address, the prefix lifetime as the neighbor cache entry lifetime, 629 and the encapsulation IP source address and UDP source port number as 630 the link-layer address. 632 Dynamic neighbor cache entries are created based on receipt of an 633 IPv6 ND message, and are garbage-collected if not used within a short 634 timescale. AERO Clients maintain dynamic neighbor cache entries for 635 each of their active correspondent Clients with lifetimes based on 636 IPv6 ND messaging constants. When an AERO Client receives a valid 637 Predirect message it creates or updates a dynamic neighbor cache 638 entry for the Predirect target network-layer and link-layer addresses 639 plus prefix length. The node then sets an "AcceptTime" variable in 640 the neighbor cache entry to ACCEPT_TIME seconds and uses this value 641 to determine whether packets received from the correspondent can be 642 accepted. When an AERO Client receives a valid Redirect message it 643 creates or updates a dynamic neighbor cache entry for the Redirect 644 target network-layer and link-layer addresses plus prefix length. 645 The Client then sets a "ForwardTime" variable in the neighbor cache 646 entry to FORWARD_TIME seconds and uses this value to determine 647 whether packets can be sent directly to the correspondent. The 648 Client also sets a "MaxRetry" variable to MAX_RETRY to limit the 649 number of keepalives sent when a correspondent may have gone 650 unreachable. 652 For dynamic neighbor cache entries, when an AERO Client receives a 653 valid NS message it (re)sets AcceptTime for the neighbor to 654 ACCEPT_TIME. When an AERO Client receives a valid solicited NA 655 message, it (re)sets ForwardTime for the neighbor to FORWARD_TIME and 656 sets MaxRetry to MAX_RETRY. When an AERO Client receives a valid 657 unsolicited NA message, it updates the correspondent's link-layer 658 addresses but DOES NOT reset AcceptTime, ForwardTime or MaxRetry. 660 It is RECOMMENDED that FORWARD_TIME be set to the default constant 661 value 30 seconds to match the default REACHABLE_TIME value specified 662 for IPv6 ND [RFC4861]. 664 It is RECOMMENDED that ACCEPT_TIME be set to the default constant 665 value 40 seconds to allow a 10 second window so that the AERO 666 redirection procedure can converge before AcceptTime decrements below 667 FORWARD_TIME. 669 It is RECOMMENDED that MAX_RETRY be set to 3 the same as described 670 for IPv6 ND address resolution in Section 7.3.3 of [RFC4861]. 672 Different values for FORWARD_TIME, ACCEPT_TIME, and MAX_RETRY MAY be 673 administratively set, if necessary, to better match the AERO link's 674 performance characteristics; however, if different values are chosen, 675 all nodes on the link MUST consistently configure the same values. 676 Most importantly, ACCEPT_TIME SHOULD be set to a value that is 677 sufficiently longer than FORWARD_TIME to allow the AERO redirection 678 procedure to converge. 680 3.9. AERO Interface Sending Algorithm 682 IP packets enter a node's AERO interface either from the network 683 layer (i.e., from a local application or the IP forwarding system), 684 or from the link layer (i.e., from the AERO tunnel virtual link). 685 Packets that enter the AERO interface from the network layer are 686 encapsulated and admitted into the AERO link, i.e., they are 687 tunnelled to an AERO interface neighbor. Packets that enter the AERO 688 interface from the link layer are either re-admitted into the AERO 689 link or delivered to the network layer where they are subject to 690 either local delivery or IP forwarding. Since each AERO node has 691 only partial information about neighbors on the link, AERO interfaces 692 may forward packets with link-local destination addresses at a layer 693 below the network layer. This means that AERO nodes act as both IP 694 routers and sub-IP layer forwarding agents. AERO interface sending 695 considerations for Clients, Servers and Relays are given below. 697 When an IP packet enters a Client's AERO interface from the network 698 layer, if the destination is covered by an ASP the Client searches 699 for a dynamic neighbor cache entry with a non-zero ForwardTime and an 700 AERO address that matches the packet's destination address. (The 701 destination address may be either an address covered by the 702 neighbor's ACP or the (link-local) AERO address itself.) If there is 703 a match, the Client uses a link-layer address in the entry as the 704 link-layer address for encapsulation then admits the packet into the 705 AERO link. If there is no match, the Client instead uses the link- 706 layer address of a neighboring Server as the link-layer address for 707 encapsulation. 709 When an IP packet enters a Server's AERO interface from the link 710 layer, if the destination is covered by an ASP the Server searches 711 for a neighbor cache entry with an AERO address that matches the 712 packet's destination address. (The destination address may be either 713 an address covered by the neighbor's ACP or the AERO address itself.) 714 If there is a match, the Server uses a link-layer address in the 715 entry as the link-layer address for encapsulation and re-admits the 716 packet into the AERO link. If there is no match, the Server instead 717 uses the link-layer address in a permanent neighbor cache entry for a 718 Relay as the link-layer address for encapsulation. 720 When an IP packet enters a Relay's AERO interface from the network 721 layer, the Relay searches its IP forwarding table for an entry that 722 is covered by an ASP and also matches the destination. If there is a 723 match, the Relay uses the link-layer address in the neighbor cache 724 entry for the next-hop Server as the link-layer address for 725 encapsulation and admits the packet into the AERO link. When an IP 726 packet enters a Relay's AERO interface from the link-layer, if the 727 destination is not a link-local address and is does not match an ASP 728 the Relay removes the packet from the AERO interface and uses IP 729 forwarding to forward the packet to the Internetwork. If the 730 destination address is a link-local address or a non-link-local 731 address that matches an ASP, and there is a more-specific ACP entry 732 in the IP forwarding table, the Relay uses the link-layer address in 733 the corresponding neighbor cache entry for the next-hop Server as the 734 link-layer address for encapsulation and re-admits the packet into 735 the AERO link. When an IP packet enters a Relay's AERO interface 736 from either the network layer or link-layer, and the packet's 737 destination address matches an ASP but there is no more-specific ACP 738 entry, the Relay drops the packet and returns an ICMP Destination 739 Unreachable message (see: Section 3.13). 741 When an AERO Server receives a packet from a Relay via the AERO 742 interface, the Server MUST NOT forward the packet back to the same or 743 a different Relay. 745 When an AERO Relay receives a packet from a Server via the AERO 746 interface, the Relay MUST NOT forward the packet back to the same 747 Server. 749 When an AERO node re-admits a packet into the AERO link without 750 involving the network layer, the node MUST NOT decrement the network 751 layer TTL/Hop-count. 753 AERO interfaces may determine the link-layer address for 754 encapsulation through consulting either the neighbor cache or the IP 755 forwarding table. IP forwarding is therefore linked to IPv6 ND via 756 the AERO address. 758 3.10. AERO Interface Encapsulation, Re-encapsulation and Decapsulation 760 AERO interfaces encapsulate IP packets according to whether they are 761 entering the AERO interface from the network layer or if they are 762 being re-admitted into the same AERO link they arrived on. This 763 latter form of encapsulation is known as "re-encapsulation". 765 The AERO interface encapsulates packets per the base tunneling 766 specifications (e.g., 767 [RFC2003][RFC2473][RFC2784][RFC4213][RFC4301][RFC5246], etc.) except 768 that it inserts a UDP header immediately following the IP 769 encapsulation header and immediately before the next header. 771 If the next header is an IPv4 or IPv6 header and the packet is an 772 ordinary data packet, no other encapsulations are necessary. For all 773 others (including IPv6 ND and DHCPv6 messages), the AERO interface 774 MUST insert an AERO shim header immediately following the UDP header 775 formatted as shown in Figure 3: 777 0 1 778 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 779 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 780 | Vers1 | Vers2 | Next Header | 781 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 783 Figure 3: AERO Shim Header 785 In the AERO shim header, "Vers1" encodes the value '0', "Vers2" 786 encodes the value '1', and "Next Header" encodes the IP protocol 787 number corresponding to the next header in the encapsulation. For 788 example, "Next Header" encodes the value '4' for an IPv4 header, '41' 789 for an IPv6 header, '44' for the IPv6 Fragment Header, '47' for GRE, 790 '50' for ESP, '51' for AH, etc. (other Next Header values are found 791 in the IANA "protocol numbers" registry). 793 During encapsulation, the AERO interface copies the "TTL/Hop Limit", 794 "Type of Service/Traffic Class" and "Congestion Experienced" values 795 in the packet's IP header into the corresponding fields in the 796 encapsulation IP header. For packets undergoing re-encapsulation, 797 the AERO interface instead copies the "TTL/Hop Limit", "Type of 798 Service/Traffic Class" and "Congestion Experienced" values in the 799 original encapsulation IP header into the corresponding fields in the 800 new encapsulation IP header (i.e., the values are transferred between 801 encapsulation headers and *not* copied from the encapsulated packet's 802 network-layer header). Note that these instructions may represent a 803 deviation from those found in the base tunneling specifications. 805 The AERO interface next sets the UDP source port to a constant value 806 that it will use in each successive packet it sends, and sets the UDP 807 length field to the length of the encapsulated packet plus 8 bytes 808 for the UDP header itself (or plus 10 bytes if an AERO shim header is 809 also included). For packets sent via a Server, the AERO interface 810 sets the UDP destination port to 8060, i.e., the IANA-registered port 811 number for AERO. For packets sent to a correspondent Client, the 812 AERO interface sets the UDP destination port to the port value stored 813 in the neighbor cache entry for this correspondent. The AERO 814 interface also sets the UDP checksum field to zero (see: 815 [RFC6935][RFC6936]) unless an integrity check is required (see: 816 Section 3.12.2). 818 The AERO interface next sets the IP protocol number in the 819 encapsulation header to 17 (i.e., the IP protocol number for UDP). 820 When IPv6 is used as the encapsulation protocol, the interface then 821 sets the flow label value in the encapsulation header the same as 822 described in [RFC6438]. When IPv4 is used as the encapsulation 823 protocol, the AERO interface sets the DF bit as discussed in 824 Section 3.12. 826 AERO interfaces decapsulate packets destined either to the node 827 itself or to a destination reached via an interface other than the 828 AERO interface the packet was received on. When the AERO interface 829 receives a UDP packet, it examines the first octet of the 830 encapsulated packet. If the most significant four bits of the first 831 octet encode the value '6' (i.e., the IP version number value for 832 IPv6) or the value '4' (i.e., the IP version number value for IPv4), 833 the AERO interface discards the encapsulation headers and accepts the 834 encapsulated packet as an ordinary IPv6 or IPv4 data packet, 835 respectively (this is often referred to as "fast path processing"). 837 If the most significant four bits encode the value '0' and the next 838 four bits encode the value '1', however, the AERO interface processes 839 the next octet as a "Next Header" field, i.e., the interface treats 840 the first two octets of the encapsulated packet as an AERO shim 841 header as shown in Figure 3 (note that the "Vers2" value is set to 1 842 to distinguish AERO encapsulations from the experimental message 843 formats specified in [RFC6706]). Further processing then proceeds 844 according to the appropriate base tunneling specification and/or 845 control message type (this is often referred to as "slow path 846 processing"). 848 3.11. AERO Interface Data Origin Authentication 850 AERO nodes employ simple data origin authentication procedures for 851 encapsulated packets they receive from other nodes on the AERO link. 852 In particular: 854 o AERO Relays and Servers accept encapsulated packets with a link- 855 layer source address that matches a permanent neighbor cache 856 entry. 858 o AERO Servers accept authentic encapsulated DHCPv6 messages, and 859 create or update a static neighbor cache entry for the source 860 based on the specific message type. 862 o AERO Servers accept encapsulated packets if there is a neighbor 863 cache entry with an AERO address that matches the packet's 864 network-layer source address and with a link-layer address that 865 matches the packet's link-layer source address. 867 o AERO Clients accept encapsulated packets if there is a static 868 neighbor cache entry with a link-layer source address that matches 869 the packet's link-layer source address. 871 o AERO Clients and Servers accept encapsulated packets if there is a 872 dynamic neighbor cache entry with an AERO address that matches the 873 packet's network-layer source address, with a link-layer address 874 that matches the packet's link-layer source address, and with a 875 non-zero AcceptTime. 877 Note that this simple data origin authentication is effective in 878 environments in which link-layer addresses cannot be spoofed. 879 Additional security mitigations may be necessary in other 880 environments. 882 3.12. AERO Interface MTU and Fragmentation 884 The AERO interface is the node's point of attachment to the AERO 885 link. AERO links over IP networks have a maximum link MTU of 64KB 886 minus the encapsulation overhead (termed here "ENCAPS"), since the 887 maximum packet size in the base IP specifications is 64KB 888 [RFC0791][RFC2460] (while IPv6 jumbograms can be up to 4GB, they are 889 considered optional for IPv6 nodes [RFC2675][RFC6434]). 