idnits 2.17.1 draft-templin-aerolink-28.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 (June 27, 2014) is 3584 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) == Unused Reference: 'RFC0768' is defined on line 1421, but no explicit reference was found in the text == Unused Reference: 'RFC0792' is defined on line 1427, but no explicit reference was found in the text == Unused Reference: 'RFC2460' is defined on line 1436, but no explicit reference was found in the text == Unused Reference: 'RFC4862' is defined on line 1460, but no explicit reference was found in the text == Unused Reference: 'RFC6434' is defined on line 1463, but no explicit reference was found in the text == Unused Reference: 'RFC0879' is defined on line 1471, but no explicit reference was found in the text == Unused Reference: 'RFC6146' is defined on line 1526, but no explicit reference was found in the text == Unused Reference: 'RFC6204' is defined on line 1530, but no explicit reference was found in the text == Unused Reference: 'RFC6691' is defined on line 1542, but no explicit reference was found in the text == Unused Reference: 'RFC6706' is defined on line 1545, but no explicit reference was found in the text == Unused Reference: 'RFC6980' is defined on line 1561, but no explicit reference was found in the text == Unused Reference: 'RFC7078' is defined on line 1564, 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) -- Obsolete informational reference (is this intentional?): RFC 879 (Obsoleted by RFC 7805, RFC 9293) -- 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 (~~), 14 warnings (==), 6 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: rfc6706 (if approved) June 27, 2014 5 Intended status: Standards Track 6 Expires: December 29, 2014 8 Transmission of IP Packets over AERO Links 9 draft-templin-aerolink-28.txt 11 Abstract 13 This document specifies the operation of IP over tunnel virtual Non- 14 Broadcast, Multiple Access (NBMA) links using Asymmetric Extended 15 Route Optimization (AERO). Nodes attached to AERO links can exchange 16 packets via trusted intermediate routers that provide forwarding 17 services to reach off-link destinations and redirection services for 18 route optimization. AERO provides an IPv6 link-local address format 19 known as the AERO address that supports operation of the IPv6 20 Neighbor Discovery (ND) protocol and links IPv6 ND to IP routing. 21 Admission control and provisioning are supported by the Dynamic Host 22 Configuration Protocol for IPv6 (DHCPv6), and node mobility is 23 naturally supported through dynamic neighbor cache updates. Although 24 IPv6 ND messaging is used in the control plane, both IPv4 and IPv6 25 are supported in the data plane. 27 Status of This Memo 29 This Internet-Draft is submitted in full conformance with the 30 provisions of BCP 78 and BCP 79. 32 Internet-Drafts are working documents of the Internet Engineering 33 Task Force (IETF). Note that other groups may also distribute 34 working documents as Internet-Drafts. The list of current Internet- 35 Drafts is at http://datatracker.ietf.org/drafts/current/. 37 Internet-Drafts are draft documents valid for a maximum of six months 38 and may be updated, replaced, or obsoleted by other documents at any 39 time. It is inappropriate to use Internet-Drafts as reference 40 material or to cite them other than as "work in progress." 42 This Internet-Draft will expire on December 29, 2014. 44 Copyright Notice 46 Copyright (c) 2014 IETF Trust and the persons identified as the 47 document authors. All rights reserved. 49 This document is subject to BCP 78 and the IETF Trust's Legal 50 Provisions Relating to IETF Documents 51 (http://trustee.ietf.org/license-info) in effect on the date of 52 publication of this document. Please review these documents 53 carefully, as they describe your rights and restrictions with respect 54 to this document. Code Components extracted from this document must 55 include Simplified BSD License text as described in Section 4.e of 56 the Trust Legal Provisions and are provided without warranty as 57 described in the Simplified BSD License. 59 Table of Contents 61 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 62 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3 63 3. Asymmetric Extended Route Optimization (AERO) . . . . . . . . 5 64 3.1. AERO Node Types . . . . . . . . . . . . . . . . . . . . . 5 65 3.2. AERO Addresses . . . . . . . . . . . . . . . . . . . . . 6 66 3.3. AERO Interface Characteristics . . . . . . . . . . . . . 7 67 3.3.1. Coordination of Multiple Underlying Interfaces . . . 9 68 3.4. AERO Interface Neighbor Cache Maintenace . . . . . . . . 10 69 3.5. AERO Interface Data Origin Authentication . . . . . . . . 11 70 3.6. AERO Interface MTU Considerations . . . . . . . . . . . . 11 71 3.7. AERO Interface Encapsulation, Re-encapsulation and 72 Decapsulation . . . . . . . . . . . . . . . . . . . . . . 13 73 3.8. AERO Router Discovery, Prefix Delegation and Address 74 Configuration . . . . . . . . . . . . . . . . . . . . . . 14 75 3.8.1. AERO Client Behavior . . . . . . . . . . . . . . . . 14 76 3.8.2. AERO Server Behavior . . . . . . . . . . . . . . . . 16 77 3.9. AERO Redirection . . . . . . . . . . . . . . . . . . . . 17 78 3.9.1. Reference Operational Scenario . . . . . . . . . . . 17 79 3.9.2. Classical Redirection Approaches . . . . . . . . . . 19 80 3.9.3. Concept of Operations . . . . . . . . . . . . . . . . 20 81 3.9.4. Message Format . . . . . . . . . . . . . . . . . . . 20 82 3.9.5. Sending Predirects . . . . . . . . . . . . . . . . . 21 83 3.9.6. Processing Predirects and Sending Redirects . . . . . 23 84 3.9.7. Re-encapsulating and Relaying Redirects . . . . . . . 24 85 3.9.8. Processing Redirects . . . . . . . . . . . . . . . . 25 86 3.9.9. Server-Oriented Redirection . . . . . . . . . . . . . 25 87 3.10. Neighbor Reachability Maintenance . . . . . . . . . . . . 25 88 3.11. Mobility Management . . . . . . . . . . . . . . . . . . . 26 89 3.12. Encapsulation Protocol Version Considerations . . . . . . 28 90 3.13. Multicast Considerations . . . . . . . . . . . . . . . . 28 91 3.14. Operation on AERO Links Without DHCPv6 Services . . . . . 29 92 3.15. Operation on Server-less AERO Links . . . . . . . . . . . 29 93 4. Implementation Status . . . . . . . . . . . . . . . . . . . . 29 94 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 29 95 6. Security Considerations . . . . . . . . . . . . . . . . . . . 30 96 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 30 97 8. References . . . . . . . . . . . . . . . . . . . . . . . . . 31 98 8.1. Normative References . . . . . . . . . . . . . . . . . . 31 99 8.2. Informative References . . . . . . . . . . . . . . . . . 32 100 Appendix A. AERO Server and Relay Interworking . . . . . . . . . 34 101 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 36 103 1. Introduction 105 This document specifies the operation of IP over tunnel virtual Non- 106 Broadcast, Multiple Access (NBMA) links using Asymmetric Extended 107 Route Optimization (AERO). The AERO link can be used for tunneling 108 to neighboring nodes on either IPv6 or IPv4 networks, i.e., AERO 109 views the IPv6 and IPv4 networks as equivalent links for tunneling. 110 Nodes attached to AERO links can exchange packets via trusted 111 intermediate routers that provide forwarding services to reach off- 112 link destinations and redirection services for route optimization 113 that addresses the requirements outlined in [RFC5522]. 115 AERO provides an IPv6 link-local address format known as the AERO 116 address that supports operation of the IPv6 Neighbor Discovery (ND) 117 [RFC4861] protocol and links IPv6 ND to IP routing. Admission 118 control and provisioning are supported by the Dynamic Host 119 Configuration Protocol for IPv6 (DHCPv6) [RFC3315], and node mobility 120 is naturally supported through dynamic neighbor cache updates. 121 Although IPv6 ND message signalling is used in the control plane, 122 both IPv4 and IPv6 are supported in the data plane. The remainder of 123 this document presents the AERO specification. 125 2. Terminology 127 The terminology in the normative references applies; the following 128 terms are defined within the scope of this document: 130 AERO link 131 a Non-Broadcast, Multiple Access (NBMA) tunnel virtual overlay 132 configured over a node's attached IPv6 and/or IPv4 networks. All 133 nodes on the AERO link appear as single-hop neighbors from the 134 perspective of the overlay IP layer. 136 AERO interface 137 a node's attachment to an AERO link. 139 AERO address 140 an IPv6 link-local address constructed as specified in Section 3.2 141 and assigned to a Client's AERO interface. 143 AERO node 144 a node that is connected to an AERO link and that participates in 145 IPv6 ND over the link. 147 AERO Client ("Client") 148 a node that assigns an AERO address on an AERO interface and 149 receives an IP prefix delegation. 151 AERO Server ("Server") 152 a node that assigns the IPv6 link-local subnet router anycast 153 address (fe80::) and an administratively provisioned IPv6 link- 154 local unicast address on an AERO interface over which it can 155 provide default forwarding and redirection services for AERO 156 Clients. 158 AERO Relay ("Relay") 159 a node that relays IP packets between Servers on the same AERO 160 link, and/or that forwards IP packets between the AERO link and 161 the native Internetwork. An AERO Relay may or may not also be 162 configured as an AERO Server. 164 ingress tunnel endpoint (ITE) 165 an AERO interface endpoint that injects tunneled packets into an 166 AERO link. 168 egress tunnel endpoint (ETE) 169 an AERO interface endpoint that receives tunneled packets from an 170 AERO link. 172 underlying network 173 a connected IPv6 or IPv4 network routing region over which AERO 174 nodes tunnel IP packets. 176 underlying interface 177 an AERO node's interface point of attachment to an underlying 178 network. 180 link-layer address 181 an IP address assigned to an AERO node's underlying interface. 182 When UDP encapsulation is used, the UDP port number is also 183 considered as part of the link-layer address. Link-layer 184 addresses are used as the encapsulation header source and 185 destination addresses. 187 network layer address 188 the source or destination address of the encapsulated IP packet. 190 end user network (EUN) 191 an internal virtual or external edge IP network that an AERO 192 Client connects to the AERO interface. 194 end user network prefix 195 an IP prefix delegated to an end user network. 197 aggregated prefix 198 an IP prefix assigned to the AERO link and from which end user 199 network prefixes are derived. (For example, and end user network 200 prefix 2001:db8:1:2::/64 is derived from the aggregated prefix 201 2001:db8::/32.) 203 Throughout the document, the simple terms "Client", "Server" and 204 "Relay" refer to "AERO Client", "AERO Server" and "AERO Relay", 205 respectively. Capitalization is used to distinguish these terms from 206 DHCPv6 client/server/relay. This is an important distinction, since 207 an AERO Server may be a DHCPv6 relay, and an AERO Relay may be a 208 DHCPv6 server. 210 The terminology of [RFC4861] (including the names of node variables 211 and protocol constants) applies to this document. Also throughout 212 the document, the term "IP" is used to generically refer to either 213 Internet Protocol version (i.e., IPv4 or IPv6). 215 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 216 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 217 document are to be interpreted as described in [RFC2119]. 219 3. Asymmetric Extended Route Optimization (AERO) 221 The following sections specify the operation of IP over Asymmetric 222 Extended Route Optimization (AERO) links: 224 3.1. AERO Node Types 226 AERO Relays relay packets between nodes connected to the same AERO 227 link and also forward packets between the AERO link and the native 228 Internetwork. The relaying process entails re-encapsulation of IP 229 packets that were received from a first AERO node and are to be 230 forwarded without modification to a second AERO node. 232 AERO Servers provide default routing services to AERO Clients. AERO 233 Servers configure a DHCPv6 relay or server function and facilitate 234 DHCPv6 Prefix Delegation (PD) exchanges. An AERO Server may also act 235 as an AERO Relay. 237 AERO Clients act as requesting routers to receive IP prefixes through 238 a DHCPv6 PD exchange via AERO Servers over the AERO link. (Each 239 client MAY associate with multiple Servers, but associating with many 240 Servers may result in excessive control message overhead.) Each IPv6 241 AERO Client receives at least a /64 IPv6 prefix delegation, and may 242 receive even shorter prefixes. Similarly, each IPv4 AERO Client 243 receives at least a /32 IPv4 prefix delegation (i.e., a singleton 244 IPv4 address), and may receive even shorter prefixes. 246 AERO Clients that act as routers sub-delegate portions of their 247 received prefix delegations to links on EUNs. End system 248 applications on AERO Clients that act as routers bind to EUN 249 interfaces (i.e., and not the AERO interface). 251 AERO Clients that act as ordinary hosts assign one or more IP 252 addresses taken from their received prefix delegations to the AERO 253 interface but DO NOT assign the delegated prefix itself to the AERO 254 interface. Instead, the Client assigns the delegated prefix to a 255 "black hole" route so that unused portions of the prefix are 256 nullified. End system applications on AERO Clients that act as hosts 257 bind directly to the AERO interface. 259 3.2. AERO Addresses 261 An AERO address is an IPv6 link-local address with an embedded IP 262 prefix and assigned to a Client's AERO interface. The AERO address 263 is formatted as follows: 265 fe80::[IP prefix] 267 For IPv6, the AERO address begins with the prefix fe80::/64 and 268 includes in its interface identifier the base prefix taken from the 269 Client's delegated IPv6 prefix. The base prefix is determined by 270 masking the delegated prefix with the prefix length. For example, if 271 the AERO Client receives the IPv6 prefix delegation: 273 2001:db8:1000:2000::/56 275 it constructs its AERO address as: 277 fe80::2001:db8:1000:2000 279 For IPv4, the AERO address begins with the prefix fe80::/96 and 280 includes in its interface identifier the base prefix taken from the 281 Client's delegated IPv4 prefix. For example, if the AERO Client 282 receives the IPv4 prefix delegation: 284 192.0.2.32/28 286 it constructs its AERO address as: 288 fe80::192.0.2.32 290 The AERO address remains stable as the Client moves between 291 topological locations, i.e., even if its link-layer addresses change. 293 NOTE: In some cases, prospective neighbors may not have a priori 294 knowledge of the Client's delegated prefix length and may therefore 295 send initial IPv6 ND messages with an AERO destination address that 296 matches the delegated prefix but does not correspond to the base 297 prefix. In that case, the Client MUST accept the address as 298 equivalent to the base address, but then use the base address as the 299 source address of any IPv6 ND message replies. For example, if the 300 Client receives the IPv6 prefix delegation 2001:db8:1000:2000::/56 301 then subsequently receives an IPv6 ND message with destination 302 address fe80::2001:db8:1000:2001, it accepts the message but uses 303 fe80::2001:db8:1000:2000 as the source address of any IPv6 ND 304 replies. 306 3.3. AERO Interface Characteristics 308 AERO interfaces use IP-in-IPv6 encapsulation [RFC2473] to exchange 309 tunneled packets with AERO neighbors attached to an underlying IPv6 310 network, and use IP-in-IPv4 encapsulation [RFC2003][RFC4213] to 311 exchange tunneled packets with AERO neighbors attached to an 312 underlying IPv4 network. AERO interfaces can also operate over 313 secured tunnel types such as IPsec [RFC4301] or TLS [RFC5246]. When 314 Network Address Translator (NAT) traversal and/or filtering middlebox 315 traversal may be necessary, a UDP header is further inserted 316 immediately above the IP encapsulation header. 318 Servers assign the address fe80:: to their AERO interfaces as a link- 319 local Subnet Router Anycast address. Servers and Relays also assign 320 a link-local address fe80::ID to support the operation of the IPv6 ND 321 protocol and the inter-Server/Relay routing system (see: Appendix A). 322 Each fe80::ID address MUST be unique among all Servers and Relays on 323 the AERO link, and MUST NOT collide with any potential AERO addresses 324 (e.g., the addresses for Servers and Relays on the link could be 325 assigned as fe80::1, fe80::2, fe80::3, etc.). Servers accept IPV6 ND 326 messages with either fe80::ID or fe80:: as the IPv6 destination 327 address, but MUST use the fe80::ID address as the IPv6 source address 328 of any IPv6 ND messages they generate. 330 When a Client does not know the fe80::ID address of a Server, it can 331 use fe80:: as a temporary destination address in IPv6 ND messages. 332 The Client may also use fe80::, e.g., as the link-local address in a 333 neighbor cache entry for a Server when the Server's fe80::ID address 334 is not known in advance. 336 When a Client enables an AERO interface, it invokes DHCPv6 PD using 337 the temporary IPv6 link-local source address 338 fe80::ffff:ffff:ffff:ffff. After the Client receives a prefix 339 delegation, it assigns the corresponding AERO address to the AERO 340 interface and deprecates the temporary address, i.e., the Client 341 invokes DHCPv6 to bootstrap the provisioning of a unique link-local 342 address before invoking IPv6 ND. 344 AERO interfaces maintain a neighbor cache and use an adaptation of 345 standard unicast IPv6 ND messaging. AERO interfaces use unicast 346 Neighbor Solicitation (NS), Neighbor Advertisement (NA), Router 347 Solicitation (RS) and Router Advertisement (RA) messages the same as 348 for any IPv6 link. AERO interfaces use two redirection message types 349 -- the first known as a Predirect message and the second being the 350 standard Redirect message (see Section 3.9). AERO links further use 351 link-local-only addressing; hence, Clients ignore any Prefix 352 Information Options (PIOs) they may receive in RA messages. 354 AERO interface Redirect/Predirect messages include Target Link-Layer 355 Address Options (TLLAOs) formatted as shown in Figure 1: 357 0 1 2 3 358 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 1 359 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 360 | Type = 2 | Length = 3 | Reserved | 361 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 362 | Link ID | Preference | UDP Port Number (or 0) | 363 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 364 | | 365 +-- --+ 366 | | 367 +-- IP Address --+ 368 | | 369 +-- --+ 370 | | 371 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 373 Figure 1: AERO Target Link-Layer Address Option (TLLAO) Format 375 In this format, Link ID is an integer value between 0 and 255 376 corresponding to an underlying interface of the target node, and 377 Preference is an integer value between 0 and 255 indicating the 378 node's preference for this underlying interface, with 0 being highest 379 preference and 255 being lowest. UDP Port Number and IP Address are 380 set to the addresses used by the target node when it sends 381 encapsulated packets over the underlying interface. When no UDP 382 encapsulation is used, UDP Port Number is set to 0. When the 383 encapsulation IP address family is IPv4, IP Address is formed as an 384 IPv4-compatible IPv6 address [RFC4291], i.e., 96 bits of leading 0's 385 followed by a 32-bit IPv4 address 387 AERO interface Redirect/Predirect messages can both update and create 388 neighbor cache entries, including link-layer address information. 389 Redirect/Predirect messages SHOULD include a Timestamp option (see 390 Section 5.3 of [RFC3971]) that other AERO nodes can use to verify the 391 message time of origin. 393 AERO interface NS/NA/RS/RA messages used for neighbor reachability 394 verification update timers in existing neighbor cache entires but do 395 not update link-layer addresses nor create new neighbor cache 396 entries. AERO interface unsolicited NA messages are used to update a 397 neighbor's cached link-layer address for the sender, e.g., following 398 a link-layer address change due to node mobility. AERO interface NS/ 399 RS messages SHOULD include a Nonce option (see Section 5.3 of 400 [RFC3971]) that the recipient echoes back in the corresponding NA/RA 401 response. 