891 IPv6 specifies a minimum link MTU of 1280 bytes [RFC2460]. This is 892 the minimum packet size the AERO interface MUST admit without 893 returning an ICMP Packet Too Big (PTB) message. Although IPv4 894 specifies a smaller minimum link MTU of 68 bytes [RFC0791], AERO 895 interfaces also observe a 1280 byte minimum for IPv4. Additionally, 896 the vast majority of links in the Internet configure an MTU of at 897 least 1500 bytes. Original source hosts have therefore become 898 conditioned to expect that IP packets up to 1500 bytes in length will 899 either be delivered to the final destination or a suitable PTB 900 message returned. However, PTB messages may be lost in the network 901 [RFC2923] resulting in failure of the IP MTU discovery mechanisms 902 [RFC1191][RFC1981]. 904 For these reasons, AERO interfaces MUST admit packets up to 1500 905 bytes in length even if some fragmentation is necessary. AERO 906 interfaces MAY admit even larger packets as long as they can be 907 accommodated without fragmentation. 909 For AERO links over IPv4, the IP ID field is only 16 bits in length, 910 meaning that fragmentation at high data rates could result in data 911 corruption due to reassembly misassociations [RFC6864][RFC4963] (see: 912 Section 3.12.2). For AERO links over both IPv4 and IPv6, studies 913 have also shown that IP fragments are dropped unconditionally over 914 some network paths [I-D.taylor-v6ops-fragdrop]. For these reasons, 915 when fragmentation is needed it is performed within the AERO 916 interface (i.e., instead of at the encapsulating IP layer) through 917 the insertion of an IPv6 Fragment Header [RFC2460]. Since the 918 Fragment Header reduces the room available for packet data, but the 919 original source has no way to control its insertion, the Fragment 920 Header length plus the length of the AERO shim header (see: 921 Section 3.10) MUST be included in the ENCAPS length even for packets 922 in which the headers do not appear. 924 The source AERO interface (i.e., the tunnel ingress) therefore sends 925 encapsulated packets to the destination AERO interface (i.e., the 926 tunnel egress) according to the following algorithm: 928 o For IP packets that are no larger than (1280-ENCAPS) bytes, the 929 tunnel ingress encapsulates the packet and admits it into the 930 tunnel without fragmentation. For IPv4 AERO links, tunnel ingress 931 sets the Don't Fragment (DF) bit to 0 so that these packets will 932 be delivered to the tunnel egress even if there is a restricting 933 link in the path, i.e., unless lost due to congestion or routing 934 errors. 936 o For IP packets that are larger than (1280-ENCAPS) bytes but no 937 larger than 1500 bytes, the tunnel ingress encapsulates the packet 938 and inserts a Fragment Header and AERO shim header above the UDP/ 939 IP encapsulation headers. Next, the tunnel ingress uses the 940 fragmentation algorithm in [RFC2460] to break the packet into two 941 non-overlapping fragments where the first fragment (including 942 ENCAPS) is no larger than 1024 bytes and the second is no larger 943 than the first. Each fragment consists of identical AERO/UDP/IP 944 encapsulation headers, followed by the Fragment Header followed by 945 the fragment of the encapsulated packet itself. The tunnel 946 ingress then admits both fragments into the tunnel, and for IPv4 947 sets the DF bit to 0 in the IP encapsulation header. These 948 fragmented encapsulated packets will be delivered to the tunnel 949 egress. 951 o For IPv4 packets that are larger than 1500 bytes and with the DF 952 bit set to 0, the tunnel ingress uses ordinary IP fragmentation to 953 break the unencapsulated packet into a minimum number of non- 954 overlapping fragments where the first fragment is no larger than 955 1024-ENCAPS and all other fragments are no larger than the first 956 fragment. The tunnel ingress then encapsulates each fragment (and 957 for IPv4 sets the DF bit to 0) then admits them into the tunnel. 958 These encapsulated fragments will be delivered to the final 959 destination via the tunnel egress. 961 o For all other IP packets, if the packet is too large to enter the 962 underlying interface following encapsulation, the tunnel ingress 963 drops the packet and returns a network-layer (L3) PTB message to 964 the original source with MTU set to the larger of 1500 bytes or 965 the underlying interface MTU minus ENCAPS. Otherwise, the tunnel 966 ingress encapsulates the packet and admits it into the tunnel 967 without fragmentation (and for IPv4 sets the DF bit to 1) and 968 translates any link-layer (L2) PTB messages it may receive from 969 the network into corresponding L3 PTB messages to send to the 970 original source as specified in Section 3.13. Since both L2 and 971 L3 PTB messages may be either lost or contain insufficient 972 information, however, it is RECOMMENDED that original sources that 973 send unfragmentable IP packets larger than 1500 bytes use 974 Packetization Layer Path MTU Discovery (PLPMTUD) [RFC4821]. 976 While sending packets according to the above algorithm, the tunnel 977 ingress MAY also send 1500 byte probe packets to determine whether 978 they can reach the tunnel egress without fragmentation. If the 979 probes succeed, the tunnel ingress can begin sending packets that are 980 no larger than 1500 bytes without fragmentation (and for IPv4 with DF 981 set to 1). Since the path MTU within the tunnel may fluctuate due to 982 routing changes, the tunnel ingress SHOULD continue to send 983 additional probes subject to rate limiting and SHOULD process any L2 984 PTB messages as an indication that the path MTU may have decreased. 985 If the path MTU within the tunnel becomes insufficient, the source 986 MUST resume fragmentation. 988 To construct a probe, the tunnel ingress prepares an NS message with 989 a Nonce option plus trailing NULL padding octets added to a length of 990 1500 bytes without including the length of the padding in the IPv6 991 Payload Length field, but with the length included in the 992 encapsulating IP header. The tunnel ingress then encapsulates the 993 padded NS message in the encapsulation headers (and for IPv4 sets DF 994 to 1) then sends the message to the tunnel egress. If the tunnel 995 egress returns a solicited NA message with a matching Nonce option, 996 the tunnel ingress deems the probe successful. 998 When the tunnel egress receives the fragments of a fragmented packet, 999 it reassembles them into a whole packet per the reassembly algorithm 1000 in [RFC2460] then discards the Fragment Header. The tunnel egress 1001 therefore MUST be capable of reassembling packets up to 1500+ENCAPS 1002 bytes in length; hence, it is RECOMMENDED that the tunnel egress be 1003 capable of reassembling at least 2KB. 1005 3.12.1. Accommodating Large IPv6 ND and DHCPv6 Messages 1007 IPv6 ND and DHCPv6 messages MUST be accommodated even if some 1008 fragmentation is necessary. These packets are therefore accommodated 1009 through a modification of the second rule in the above algorithm as 1010 follows: 1012 o For IPv6 ND and DHCPv6 messages that are larger than (1280-ENCAPS) 1013 bytes, the tunnel ingress encapsulates the packet and inserts a 1014 Fragment Header and AERO shim header above the UDP/IP 1015 encapsulation headers. Next, the tunnel ingress uses the 1016 fragmentation algorithm in [RFC2460] to break the packet into a 1017 minimum number of non-overlapping fragments where the first 1018 fragment (including ENCAPS) is no larger than 1024 bytes and the 1019 remaining fragments are no larger than the first. The tunnel 1020 ingress then encapsulates each fragment (and for IPv4 sets the DF 1021 bit to 0) then admits them into the tunnel. 1023 IPv6 ND and DHCPv6 messages that exceed the minimum reassembly size 1024 listed above rarely occur in the modern era, however the tunnel 1025 egress SHOULD be able to reassemble them if they do. This means that 1026 the tunnel egress SHOULD include a configuration knob allowing the 1027 operator to set a larger reassembly buffer size if large IPv6ND and 1028 DHCPv6 messages become more common in the future. 1030 The tunnel ingress can send large IPv6 ND and DHCPv6 messages without 1031 fragmentation if there is assurance that large packets can traverse 1032 the tunnel without fragmentation. The tunnel ingress MAY send probe 1033 packets of 1500 bytes or larger as specified above to determine a 1034 size for which fragmentation can be avoided. 1036 3.12.2. Integrity 1038 When fragmentation is needed, there must be assurance that reassembly 1039 can be safely conducted without incurring data corruption. Sources 1040 of corruption can include implementation errors, memory errors and 1041 misassociation of fragments from a first datagram with fragments of 1042 another datagram. The first two conditions (implementation and 1043 memory errors) are mitigated by modern systems and implementations 1044 that have demonstrated integrity through decades of operational 1045 practice. The third condition (reassembly misassociations) must be 1046 accounted for by AERO. 1048 The AERO fragmentation procedure described in the above algorithms 1049 uses the IPv6 Fragment Header and reuses standard IPv6 fragmentation 1050 and reassembly code. Since the Fragment Header includes a 32-bit ID 1051 field, there would need to be 2^32 packets alive in the network 1052 before a second packet with a duplicate ID enters the system with the 1053 (remote) possibility for a reassembly misassociation. For 1280 byte 1054 packets, and for a maximum network lifetime value of 60 1055 seconds[RFC2460], this means that the tunnel ingress would need to 1056 produce ~(7 *10^12) bits/sec in order for a duplication event to be 1057 possible. This exceeds the bandwidth of data link technologies of 1058 the modern era, but not necessarily so going forward into the future. 1059 Although typical wireless data links used by AERO Clients support 1060 vastly lower data rates, the aggregate data rates between AERO 1061 Servers and Relays may be substantial. However, high speed data 1062 links in the network core are expected to configure larger MTUs, 1063 e.g., 4KB, 8KB or even larger. Hence, no integrity check is included 1064 to cover the AERO fragmentation and reassembly procedures. 1066 When the tunnel ingress sends an IPv4-encapsulated packet with the DF 1067 bit set to 0 in the above algorithms, there is a chance that the 1068 packet may be fragmented by an IPv4 router somewhere within the 1069 tunnel. Since the largest such packet is only 1280 bytes, however, 1070 it is very likely that the packet will traverse the tunnel without 1071 incurring a restricting link. Even when a link within the tunnel 1072 configures an MTU smaller than 1280 bytes, it is very likely that it 1073 does so due to limited performance characteristics [RFC3819]. This 1074 means that the tunnel would not be able to convey fragmented 1075 IPv4-encapsulated packets fast enough to produce reassembly 1076 misassociations, as discussed above. However, AERO must also account 1077 for the possibility of tunnel paths that include "poorly managed" 1078 IPv4 link MTUs. 1080 Since the IPv4 header includes only a 16-bit ID field, there would 1081 only need to be 2^16 packets alive in the network before a second 1082 packet with a duplicate ID enters the system. For 1280 byte packets, 1083 and for a maximum network lifetime value of 120 seconds[RFC0791], 1084 this means that the tunnel ingress would only need to produce ~(5 1085 *10^6) bits/sec in order for a duplication event to be possible - a 1086 value that is well within range for many modern wired and wireless 1087 data link technologies. 1089 Therefore, if there is strong operational assurance that no IPv4 1090 links capable of supporting data rates of 5Mbps or more configure an 1091 MTU smaller than 1280 the tunnel ingress MAY omit an integrity check 1092 for the IPv4 fragmentation and reassembly procedures; otherwise, the 1093 tunnel ingress SHOULD include an integrity check. When an upper 1094 layer encapsulation (e.g., IPsec) already includes an integrity 1095 check, the tunnel ingress need not include an additional check. 1096 Otherwise, the tunnel ingress calculates the UDP checksum over the 1097 encapsulated packet and writes the value into the UDP encapsulation 1098 header, i.e., instead of writing the value 0. The tunnel egress will 1099 then verify the UDP checksum and discard the packet if the checksum 1100 is incorrect. 1102 3.13. AERO Interface Error Handling 1104 When an AERO node admits encapsulated packets into the AERO 1105 interface, it may receive link-layer (L2) or network-layer (L3) error 1106 indications. 1108 An L2 error indication is an ICMP error message generated by a router 1109 on the path to the neighbor or by the neighbor itself. The message 1110 includes an IP header with the address of the node that generated the 1111 error as the source address and with the link-layer address of the 1112 AERO node as the destination address. 1114 The IP header is followed by an ICMP header that includes an error 1115 Type, Code and Checksum. For ICMPv6 [RFC4443], the error Types 1116 include "Destination Unreachable", "Packet Too Big (PTB)", "Time 1117 Exceeded" and "Parameter Problem". For ICMPv4 [RFC0792], the error 1118 Types include "Destination Unreachable", "Fragmentation Needed" (a 1119 Destination Unreachable Code that is analogous to the ICMPv6 PTB), 1120 "Time Exceeded" and "Parameter Problem". 1122 The ICMP header is followed by the leading portion of the packet that 1123 generated the error, also known as the "packet-in-error". For 1124 ICMPv6, [RFC4443] specifies that the packet-in-error includes: "As 1125 much of invoking packet as possible without the ICMPv6 packet 1126 exceeding the minimum IPv6 MTU" (i.e., no more than 1280 bytes). For 1127 ICMPv4, [RFC0792] specifies that the packet-in-error includes: 1128 "Internet Header + 64 bits of Original Data Datagram", however 1129 [RFC1812] Section 4.3.2.3 updates this specification by stating: "the 1130 ICMP datagram SHOULD contain as much of the original datagram as 1131 possible without the length of the ICMP datagram exceeding 576 1132 bytes". 