403 3.3.1. Coordination of Multiple Underlying Interfaces 405 AERO interfaces may be configured over multiple underlying 406 interfaces. For example, common handheld devices have both wireless 407 local area network ("WLAN") and cellular wireless links. These links 408 are typically used "one at a time" with low-cost WLAN preferred and 409 highly-available cellular wireless as a standby. In a more complex 410 example, aircraft frequently have many wireless data link types (e.g. 411 satellite-based, terrestrial, air-to-air directional, etc.) with 412 diverse performance and cost properties. 414 If a Client's multiple underlying interfaces are used "one at a time" 415 (i.e., all other interfaces are in standby mode while one interface 416 is active), then Predirect/Redirect messages include only a single 417 TLLAO with Link ID set to 0. 419 If the Client has multiple active underlying interfaces, then from 420 the perspective of IPv6 ND it would appear to have a single link- 421 local address with multiple link-layer addresses. In that case, 422 Predirect/Redirect messages MAY include multiple TLLAOs -- each with 423 a different Link ID that corresponds to an underlying interface of 424 the Client. Further details on coordination of multiple active 425 underlying interfaces are outside the scope of this specification. 427 3.4. AERO Interface Neighbor Cache Maintenace 429 Each AERO interface maintains a conceptual neighbor cache that 430 includes an entry for each neighbor it communicates with on the AERO 431 link, the same as for any IPv6 interface [RFC4861]. Neighbor cache 432 entries are created and maintained as follows: 434 When an AERO Server relays a DHCPv6 Reply message to an AERO Client, 435 it creates or updates a neighbor cache entry for the Client based on 436 the AERO address corresponding to the Client's prefix delegation as 437 the network-layer address and with the Client's encapsulation IP 438 address and UDP port number as the link-layer address. 440 When an AERO Client receives a DHCPv6 Reply message from an AERO 441 Server, it creates or updates a neighbor cache entry for the Server 442 based on the Reply message link-local source address as the network- 443 layer address, and the encapsulation IP source address and UDP source 444 port number as the link-layer address. 446 When an AERO Client receives a valid Predirect message it creates or 447 updates a neighbor cache entry for the Predirect target network-layer 448 and link-layer addresses, and also creates an IP forwarding table 449 entry for the predirected (source) prefix. The node then sets an 450 "AcceptTime" variable for the neighbor and uses this value to 451 determine whether messages received from the predirected neighbor can 452 be accepted. 454 When an AERO Client receives a valid Redirect message it creates or 455 updates a neighbor cache entry for the Redirect target network-layer 456 and link-layer addresses, and also creates an IP forwarding table 457 entry for the redirected (destination) prefix. The node then sets a 458 "ForwardTime" variable for the neighbor and uses this value to 459 determine whether packets can be sent directly to the redirected 460 neighbor. The node also maintains a constant value MAX_RETRY to 461 limit the number of keepalives sent when a neighbor may have gone 462 unreachable. 464 When an AERO Client receives a valid NS message it (re)sets 465 AcceptTime for the neighbor to ACCEPT_TIME. 467 When an AERO Client receives a valid solicited NA message, it 468 (re)sets ForwardTime for the neighbor to FORWARD_TIME. (When an AERO 469 Client receives a valid unsolicited NA message, it updates the 470 neighbor's link-layer address but DOES NOT reset ForwardTime.) 472 It is RECOMMENDED that FORWARD_TIME be set to the default constant 473 value 30 seconds to match the default REACHABLE_TIME value specified 474 for IPv6 ND [RFC4861]. 476 It is RECOMMENDED that ACCEPT_TIME be set to the default constant 477 value 40 seconds to allow a 10 second window so that the AERO 478 redirection procedure can converge before AcceptTime decrements below 479 FORWARD_TIME. 481 It is RECOMMENDED that MAX_RETRY be set to 3 the same as described 482 for IPv6 ND address resolution in Section 7.3.3 of [RFC4861]. 484 Different values for FORWARD_TIME, ACCEPT_TIME, and MAX_RETRY MAY be 485 administratively set, if necessary, to better match the AERO link's 486 performance characteristics; however, if different values are chosen, 487 all nodes on the link MUST consistently configure the same values. 488 In particular, ACCEPT_TIME SHOULD be set to a value that is 489 sufficiently longer than FORWARD_TIME to allow the AERO redirection 490 procedure to converge. 492 3.5. AERO Interface Data Origin Authentication 494 AERO nodes use a simple data origin authentication for encapsulated 495 packets they receive from other nodes. In particular, AERO nodes 496 accept encapsulated packets with a link-layer source address 497 belonging to one of their current AERO Servers and accept 498 encapsulated packets with a link-layer source address that is correct 499 for the network-layer source address. 501 The AERO node considers the link-layer source address correct for the 502 network-layer source address if there is an IP forwarding table entry 503 that matches the network-layer source address as well as a neighbor 504 cache entry corresponding to the next hop that includes the link- 505 layer address and AcceptTime is non-zero. 507 Note that this simple data origin authentication only applies to 508 environments in which link-layer addresses cannot be spoofed. 509 Additional security mitigations may be necessary in other 510 environments. 512 3.6. AERO Interface MTU Considerations 514 The AERO link Maximum Transmission Unit (MTU) is 64KB minus the 515 encapsulation overhead for IPv4 as the link-layer [RFC0791] and 4GB 516 minus the encapsulation overhead for IPv6 as the link layer 517 [RFC2675]. This is the most that IPv4 and IPv6 (respectively) can 518 convey within the constraints of protocol constants, but actual sizes 519 available for tunneling will frequently be much smaller. 521 The base tunneling specifications for IPv4 and IPv6 typically set a 522 static MTU on the tunnel interface to 1500 bytes minus the 523 encapsulation overhead or smaller still if the tunnel is likely to 524 incur additional encapsulations on the path. This can result in path 525 MTU related black holes when packets that are too large to be 526 accommodated over the AERO link are dropped, but the resulting ICMP 527 Packet Too Big (PTB) messages are lost on the return path. As a 528 result, AERO nodes use the following MTU mitigations to accommodate 529 larger packets. 531 AERO nodes set their AERO interface MTU to the larger of the 532 underlying interface MTU minus the encapsulation overhead, and 1500 533 bytes. (If there are multiple underlying interfaces, the node sets 534 the AERO interface MTU according to the largest underlying interface 535 MTU, or 64KB /4G minus the encapsulation overhead if the largest MTU 536 cannot be determined.) AERO nodes optionally cache other per- 537 neighbor MTU values in the underlying IP path MTU discovery cache 538 initialized to the underlying interface MTU. 540 AERO nodes admit packets that are no larger than 1280 bytes minus the 541 encapsulation overhead (*) as well as packets that are larger than 542 1500 bytes into the tunnel without fragmentation, i.e., as long as 543 they are no larger than the AERO interface MTU before encapsulation 544 and also no larger than the cached per-neighbor MTU following 545 encapsulation. For IPv4, the node sets the "Don't Fragment" (DF) bit 546 to 0 for packets no larger than 1280 bytes minus the encapsulation 547 overhead (*) and sets the DF bit to 1 for packets larger than 1500 548 bytes. If a large packet is lost in the path, the node may 549 optionally cache the MTU reported in the resulting PTB message or may 550 ignore the message, e.g., if there is a possibility that the message 551 is spurious. 553 For packets destined to an AERO node that are larger than 1280 bytes 554 minus the encapsulation overhead (*) but no larger than 1500 bytes, 555 the node uses IP fragmentation to fragment the encapsulated packet 556 into two pieces (where the first fragment contains 1024 bytes of the 557 original IP packet) then admits the fragments into the tunnel. If 558 the link-layer protocol is IPv4, the node admits each fragment into 559 the tunnel with DF set to 0 and subject to rate limiting to avoid 560 reassembly errors [RFC4963][RFC6864]. For both IPv4 and IPv6, the 561 node also sends a 1500 byte probe message (**) to the neighbor, 562 subject to rate limiting. 564 To construct a probe, the node prepares an NS message with a Nonce 565 option plus trailing padding octets added to a length of 1500 bytes 566 without including the length of the padding in the IPv6 Payload 567 Length field. The node then encapsulates the NS in the encapsulation 568 headers (while including the length of the padding in the 569 encapsulation header length fields), sets DF to 1 (for IPv4) and 570 sends the padded NS message to the neighbor. If the neighbor returns 571 an NA message with a correct Nonce value, the node may then send 572 whole packets within this size range and (for IPv4) relax the rate 573 limiting requirement. (Note that the trailing padding SHOULD NOT be 574 included within the Nonce option itself but rather as padding beyond 575 the last option in the NS message; otherwise, the (large) Nonce 576 option would be echoed back in the solicited NA message and may be 577 lost at a link with a small MTU along the reverse path.) 579 AERO nodes MUST be capable of reassembling packets up to 1500 bytes 580 plus the encapsulation overhead length. It is therefore RECOMMENDED 581 that AERO nodes be capable of reassembling at least 2KB. 583 (*) Note that if it is known without probing that the minimum Path 584 MTU to an AERO node is MINMTU bytes (where 1280 < MINMTU < 1500) then 585 MINMTU can be used instead of 1280 in the fragmentation threshold 586 considerations listed above. 588 (**) It is RECOMMENDED that no probes smaller than 1500 bytes be used 589 for MTU probing purposes, since smaller probes may be fragmented if 590 there is a nested tunnel somewhere on the path to the neighbor. 591 Probe sizes larger than 1500 bytes MAY be used, but may be 592 unnecessary since original sources are expected to implement 593 [RFC4821] when sending large packets. 595 3.7. AERO Interface Encapsulation, Re-encapsulation and Decapsulation 597 AERO interfaces encapsulate IP packets according to whether they are 598 entering the AERO interface for the first time or if they are being 599 forwarded out the same AERO interface that they arrived on. This 600 latter form of encapsulation is known as "re-encapsulation". 602 AERO interfaces encapsulate packets per the specifications in 603 [RFC2003][RFC2473][RFC4213][RFC4301][RFC5246] except that the 604 interface copies the "TTL/Hop Limit", "Type of Service/Traffic Class" 605 and "Congestion Experienced" values in the packet's IP header into 606 the corresponding fields in the encapsulation header. For packets 607 undergoing re-encapsulation, the AERO interface instead copies the 608 "TTL/Hop Limit", "Type of Service/Traffic Class" and "Congestion 609 Experienced" values in the original encapsulation header into the 610 corresponding fields in the new encapsulation header (i.e., the 611 values are transferred between encapsulation headers and *not* copied 612 from the encapsulated packet's network-layer header). 614 When AERO UDP encapsulation is used, the AERO interface encapsulates 615 the packet per the specifications in [RFC2003][RFC2473][RFC4213] 616 except that it inserts a UDP header between the encapsulation header 617 and the packet's IP header. The AERO interface sets the UDP source 618 port to a constant value that it will use in each successive packet 619 it sends, sets the UDP checksum field to zero (see: 621 [RFC6935][RFC6936]) and sets the UDP length field to the length of 622 the IP packet plus 8 bytes for the UDP header itself. For packets 623 sent via a Server, the AERO interface sets the UDP destination port 624 to 8060 (i.e., the IANA-registered port number for AERO) when AERO- 625 only encapsulation is used. For packets sent to a neighboring 626 Client, the AERO interface sets the UDP destination port to the port 627 value stored in the neighbor cache entry for this neighbor. 629 The AERO interface next sets the IP protocol number in the 630 encapsulation header to the appropriate value for the first protocol 631 layer within the encapsulation (e.g., IPv4, IPv6, UDP, IPsec, etc.). 632 When IPv6 is used as the encapsulation protocol, the interface then 633 sets the flow label value in the encapsulation header the same as 634 described in [RFC6438]. When IPv4 is used as the encapsulation 635 protocol, the AERO interface sets the DF bit as discussed in 636 Section 3.6. 638 AERO interfaces decapsulate packets destined either to the node 639 itself or to a destination reached via an interface other than the 640 receiving AERO interface. When AERO UDP encapsulation is used (i.e., 641 when a UDP header with destination port 8060 is present) the 642 interface examines the first octet of the encapsulated packet. If 643 the most significant four bits of the first octet encode the value 644 '0110' (i.e., the version number value for IPv6) or the value '0100' 645 (i.e., the version number value for IPv4), the packet is accepted and 646 the encapsulating UDP header is discarded; otherwise, the packet is 647 discarded. 649 Further decapsulation then proceeds according to the appropriate 650 tunnel type [RFC2003][RFC2473][RFC4213][RFC4301][RFC5246]. 652 3.8. AERO Router Discovery, Prefix Delegation and Address Configuration 654 3.8.1. AERO Client Behavior 656 AERO Clients discover the link-layer addresses of AERO Servers via 657 static configuration, or through an automated means such as DNS name 658 resolution. In the absence of other information, the Client resolves 659 the Fully-Qualified Domain Name (FQDN) "linkupnetworks.domainname", 660 where "domainname" is the DNS domain appropriate for the Client's 661 attached underlying network. After discovering the link-layer 662 addresses, the Client associates with one or more of the 663 corresponding Servers. 665 To associate with a Server, the Client acts as a requesting router to 666 request an IP prefix through DHCPv6 PD [RFC3315][RFC3633][RFC6355] 667 using fe80::ffff:ffff:ffff:ffff as the IPv6 source address (see 668 Section 3.3), 'All_DHCP_Relay_Agents_and_Servers' as the IPv6 669 destination address and the link-layer address of the Server as the 670 link-layer destination address. The Client includes a DHCPv6 Unique 671 Identifier (DUID) in the Client Identifier option of its DHCPv6 672 messages (as well as a DHCPv6 authentication option if necessary) to 673 identify itself to the DHCPv6 server. If the Client is pre- 674 provisioned with an IP prefix associated with the AERO service, it 675 MAY also include the prefix in its DHCPv6 PD Request to indicate its 676 preferred prefix to the DHCPv6 server. The Client then sends the 677 encapsulated DHCPv6 request via an underlying interface. 679 When the Client receives its prefix delegation via a Reply from the 680 DHCPv6 server, it creates a neighbor cache entry with the Server's 681 link-local address (i.e., fe80::ID) as the network-layer address and 682 the Server's encapsulation address as the link-layer addresses. 683 Next, the Client assigns the AERO address derived from the delegated 684 prefix to the AERO interface and sub-delegates the prefix to nodes 685 and links within its attached EUNs (the AERO address thereafter 686 remains stable as the Client moves). The Client also sets both 687 AcceptTime and ForwardTime for each Server to the constant value 688 REACHABLE_TIME. The Client further renews its prefix delegation by 689 performing DHCPv6 Renew/Reply exchanges with its AERO address as the 690 IPv6 source address, 'All_DHCP_Relay_Agents_and_Servers' as the IPv6 691 destination address, the link-layer address of a Server as the link- 692 layer destination address and the same DUID and authentication 693 information. If the Client wishes to associate with multiple 694 Servers, it can perform DHCPv6 Renew/Reply exchanges via each of the 695 Servers. 697 The Client then sends an RS message to each of its associated Servers 698 to receive an RA message with a default router lifetime and any other 699 link-specific parameters. When the Client receives an RA message, it 700 configures or updates a default route according to the default router 701 lifetime but ignores any Prefix Information Options (PIOs) included 702 in the RA message since the AERO link is link-local-only. The Client 703 further ignores any RS messages it might receive, since only Servers 704 may process RS messages. 706 The Client then sends periodic RS messages to each Server before 707 AcceptTime and ForwardTime expire to obtain new RA messages for 708 Neighbor Unreachability Detection (NUD), to refresh any network 709 state, and to update the default router lifetime and any other link- 710 specific parameters. When the Client receives a new RA message, it 711 resets AcceptTime and ForwardTime to REACHABLE_TIME. The Client can 712 also forward IP packets destined to networks beyond its local EUNs 713 via a Server as a default router. The Server may in turn return a 714 redirection message informing the Client of a neighbor on the AERO 715 link that is topologically closer to the final destination (see 716 Section 3.9). 718 Note that, since the Client's AERO address is configured from the 719 unique DHCPv6 prefix delegation it receives, there is no need for 720 Duplicate Address Detection (DAD) on AERO links. Other nodes 721 maliciously attempting to hijack an authorized Client's AERO address 722 will be denied access to the network by the DHCPv6 server due to an 723 unacceptable link-layer address and/or security parameters (see: 724 Security Considerations). 726 3.8.2. AERO Server Behavior 728 AERO Servers configure a DHCPv6 relay function on their AERO links. 729 AERO Servers arrange to add their encapsulation layer IP addresses 730 (i.e., their link-layer addresses) to the DNS resource records for 731 the FQDN "linkupnetworks.domainname" before entering service. 733 When an AERO Server relays a prospective Client's DHCPv6 PD messages 734 to the DHCPv6 server, it wraps each message in a "Relay-forward" 735 message per [RFC3315] and includes a DHCPv6 Interface Identifier 736 option that encodes a value that identifies the AERO link to the 737 DHCPv6 server. Without creating internal state, the Server then 738 includes the Client's link-layer address in a DHCPv6 Client Link 739 Layer Address Option (CLLAO) [RFC6939] with the link-layer address 740 format shown in Figure 1 (i.e., Link ID followed by Preference 741 followed by UDP Port Number followed by IP Address). The Server sets 742 the CLLAO 'option-length' field to 22 (2 plus the length of the link- 743 layer address) and sets the 'link-layer type' field to TBD (see: IANA 744 Considerations). The Server finally includes a DHCPv6 Echo Request 745 Option (ERO) [RFC4994] that encodes the option code for the CLLAO in 746 a 'requested-option-code-n' field, then relays the message to the 747 DHCPv6 server. The CLLAO information will therefore subsequently be 748 echoed back in the DHCPv6 server's "Relay-reply" message. 750 When the DHCPv6 server issues the prefix delegation in a "Relay- 751 reply" message via the AERO Server (acting as a DHCPv6 relay), the 752 Server obtains the Client's link-layer address from the echoed CLLAO 753 option and also obtains the Client's delegated prefix from the 754 message. The Server then creates a neighbor cache entry for the 755 Client's AERO address with the Client's link-layer address as the 756 link-layer address for the neighbor cache entry. The neighbor cache 757 entry is created with both AcceptTime and ForwardTime set to 758 REACHABLE_TIME, since the Client will continue to send RS messages 759 within REACHABLE_TIME seconds as long as it wishes to remain 760 associated with this Server. 762 The Server also configures an IP forwarding table entry that lists 763 the Client's AERO address as the next hop toward the delegated IP 764 prefix with a lifetime derived from the DHCPv6 lease lifetime. The 765 Server finally injects the Client's prefix as an IP route into the 766 inter-Server/Relay routing system (see: Appendix A) then relays the 767 DHCPv6 message to the Client while using fe80::ID as the IPv6 source 768 address, the link-local address found in the "peer address" field of 769 the Relay-reply message as the IPv6 destination address, and the 770 Client's link-layer address as the destination link-layer address. 