1134 The L2 error message format is shown in Figure 4: 1136 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1137 ~ ~ 1138 | L2 IP Header of | 1139 | error message | 1140 ~ ~ 1141 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1142 | L2 ICMP Header | 1143 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1144 ~ ~ P 1145 | IP and other encapsulation | a 1146 | headers of original L3 packet | c 1147 ~ ~ k 1148 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ e 1149 ~ ~ t 1150 | IP header of | 1151 | original L3 packet | i 1152 ~ ~ n 1153 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1154 ~ ~ e 1155 | Upper layer headers and | r 1156 | leading portion of body | r 1157 | of the original L3 packet | o 1158 ~ ~ r 1159 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1161 Figure 4: AERO Interface L2 Error Message Format 1163 The AERO node rules for processing these L2 error messages is as 1164 follows: 1166 o When an AERO node receives an L2 Parameter Problem message, it 1167 processes the message the same as described as for ordinary ICMP 1168 errors in the normative references [RFC0792][RFC4443]. 1170 o When an AERO node receives persistent L2 IPv4 Time Exceeded 1171 messages, the IP ID field may be wrapping before earlier fragments 1172 have been processed. In that case, the node SHOULD begin 1173 including IPv4 integrity checks (see: Section 3.12.2). 1175 o When an AERO Client receives persistent L2 Destination Unreachable 1176 messages in response to tunneled packets that it sends to one of 1177 its dynamic neighbor correspondents, the Client SHOULD test the 1178 path to the correspondent using Neighbor Unreachability Detection 1179 (NUD) (see Section 3.16). If NUD fails, the Client SHOULD set 1180 ForwardTime for the corresponding dynamic neighbor cache entry to 1181 0 and allow future packets destined to the correspondent to flow 1182 through a Server. 1184 o When an AERO Client receives persistent L2 Destination Unreachable 1185 messages in response to tunneled packets that it sends to one of 1186 its static neighbor Servers, the Client SHOULD test the path to 1187 the Server using NUD. If NUD fails, the Client SHOULD delete the 1188 neighbor cache entry and attempt to associate with a new Server. 1190 o When an AERO Server receives persistent L2 Destination Unreachable 1191 messages in response to tunneled packets that it sends to one of 1192 its static neighbor Clients, the Server SHOULD test the path to 1193 the Client using NUD. If NUD fails, the Server SHOULD cancel the 1194 DHCPv6 PD lease for the Client's ACP, withdraw its route for the 1195 ACP from the AERO routing system and delete the neighbor cache 1196 entry (see Section 3.16 and Section 3.17). 1198 o When an AERO Relay or Server receives an L2 Destination 1199 Unreachable message in response to a tunneled packet that it sends 1200 to one of its permanent neighbors, it discards the message since 1201 the routing system is likely in a temporary transitional state 1202 that will soon re-converge. 1204 o When an AERO node receives an L2 PTB message, it translates the 1205 message into an L3 PTB message if possible (*) and forwards the 1206 message toward the original source as described below. 1208 To translate an L2 PTB message to an L3 PTB message, the AERO node 1209 first caches the MTU field value of the L2 ICMP header. The node 1210 next discards the L2 IP and ICMP headers, and also discards the 1211 encapsulation headers of the original L3 packet. Next the node 1212 encapsulates the included segment of the original L3 packet in an L3 1213 IP and ICMP header, and sets the ICMP header Type and Code values to 1214 appropriate values for the L3 IP protocol. In the process, the node 1215 writes the maximum of 1500 bytes and (L2 MTU - ENCAPS) into the MTU 1216 field of the L3 ICMP header. 1218 The node next writes the IP source address of the original L3 packet 1219 as the destination address of the L3 PTB message and determines the 1220 next hop to the destination. If the next hop is reached via the AERO 1221 interface, the node uses the IPv6 address "::" or the IPv4 address 1222 "0.0.0.0" as the IP source address of the L3 PTB message. Otherwise, 1223 the node uses one of its non link-local addresses as the source 1224 address of the L3 PTB message. The node finally calculates the ICMP 1225 checksum over the L3 PTB message and writes the Checksum in the 1226 corresponding field of the L3 ICMP header. The L3 PTB message 1227 therefore is formatted as follows: 1229 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1230 ~ ~ 1231 | L3 IP Header of | 1232 | error message | 1233 ~ ~ 1234 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1235 | L3 ICMP Header | 1236 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1237 ~ ~ p 1238 | IP header of | k 1239 | original L3 packet | t 1240 ~ ~ 1241 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ i 1242 ~ ~ n 1243 | Upper layer headers and | 1244 | leading portion of body | e 1245 | of the original L3 packet | r 1246 ~ ~ r 1247 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1249 Figure 5: AERO Interface L3 Error Message Format 1251 After the node has prepared the L3 PTB message, it either forwards 1252 the message via a link outside of the AERO interface without 1253 encapsulation, or encapsulates and forwards the message to the next 1254 hop via the AERO interface. 1256 When an AERO Relay receives an L3 packet for which the destination 1257 address is covered by an ASP, if there is no more-specific routing 1258 information for the destination the Relay drops the packet and 1259 returns an L3 Destination Unreachable message. The Relay first 1260 writes the IP source address of the original L3 packet as the 1261 destination address of the L3 Destination Unreachable message and 1262 determines the next hop to the destination. If the next hop is 1263 reached via the AERO interface, the Relay uses the IPv6 address "::" 1264 or the IPv4 address "0.0.0.0" as the IP source address of the L3 1265 Destination Unreachable message and forwards the message to the next 1266 hop within the AERO interface. Otherwise, the Relay uses one of its 1267 non link-local addresses as the source address of the L3 Destination 1268 Unreachable message and forwards the message via a link outside the 1269 AERO interface. 1271 When an AERO node receives any L3 error message via the AERO 1272 interface, it examines the destination address in the L3 IP header of 1273 the message. If the next hop toward the destination address of the 1274 error message is via the AERO interface, the node re-encapsulates and 1275 forwards the message to the next hop within the AERO interface. 1276 Otherwise, if the source address in the L3 IP header of the message 1277 is the IPv6 address "::" or the IPv4 address "0.0.0.0", the node 1278 writes one of its non link-local addresses as the source address of 1279 the L3 message and recalculates the IP and/or ICMP checksums. The 1280 node finally forwards the message via a link outside of the AERO 1281 interface. 1283 (*) Note that in some instances the packet-in-error field of an L2 1284 PTB message may not include enough information for translation to an 1285 L3 PTB message. In that case, the AERO interface simply discards the 1286 L2 PTB message. It can therefore be said that translation of L2 PTB 1287 messages to L3 PTB messages can provide a useful optimization when 1288 possible, but is not critical for sources that correctly use PLPMTUD. 1290 3.14. AERO Router Discovery, Prefix Delegation and Address 1291 Configuration 1293 3.14.1. AERO DHCPv6 Service Model 1295 Each AERO Server configures a DHCPv6 server function to facilitate PD 1296 requests from Clients. Each Server is pre-configured with an 1297 identical list of ACP-to-Client ID mappings for all Clients enrolled 1298 in the AERO system, as well as any information necessary to 1299 authenticate Clients. The configuration information is maintained by 1300 a central administrative authority for the AERO link and securely 1301 propagated to all Servers whenever a new Client is enrolled or an 1302 existing Client is withdrawn. 1304 With these identical configurations, each Server can function 1305 independently of all other Servers, including the maintenance of 1306 active leases. Therefore, no Server-to-Server DHCPv6 state 1307 synchronization is necessary, and Clients can optionally hold 1308 separate leases for the same ACP from multiple Servers. 1310 In this way, Clients can easily associate with multiple Servers, and 1311 can receive new leases from new Servers before deprecating leases 1312 held through old Servers. This enables a graceful "make-before- 1313 break" capability. 1315 3.14.2. AERO Client Behavior 1317 AERO Clients discover the link-layer addresses of AERO Servers via 1318 static configuration, or through an automated means such as DNS name 1319 resolution. In the absence of other information, the Client resolves 1320 the FQDN "linkupnetworks.[domainname]" where "linkupnetworks" is a 1321 constant text string and "[domainname]" is the connection-specific 1322 DNS suffix for the Client's underlying network connection (e.g., 1323 "example.com"). After discovering the link-layer addresses, the 1324 Client associates with one or more of the corresponding Servers. 1326 To associate with a Server, the Client acts as a requesting router to 1327 request an ACP through a DHCPv6 PD exchange[RFC3315][RFC3633] in 1328 which the Client's Solicit/Request messages use the IPv6 1329 "unspecified" address (i.e., "::") as the IPv6 source address, 1330 'All_DHCP_Relay_Agents_and_Servers' as the IPv6 destination address 1331 and the link-layer address of the Server as the link-layer 1332 destination address. The Client also includes a Client Identifier 1333 option with a DHCP Unique Identifier (DUID) plus any necessary 1334 authentication options to identify itself to the DHCPv6 server, and 1335 includes a Client Link Layer Address Option (CLLAO) [RFC6939] with 1336 the format shown in Figure 6: 1338 0 1 2 3 1339 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 1340 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1341 | OPTION_CLIENT_LINKLAYER_ADDR | option-length | 1342 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1343 | link-layer type (16 bits) | Link ID | Preference | 1344 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1346 Figure 6: AERO Client Link-Layer Address Option (CLLAO) Format 1348 The Client sets the CLLAO 'option-length' field to 4 and sets the 1349 'link-layer type' field to TBD1 (see: IANA Considerations), then 1350 includes appropriate Link ID and Preference values for the underlying 1351 interface over which the Solicit/Request will be issued (note that 1352 these are the same values that would be included in a TLLAO as shown 1353 in Figure 2). If the Client is pre-provisioned with an ACP 1354 associated with the AERO service, it MAY also include the ACP in the 1355 Solicit/Request message Identity Association (IA) option to indicate 1356 its preferred ACP to the DHCPv6 server. The Client then sends the 1357 encapsulated DHCPv6 request via the underlying interface. 1359 When the Client receives its ACP and the set of ASPs via a DHCPv6 1360 Reply from the AERO Server, it creates a static neighbor cache entry 1361 with the Server's link-local address as the network-layer address and 1362 the Server's encapsulation address as the link-layer address. The 1363 Client then records the lifetime for the ACP in the neighbor cache 1364 entry and marks the neighbor cache entry as "default", i.e., the 1365 Client considers the Server as a default router. (Note that the AERO 1366 Server and default router functions can be separated such that the 1367 control plane and data plane can be maintained on different platforms 1368 if desired.) If the Reply message contains a Vendor-Specific 1369 Information Option (see: Section 3.14.3) the Client also caches each 1370 ASP in the option. 1372 The Client then applies the AERO address to the AERO interface and 1373 sub-delegates the ACP to nodes and links within its attached EUNs 1374 (the AERO address thereafter remains stable as the Client moves). 1375 The Client also assigns a default IP route to the AERO interface as a 1376 route-to-interface, i.e., with no explicit next-hop. The next hop 1377 will then be determined after a packet has been submitted to the AERO 1378 interface by inspecting the neighbor cache (see above). 1380 On some platforms (e.g., popular cell phone operating systems), the 1381 act of assigning a default IPv6 route may not be permitted from a 1382 user application due to security policy. Typically, those platforms 1383 include a TUN/TAP interface that acts as a point-to-point conduit 1384 between user applications and the AERO interface. In that case, the 1385 Client can instead generate a "synthesized RA" message. The message 1386 conforms to [RFC4861] and is prepared as follows: 1388 o the IPv6 source address is fe80:: 1390 o the IPv6 destination address is all-nodes multicast 1392 o the Router Lifetime is set to a time that is no longer than the 1393 ACP DHCPv6 lifetime 1395 o the message does not include a Source Link Layer Address Option 1396 (SLLAO) 1398 o the message includes a Prefix Information Option (PIO) with a /64 1399 prefix taken from the ACP as the prefix for autoconfiguration 1401 The Client then sends the synthesized RA message via the TUN/TAP 1402 interface, where the operating system kernel will interpret it as 1403 though it were generated by an actual router. The operating system 1404 will then install a default route and use StateLess Address 1405 AutoConfiguration (SLAAC) to configure an IPv6 address on the TUN/TAP 1406 interface. Methods for similarly installing an IPv4 default route 1407 and IPv4 address on the TUN/TAP interface are based on synthesized 1408 DHCPv4 messages [RFC2131]. Note that in this method, the Client 1409 appears as a mobility proxy for applications that bind to the (point- 1410 to-point) TUN/TAP interface. The arrangement can be likened to a 1411 Proxy AERO scenario in which the mobile node and Client are located 1412 within the same physical platform (see Section 3.22 for further 1413 details on Proxy AERO). 1415 The Client subsequently renews its ACP delegation through each of its 1416 Servers by performing DHCPv6 Renew/Reply exchanges with its AERO 1417 address as the IPv6 source address, 1418 'All_DHCP_Relay_Agents_and_Servers' as the IPv6 destination address, 1419 the link-layer address of a Server as the link-layer destination 1420 address and the same Client identifier, authentication options and 1421 CLLAO option as was used in the initial PD request. Note that if the 1422 Client does not issue a DHCPv6 Renew before the Server has terminated 1423 the lease (e.g., if the Client has been out of touch with the Server 1424 for a considerable amount of time), the Server's Reply will report 1425 NoBinding and the Client must re-initiate the DHCPv6 PD procedure. 1426 If the Client sends synthesized RA and/or DHCPv4 messages (see 1427 above), it also sends a new synthesized message when issuing a DHCPv6 1428 Renew or when re-initiating the DHCPv6 PD procedure. 1430 Since the Client's AERO address is configured from the unique ACP 1431 delegation it receives, there is no need for Duplicate Address 1432 Detection (DAD) on AERO links. Other nodes maliciously attempting to 1433 hijack an authorized Client's AERO address will be denied access to 1434 the network by the DHCPv6 server due to an unacceptable link-layer 1435 address and/or security parameters (see: Security Considerations). 1437 AERO Clients ignore the IP address and UDP port number in any S/TLLAO 1438 options in ND messages they receive directly from another AERO 1439 Client, but examine the Link ID and Preference values to match the 1440 message with the correct link-layer address information. 1442 3.14.3. AERO Server Behavior 1444 AERO Servers configure a DHCPv6 server function on their AERO links. 1445 AERO Servers arrange to add their encapsulation layer IP addresses 1446 (i.e., their link-layer addresses) to the DNS resource records for 1447 the FQDN "linkupnetworks.[domainname]" before entering service. 1449 When an AERO Server receives a prospective Client's DHCPv6 PD 1450 Solicit/Request message, it first authenticates the message. If 1451 authentication succeeds, the Server determines the correct ACP to 1452 delegate to the Client by matching the Client's DUID within an online 1453 directory service (e.g., LDAP). The Server then delegates the ACP 1454 and creates a static neighbor cache entry for the Client's AERO 1455 address with lifetime set to no more than the lease lifetime and the 1456 Client's link-layer address as the link-layer address for the Link ID 1457 specified in the CLLAO option. The Server then creates an IP 1458 forwarding table entry so that the AERO routing system will propagate 1459 the ACP to all Relays (see: Section 3.7). Finally, the Server sends 1460 a DHCPv6 Reply message to the Client while using fe80::ID as the IPv6 1461 source address, the Client's AERO address as the IPv6 destination 1462 address, and the Client's link-layer address as the destination link- 1463 layer address. The Server also includes a Server Unicast option with 1464 server-address set to fe80::ID so that all future Client/Server 1465 transactions will be link-local-only unicast over the AERO link. 1467 When the Server sends the DHCPv6 Reply message, it also includes a 1468 DHCPv6 Vendor-Specific Information Option with 'enterprise-number' 1469 set to "TBD2" (see: IANA Considerations). The option is formatted as 1470 shown in[RFC3315] and with the AERO enterprise-specific format shown 1471 in Figure 7: 1473 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 1474 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1475 | OPTION_VENDOR_OPTS | option-len | 1476 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1477 | enterprise-number ("TBD2") | 1478 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1479 | Reserved | Prefix Length | 1480 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1481 | | 1482 + ASP (1) + 1483 | | 1484 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1485 | Reserved | Prefix Length | 1486 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1487 | | 1488 + ASP (2) + 1489 | | 1490 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1491 | Reserved | Prefix Length | 1492 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1493 | | 1494 + ASP (3) + 1495 | | 1496 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1497 . (etc.) . 1498 . . 1499 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1501 Figure 7: AERO Vendor-Specific Information Option 1503 Per Figure 7, the option includes one or more ASP. The ASP field 1504 contains the IP prefix as it would appear in the interface identifier 1505 portion of the corresponding AERO address (see: Section 3.3). For 1506 IPv6, valid values for the Prefix Length field are 0 through 64; for 1507 IPv4, valid values are 0 through 32. 1509 After the initial DHCPv6 PD exchange, the AERO Server maintains the 1510 neighbor cache entry for the Client until the lease lifetime expires. 1511 If the Client issues a Renew/Reply exchange, the Server extends the 1512 lifetime. If the Client issues a Release/Reply, or if the Client 1513 does not issue a Renew/Reply before the lifetime expires, the Server 1514 deletes the neighbor cache entry for the Client and withdraws the IP 1515 route from the AERO routing system. 1517 3.15. AERO Intradomain Route Optimization 1519 When a source Client forwards packets to a prospective correspondent 1520 Client within the same AERO link domain (i.e., one for which the 1521 packet's destination address is covered by an ASP), the source Client 1522 initiates an intra-domain AERO route optimization procedure. The 1523 procedure is based on an exchange of IPv6 ND messages using a chain 1524 of AERO Servers and Relays as a trust basis. This procedure is in 1525 contrast to the Return Routability procedure required for route 1526 optimization to a correspondent Client located in the Internet as 1527 described in Section 3.24. The following sections specify the AERO 1528 intradomain route optimization procedure. 1530 3.15.1. Reference Operational Scenario 1532 Figure 8 depicts the AERO intradomain route optimization reference 1533 operational scenario, using IPv6 addressing as the example (while not 1534 shown, a corresponding example for IPv4 addressing can be easily 1535 constructed). The figure shows an AERO Relay ('R1'), two AERO 1536 Servers ('S1', 'S2'), two AERO Clients ('C1', 'C2') and two ordinary 1537 IPv6 hosts ('H1', 'H2'): 1539 +--------------+ +--------------+ +--------------+ 1540 | Server S1 | | Relay R1 | | Server S2 | 1541 +--------------+ +--------------+ +--------------+ 1542 fe80::2 fe80::1 fe80::3 1543 L2(S1) L2(R1) L2(S2) 1544 | | | 1545 X-----+-----+------------------+-----------------+----+----X 1546 | AERO Link | 1547 L2(A) L2(B) 1548 fe80::2001:db8:0:0 fe80::2001:db8:1:0 1549 +--------------+ +--------------+ 1550 |AERO Client C1| |AERO Client C2| 1551 +--------------+ +--------------+ 1552 2001:DB8:0::/48 2001:DB8:1::/48 1553 | | 1554 .-. .-. 1555 ,-( _)-. 2001:db8:0::1 2001:db8:1::1 ,-( _)-. 1556 .-(_ IP )-. +---------+ +---------+ .-(_ IP )-. 1557 (__ EUN )--| Host H1 | | Host H2 |--(__ EUN ) 1558 `-(______)-' +---------+ +---------+ `-(______)-' 1560 Figure 8: AERO Reference Operational Scenario 1562 In Figure 8, Relay ('R1') assigns the address fe80::1 to its AERO 1563 interface with link-layer address L2(R1), Server ('S1') assigns the 1564 address fe80::2 with link-layer address L2(S1),and Server ('S2') 1565 assigns the address fe80::3 with link-layer address L2(S2). Servers 1566 ('S1') and ('S2') next arrange to add their link-layer addresses to a 1567 published list of valid Servers for the AERO link. 1569 AERO Client ('C1') receives the ACP 2001:db8:0::/48 in a DHCPv6 PD 1570 exchange via AERO Server ('S1') then assigns the address 1571 fe80::2001:db8:0:0 to its AERO interface with link-layer address 1572 L2(C1). Client ('C1') configures a default route and neighbor cache 1573 entry via the AERO interface with next-hop address fe80::2 and link- 1574 layer address L2(S1), then sub-delegates the ACP to its attached 1575 EUNs. IPv6 host ('H1') connects to the EUN, and configures the 1576 address 2001:db8:0::1. 1578 AERO Client ('C2') receives the ACP 2001:db8:1::/48 in a DHCPv6 PD 1579 exchange via AERO Server ('S2') then assigns the address 1580 fe80::2001:db8:1:0 to its AERO interface with link-layer address 1581 L2(C2). Client ('C2') configures a default route and neighbor cache 1582 entry via the AERO interface with next-hop address fe80::3 and link- 1583 layer address L2(S2), then sub-delegates the ACP to its attached 1584 EUNs. IPv6 host ('H1') connects to the EUN, and configures the 1585 address 2001:db8:1::1. 1587 3.15.2. Concept of Operations 1589 Again, with reference to Figure 8, when source host ('H1') sends a 1590 packet to destination host ('H2'), the packet is first forwarded over 1591 the source host's attached EUN to Client ('C1'). Client ('C1') then 1592 forwards the packet via its AERO interface to Server ('S1') and also 1593 sends a Predirect message toward Client ('C2') via Server ('S1'). 1594 Server ('S1') then re-encapsulates and forwards both the packet and 1595 the Predirect message out the same AERO interface toward Client 1596 ('C2') via Relay ('R1'). 1598 When Relay ('R1') receives the packet and Predirect message, it 1599 consults its forwarding table to discover Server ('S2') as the next 1600 hop toward Client ('C2'). Relay ('R1') then forwards both the packet 1601 and the Predirect message to Server ('S2'), which then forwards them 1602 to Client ('C2'). 1604 After Client ('C2') receives the Predirect message, it process the 1605 message and returns a Redirect message toward Client ('C1') via 1606 Server ('S2'). During the process, Client ('C2') also creates or 1607 updates a dynamic neighbor cache entry for Client ('C1'). 1609 When Server ('S2') receives the Redirect message, it re-encapsulates 1610 the message and forwards it on to Relay ('R1'), which forwards the 1611 message on to Server ('S1') which forwards the message on to Client 1612 ('C1'). After Client ('C1') receives the Redirect message, it 1613 processes the message and creates or updates a dynamic neighbor cache 1614 entry for Client ('C2'). 1616 Following the above Predirect/Redirect message exchange, forwarding 1617 of packets from Client ('C1') to Client ('C2') without involving any 1618 intermediate nodes is enabled. The mechanisms that support this 1619 exchange are specified in the following sections. 1621 3.15.3. Message Format 1623 AERO Redirect/Predirect messages use the same format as for ICMPv6 1624 Redirect messages depicted in Section 4.5 of [RFC4861], but also 1625 include a new "Prefix Length" field taken from the low-order 8 bits 1626 of the Redirect message Reserved field. For IPv6, valid values for 1627 the Prefix Length field are 0 through 64; for IPv4, valid values are 1628 0 through 32. The Redirect/Predirect messages are formatted as shown 1629 in Figure 9: 1631 0 1 2 3 1632 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 1633 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1634 | Type (=137) | Code (=0/1) | Checksum | 1635 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1636 | Reserved | Prefix Length | 1637 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1638 | | 1639 + + 1640 | | 1641 + Target Address + 1642 | | 1643 + + 1644 | | 1645 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1646 | | 1647 + + 1648 | | 1649 + Destination Address + 1650 | | 1651 + + 1652 | | 1653 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1654 | Options ... 1655 +-+-+-+-+-+-+-+-+-+-+-+- 1657 Figure 9: AERO Redirect/Predirect Message Format 1659 3.15.4. Sending Predirects 1661 When a Client forwards a packet with a source address from one of its 1662 ACPs toward a destination address covered by an ASP (i.e., toward 1663 another AERO Client connected to the same AERO link), the source 1664 Client MAY send a Predirect message forward toward the destination 1665 Client via the Server. 1667 In the reference operational scenario, when Client ('C1') forwards a 1668 packet toward Client ('C2'), it MAY also send a Predirect message 1669 forward toward Client ('C2'), subject to rate limiting (see 1670 Section 8.2 of [RFC4861]). Client ('C1') prepares the Predirect 1671 message as follows: 1673 o the link-layer source address is set to 'L2(C1)' (i.e., the link- 1674 layer address of Client ('C1')). 1676 o the link-layer destination address is set to 'L2(S1)' (i.e., the 1677 link-layer address of Server ('S1')). 1679 o the network-layer source address is set to fe80::2001:db8:0:0 1680 (i.e., the AERO address of Client ('C1')). 1682 o the network-layer destination address is set to fe80::2001:db8:1:0 1683 (i.e., the AERO address of Client ('C2')). 1685 o the Type is set to 137. 1687 o the Code is set to 1 to indicate "Predirect". 1689 o the Prefix Length is set to the length of the prefix to be 1690 assigned to the Target Address. 1692 o the Target Address is set to fe80::2001:db8:0:0 (i.e., the AERO 1693 address of Client ('C1')). 1695 o the Destination Address is set to the source address of the 1696 originating packet that triggered the Predirection event. (If the 1697 originating packet is an IPv4 packet, the address is constructed 1698 in IPv4-compatible IPv6 address format). 1700 o the message includes one or more TLLAOs with Link ID and 1701 Preference set to appropriate values for Client ('C1')'s 1702 underlying interfaces, and with UDP Port Number and IP Address set 1703 to 0'. 1705 o the message SHOULD include a Timestamp option and a Nonce option. 1707 o the message includes a Redirected Header Option (RHO) that 1708 contains the originating packet truncated if necessary to ensure 1709 that at least the network-layer header is included but the size of 1710 the message does not exceed 1280 bytes. 1712 Note that the act of sending Predirect messages is cited as "MAY", 1713 since Client ('C1') may have advanced knowledge that the direct path 1714 to Client ('C2') would be unusable or otherwise undesirable. If the 1715 direct path later becomes unusable after the initial route 1716 optimization, Client ('C1') simply allows packets to again flow 1717 through Server ('S1'). 1719 3.15.5. Re-encapsulating and Relaying Predirects 1721 When Server ('S1') receives a Predirect message from Client ('C1'), 1722 it first verifies that the TLLAOs in the Predirect are a proper 1723 subset of the Link IDs in Client ('C1')'s neighbor cache entry. If 1724 the Client's TLLAOs are not acceptable, Server ('S1') discards the 1725 message. Otherwise, Server ('S1') validates the message according to 1726 the ICMPv6 Redirect message validation rules in Section 8.1 of 1728 [RFC4861], except that the Predirect has Code=1. Server ('S1') also 1729 verifies that Client ('C1') is authorized to use the Prefix Length in 1730 the Predirect when applied to the AERO address in the network-layer 1731 source address by searching for the AERO address in the neighbor 1732 cache. If validation fails, Server ('S1') discards the Predirect; 1733 otherwise, it copies the correct UDP Port numbers and IP Addresses 1734 for Client ('C1')'s links into the (previously empty) TLLAOs. 