772 Servers respond to NS/RS messages from Clients on their AERO 773 interfaces by returning an NA/RA message. The Server SHOULD NOT 774 include PIOs in the RA messages it sends to Clients, since the Client 775 will ignore any such options. When the Server receives an NS/RS 776 message from the Client, it resets AcceptTime and ForwardTime to 777 REACHABLE_TIME. 779 Servers ignore any RA messages they may receive from a Client, but 780 they MAY examine RA messages received from other Servers for 781 consistency verification purposes. Servers do not send NS messages 782 for the purpose of updating Client neighbor cache timers, since 783 Clients are responsible for refreshing neighbor cache state. 785 When the Server forwards a packet via the same AERO interface on 786 which it arrived, it initiates an AERO route optimization procedure 787 as specified in Section 3.9. 789 3.9. AERO Redirection 791 3.9.1. Reference Operational Scenario 793 Figure 2 depicts the AERO redirection reference operational scenario, 794 using IPv6 addressing as the example (while not shown, a 795 corresponding example for IPv4 addressing can be easily constructed). 796 The figure shows an AERO Server('A'), two AERO Clients ('B', 'C') and 797 three ordinary IPv6 hosts ('D', 'E', 'F'): 799 .-(::::::::) 800 .-(:::: IP ::::)-. +-------------+ 801 (:: Internetwork ::)--| Host F | 802 `-(::::::::::::)-' +-------------+ 803 `-(::::::)-' 2001:db8:2::1 804 | 805 +--------------+ 806 | AERO Server A| 807 | (D->B; E->C) | 808 +--------------+ 809 fe80::ID 810 L2(A) 811 | 812 X-----+-----------+-----------+--------X 813 | AERO Link | 814 L2(B) L2(C) 815 fe80::2001:db8:0:0 fe80::2001:db8:1:0 .-. 816 +--------------+ +--------------+ ,-( _)-. 817 | AERO Client B| | AERO Client C| .-(_ IP )-. 818 | (default->A) | | (default->A) |--(__ EUN ) 819 +--------------+ +--------------+ `-(______)-' 820 2001:DB8:0::/48 2001:DB8:1::/48 | 821 | 2001:db8:1::1 822 .-. +-------------+ 823 ,-( _)-. 2001:db8:0::1 | Host E | 824 .-(_ IP )-. +-------------+ +-------------+ 825 (__ EUN )--| Host D | 826 `-(______)-' +-------------+ 828 Figure 2: AERO Reference Operational Scenario 830 In Figure 2, AERO Server ('A') connects to the AERO link and connects 831 to the IP Internetwork, either directly or via an AERO Relay (not 832 shown). Server ('A') assigns the address fe80::ID to its AERO 833 interface with link-layer address L2(A). Server ('A') next arranges 834 to add L2(A) to a published list of valid Servers for the AERO link. 836 AERO Client ('B') receives the prefix 2001:db8:0::/48 in a DHCPv6 PD 837 exchange via AERO Server ('A') then assigns the address 838 fe80::2001:db8:0:0 to its AERO interface with link-layer address 839 L2(B). Client ('B') configures a default route and neighbor cache 840 entry via the AERO interface with next-hop address fe80::ID and link- 841 layer address L2(A), then sub-delegates the prefix 2001:db8:0::/48 to 842 its attached EUNs. IPv6 host ('D') connects to the EUN, and 843 configures the address 2001:db8:0::1. 845 AERO Client ('C') receives the prefix 2001:db8:1::/48 in a DHCPv6 PD 846 exchange via AERO Server ('A') then assigns the address 847 fe80::2001:db8:1:0 to its AERO interface with link-layer address 848 L2(C). Client ('C') configures a default route and neighbor cache 849 entry via the AERO interface with next-hop address fe80::ID and link- 850 layer address L2(A), then sub-delegates the prefix 2001:db8:1::/48 to 851 its attached EUNs. IPv6 host ('E') connects to the EUN, and 852 configures the address 2001:db8:1::1. 854 Finally, IPv6 host ('F') connects to a network outside of the AERO 855 link domain. Host ('F') configures its IPv6 interface in a manner 856 specific to its attached IPv6 link, and assigns the address 857 2001:db8:2::1 to its IPv6 link interface. 859 3.9.2. Classical Redirection Approaches 861 With reference to Figure 2, when the source host ('D') sends a packet 862 to destination host ('E'), the packet is first forwarded via the EUN 863 to AERO Client ('B'). Client ('B') then forwards the packet over its 864 AERO interface to AERO Server ('A'), which then re-encapsulates and 865 forwards the packet to AERO Client ('C'), where the packet is finally 866 forwarded to destination host ('E'). When Server ('A') re- 867 encapsulates and forwards the packet back out on its advertising AERO 868 interface, it must arrange to redirect Client ('B') toward Client 869 ('C') as a better next-hop node on the AERO link that is closer to 870 the final destination. However, this redirection process applied to 871 AERO interfaces must be more carefully orchestrated than on ordinary 872 links since the parties may be separated by potentially many 873 underlying network routing hops. 875 Consider a first alternative in which Server ('A') informs Client 876 ('B') only and does not inform Client ('C') (i.e., "classical 877 redirection"). In that case, Client ('C') has no way of knowing that 878 Client ('B') is authorized to forward packets from the claimed source 879 address, and it may simply elect to drop the packets. Also, Client 880 ('B') has no way of knowing whether Client ('C') is performing some 881 form of source address filtering that would reject packets arriving 882 from a node other than a trusted default router, nor whether Client 883 ('C') is even reachable via a direct path that does not involve 884 Server ('A'). 886 Consider a second alternative in which Server ('A') informs both 887 Client ('B') and Client ('C') separately, via independent redirection 888 control messages (i.e., "augmented redirection"). In that case, if 889 Client ('B') receives the redirection control message but Client 890 ('C') does not, subsequent packets sent by Client ('B') could be 891 dropped due to filtering since Client ('C') would not have a route to 892 verify the claimed source address. Also, if Client ('C') receives 893 the redirection control message but Client ('B') does not, subsequent 894 packets sent in the reverse direction by Client ('C') would be lost. 896 Since both of these alternatives have shortcomings, a new redirection 897 technique (i.e., "AERO redirection") is needed. 899 3.9.3. Concept of Operations 901 Again, with reference to Figure 2, when source host ('D') sends a 902 packet to destination host ('E'), the packet is first forwarded over 903 the source host's attached EUN to Client ('B'), which then forwards 904 the packet via its AERO interface to Server ('A'). 906 Server ('A') then re-encapsulates and forwards the packet out the 907 same AERO interface toward Client ('C') and also sends an AERO 908 "Predirect" message forward to Client ('C') as specified in 909 Section 3.9.5. The Predirect message includes Client ('B')'s 910 network- and link-layer addresses as well as information that Client 911 ('C') can use to determine the IP prefix used by Client ('B') . After 912 Client ('C') receives the Predirect message, it process the message 913 and returns an AERO Redirect message destined for Client ('B') via 914 Server ('A') as specified in Section 3.9.6. During the process, 915 Client ('C') also creates or updates a neighbor cache entry for 916 Client ('B') and creates an IP forwarding table entry for Client 917 ('B')'s prefix. 919 When Server ('A') receives the Redirect message, it re-encapsulates 920 the message and forwards it on to Client ('B') as specified in 921 Section 3.9.7. The message includes Client ('C')'s network- and 922 link-layer addresses as well as information that Client ('B') can use 923 to determine the IP prefix used by Client ('C'). After Client ('B') 924 receives the Redirect message, it processes the message as specified 925 in Section 3.9.8. During the process, Client ('B') also creates or 926 updates a neighbor cache entry for Client ('C') and creates an IP 927 forwarding table entry for Client ('C')'s prefix. 929 Following the above Predirect/Redirect message exchange, forwarding 930 of packets from Client ('B') to Client ('C') without involving Server 931 ('A) as an intermediary is enabled. The mechanisms that support this 932 exchange are specified in the following sections. 934 3.9.4. Message Format 936 AERO Redirect/Predirect messages use the same format as for ICMPv6 937 Redirect messages depicted in Section 4.5 of [RFC4861], but also 938 include a new "Prefix Length" field taken from the low-order 8 bits 939 of the Redirect message Reserved field. (For IPv6, valid values for 940 the Prefix Length field are 0 through 64; for IPv4, valid values are 941 0 through 32.) The Redirect/Predirect messages are formatted as 942 shown in Figure 3: 944 0 1 2 3 945 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 1 946 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 947 | Type (=137) | Code (=0/1) | Checksum | 948 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 949 | Reserved | Prefix Length | 950 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 951 | | 952 + + 953 | | 954 + Target Address + 955 | | 956 + + 957 | | 958 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 959 | | 960 + + 961 | | 962 + Destination Address + 963 | | 964 + + 965 | | 966 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 967 | Options ... 968 +-+-+-+-+-+-+-+-+-+-+-+- 970 Figure 3: AERO Redirect/Predirect Message Format 972 3.9.5. Sending Predirects 974 When a Server forwards a packet from one of its associated Clients 975 toward another AERO Client connected to the same AERO link, the 976 Server sends a Predirect message forward toward the destination 977 Client instead of sending a Redirect message back to the source 978 Client. 980 In the reference operational scenario, when Server ('A') forwards a 981 packet sent by Client ('B') toward Client ('C'), it also sends a 982 Predirect message forward toward Client ('C'), subject to rate 983 limiting (see Section 8.2 of [RFC4861]). Server ('A') prepares the 984 Predirect message as follows: 986 o the link-layer source address is set to 'L2(A)' (i.e., the link- 987 layer address of Server ('A')). 989 o the link-layer destination address is set to 'L2(C)' (i.e., the 990 link-layer address of Client ('C')). 992 o the network-layer source address is set to fe80::2001:db8:0:0 993 (i.e., the AERO address of Client ('B')). 995 o the network-layer destination address is set to fe80::2001:db8:1:0 996 (i.e., the AERO address of Client ('C')). 998 o the Type is set to 137. 1000 o the Code is set to 1 to indicate "Predirect". 