1736 Server ('S1') then examines the network-layer destination address of 1737 the Predirect to determine the next hop toward Client ('C2') by 1738 searching for the AERO address in the neighbor cache. Since Client 1739 ('C2') is not one of its neighbors, Server ('S1') re-encapsulates the 1740 Predirect and relays it via Relay ('R1') by changing the link-layer 1741 source address of the message to 'L2(S1)' and changing the link-layer 1742 destination address to 'L2(R1)'. Server ('S1') finally forwards the 1743 re-encapsulated message to Relay ('R1') without decrementing the 1744 network-layer TTL/Hop Limit field. 1746 When Relay ('R1') receives the Predirect message from Server ('S1') 1747 it determines that Server ('S2') is the next hop toward Client ('C2') 1748 by consulting its forwarding table. Relay ('R1') then re- 1749 encapsulates the Predirect while changing the link-layer source 1750 address to 'L2(R1)' and changing the link-layer destination address 1751 to 'L2(S2)'. Relay ('R1') then relays the Predirect via Server 1752 ('S2'). 1754 When Server ('S2') receives the Predirect message from Relay ('R1') 1755 it determines that Client ('C2') is a neighbor by consulting its 1756 neighbor cache. Server ('S2') then re-encapsulates the Predirect 1757 while changing the link-layer source address to 'L2(S2)' and changing 1758 the link-layer destination address to 'L2(C2)'. Server ('S2') then 1759 forwards the message to Client ('C2'). 1761 3.15.6. Processing Predirects and Sending Redirects 1763 When Client ('C2') receives the Predirect message, it accepts the 1764 Predirect only if the message has a link-layer source address of one 1765 of its Servers (e.g., L2(S2)). Client ('C2') further accepts the 1766 message only if it is willing to serve as a redirection target. 1767 Next, Client ('C2') validates the message according to the ICMPv6 1768 Redirect message validation rules in Section 8.1 of [RFC4861], except 1769 that it accepts the message even though Code=1 and even though the 1770 network-layer source address is not that of it's current first-hop 1771 router. 1773 In the reference operational scenario, when Client ('C2') receives a 1774 valid Predirect message, it either creates or updates a dynamic 1775 neighbor cache entry that stores the Target Address of the message as 1776 the network-layer address of Client ('C1') , stores the link-layer 1777 addresses found in the TLLAOs as the link-layer addresses of Client 1778 ('C1') and stores the Prefix Length as the length to be applied to 1779 the network-layer address for forwarding purposes. Client ('C2') 1780 then sets AcceptTime for the neighbor cache entry to ACCEPT_TIME. 1782 After processing the message, Client ('C2') prepares a Redirect 1783 message response as follows: 1785 o the link-layer source address is set to 'L2(C2)' (i.e., the link- 1786 layer address of Client ('C2')). 1788 o the link-layer destination address is set to 'L2(S2)' (i.e., the 1789 link-layer address of Server ('S2')). 1791 o the network-layer source address is set to fe80::2001:db8:1:0 1792 (i.e., the AERO address of Client ('C2')). 1794 o the network-layer destination address is set to fe80::2001:db8:0:0 1795 (i.e., the AERO address of Client ('C1')). 1797 o the Type is set to 137. 1799 o the Code is set to 0 to indicate "Redirect". 1801 o the Prefix Length is set to the length of the prefix to be applied 1802 to the Target Address. 1804 o the Target Address is set to fe80::2001:db8:1:0 (i.e., the AERO 1805 address of Client ('C2')). 1807 o the Destination Address is set to the destination address of the 1808 originating packet that triggered the Redirection event. (If the 1809 originating packet is an IPv4 packet, the address is constructed 1810 in IPv4-compatible IPv6 address format). 1812 o the message includes one or more TLLAOs with Link ID and 1813 Preference set to appropriate values for Client ('C2')'s 1814 underlying interfaces, and with UDP Port Number and IP Address set 1815 to '0'. 1817 o the message SHOULD include a Timestamp option and MUST echo the 1818 Nonce option received in the Predirect (i.e., if a Nonce option is 1819 included). 1821 o the message includes as much of the RHO copied from the 1822 corresponding AERO Predirect message as possible such that at 1823 least the network-layer header is included but the size of the 1824 message does not exceed 1280 bytes. 1826 After Client ('C2') prepares the Redirect message, it sends the 1827 message to Server ('S2'). 1829 3.15.7. Re-encapsulating and Relaying Redirects 1831 When Server ('S2') receives a Redirect message from Client ('C2'), it 1832 first verifies that the TLLAOs in the Redirect are a proper subset of 1833 the Link IDs in Client ('C2')'s neighbor cache entry. If the 1834 Client's TLLAOs are not acceptable, Server ('S2') discards the 1835 message. Otherwise, Server ('S2') validates the message according to 1836 the ICMPv6 Redirect message validation rules in Section 8.1 of 1837 [RFC4861]. Server ('S2') also verifies that Client ('C2') is 1838 authorized to use the Prefix Length in the Redirect when applied to 1839 the AERO address in the network-layer source address by searching for 1840 the AERO address in the neighbor cache. If validation fails, Server 1841 ('S2') discards the Predirect; otherwise, it copies the correct UDP 1842 Port numbers and IP Addresses for Client ('C2')'s links into the 1843 (previously empty) TLLAOs. 1845 Server ('S2') then examines the network-layer destination address of 1846 the Predirect to determine the next hop toward Client ('C2') by 1847 searching for the AERO address in the neighbor cache. Since Client 1848 ('C2') is not a neighbor, Server ('S2') re-encapsulates the Predirect 1849 and relays it via Relay ('R1') by changing the link-layer source 1850 address of the message to 'L2(S2)' and changing the link-layer 1851 destination address to 'L2(R1)'. Server ('S2') finally forwards the 1852 re-encapsulated message to Relay ('R1') without decrementing the 1853 network-layer TTL/Hop Limit field. 1855 When Relay ('R1') receives the Predirect message from Server ('S2') 1856 it determines that Server ('S1') is the next hop toward Client ('C1') 1857 by consulting its forwarding table. Relay ('R1') then re- 1858 encapsulates the Predirect while changing the link-layer source 1859 address to 'L2(R1)' and changing the link-layer destination address 1860 to 'L2(S1)'. Relay ('R1') then relays the Predirect via Server 1861 ('S1'). 1863 When Server ('S1') receives the Predirect message from Relay ('R1') 1864 it determines that Client ('C1') is a neighbor by consulting its 1865 neighbor cache. Server ('S1') then re-encapsulates the Predirect 1866 while changing the link-layer source address to 'L2(S1)' and changing 1867 the link-layer destination address to 'L2(C1)'. Server ('S1') then 1868 forwards the message to Client ('C1'). 1870 3.15.8. Processing Redirects 1872 When Client ('C1') receives the Redirect message, it accepts the 1873 message only if it has a link-layer source address of one of its 1874 Servers (e.g., ''L2(S1)'). Next, Client ('C1') validates the message 1875 according to the ICMPv6 Redirect message validation rules in 1876 Section 8.1 of [RFC4861], except that it accepts the message even 1877 though the network-layer source address is not that of it's current 1878 first-hop router. Following validation, Client ('C1') then processes 1879 the message as follows. 1881 In the reference operational scenario, when Client ('C1') receives 1882 the Redirect message, it either creates or updates a dynamic neighbor 1883 cache entry that stores the Target Address of the message as the 1884 network-layer address of Client ('C2'), stores the link-layer 1885 addresses found in the TLLAOs as the link-layer addresses of Client 1886 ('C2') and stores the Prefix Length as the length to be applied to 1887 the network-layer address for forwarding purposes. Client ('C1') 1888 then sets ForwardTime for the neighbor cache entry to FORWARD_TIME. 1890 Now, Client ('C1') has a neighbor cache entry with a valid 1891 ForwardTime value, while Client ('C2') has a neighbor cache entry 1892 with a valid AcceptTime value. Thereafter, Client ('C1') may forward 1893 ordinary network-layer data packets directly to Client ('C2') without 1894 involving any intermediate nodes, and Client ('C2') can verify that 1895 the packets came from an acceptable source. (In order for Client 1896 ('C2') to forward packets to Client ('C1'), a corresponding 1897 Predirect/Redirect message exchange is required in the reverse 1898 direction; hence, the mechanism is asymmetric.) 1900 3.15.9. Server-Oriented Redirection 1902 In some environments, the Server nearest the target Client may need 1903 to serve as the redirection target, e.g., if direct Client-to-Client 1904 communications are not possible. In that case, the Server prepares 1905 the Redirect message the same as if it were the destination Client 1906 (see: Section 3.15.6), except that it writes its own link-layer 1907 address in the TLLAO option. The Server must then maintain a dynamic 1908 neighbor cache entry for the redirected source Client. 1910 3.16. Neighbor Unreachability Detection (NUD) 1912 AERO nodes perform Neighbor Unreachability Detection (NUD) by sending 1913 unicast NS messages to elicit solicited NA messages from neighbors 1914 the same as described in [RFC4861]. NUD is performed either 1915 reactively in response to persistent L2 errors (see Section 3.13) or 1916 proactively to refresh existing neighbor cache entries. 1918 When an AERO node sends an NS/NA message, it MUST use its link-local 1919 address as the IPv6 source address and the link-local address of the 1920 neighbor as the IPv6 destination address. When an AERO node receives 1921 an NS message or a solicited NA message, it accepts the message if it 1922 has a neighbor cache entry for the neighbor; otherwise, it ignores 1923 the message. 1925 When a source Client is redirected to a target Client it SHOULD 1926 proactively test the direct path by sending an initial NS message to 1927 elicit a solicited NA response. While testing the path, the source 1928 Client can optionally continue sending packets via the Server, 1929 maintain a small queue of packets until target reachability is 1930 confirmed, or (optimistically) allow packets to flow directly to the 1931 target. The source Client SHOULD thereafter continue to proactively 1932 test the direct path to the target Client (see Section 7.3 of 1933 [RFC4861]) periodically in order to keep dynamic neighbor cache 1934 entries alive. 1936 In particular, while the source Client is actively sending packets to 1937 the target Client it SHOULD also send NS messages separated by 1938 RETRANS_TIMER milliseconds in order to receive solicited NA messages. 1939 If the source Client is unable to elicit a solicited NA response from 1940 the target Client after MAX_RETRY attempts, it SHOULD set ForwardTime 1941 to 0 and resume sending packets via one of its Servers. Otherwise, 1942 the source Client considers the path usable and SHOULD thereafter 1943 process any link-layer errors as a hint that the direct path to the 1944 target Client has either failed or has become intermittent. 1946 When a target Client receives an NS message from a source Client, it 1947 resets AcceptTime to ACCEPT_TIME if a neighbor cache entry exists; 1948 otherwise, it discards the NS message. If ForwardTime is non-zero, 1949 the target Client then sends a solicited NA message to the link-layer 1950 address of the source Client; otherwise, it sends the solicited NA 1951 message to the link-layer address of one of its Servers. 1953 When a source Client receives a solicited NA message from a target 1954 Client, it resets ForwardTime to FORWARD_TIME if a neighbor cache 1955 entry exists; otherwise, it discards the NA message. 1957 When ForwardTime for a dynamic neighbor cache entry expires, the 1958 source Client resumes sending any subsequent packets via a Server and 1959 may (eventually) attempt to re-initiate the AERO redirection process. 1960 When AcceptTime for a dynamic neighbor cache entry expires, the 1961 target Client discards any subsequent packets received directly from 1962 the source Client. When both ForwardTime and AcceptTime for a 1963 dynamic neighbor cache entry expire, the Client deletes the neighbor 1964 cache entry. 1966 3.17. Mobility Management 1968 3.17.1. Announcing Link-Layer Address Changes 1970 When a Client needs to change its link-layer address, e.g., due to a 1971 mobility event, it performs an immediate DHCPv6 Rebind/Reply exchange 1972 via each of its Servers using the new link-layer address as the 1973 source and with a CLLAO that includes the correct Link ID and 1974 Preference values. If authentication succeeds, the Server then 1975 update its neighbor cache and sends a DHCPv6 Reply. Note that if the 1976 Client does not issue a DHCPv6 Rebind before the lease lifetime 1977 expires (e.g., if the Client has been out of touch with the Server 1978 for a considerable amount of time), the Server's Reply will report 1979 NoBinding and the Client must re-initiate the DHCPv6 PD procedure. 1981 Next, the Client sends unsolicited NA messages to each of its 1982 correspondent Client neighbors using the same procedures as specified 1983 in Section 7.2.6 of [RFC4861], except that it sends the messages as 1984 unicast to each neighbor via a Server instead of multicast. In this 1985 process, the Client should send no more than 1986 MAX_NEIGHBOR_ADVERTISEMENT messages separated by no less than 1987 RETRANS_TIMER seconds to each neighbor. 1989 With reference to Figure 8, when Client ('C2') needs to change its 1990 link-layer address it sends unicast unsolicited NA messages to Client 1991 ('C1') via Server ('S2') as follows: 1993 o the link-layer source address is set to 'L2(C2)' (i.e., the link- 1994 layer address of Client ('C2')). 1996 o the link-layer destination address is set to 'L2(S2)' (i.e., the 1997 link-layer address of Server ('S2')). 1999 o the network-layer source address is set to fe80::2001:db8:1:0 2000 (i.e., the AERO address of Client ('C2')). 2002 o the network-layer destination address is set to fe80::2001:db8:0:0 2003 (i.e., the AERO address of Client ('C1')). 2005 o the Type is set to 136. 2007 o the Code is set to 0. 2009 o the Solicited flag is set to 0. 2011 o the Override flag is set to 1. 2013 o the Target Address is set to fe80::2001:db8:1:0 (i.e., the AERO 2014 address of Client ('C2')). 2016 o the message includes one or more TLLAOs with Link ID and 2017 Preference set to appropriate values for Client ('C2')'s 2018 underlying interfaces, and with UDP Port Number and IP Address set 2019 to '0'. 2021 o the message SHOULD include a Timestamp option. 2023 When Server ('S1') receives the NA message, it relays the message in 2024 the same way as described for relaying Redirect messages in 2025 Section 3.15.7. In particular, Server ('S1') copies the correct UDP 2026 port numbers and IP addresses into the TLLAOs, changes the link-layer 2027 source address to its own address, changes the link-layer destination 2028 address to the address of Relay ('R1'), then forwards the NA message 2029 via the relaying chain the same as for a Redirect. 2031 When Client ('C1') receives the NA message, it accepts the message 2032 only if it already has a neighbor cache entry for Client ('C2') then 2033 updates the link-layer addresses for Client ('C2') based on the 2034 addresses in the TLLAOs. Next, Client ('C1') SHOULD initiate the NUD 2035 procedures specified in Section 3.16 to provide Client ('C2') with an 2036 indication that the link-layer source address has been updated, and 2037 to refresh ('C2')'s AcceptTime and ('C1')'s ForwardTime timers. 2039 If Client ('C2') receives an NS message from Client ('C1') indicating 2040 that an unsolicited NA has updated its neighbor cache, Client ('C2') 2041 need not send additional unsolicited NAs. If Client ('C2')'s 2042 unsolicited NA messages are somehow lost, however, Client ('C1') will 2043 soon learn of the mobility event via NUD. 2045 3.17.2. Bringing New Links Into Service 2047 When a Client needs to bring a new underlying interface into service 2048 (e.g., when it activates a new data link), it performs an immediate 2049 Rebind/Reply exchange via each of its Servers using the new link- 2050 layer address as the source address and with a CLLAO that includes 2051 the new Link ID and Preference values. If authentication succeeds, 2052 the Server then updates its neighbor cache and sends a DHCPv6 Reply. 2053 The Client MAY then send unsolicited NA messages to each of its 2054 correspondent Clients to inform them of the new link-layer address as 2055 described in Section 3.17.1. 2057 3.17.3. Removing Existing Links from Service 2059 When a Client needs to remove an existing underlying interface from 2060 service (e.g., when it de-activates an existing data link), it 2061 performs an immediate Rebind/Reply exchange via each of its Servers 2062 over any available link with a CLLAO that includes the deprecated 2063 Link ID and a Preference value of 0. If authentication succeeds, the 2064 Server then updates its neighbor cache and sends a DHCPv6 Reply. The 2065 Client SHOULD then send unsolicited NA messages to each of its 2066 correspondent Clients to inform them of the deprecated link-layer 2067 address as described in Section 3.17.1. 2069 3.17.4. Moving to a New Server 2071 When a Client associates with a new Server, it performs the Client 2072 procedures specified in Section 3.14.2. 2074 When a Client disassociates with an existing Server, it sends a 2075 DHCPv6 Release message via a new Server to the unicast link-local 2076 network layer address of the old Server. The new Server then writes 2077 its own link-layer address in the DHCPv6 release message IP source 2078 address and forwards the message to the old Server. 2080 When the old Server receives the DHCPv6 Release, it first 2081 authenticates the message. The Server then resets the Client's 2082 neighbor cache entry lifetime to 5 seconds, rewrites the link-layer 2083 address in the neighbor cache entry to the address of the new Server, 2084 then returns a DHCPv6 Reply message to the Client via the old Server. 2085 When the lifetime expires, the old Server withdraws the IP route from 2086 the AERO routing system and deletes the neighbor cache entry for the 2087 Client. The Client can then use the Reply message to verify that the 2088 termination signal has been processed, and can delete both the 2089 default route and the neighbor cache entry for the old Server. (Note 2090 that since Release/Reply messages may be lost in the network the 2091 Client MUST retry until it gets Reply indicating that the Release was 2092 successful.) 2094 Clients SHOULD NOT move rapidly between Servers in order to avoid 2095 causing excessive oscillations in the AERO routing system. Such 2096 oscillations could result in intermittent reachability for the Client 2097 itself, while causing little harm to the network. Examples of when a 2098 Client might wish to change to a different Server include a Server 2099 that has gone unreachable, topological movements of significant 2100 distance, etc. 2102 3.18. Encapsulation Protocol Version Considerations 2104 A source Client may connect only to an IPvX underlying network, while 2105 the target Client connects only to an IPvY underlying network. In 2106 that case, the target and source Clients have no means for reaching 2107 each other directly (since they connect to underlying networks of 2108 different IP protocol versions) and so must ignore any redirection 2109 messages and continue to send packets via the Server. 2111 3.19. Multicast Considerations 2113 When the underlying network does not support multicast, AERO nodes 2114 map IPv6 link-scoped multicast addresses (including 2115 'All_DHCP_Relay_Agents_and_Servers') to the link-layer address of a 2116 Server. 2118 When the underlying network supports multicast, AERO nodes use the 2119 multicast address mapping specification found in [RFC2529] for IPv4 2120 underlying networks and use a direct multicast mapping for IPv6 2121 underlying networks. (In the latter case, "direct multicast mapping" 2122 means that if the IPv6 multicast destination address of the 2123 encapsulated packet is "M", then the IPv6 multicast destination 2124 address of the encapsulating header is also "M".) 2126 3.20. Operation on AERO Links Without DHCPv6 Services 2128 When Servers on the AERO link do not provide DHCPv6 services, 2129 operation can still be accommodated through administrative 2130 configuration of ACPs on AERO Clients. In that case, administrative 2131 configurations of AERO interface neighbor cache entries on both the 2132 Server and Client are also necessary. However, this may interfere 2133 with the ability for Clients to dynamically change to new Servers, 2134 and can expose the AERO link to misconfigurations unless the 2135 administrative configurations are carefully coordinated. 2137 3.21. Operation on Server-less AERO Links 2139 In some AERO link scenarios, there may be no Servers on the link and/ 2140 or no need for Clients to use a Server as an intermediary trust 2141 anchor. In that case, each Client acts as a Server unto itself to 2142 establish neighbor cache entries by performing direct Client-to- 2143 Client IPv6 ND message exchanges, and some other form of trust basis 2144 must be applied so that each Client can verify that the prospective 2145 neighbor is authorized to use its claimed ACP. 2147 When there is no Server on the link, Clients must arrange to receive 2148 ACPs and publish them via a secure alternate prefix delegation 2149 authority through some means outside the scope of this document. 2151 3.22. Proxy AERO 2153 Proxy Mobile IPv6 (PMIPv6) [RFC5213][RFC5844][RFC5949] presents a 2154 localized mobility management scheme for use within an access network 2155 domain. It is typically used in WiFi and cellular wireless access 2156 networks, and allows Mobile Nodes (MNs) to receive and retain an IP 2157 address that remains stable within the access network domain without 2158 needing to implement any special mobility protocols. In the PMIPv6 2159 architecture, access network devices known as Mobility Access 2160 Gateways (MAGs) provide MNs with an access link abstraction and 2161 receive prefixes for the MNs from a Local Mobility Anchor (LMA). 2163 In a proxy AERO domain, a proxy AERO Client (acting as a MAG) can 2164 similarly provide proxy services for MNs that do not participate in 2165 AERO messaging. The proxy Client presents an access link abstraction 2166 to MNs, and performs DHCPv6 PD exchanges over the AERO interface with 2167 an AERO Server (acting as an LMA) to receive ACPs for address 2168 provisioning of new MNs that come onto an access link. This scheme 2169 assumes that proxy Clients act as fixed (non-mobile) infrastructure 2170 elements under the same administrative trust basis as for Relays and 2171 Servers. 2173 When an MN comes onto an access link within a proxy AERO domain for 2174 the first time, the proxy Client authenticates the MN and obtains a 2175 unique identifier that it can use as a DHCPv6 DUID then issues a 2176 DHCPv6 PD Request to its Server. When the Server delegates an ACP, 2177 the proxy Client creates an AERO address for the MN and assigns the 2178 ACP to the MN's access link. The proxy Client then configures itself 2179 as a default router for the MN and provides address autoconfiguration 2180 services (e.g., SLAAC, DHCPv6, DHCPv4, etc.) for provisioning MN 2181 addresses from the ACP over the access link. Since the proxy Client 2182 may serve many such MNs simultaneously, it may receive multiple ACP 2183 prefix delegations and configure multiple AERO addresses, i.e., one 2184 for each MN. 2186 When two MNs are associated with the same proxy Client, the Client 2187 can forward traffic between the MNs without involving a Server since 2188 it configures the AERO addresses of both MNs and therefore also has 2189 the necessary routing information. When two MNs are associated with 2190 different proxy Clients, the source MN's Client can initiate standard 2191 AERO route optimization to discover a direct path to the target MN's 2192 Client through the exchange of Predirect/Redirect messages. 2194 When an MN in a proxy AERO domain leaves an access link provided by 2195 an old proxy Client, the MN issues an access link-specific "leave" 2196 message that informs the old Client of the link-layer address of a 2197 new Client on the planned new access link. This is known as a 2198 "predictive handover". When an MN comes onto an access link provided 2199 by a new proxy Client, the MN issues an access link-specific "join" 2200 message that informs the new Client of the link-layer address of the 2201 old Client on the actual old access link. This is known as a 2202 "reactive handover". 2204 Upon receiving a predictive handover indication, the old proxy Client 2205 sends a DHCPv6 PD Request message directly to the new Client and 2206 queues any arriving data packets addressed to the departed MN. The 2207 Request message includes the MN's ID as the DUID, the ACP in an IA_PD 2208 option, the AERO address derived from the MN's ACP as the network- 2209 layer source address, 'All_DHCP_Relay_Agents_and_Servers' as the 2210 network-layer destination address, the old Client's address as the 2211 link-layer source address and the new Client's address as the link- 2212 layer destination address. When the new Client receives the Request 2213 message, it changes the link-layer source address to its own address, 2214 changes the link-layer destination address to the address of its 2215 Server, and forwards the message to the Server. At the same time, 2216 the new Client creates access link state for the ACP in anticipation 2217 of the MN's arrival (while queuing any data packets until the MN 2218 arrives), creates a neighbor cache entry for the old Client with 2219 AcceptTime set to ACCEPT_TIME, then sends a Redirect message back to 2220 the old Client. When the old Client receives the Redirect message, 2221 it creates a neighbor cache entry for new Client with ForwardTime set 2222 to FORWARD_TIME, then forwards any queued data packets to the new 2223 Client. At the same time, the old Client sends a DHCPv6 PD Release 2224 message to its Server. Finally, the old Client sends unsolicited NA 2225 messages to any of the ACP's correspondents with a TLLAO containing 2226 the link-layer address of the new Client. This follows the procedure 2227 specified in Section 3.17.1, except that it is the old Client and not 2228 the Server that supplies the link-layer address. 2230 Upon receiving a reactive handover indication, the new proxy Client 2231 creates access link state for the MN's ACP, sends a DHCPv6 PD Request 2232 message to its Server, and sends a DHCPv6 PD Release message directly 2233 to the old Client. The Release message includes the MN's ID as the 2234 DUID, the ACP in an IA_PD option, the AERO address derived from the 2235 MN's ACP as the network-layer source address, 2236 'All_DHCP_Relay_Agents_and_Servers' as the network-layer destination 2237 address, the new Client's address as the link-layer source address 2238 and the old Client's address as the link-layer destination address. 2239 When the old Client receives the Release message, it changes the 2240 link-layer source address to its own address, changes the link-layer 2241 destination address to the address of its Server, and forwards the 2242 message to the Server. At the same time, the old Client sends a 2243 Predirect message back to the new Client and queues any arriving data 2244 packets addressed to the departed MN. When the new Client receives 2245 the Predirect, it creates a neighbor cache entry for the old Client 2246 with AcceptTime set to ACCEPT_TIME, then sends a Redirect message 2247 back to the old Client. When the old Client receives the Redirect 2248 message, it creates a neighbor cache entry for the new Client with 2249 ForwardTime set to FORWARD_TIME, then forwards any queued data 2250 packets to the new Client. Finally, the old Client sends unsolicited 2251 NA messages to correspondents the same as for the predictive case. 2253 When a Server processes a DHCPv6 Request message, it creates a 2254 neighbor cache entry for this ACP if none currently exists. If a 2255 neighbor cache entry already exists, however, the Server changes the 2256 link-layer address to the address of the new proxy Client (this 2257 satisfies the case of both the old Client and new Client using the 2258 same Server). 