1002 o the Prefix Length is set to the length of the prefix to be applied 1003 to the Target Address. 1005 o the Target Address is set to fe80::2001:db8:0:0 (i.e., the AERO 1006 address of Client ('B')). 1008 o the Destination Address is set to the source address of the 1009 originating packet that triggered the Predirection event. (If the 1010 originating packet is an IPv4 packet, the address is constructed 1011 in IPv4-compatible IPv6 address format). 1013 o the message includes a TLLAO with Link ID and Preference set to 1014 appropriate values for Client ('B')'s underlying interface, and 1015 with UDP Port Number and IP Address set to 'L2(B)'. 1017 o the message includes a Timestamp option. 1019 o the message includes a Redirected Header Option (RHO) that 1020 contains the originating packet truncated to ensure that at least 1021 the network-layer header is included but the size of the message 1022 does not exceed 1280 bytes. 1024 Note that the reference operational scenario applies to the case when 1025 the source and destination Clients are associated with the same 1026 Server. When the source and destination Clients are associated with 1027 different Servers, the source Client's Server forwards the packets 1028 and Predirect messages to a Relay which in turn forwards them toward 1029 the destination Client. In that case, the Server sets the Predirect 1030 link-layer destination address to the link-layer address of the 1031 Relay. 1033 Servers therefore require knowledge of all aggregated IP prefixes 1034 associated with the AERO link so that they can determine when a 1035 prospective destination Client is on-link. See Appendix A for a 1036 discussion of AERO Server/Relay interworking. 1038 3.9.6. Processing Predirects and Sending Redirects 1040 When Client ('C') receives a Predirect message, it accepts the 1041 message only if the message has a link-layer source address of the 1042 Server, i.e. 'L2(A)'. Client ('C') further accepts the message only 1043 if it is willing to serve as a redirection target. Next, Client 1044 ('C') validates the message according to the ICMPv6 Redirect message 1045 validation rules in Section 8.1 of [RFC4861], except that it accepts 1046 the message even though the network-layer source address is not that 1047 of it's current first-hop router. 1049 In the reference operational scenario, when Client ('C') receives a 1050 valid Predirect message, it either creates or updates a neighbor 1051 cache entry that stores the Target Address of the message as the 1052 network-layer address of Client ('B') and stores the link-layer 1053 address found in the TLLAO as the link-layer address(es) of Client 1054 ('B'). Client ('C') then sets AcceptTime for the neighbor cache 1055 entry to ACCEPT_TIME. Next, Client ('C') applies the Prefix Length 1056 to the Destination Address and records the resulting prefix in its IP 1057 forwarding table. 1059 After processing the message, Client ('C') prepares a Redirect 1060 message response as follows: 1062 o the link-layer source address is set to 'L2(C)' (i.e., the link- 1063 layer address of Client ('C')). 1065 o the link-layer destination address is set to 'L2(A)' (i.e., the 1066 link-layer address of Server ('A')). 1068 o the network-layer source address is set to fe80::2001:db8:1:0 1069 (i.e., the AERO address of Client ('C')). 1071 o the network-layer destination address is set to fe80::2001:db8:0:0 1072 (i.e., the AERO address of Client ('B')). 1074 o the Type is set to 137. 1076 o the Code is set to 0 to indicate "Redirect". 1078 o the Prefix Length is set to the length of the prefix to be applied 1079 to the Target Address. 1081 o the Target Address is set to fe80::2001:db8:1:0 (i.e., the AERO 1082 address of Client ('C')). 1084 o the Destination Address is set to the destination address of the 1085 originating packet that triggered the Redirection event. (If the 1086 originating packet is an IPv4 packet, the address is constructed 1087 in IPv4-compatible IPv6 address format). 1089 o the message includes a TLLAO with Link ID and Preference set to 1090 appropriate values for Client ('C')'s underlying interface, and 1091 with UDP Port Number and IP Address set to '0'. 1093 o the message includes a Timestamp option. 1095 o the message includes as much of the RHO copied from the 1096 corresponding AERO Predirect message as possible such that at 1097 least the network-layer header is included but the size of the 1098 message does not exceed 1280 bytes. 1100 After Client ('C') prepares the Redirect message, it sends the 1101 message to Server ('A'). 1103 3.9.7. Re-encapsulating and Relaying Redirects 1105 When Server ('A') receives a Redirect message from Client ('C'), it 1106 validates the message according to the ICMPv6 Redirect message 1107 validation rules in Section 8.1 of [RFC4861] and also verifies that 1108 Client ('C') is authorized to use the Prefix Length in the Redirect 1109 message when applied to the AERO address in the network-layer source 1110 of the Redirect message by searching for the AERO address' embedded 1111 prefix in the IP routing table. If validation fails, Server ('A') 1112 discards the message; otherwise, it copies the correct UDP Port 1113 number and IP Address for Client ('C') into the (previously empty) 1114 TLLAO. 1116 Server ('A') then examines the network-layer destination address of 1117 the message to determine the next hop toward the prefix of Client 1118 ('B') by searching for the AERO address' embedded prefix in the IP 1119 routing table. If the next hop is reached via the AERO interface, 1120 Server ('A') re-encapsulates the Redirect and relays it on to Client 1121 ('B') by changing the link-layer source address of the message to 1122 'L2(A)' and changing the link-layer destination address to 'L2(B)'. 1123 Server ('A') finally forwards the re-encapsulated message to Client 1124 ('B') without decrementing the network-layer TTL/Hop Limit field. 1126 While not shown in Figure 2, AERO Relays relay Redirect and Predirect 1127 messages in exactly this same fashion described above (see: 1128 Appendix A). 1130 3.9.8. Processing Redirects 1132 When Client ('B') receives the Redirect message, it accepts the 1133 message only if it has a link-layer source address of the Server, 1134 i.e. 'L2(A)'. Next, Client ('B') validates the message according to 1135 the ICMPv6 Redirect message validation rules in Section 8.1 of 1136 [RFC4861], except that it accepts the message even though the 1137 network-layer source address is not that of it's current first-hop 1138 router. Following validation, Client ('B') then processes the 1139 message as follows. 1141 In the reference operational scenario, when Client ('B') receives the 1142 Redirect message, it either creates or updates a neighbor cache entry 1143 that stores the Target Address of the message as the network-layer 1144 address of Client ('C') and stores the link-layer address found in 1145 the TLLAO as the link-layer address of Client ('C'). Client ('B') 1146 then sets the neighbor cache entry ForwardTime variable with timeout 1147 value FORWARD_TIME. Next, Client ('B') applies the Prefix Length to 1148 the Destination Address and records the resulting IP prefix in its IP 1149 forwarding table. 1151 Now, Client ('B') has an IP forwarding table entry for Client('C')'s 1152 prefix and a neighbor cache entry with a valid ForwardTime value, 1153 while Client ('C') has an IP forwarding table entry for Client 1154 ('B')'s prefix with a valid AcceptTime value. Thereafter, Client 1155 ('B') may forward ordinary network-layer data packets directly to 1156 Client ("C") without involving Server ('A') and Client ('C') can 1157 verify that the packets came from an acceptable source. (In order 1158 for Client ('C') to forward packets to Client ('B') a corresponding 1159 Predirect/Redirect message exchange is required in the reverse 1160 direction.) 1162 3.9.9. Server-Oriented Redirection 1164 In some environments, the Server nearest the destination Client may 1165 need to serve as the redirection target, e.g., if direct Client-to- 1166 Client communications are not possible. In that case, the Server 1167 prepares the Redirect message the same as if it were the destination 1168 Client (see: Section 3.9.6), except that it writes its own link-layer 1169 address in the TLLAO option. 1171 3.10. Neighbor Reachability Maintenance 1173 AERO nodes send unicast NS messages to elicit solicited NA messages 1174 from neighbors the same as described for Neighbor Unreachability 1175 Detection (NUD) in [RFC4861]. When an AERO node sends an NS/NA 1176 message, it MUST use its link-local address as the IPv6 source 1177 address and the link-local address of the neighbor as the IPv6 1178 destination address. When an AERO node receives an NS message or a 1179 solicited NA message, it accepts the message if it has a neighbor 1180 cache entry for the neighbor; otherwise, it ignores the message. 1182 When a source Client is redirected to a target Client it SHOULD test 1183 the direct path by sending an initial NS message to elicit a 1184 solicited NA response. While testing the path, the source Client can 1185 optionally continue sending packets via the Server, maintain a small 1186 queue of packets until target reachability is confirmed, or 1187 (optimistically) allow packets to flow directly to the target. The 1188 source Client SHOULD thereafter continue to test the direct path to 1189 the target Client (see Section 7.3 of [RFC4861]) periodically in 1190 order to keep neighbor cache entries alive. 1192 In particular, while the source Client is actively sending packets to 1193 the target Client it SHOULD also send NS messages separated by 1194 RETRANS_TIMER milliseconds in order to receive solicited NA messages. 1195 If the source Client is unable to elicit a solicited NA response from 1196 the target Client after MAX_RETRY attempts, it SHOULD set ForwardTime 1197 to 0 and resume sending packets via the Server which may or may not 1198 result in a new redirection event. Otherwise, the source Client 1199 considers the path usable and SHOULD thereafter process any link- 1200 layer errors as a hint that the direct path to the target Client has 1201 either failed or has become intermittent. 1203 When a target Client receives an NS message from a source Client, it 1204 resets AcceptTime to ACCEPT_TIME if a neighbor cache entry exists; 1205 otherwise, it discards the NS message. 