2260 When a Server processes a DHCPv6 Release message, it resets the 2261 neighbor cache entry lifetime for this ACP to 5 seconds if the cached 2262 link-layer address matches the old proxy Client's address. 2263 Otherwise, the Server ignores the Release message (this satisfies the 2264 case of both the old Client and new Client using the same Server). 2266 When a correspondent Client receives an unsolicited NA message, it 2267 changes the link-layer address for the ACP's neighbor cache entry to 2268 the address of the new proxy Client. The correspondent Client then 2269 issues a Predirect/Redirect exchange to establish a new neighbor 2270 cache entry in the new Client. 2272 From an architectural perspective, in addition to the use of DHCPv6 2273 PD and IPv6 ND signaling the AERO approach differs from PMIPv6 in its 2274 use of the NBMA virtual link model instead of point-to-point tunnels. 2275 This provides a more agile interface for Client/Server and Client/ 2276 Client coordinations, and also facilitates simple route optimization. 2277 The AERO routing system is also arranged in such a fashion that 2278 Clients get the same service from any Server they happen to associate 2279 with. This provides a natural fault tolerance and load balancing 2280 capability such as desired for distributed mobility management. 2282 3.23. Extending AERO Links Through Security Gateways 2284 When an enterprise mobile device moves from a campus LAN connection 2285 to a public Internet link, it must re-enter the enterprise via a 2286 security gateway that has both a physical interface connection to the 2287 Internet and a physical interface connection to the enterprise 2288 internetwork. This most often entails the establishment of a Virtual 2289 Private Network (VPN) link over the public Internet from the mobile 2290 device to the security gateway. During this process, the mobile 2291 device supplies the security gateway with its public Internet address 2292 as the link-layer address for the VPN. The mobile device then acts 2293 as an AERO Client to negotiate with the security gateway to obtain 2294 its ACP. 2296 In order to satisfy this need, the security gateway also operates as 2297 an AERO Server with support for AERO Client proxying. In particular, 2298 when a mobile device (i.e., the Client) connects via the security 2299 gateway (i.e., the Server), the Server provides the Client with an 2300 ACP in a DHCPv6 PD exchange the same as if it were attached to an 2301 enterprise campus access link. The Server then replaces the Client's 2302 link-layer source address with the Server's enterprise-facing link- 2303 layer address in all AERO messages the Client sends toward neighbors 2304 on the AERO link. The AERO messages are then delivered to other 2305 devices on the AERO link as if they were originated by the security 2306 gateway instead of by the AERO Client. In the reverse direction, the 2307 AERO messages sourced by devices within the enterprise network can be 2308 forwarded to the security gateway, which then replaces the link-layer 2309 destination address with the Client's link-layer address and replaces 2310 the link-layer source address with its own (Internet-facing) link- 2311 layer address. 2313 After receiving the ACP, the Client can send IP packets that use an 2314 address taken from the ACP as the network layer source address, the 2315 Client's link-layer address as the link-layer source address, and the 2316 Server's Internet-facing link-layer address as the link-layer 2317 destination address. The Server will then rewrite the link-layer 2318 source address with the Server's own enterprise-facing link-layer 2319 address and rewrite the link-layer destination address with the 2320 target AERO node's link-layer address, and the packets will enter the 2321 enterprise network as though they were sourced from a device located 2322 within the enterprise. In the reverse direction, when a packet 2323 sourced by a node within the enterprise network uses a destination 2324 address from the Client's ACP, the packet will be delivered to the 2325 security gateway which then rewrites the link-layer destination 2326 address to the Client's link-layer address and rewrites the link- 2327 layer source address to the Server's Internet-facing link-layer 2328 address. The Server then delivers the packet across the VPN to the 2329 AERO Client. In this way, the AERO virtual link is essentially 2330 extended *through* the security gateway to the point at which the VPN 2331 link and AERO link are effectively grafted together by the link-layer 2332 address rewriting performed by the security gateway. All AERO 2333 messaging services (including route optimization and mobility 2334 signaling) are therefore extended to the Client. 2336 In order to support this virtual link grafting, the security gateway 2337 (acting as an AERO Server) must keep static neighbor cache entries 2338 for all of its associated Clients located on the public Internet. 2339 The neighbor cache entry is keyed by the AERO Client's AERO address 2340 the same as if the Client were located within the enterprise 2341 internetwork. The neighbor cache is then managed in all ways as 2342 though the Client were an ordinary AERO Client. This includes the 2343 AERO IPv6 ND messaging signaling for Route Optimization and Neighbor 2344 Unreachability Detection. 2346 Note that the main difference between a security gateway acting as an 2347 AERO Server and an enterprise-internal AERO Server is that the 2348 security gateway has at least one enterprise-internal physical 2349 interface and at least one public Internet physical interface. 2350 Conversely, the enterprise-internal AERO Server has only enterprise- 2351 internal physical interfaces. For this reason security gateway 2352 proxying is needed to ensure that the public Internet link-layer 2353 addressing space is kept separate from the enterprise-internal link- 2354 layer addressing space. This is afforded through a natural extension 2355 of the security association caching already performed for each VPN 2356 client by the security gateway. 2358 3.24. Extending IPv6 AERO Links to the Internet 2360 When an IPv6 host ('H1') with an address from an ACP owned by AERO 2361 Client ('C1') sends packets to a correspondent IPv6 host ('H2'), the 2362 packets eventually arrive at the IPv6 router that owns ('H2')s 2363 prefix. This IPv6 router may or may not be an AERO Client ('C2') 2364 either within the same home network as ('C1') or in a different home 2365 network. 2367 If Client ('C1') is currently located outside the boundaries of its 2368 home network, it will connect back into the home network via a 2369 security gateway acting as an AERO Server. The packets sent by 2370 ('H1') via ('C1') will then be forwarded through the security gateway 2371 then through the home network and finally to ('C2') where they will 2372 be delivered to ('H2'). This could lead to sub-optimal performance 2373 when ('C2') could instead be reached via a more direct route without 2374 involving the security gateway. 2376 Consider the case when host ('H1') has the IPv6 address 2377 2001:db8:1::1, and Client ('C1') has the ACP 2001:db8:1::/64 with 2378 underlying IPv6 Internet address of 2001:db8:1000::1. Also, host 2379 ('H2') has the IPv6 address 2001:db8:2::1, and Client ('C2') has the 2380 ACP 2001:db8:2::/64 with underlying IPv6 address of 2001:db8:2000::1. 2381 Client ('C1') can determine whether 'C2' is indeed also an AERO 2382 Client willing to serve as a route optimization correspondent by 2383 resolving the AAAA records for the DNS FQDN that matches ('H2')s 2384 prefix, i.e.: 2386 '0.0.0.0.2.0.0.0.8.b.d.0.1.0.0.2.aero.linkupnetworks.net' 2388 If ('C2') is indeed a candidate correspondent, the FQDN lookup will 2389 return a PTR resource record that contains the domain name for the 2390 AERO link that manages ('C2')s ASP. Client ('C1') can then attempt 2391 route optimization using an approach similar to the Return 2392 Routability procedure specified for Mobile IPv6 (MIPv6) [RFC6275]. 2393 In order to support this process, both Clients MUST intercept and 2394 decapsulate packets that have a subnet router anycast address 2395 corresponding to any of the /64 prefixes covered by their respective 2396 ACPs. 2398 To initiate the process, Client ('C1') creates a specially-crafted 2399 encapsulated AERO Predirect message that will be routed through its 2400 home network then through ('C2')s home network and finally to ('C2') 2401 itself. Client ('C1') prepares the initial message in the exchange 2402 as follows: 2404 o The encapsulating IPv6 header source address is set to 2405 2001:db8:1:: (i.e., the IPv6 subnet router anycast address for 2406 ('C1')s ACP) 2408 o The encapsulating IPv6 header destination address is set to 2409 2001:db8:2:: (i.e., the IPv6 subnet router anycast address for 2410 ('C2')s ACP) 2412 o The encapsulating IPv6 header is followed by a UDP header with 2413 source and destination port set to 8060 2415 o The encapsulated IPv6 header source address is set to 2416 fe80::2001:db8:1:0 (i.e., the AERO address for ('C1')) 2418 o The encapsulated IPv6 header destination address is set to 2419 fe80::2001:db8:2:0 (i.e., the AERO address for ('C2')) 2421 o The encapsulated AERO Predirect message includes all of the 2422 securing information that would occur in a MIPv6 "Home Test Init" 2423 message (format TBD) 2425 Client ('C1') then further encapsulates the message in the 2426 encapsulating headers necessary to convey the packet to the security 2427 gateway (e.g., through IPsec encapsulation) so that the message now 2428 appears "double-encapsulated". ('C1') then sends the message to the 2429 security gateway, which re-encapsulates and forwards it over the home 2430 network from where it will eventually reach ('C2'). 2432 At the same time, ('C1') creates and sends a second encapsulated AERO 2433 Predirect message that will be routed through the IPv6 Internet 2434 without involving the security gateway. Client ('C1') prepares the 2435 message as follows: 2437 o The encapsulating IPv6 header source address is set to 2438 2001:db8:1000:1 (i.e., the Internet IPv6 address of ('C1')) 2440 o The encapsulating IPv6 header destination address is set to 2441 2001:db8:2:: (i.e., the IPv6 subnet router anycast address for 2442 ('C2')s ACP) 2444 o The encapsulating IPv6 header is followed by a UDP header with 2445 source and destination port set to 8060 2447 o The encapsulated IPv6 header source address is set to 2448 fe80::2001:db8:1:0 (i.e., the AERO address for ('C1')) 2450 o The encapsulated IPv6 header destination address is set to 2451 fe80::2001:db8:2:0 (i.e., the AERO address for ('C2')) 2453 o The encapsulated AERO Predirect message includes all of the 2454 securing information that would occur in a MIPv6 "Care-of Test 2455 Init" message (format TBD) 2457 ('C2') will receive both Predirect messages through its home network 2458 then return a corresponding Redirect for each of the Predirect 2459 messages with the source and destination addresses in the inner and 2460 outer headers reversed. The first message includes all of the 2461 securing information that would occur in a MIPv6 "Home Test" message, 2462 while the second message includes all of the securing information 2463 that would occur in a MIPv6 "Care-of Test" message (formats TBD). 2465 When ('C1') receives the Redirect messages, it performs the necessary 2466 security procedures per the MIPv6 specification. It then prepares an 2467 encapsulated NS message that includes the same source and destination 2468 addresses as for the "Care-of Test Init" Predirect message, and 2469 includes all of the securing information that would occur in a MIPv6 2470 "Binding Update" message (format TBD) and sends the message to 2471 ('C2'). 2473 When ('C2') receives the NS message, if the securing information is 2474 correct it creates or updates a neighbor cache entry for ('C1') with 2475 fe80::2001:db8:1:0 as the network-layer address, 2001:db8:1000::1 as 2476 the link-layer address and with AcceptTime set to ACCEPT_TIME. 2477 ('C2') then sends an encapsulated NA message back to ('C1') that 2478 includes the same source and destination addresses as for the "Care- 2479 of Test" Redirect message, and includes all of the securing 2480 information that would occur in a MIPv6 "Binding Acknowledgement" 2481 message (format TBD) and sends the message to ('C1'). 2483 When ('C1') receives the NA message, it creates or updates a neighbor 2484 cache entry for ('C2') with fe80::2001:db8:2:0 as the network-layer 2485 address and 2001:db8:2:: as the link-layer address and with 2486 ForwardTime set to FORWARD_TIME, thus completing the route 2487 optimization in the forward direction. 2489 ('C1') subsequently forwards encapsulated packets with outer source 2490 address 2001:db8:1000::1, with outer destination address 2491 2001:db8:2::, with inner source address taken from the 2001:db8:1::, 2492 and with inner destination address taken from 2001:db8:2:: due to the 2493 fact that it has a securely-established neighbor cache entry with 2494 non-zero ForwardTime. ('C2') subsequently accepts any such 2495 encapsulated packets due to the fact that it has a securely- 2496 established neighbor cache entry with non-zero AcceptTime. 2498 In order to keep neighbor cache entries alive, ('C1') periodically 2499 sends additional NS messages to ('C2') and receives any NA responses. 2500 If ('C1') moves to a different point of attachment after the initial 2501 route optimization, it sends a new secured NS message to ('C2') as 2502 above to update ('C2')s neighbor cache. 2504 If ('C2') has packets to send to ('C1'), it performs a corresponding 2505 route optimization in the opposite direction following the same 2506 procedures described above. In the process, the already-established 2507 unidirectional neighbor cache entries within ('C1') and ('C2') are 2508 updated to include the now-bidirectional information. In particular, 2509 the AcceptTime and ForwardTime variables for both neighbor cache 2510 entries are updated to non-zero values, and the link-layer address 2511 for ('C1')s neighbor cache entry for ('C2') is reset to 2512 2001:db8:2000::1. 