1207 When a source Client receives a solicited NA message from a target 1208 Client, it resets ForwardTime to FORWARD_TIME if a neighbor cache 1209 entry exists; otherwise, it discards the NA message. 1211 When ForwardTime for a neighbor cache entry expires, the source 1212 Client resumes sending any subsequent packets via the Server and may 1213 (eventually) receive a new Redirect message. When AcceptTime for a 1214 neighbor cache entry expires, the target Client discards any 1215 subsequent packets received directly from the source Client. When 1216 both ForwardTime and AcceptTime for a neighbor cache entry expire, 1217 the Client deletes both the neighbor cache entry and the 1218 corresponding IP forwarding table entry. 1220 3.11. Mobility Management 1222 When a Client needs to change its link-layer address, e.g., due to a 1223 mobility event, it performs an immediate DHCPv6 Renew/Reply via each 1224 of its Servers using the new link-layer address as the source. The 1225 DHCPv6 Server will re-authenticate the Client and (assuming 1226 authentication succeeds) the DHCPv6 Renew/Reply exchange will update 1227 each Server's neighbor cache. 1229 Next, the Client sends unsolicited NA messages to each of its active 1230 neighbors using the same procedures as specified in Section 7.2.6 of 1231 [RFC4861], except that it sends the messages as unicast to each 1232 neighbor via a Server instead of multicast. In this process, the 1233 Client should send no more than MAX_NEIGHBOR_ADVERTISEMENT messages 1234 separated by no less than RETRANS_TIMER seconds to each neighbor. 1236 With reference to Figure 2, Client ('C') sends unicast unsolicited NA 1237 messages to Client ('B') via Server ('A') as follows: 1239 o the link-layer source address is set to 'L2(C)' (i.e., the link- 1240 layer address of Client ('C')). 1242 o the link-layer destination address is set to 'L2(A)' (i.e., the 1243 link-layer address of Server ('A')). 1245 o the network-layer source address is set to fe80::2001:db8:1:0 1246 (i.e., the AERO address of Client ('C')). 1248 o the network-layer destination address is set to fe80::2001:db8:0:0 1249 (i.e., the AERO address of Client ('B')). 1251 o the Type is set to 136. 1253 o the Code is set to 0. 1255 o the Solicited flag is set to 0. 1257 o the Override flag is set to 1. 1259 o the Target Address is set to fe80::2001:db8:1:0 (i.e., the AERO 1260 address of Client ('C')). 1262 o the message includes a TLLAO with Link ID and Preference set to 1263 appropriate values for Client ('C')'s underlying interface, and 1264 with UDP Port Number and IP Address set to '0'. 1266 o the message includes a Timestamp option. 1268 When Server ('A') receives the NA message, it relays the message in 1269 the same way as described for relaying Redirect messages in 1270 Section 3.9.7. In particular, Server ('A') copies the correct UDP 1271 port number and IP address into the TLLAO, changes the link-layer 1272 source address to its own address, changes the link-layer destination 1273 address to the address of Client ('B'), then forwards the NA message 1274 based on an IP route matching the AERO address in the network-layer 1275 destination address. When Client ('B') receives the NA message, it 1276 accepts the message only if it already has a neighbor cache entry for 1277 Client ('C') then updates the link-layer address for Client ('C') 1278 based on the address in the TLLAO. However, Client ('B') MUST NOT 1279 update ForwardTime since it has no way of knowing whether Client 1280 ('C') has updated AcceptTime. 1282 When a Client associates with a new Server, it issues a new DHCPv6 1283 Renew message via the new Server as the DHCPv6 relay. The new Server 1284 then relays the message to the DHCPv6 server and processes the 1285 resulting exchange. After the Client receives the resulting DHCPv6 1286 Reply message, it sends an RS message to the new Server to receive a 1287 new RA message. 1289 When a Client disassociates with an existing Server, it sends a 1290 "terminating RS" message to the old Server. The terminating RS 1291 message is prepared exactly the same as for an ordinary RS message, 1292 except that the Code field contains the value '1'. When the old 1293 Server receives the terminating RS message, it withdraws the IP route 1294 from the routing system and deletes the neighbor cache entry and IP 1295 forwarding table entry for the Client. The old Server then returns 1296 an RA message with default router lifetime set to 0 which the Client 1297 can use to verify that the termination signal has been processed. 1298 The client then deletes both the default route and the neighbor cache 1299 entry for the old Server. (Note that the Client and the old Server 1300 MAY impose a small delay before deleting the neighbor cache and IP 1301 forwarding table entries so that any packets already in the system 1302 can still be delivered to the Client.) 1304 3.12. Encapsulation Protocol Version Considerations 1306 A source Client may connect only to an IPvX underlying network, while 1307 the target Client connects only to an IPvY underlying network. In 1308 that case, the target and source Clients have no means for reaching 1309 each other directly (since they connect to underlying networks of 1310 different IP protocol versions) and so must ignore any redirection 1311 messages and continue to send packets via the Server. 1313 3.13. Multicast Considerations 1315 When the underlying network does not support multicast, AERO nodes 1316 map IPv6 link-scoped multicast addresses (including 1317 'All_DHCP_Relay_Agents_and_Servers') to the link-layer address of a 1318 Server. 1320 When the underlying network supports multicast, AERO nodes use the 1321 multicast address mapping specification found in [RFC2529] for IPv4 1322 underlying networks and use a direct multicast mapping for IPv6 1323 underlying networks. (In the latter case, "direct multicast mapping" 1324 means that if the IPv6 multicast destination address of the 1325 encapsulated packet is "M", then the IPv6 multicast destination 1326 address of the encapsulating header is also "M".) 1328 3.14. Operation on AERO Links Without DHCPv6 Services 1330 When the AERO link does not provide DHCPv6 services, operation can 1331 still be accommodated through administrative configuration of 1332 prefixes on AERO Clients. In that case, administrative 1333 configurations of IP routes and AERO interface neighbor cache entries 1334 on both the Server and Client are also necessary. However, this may 1335 preclude the ability for Clients to dynamically change to new 1336 Servers, and can expose the AERO link to misconfigurations unless the 1337 administrative configurations are carefully coordinated. 1339 3.15. Operation on Server-less AERO Links 1341 In some AERO link scenarios, there may be no Servers on the link and/ 1342 or no need for Clients to use a Server as an intermediary trust 1343 anchor. In that case, each Client acts as a Server unto itself to 1344 establish neighbor cache entries and IP forwarding table entries by 1345 performing direct Client-to-Client Predirect/Redirect exchanges, and 1346 some other form of trust basis must be applied so that each Client 1347 can verify that the prospective neighbor is authorized to use its 1348 claimed prefix. 1350 When there is no Server on the link, Clients must arrange to receive 1351 prefix delegations and publish the delegations via a secure alternate 1352 prefix delegation authority through some means outside the scope of 1353 this document. 1355 4. Implementation Status 1357 An application-layer implementation is in progress. 1359 5. IANA Considerations 1361 The IANA is instructed to assign a new 2-octet Hardware Type number 1362 for AERO in the "arp-parameters" registry per Section 2 of [RFC5494]. 1363 The number is assigned from the 2-octet Unassigned range with 1364 Hardware Type "AERO" and with this document as the reference. 1366 6. Security Considerations 1368 AERO link security considerations are the same as for standard IPv6 1369 Neighbor Discovery [RFC4861] except that AERO improves on some 1370 aspects. In particular, AERO uses a trust basis between Clients and 1371 Servers, where the Clients only engage in the AERO mechanism when it 1372 is facilitated by a trust anchor. AERO also uses DHCPv6 1373 authentication for Client authentication and network admission 1374 control. 1376 AERO links must be protected against link-layer address spoofing 1377 attacks in which an attacker on the link pretends to be a trusted 1378 neighbor. Links that provide link-layer securing mechanisms (e.g., 1379 IEEE 802.1X WLANs) and links that provide physical security (e.g., 1380 enterprise network wired LANs) provide a first line of defense that 1381 is often sufficient. In other instances, additional securing 1382 mechanisms such as Secure Neighbor Discovery (SeND) [RFC3971], IPsec 1383 [RFC4301] or TLS [RFC5246] may be necessary. 1385 AERO Clients MUST ensure that their connectivity is not used by 1386 unauthorized nodes on EUNs to gain access to a protected network, 1387 i.e., AERO Clients that act as routers MUST NOT provide routing 1388 services for unauthorized nodes. (This concern is no different than 1389 for ordinary hosts that receive an IP address delegation but then 1390 "share" the address with unauthorized nodes via a NAT function.) 1392 On some AERO links, establishment and maintenance of a direct path 1393 between neighbors requires secured coordination such as through the 1394 Internet Key Exchange (IKEv2) protocol [RFC5996] to establish a 1395 security association. 1397 7. Acknowledgements 1399 Discussions both on IETF lists and in private exchanges helped shape 1400 some of the concepts in this work. Individuals who contributed 1401 insights include Mikael Abrahamsson, Fred Baker, Stewart Bryant, 1402 Brian Carpenter, Wojciech Dec, Brian Haberman, Joel Halpern, Sascha 1403 Hlusiak, Lee Howard, Joe Touch and Bernie Volz. Members of the IESG 1404 also provided valuable input during their review process that greatly 1405 improved the document. Special thanks go to Stewart Bryant, Joel 1406 Halpern and Brian Haberman for their shepherding guidance. 1408 This work has further been encouraged and supported by Boeing 1409 colleagues including Keith Bartley, Dave Bernhardt, Cam Brodie, 1410 Balaguruna Chidambaram, Wen Fang, Anthony Gregory, Jeff Holland, Ed 1411 King, Gen MacLean, Kent Shuey, Mike Slane, Julie Wulff, Yueli Yang, 1412 and other members of the BR&T and BIT mobile networking teams. 1414 Earlier works on NBMA tunneling approaches are found in 1415 [RFC2529][RFC5214][RFC5569]. 1417 8. References 1419 8.1. Normative References 1421 [RFC0768] Postel, J., "User Datagram Protocol", STD 6, RFC 768, 1422 August 1980. 1424 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, September 1425 1981. 1427 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 1428 RFC 792, September 1981. 1430 [RFC2003] Perkins, C., "IP Encapsulation within IP", RFC 2003, 1431 October 1996. 1433 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1434 Requirement Levels", BCP 14, RFC 2119, March 1997. 1436 [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 1437 (IPv6) Specification", RFC 2460, December 1998. 1439 [RFC2473] Conta, A. and S. Deering, "Generic Packet Tunneling in 1440 IPv6 Specification", RFC 2473, December 1998. 1442 [RFC3315] Droms, R., Bound, J., Volz, B., Lemon, T., Perkins, C., 1443 and M. Carney, "Dynamic Host Configuration Protocol for 1444 IPv6 (DHCPv6)", RFC 3315, July 2003. 1446 [RFC3633] Troan, O. and R. Droms, "IPv6 Prefix Options for Dynamic 1447 Host Configuration Protocol (DHCP) version 6", RFC 3633, 1448 December 2003. 1450 [RFC3971] Arkko, J., Kempf, J., Zill, B., and P. Nikander, "SEcure 1451 Neighbor Discovery (SEND)", RFC 3971, March 2005. 1453 [RFC4213] Nordmark, E. and R. Gilligan, "Basic Transition Mechanisms 1454 for IPv6 Hosts and Routers", RFC 4213, October 2005. 1456 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 1457 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 1458 September 2007. 1460 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 1461 Address Autoconfiguration", RFC 4862, September 2007. 1463 [RFC6434] Jankiewicz, E., Loughney, J., and T. Narten, "IPv6 Node 1464 Requirements", RFC 6434, December 2011. 1466 8.2. Informative References 1468 [IRON] Templin, F., "The Internet Routing Overlay Network 1469 (IRON)", Work in Progress, June 2012. 1471 [RFC0879] Postel, J., "TCP maximum segment size and related topics", 1472 RFC 879, November 1983. 1474 [RFC1930] Hawkinson, J. and T. Bates, "Guidelines for creation, 1475 selection, and registration of an Autonomous System (AS)", 1476 BCP 6, RFC 1930, March 1996. 1478 [RFC2529] Carpenter, B. and C. Jung, "Transmission of IPv6 over IPv4 1479 Domains without Explicit Tunnels", RFC 2529, March 1999. 1481 [RFC2675] Borman, D., Deering, S., and R. Hinden, "IPv6 Jumbograms", 1482 RFC 2675, August 1999. 1484 [RFC4271] Rekhter, Y., Li, T., and S. Hares, "A Border Gateway 1485 Protocol 4 (BGP-4)", RFC 4271, January 2006. 1487 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 1488 Architecture", RFC 4291, February 2006. 1490 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 1491 Internet Protocol", RFC 4301, December 2005. 1493 [RFC4821] Mathis, M. and J. Heffner, "Packetization Layer Path MTU 1494 Discovery", RFC 4821, March 2007. 1496 [RFC4963] Heffner, J., Mathis, M., and B. Chandler, "IPv4 Reassembly 1497 Errors at High Data Rates", RFC 4963, July 2007. 1499 [RFC4994] Zeng, S., Volz, B., Kinnear, K., and J. Brzozowski, 1500 "DHCPv6 Relay Agent Echo Request Option", RFC 4994, 1501 September 2007. 1503 [RFC5214] Templin, F., Gleeson, T., and D. Thaler, "Intra-Site 1504 Automatic Tunnel Addressing Protocol (ISATAP)", RFC 5214, 1505 March 2008. 1507 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 1508 (TLS) Protocol Version 1.2", RFC 5246, August 2008. 1510 [RFC5494] Arkko, J. and C. Pignataro, "IANA Allocation Guidelines 1511 for the Address Resolution Protocol (ARP)", RFC 5494, 1512 April 2009. 1514 [RFC5522] Eddy, W., Ivancic, W., and T. Davis, "Network Mobility 1515 Route Optimization Requirements for Operational Use in 1516 Aeronautics and Space Exploration Mobile Networks", RFC 1517 5522, October 2009. 1519 [RFC5569] Despres, R., "IPv6 Rapid Deployment on IPv4 1520 Infrastructures (6rd)", RFC 5569, January 2010. 1522 [RFC5996] Kaufman, C., Hoffman, P., Nir, Y., and P. Eronen, 1523 "Internet Key Exchange Protocol Version 2 (IKEv2)", RFC 1524 5996, September 2010. 1526 [RFC6146] Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful 1527 NAT64: Network Address and Protocol Translation from IPv6 1528 Clients to IPv4 Servers", RFC 6146, April 2011. 1530 [RFC6204] Singh, H., Beebee, W., Donley, C., Stark, B., and O. 1531 Troan, "Basic Requirements for IPv6 Customer Edge 1532 Routers", RFC 6204, April 2011. 1534 [RFC6355] Narten, T. and J. Johnson, "Definition of the UUID-Based 1535 DHCPv6 Unique Identifier (DUID-UUID)", RFC 6355, August 1536 2011. 1538 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 1539 for Equal Cost Multipath Routing and Link Aggregation in 1540 Tunnels", RFC 6438, November 2011. 1542 [RFC6691] Borman, D., "TCP Options and Maximum Segment Size (MSS)", 1543 RFC 6691, July 2012. 1545 [RFC6706] Templin, F., "Asymmetric Extended Route Optimization 1546 (AERO)", RFC 6706, August 2012. 1548 [RFC6864] Touch, J., "Updated Specification of the IPv4 ID Field", 1549 RFC 6864, February 2013. 1551 [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and 1552 UDP Checksums for Tunneled Packets", RFC 6935, April 2013. 1554 [RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement 1555 for the Use of IPv6 UDP Datagrams with Zero Checksums", 1556 RFC 6936, April 2013. 1558 [RFC6939] Halwasia, G., Bhandari, S., and W. Dec, "Client Link-Layer 1559 Address Option in DHCPv6", RFC 6939, May 2013. 1561 [RFC6980] Gont, F., "Security Implications of IPv6 Fragmentation 1562 with IPv6 Neighbor Discovery", RFC 6980, August 2013. 1564 [RFC7078] Matsumoto, A., Fujisaki, T., and T. Chown, "Distributing 1565 Address Selection Policy Using DHCPv6", RFC 7078, January 1566 2014. 1568 Appendix A. AERO Server and Relay Interworking 1570 Figure 2 depicts a reference AERO operational scenario with a single 1571 Server on the AERO link. In order to support scaling to larger 1572 numbers of nodes, the AERO link can deploy multiple Servers and 1573 Relays, e.g., as shown in Figure 4. 1575 .-(::::::::) 1576 .-(:::: IP ::::)-. 1577 (:: Internetwork ::) 1578 `-(::::::::::::)-' 1579 `-(::::::)-' 1580 | 1581 +--------------+ +------+-------+ +--------------+ 1582 |AERO Server C | | AERO Relay D | |AERO Server E | 1583 | (default->D) | | (A->C; G->E) | | (default->D) | 1584 | (A->B) | +-------+------+ | (G->F) | 1585 +-------+------+ | +------+-------+ 1586 | | | 1587 X---+---+-------------------+------------------+---+---X 1588 | AERO Link | 1589 +-----+--------+ +--------+-----+ 1590 |AERO Client B | |AERO Client F | 1591 | (default->C) | | (default->E) | 1592 +--------------+ +--------------+ 1593 .-. .-. 1594 ,-( _)-. ,-( _)-. 1595 .-(_ IP )-. .-(_ IP )-. 1596 (__ EUN ) (__ EUN ) 1597 `-(______)-' `-(______)-' 1598 | | 1599 +--------+ +--------+ 1600 | Host A | | Host G | 1601 +--------+ +--------+ 1603 Figure 4: AERO Server/Relay Interworking 1605 In this example, Client ('B') associates with Server ('C'), while 1606 Client ('F') associates with Server ('E'). Furthermore, Servers 1607 ('C') and ('E') do not associate with each other directly, but rather 1608 have an association with Relay ('D') (i.e., a router that has full 1609 topology information concerning its associated Servers and their 1610 Clients). Relay ('D') connects to the AERO link, and also connects 1611 to the native IP Internetwork. 1613 When source host ('A') sends a packet toward destination host ('G'), 1614 IP forwarding directs the packet through the EUN to Client ('B'), 1615 which forwards the packet to Server ('C'). Server ('C') forwards 1616 both the packet and a Predirect message through Relay ('D'). Relay 1617 ('D') then forwards both the original packet and Predirect to Server 1618 ('E'). When Server ('E') receives the packet and Predirect message, 1619 it forwards them to Client ('F'). 1621 After processing the Predirect message, Client ('F') sends a Redirect 1622 message to Server ('E'). Server ('E'), in turn, forwards the message 1623 through Relay ('D') to Server ('C'). When Server ('C') receives the 1624 Redirect message, it forwards the message to Client ('B') informing 1625 it that host 'G's EUN can be reached via Client ('F'), thus 1626 completing the AERO redirection. 1628 The network-layer routing information shared between Servers and 1629 Relays must be carefully coordinated. In particular, Relays require 1630 full topology information, while individual Servers only require 1631 partial topology information, i.e., they only need to know the set of 1632 aggregated prefixes associated with the AERO link and the EUN 1633 prefixes associated with their current set of associated Clients. 1634 This can be accomplished in a number of ways, but a prominent example 1635 is through the use of an internal instance of the Border Gateway 1636 Protocol (BGP) [RFC4271] coordinated between Servers and Relays. 1637 This internal BGP instance does not interact with the public Internet 1638 BGP instance; therefore, the AERO link is presented to the IP 1639 Internetwork as a small set of aggregated prefixes as opposed to the 1640 full set of individual Client prefixes. 1642 In a reference BGP arrangement, each AERO Server is configured as an 1643 Autonomous System Border Router (ASBR) for a stub Autonomous System 1644 (AS) (possibly using a private AS Number (ASN) [RFC1930]), and each 1645 Server further peers with each Relay but does not peer with other 1646 Servers. Each Server maintains a working set of associated Clients, 1647 and dynamically announces new Client prefixes and withdraws departed 1648 Client prefixes in its BGP updates. The Relays therefore discover 1649 the full topology of the AERO link in terms of the working set of 1650 Clients associated with each Server. Since Clients are expected to 1651 remain associated with their current set of Servers for extended 1652 timeframes, the amount of BGP control messaging between Servers and 1653 Relays should be minimal. However, Servers SHOULD dampen any route 1654 oscillations caused by impatient Clients that repeatedly associate 1655 and disassociate with the Server. 1657 See [IRON] for further architectural discussion. 1659 Author's Address 1661 Fred L. Templin (editor) 1662 Boeing Research & Technology 1663 P.O. Box 3707 1664 Seattle, WA 98124 1665 USA 1667 Email: fltemplin@acm.org