2514 Note that two AERO Clients can use full security protocol messaging 2515 instead of Return Routability, e.g., if strong authentication and/or 2516 confidentiality are desired. In that case, security protocol key 2517 exchanges such as specified for MOBIKE [RFC4555] would be used to 2518 establish security associations and neighbor cache entries between 2519 the AERO clients. Thereafter, AERO NS/NA messaging can be used to 2520 maintain neighbor cache entries, test reachability, and to announce 2521 mobility events. If reachability testing fails, e.g., if both 2522 Clients move at roughly the same time, the Clients can tear down the 2523 security association and neighbor cache entries and again allow 2524 packets to flow through their home network. 2526 4. Implementation Status 2528 An application-layer implementation is in progress. 2530 5. IANA Considerations 2532 IANA is instructed to assign a new 2-octet Hardware Type number 2533 "TBD1" for AERO in the "arp-parameters" registry per Section 2 of 2534 [RFC5494]. The number is assigned from the 2-octet Unassigned range 2535 with Hardware Type "AERO" and with this document as the reference. 2537 IANA is instructed to assign a 4-octet Enterprise Number "TBD2" for 2538 AERO in the "enterprise-numbers" registry per [RFC3315]. 2540 6. Security Considerations 2542 AERO link security considerations are the same as for standard IPv6 2543 Neighbor Discovery [RFC4861] except that AERO improves on some 2544 aspects. In particular, AERO uses a trust basis between Clients and 2545 Servers, where the Clients only engage in the AERO mechanism when it 2546 is facilitated by a trust anchor. Unless there is some other means 2547 of authenticating the Client's identity (e.g., link-layer security), 2548 AERO nodes SHOULD also use DHCPv6 securing services (e.g., DHCPv6 2549 authentication, Secure DHCPv6 [I-D.ietf-dhc-sedhcpv6], etc.) for 2550 Client authentication and network admission control. 2552 AERO Redirect, Predirect and unsolicited NA messages SHOULD include a 2553 Timestamp option (see Section 5.3 of [RFC3971]) that other AERO nodes 2554 can use to verify the message time of origin. AERO Predirect, NS and 2555 RS messages SHOULD include a Nonce option (see Section 5.3 of 2556 [RFC3971]) that recipients echo back in corresponding responses. 2558 AERO links must be protected against link-layer address spoofing 2559 attacks in which an attacker on the link pretends to be a trusted 2560 neighbor. Links that provide link-layer securing mechanisms (e.g., 2561 IEEE 802.1X WLANs) and links that provide physical security (e.g., 2562 enterprise network wired LANs) provide a first line of defense that 2563 is often sufficient. In other instances, additional securing 2564 mechanisms such as Secure Neighbor Discovery (SeND) [RFC3971], IPsec 2565 [RFC4301] or TLS [RFC5246] may be necessary. 2567 AERO Clients MUST ensure that their connectivity is not used by 2568 unauthorized nodes on their EUNs to gain access to a protected 2569 network, i.e., AERO Clients that act as routers MUST NOT provide 2570 routing services for unauthorized nodes. (This concern is no 2571 different than for ordinary hosts that receive an IP address 2572 delegation but then "share" the address with unauthorized nodes via a 2573 NAT function.) 2575 On some AERO links, establishment and maintenance of a direct path 2576 between neighbors requires secured coordination such as through the 2577 Internet Key Exchange (IKEv2) protocol [RFC5996] to establish a 2578 security association. 2580 7. Acknowledgements 2582 Discussions both on IETF lists and in private exchanges helped shape 2583 some of the concepts in this work. Individuals who contributed 2584 insights include Mikael Abrahamsson, Mark Andrews, Fred Baker, 2585 Stewart Bryant, Brian Carpenter, Wojciech Dec, Ralph Droms, Sri 2586 Gundavelli, Brian Haberman, Joel Halpern, Sascha Hlusiak, Lee Howard, 2587 Andre Kostur, Ted Lemon, Joe Touch and Bernie Volz. Members of the 2588 IESG also provided valuable input during their review process that 2589 greatly improved the document. Special thanks go to Stewart Bryant, 2590 Joel Halpern and Brian Haberman for their shepherding guidance. 2592 This work has further been encouraged and supported by Boeing 2593 colleagues including Keith Bartley, Dave Bernhardt, Cam Brodie, 2594 Balaguruna Chidambaram, Claudiu Danilov, Wen Fang, Anthony Gregory, 2595 Jeff Holland, Ed King, Gen MacLean, Kent Shuey, Brian Skeen, Mike 2596 Slane, Julie Wulff, Yueli Yang, and other members of the BR&T and BIT 2597 mobile networking teams. 2599 Earlier works on NBMA tunneling approaches are found in 2600 [RFC2529][RFC5214][RFC5569]. 2602 Many of the constructs presented in this second edition of AERO are 2603 based on the author's earlier works, including: 2605 o The Internet Routing Overlay Network (IRON) 2606 [RFC6179][I-D.templin-ironbis] 2608 o Virtual Enterprise Traversal (VET) 2609 [RFC5558][I-D.templin-intarea-vet] 2611 o The Subnetwork Encapsulation and Adaptation Layer (SEAL) 2612 [RFC5320][I-D.templin-intarea-seal] 2614 o AERO, First Edition [RFC6706] 2616 Finally, Satoru Matsushima and Ryuji Wakikawa also discuss BGP in 2617 their work. Behcet Sarikaya made comments on the dmm list. 2619 8. References 2621 8.1. Normative References 2623 [RFC0768] Postel, J., "User Datagram Protocol", STD 6, RFC 768, 2624 August 1980. 2626 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, September 2627 1981. 2629 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 2630 RFC 792, September 1981. 2632 [RFC2003] Perkins, C., "IP Encapsulation within IP", RFC 2003, 2633 October 1996. 2635 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 2636 Requirement Levels", BCP 14, RFC 2119, March 1997. 2638 [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 2639 (IPv6) Specification", RFC 2460, December 1998. 2641 [RFC2473] Conta, A. and S. Deering, "Generic Packet Tunneling in 2642 IPv6 Specification", RFC 2473, December 1998. 2644 [RFC3315] Droms, R., Bound, J., Volz, B., Lemon, T., Perkins, C., 2645 and M. Carney, "Dynamic Host Configuration Protocol for 2646 IPv6 (DHCPv6)", RFC 3315, July 2003. 2648 [RFC3633] Troan, O. and R. Droms, "IPv6 Prefix Options for Dynamic 2649 Host Configuration Protocol (DHCP) version 6", RFC 3633, 2650 December 2003. 2652 [RFC3971] Arkko, J., Kempf, J., Zill, B., and P. Nikander, "SEcure 2653 Neighbor Discovery (SEND)", RFC 3971, March 2005. 2655 [RFC4213] Nordmark, E. and R. Gilligan, "Basic Transition Mechanisms 2656 for IPv6 Hosts and Routers", RFC 4213, October 2005. 2658 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 2659 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 2660 September 2007. 2662 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 2663 Address Autoconfiguration", RFC 4862, September 2007. 2665 [RFC6434] Jankiewicz, E., Loughney, J., and T. Narten, "IPv6 Node 2666 Requirements", RFC 6434, December 2011. 2668 8.2. Informative References 2670 [I-D.ietf-dhc-sedhcpv6] 2671 Jiang, S., Shen, S., Zhang, D., and T. Jinmei, "Secure 2672 DHCPv6", draft-ietf-dhc-sedhcpv6-04 (work in progress), 2673 September 2014. 2675 [I-D.templin-intarea-seal] 2676 Templin, F., "The Subnetwork Encapsulation and Adaptation 2677 Layer (SEAL)", draft-templin-intarea-seal-68 (work in 2678 progress), January 2014. 2680 [I-D.templin-intarea-vet] 2681 Templin, F., "Virtual Enterprise Traversal (VET)", draft- 2682 templin-intarea-vet-40 (work in progress), May 2013. 2684 [I-D.templin-ironbis] 2685 Templin, F., "The Interior Routing Overlay Network 2686 (IRON)", draft-templin-ironbis-16 (work in progress), 2687 March 2014. 2689 [I-D.vandevelde-idr-remote-next-hop] 2690 Velde, G., Patel, K., Rao, D., Raszuk, R., and R. Bush, 2691 "BGP Remote-Next-Hop", draft-vandevelde-idr-remote-next- 2692 hop-08 (work in progress), October 2014. 2694 [RFC0879] Postel, J., "TCP maximum segment size and related topics", 2695 RFC 879, November 1983. 2697 [RFC1035] Mockapetris, P., "Domain names - implementation and 2698 specification", STD 13, RFC 1035, November 1987. 2700 [RFC1191] Mogul, J. and S. Deering, "Path MTU discovery", RFC 1191, 2701 November 1990. 2703 [RFC1812] Baker, F., "Requirements for IP Version 4 Routers", RFC 2704 1812, June 1995. 2706 [RFC1930] Hawkinson, J. and T. Bates, "Guidelines for creation, 2707 selection, and registration of an Autonomous System (AS)", 2708 BCP 6, RFC 1930, March 1996. 2710 [RFC1981] McCann, J., Deering, S., and J. Mogul, "Path MTU Discovery 2711 for IP version 6", RFC 1981, August 1996. 2713 [RFC2131] Droms, R., "Dynamic Host Configuration Protocol", RFC 2714 2131, March 1997. 2716 [RFC2529] Carpenter, B. and C. Jung, "Transmission of IPv6 over IPv4 2717 Domains without Explicit Tunnels", RFC 2529, March 1999. 2719 [RFC2675] Borman, D., Deering, S., and R. Hinden, "IPv6 Jumbograms", 2720 RFC 2675, August 1999. 2722 [RFC2784] Farinacci, D., Li, T., Hanks, S., Meyer, D., and P. 2723 Traina, "Generic Routing Encapsulation (GRE)", RFC 2784, 2724 March 2000. 2726 [RFC2923] Lahey, K., "TCP Problems with Path MTU Discovery", RFC 2727 2923, September 2000. 2729 [RFC3596] Thomson, S., Huitema, C., Ksinant, V., and M. Souissi, 2730 "DNS Extensions to Support IP Version 6", RFC 3596, 2731 October 2003. 2733 [RFC3819] Karn, P., Bormann, C., Fairhurst, G., Grossman, D., 2734 Ludwig, R., Mahdavi, J., Montenegro, G., Touch, J., and L. 2735 Wood, "Advice for Internet Subnetwork Designers", BCP 89, 2736 RFC 3819, July 2004. 2738 [RFC4271] Rekhter, Y., Li, T., and S. Hares, "A Border Gateway 2739 Protocol 4 (BGP-4)", RFC 4271, January 2006. 2741 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 2742 Architecture", RFC 4291, February 2006. 2744 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 2745 Internet Protocol", RFC 4301, December 2005. 2747 [RFC4443] Conta, A., Deering, S., and M. Gupta, "Internet Control 2748 Message Protocol (ICMPv6) for the Internet Protocol 2749 Version 6 (IPv6) Specification", RFC 4443, March 2006. 2751 [RFC4555] Eronen, P., "IKEv2 Mobility and Multihoming Protocol 2752 (MOBIKE)", RFC 4555, June 2006. 2754 [RFC4592] Lewis, E., "The Role of Wildcards in the Domain Name 2755 System", RFC 4592, July 2006. 2757 [RFC4821] Mathis, M. and J. Heffner, "Packetization Layer Path MTU 2758 Discovery", RFC 4821, March 2007. 2760 [RFC4963] Heffner, J., Mathis, M., and B. Chandler, "IPv4 Reassembly 2761 Errors at High Data Rates", RFC 4963, July 2007. 2763 [RFC4994] Zeng, S., Volz, B., Kinnear, K., and J. Brzozowski, 2764 "DHCPv6 Relay Agent Echo Request Option", RFC 4994, 2765 September 2007. 2767 [RFC5213] Gundavelli, S., Leung, K., Devarapalli, V., Chowdhury, K., 2768 and B. Patil, "Proxy Mobile IPv6", RFC 5213, August 2008. 2770 [RFC5214] Templin, F., Gleeson, T., and D. Thaler, "Intra-Site 2771 Automatic Tunnel Addressing Protocol (ISATAP)", RFC 5214, 2772 March 2008. 2774 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 2775 (TLS) Protocol Version 1.2", RFC 5246, August 2008. 2777 [RFC5320] Templin, F., "The Subnetwork Encapsulation and Adaptation 2778 Layer (SEAL)", RFC 5320, February 2010. 2780 [RFC5494] Arkko, J. and C. Pignataro, "IANA Allocation Guidelines 2781 for the Address Resolution Protocol (ARP)", RFC 5494, 2782 April 2009. 2784 [RFC5522] Eddy, W., Ivancic, W., and T. Davis, "Network Mobility 2785 Route Optimization Requirements for Operational Use in 2786 Aeronautics and Space Exploration Mobile Networks", RFC 2787 5522, October 2009. 2789 [RFC5558] Templin, F., "Virtual Enterprise Traversal (VET)", RFC 2790 5558, February 2010. 2792 [RFC5569] Despres, R., "IPv6 Rapid Deployment on IPv4 2793 Infrastructures (6rd)", RFC 5569, January 2010. 2795 [RFC5720] Templin, F., "Routing and Addressing in Networks with 2796 Global Enterprise Recursion (RANGER)", RFC 5720, February 2797 2010. 2799 [RFC5844] Wakikawa, R. and S. Gundavelli, "IPv4 Support for Proxy 2800 Mobile IPv6", RFC 5844, May 2010. 2802 [RFC5949] Yokota, H., Chowdhury, K., Koodli, R., Patil, B., and F. 2803 Xia, "Fast Handovers for Proxy Mobile IPv6", RFC 5949, 2804 September 2010. 2806 [RFC5996] Kaufman, C., Hoffman, P., Nir, Y., and P. Eronen, 2807 "Internet Key Exchange Protocol Version 2 (IKEv2)", RFC 2808 5996, September 2010. 2810 [RFC6146] Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful 2811 NAT64: Network Address and Protocol Translation from IPv6 2812 Clients to IPv4 Servers", RFC 6146, April 2011. 2814 [RFC6179] Templin, F., "The Internet Routing Overlay Network 2815 (IRON)", RFC 6179, March 2011. 2817 [RFC6204] Singh, H., Beebee, W., Donley, C., Stark, B., and O. 2818 Troan, "Basic Requirements for IPv6 Customer Edge 2819 Routers", RFC 6204, April 2011. 2821 [RFC6275] Perkins, C., Johnson, D., and J. Arkko, "Mobility Support 2822 in IPv6", RFC 6275, July 2011. 2824 [RFC6355] Narten, T. and J. Johnson, "Definition of the UUID-Based 2825 DHCPv6 Unique Identifier (DUID-UUID)", RFC 6355, August 2826 2011. 2828 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 2829 for Equal Cost Multipath Routing and Link Aggregation in 2830 Tunnels", RFC 6438, November 2011. 2832 [RFC6691] Borman, D., "TCP Options and Maximum Segment Size (MSS)", 2833 RFC 6691, July 2012. 2835 [RFC6706] Templin, F., "Asymmetric Extended Route Optimization 2836 (AERO)", RFC 6706, August 2012. 2838 [RFC6864] Touch, J., "Updated Specification of the IPv4 ID Field", 2839 RFC 6864, February 2013. 2841 [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and 2842 UDP Checksums for Tunneled Packets", RFC 6935, April 2013. 2844 [RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement 2845 for the Use of IPv6 UDP Datagrams with Zero Checksums", 2846 RFC 6936, April 2013. 2848 [RFC6939] Halwasia, G., Bhandari, S., and W. Dec, "Client Link-Layer 2849 Address Option in DHCPv6", RFC 6939, May 2013. 2851 [RFC6980] Gont, F., "Security Implications of IPv6 Fragmentation 2852 with IPv6 Neighbor Discovery", RFC 6980, August 2013. 2854 [RFC7078] Matsumoto, A., Fujisaki, T., and T. Chown, "Distributing 2855 Address Selection Policy Using DHCPv6", RFC 7078, January 2856 2014. 2858 [TUNTAP] Wikipedia, W., "http://en.wikipedia.org/wiki/TUN/TAP", 2859 October 2014. 2861 Author's Address 2863 Fred L. Templin (editor) 2864 Boeing Research & Technology 2865 P.O. Box 3707 2866 Seattle, WA 98124 2867 USA 2869 Email: fltemplin@acm.org