idnits 2.17.1 draft-templin-6man-aero-46.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (25 April 2022) is 729 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Unused Reference: 'RFC3971' is defined on line 4333, but no explicit reference was found in the text == Unused Reference: 'RFC3972' is defined on line 4338, but no explicit reference was found in the text == Unused Reference: 'RFC4191' is defined on line 4342, but no explicit reference was found in the text == Unused Reference: 'RFC7739' is defined on line 4374, but no explicit reference was found in the text == Unused Reference: 'BGP' is defined on line 4395, but no explicit reference was found in the text == Unused Reference: 'I-D.bonica-6man-crh-helper-opt' is defined on line 4405, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-intarea-frag-fragile' is defined on line 4412, but no explicit reference was found in the text == Unused Reference: 'I-D.templin-6man-dhcpv6-ndopt' is defined on line 4443, but no explicit reference was found in the text == Unused Reference: 'I-D.templin-v6ops-pdhost' is defined on line 4477, but no explicit reference was found in the text == Unused Reference: 'OVPN' is defined on line 4491, but no explicit reference was found in the text == Unused Reference: 'RFC2003' is defined on line 4506, but no explicit reference was found in the text == Unused Reference: 'RFC2004' is defined on line 4510, but no explicit reference was found in the text == Unused Reference: 'RFC2983' is defined on line 4527, but no explicit reference was found in the text == Unused Reference: 'RFC3168' is defined on line 4531, but no explicit reference was found in the text == Unused Reference: 'RFC3330' is defined on line 4536, but no explicit reference was found in the text == Unused Reference: 'RFC4122' is defined on line 4545, but no explicit reference was found in the text == Unused Reference: 'RFC4982' is defined on line 4594, but no explicit reference was found in the text == Unused Reference: 'RFC6139' is defined on line 4636, but no explicit reference was found in the text == Unused Reference: 'RFC6273' is defined on line 4656, but no explicit reference was found in the text == Unused Reference: 'RFC6355' is defined on line 4665, but no explicit reference was found in the text == Unused Reference: 'RFC6438' is defined on line 4670, but no explicit reference was found in the text == Unused Reference: 'RFC6935' is defined on line 4679, but no explicit reference was found in the text == Unused Reference: 'RFC6936' is defined on line 4684, but no explicit reference was found in the text == Outdated reference: A later version (-74) exists of draft-templin-6man-omni-60 == Outdated reference: A later version (-31) exists of draft-bonica-6man-comp-rtg-hdr-27 == Outdated reference: A later version (-13) exists of draft-ietf-intarea-tunnels-10 == Outdated reference: A later version (-30) exists of draft-ietf-ipwave-vehicular-networking-28 == Outdated reference: A later version (-26) exists of draft-ietf-rtgwg-atn-bgp-17 -- Obsolete informational reference (is this intentional?): RFC 3330 (Obsoleted by RFC 5735) -- Obsolete informational reference (is this intentional?): RFC 6106 (Obsoleted by RFC 8106) -- Obsolete informational reference (is this intentional?): RFC 6347 (Obsoleted by RFC 9147) Summary: 0 errors (**), 0 flaws (~~), 29 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group F. L. Templin, Ed. 3 Internet-Draft Boeing Research & Technology 4 Intended status: Informational 25 April 2022 5 Expires: 27 October 2022 7 Automatic Extended Route Optimization (AERO) 8 draft-templin-6man-aero-46 10 Abstract 12 This document specifies an Automatic Extended Route Optimization 13 (AERO) service for IP internetworking over Overlay Multilink Network 14 (OMNI) interfaces. AERO/OMNI use an IPv6 link-local address format 15 that supports operation of the IPv6 Neighbor Discovery (IPv6 ND) 16 protocol. Prefix delegation/registration services are employed for 17 network admission and to manage the IP forwarding and routing 18 systems. Secure multilink operation, mobility management, multicast, 19 traffic path selection and route optimization are naturally supported 20 through dynamic neighbor cache updates. AERO is a widely-applicable 21 mobile internetworking service especially well-suited to aviation 22 services, intelligent transportation systems, mobile end user devices 23 and many other applications. 25 Status of This Memo 27 This Internet-Draft is submitted in full conformance with the 28 provisions of BCP 78 and BCP 79. 30 Internet-Drafts are working documents of the Internet Engineering 31 Task Force (IETF). Note that other groups may also distribute 32 working documents as Internet-Drafts. The list of current Internet- 33 Drafts is at https://datatracker.ietf.org/drafts/current/. 35 Internet-Drafts are draft documents valid for a maximum of six months 36 and may be updated, replaced, or obsoleted by other documents at any 37 time. It is inappropriate to use Internet-Drafts as reference 38 material or to cite them other than as "work in progress." 40 This Internet-Draft will expire on 27 October 2022. 42 Copyright Notice 44 Copyright (c) 2022 IETF Trust and the persons identified as the 45 document authors. All rights reserved. 47 This document is subject to BCP 78 and the IETF Trust's Legal 48 Provisions Relating to IETF Documents (https://trustee.ietf.org/ 49 license-info) in effect on the date of publication of this document. 50 Please review these documents carefully, as they describe your rights 51 and restrictions with respect to this document. Code Components 52 extracted from this document must include Revised BSD License text as 53 described in Section 4.e of the Trust Legal Provisions and are 54 provided without warranty as described in the Revised BSD License. 56 Table of Contents 58 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 59 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 6 60 3. Automatic Extended Route Optimization (AERO) . . . . . . . . 15 61 3.1. AERO Node Types . . . . . . . . . . . . . . . . . . . . . 15 62 3.2. The AERO Service over OMNI Links . . . . . . . . . . . . 16 63 3.2.1. AERO/OMNI Reference Model . . . . . . . . . . . . . . 16 64 3.2.2. Addressing and Node Identification . . . . . . . . . 20 65 3.2.3. AERO Routing System . . . . . . . . . . . . . . . . . 21 66 3.2.4. Segment Routing Topologies (SRTs) . . . . . . . . . . 23 67 3.2.5. Segment Routing For OMNI Link Selection . . . . . . . 23 68 3.3. OMNI Interface Characteristics . . . . . . . . . . . . . 24 69 3.4. OMNI Interface Initialization . . . . . . . . . . . . . . 26 70 3.4.1. AERO Proxy/Server and Relay Behavior . . . . . . . . 27 71 3.4.2. AERO Client Behavior . . . . . . . . . . . . . . . . 27 72 3.4.3. AERO Host Behavior . . . . . . . . . . . . . . . . . 28 73 3.4.4. AERO Gateway Behavior . . . . . . . . . . . . . . . . 28 74 3.5. OMNI Interface Neighbor Cache Maintenance . . . . . . . . 28 75 3.5.1. OMNI ND Messages . . . . . . . . . . . . . . . . . . 30 76 3.5.2. OMNI Neighbor Advertisement Message Flags . . . . . . 32 77 3.5.3. OMNI Neighbor Window Synchronization . . . . . . . . 33 78 3.6. OMNI Interface Encapsulation and Fragmentation . . . . . 33 79 3.7. OMNI Interface Decapsulation . . . . . . . . . . . . . . 36 80 3.8. OMNI Interface Data Origin Authentication . . . . . . . . 36 81 3.9. OMNI Interface MTU . . . . . . . . . . . . . . . . . . . 37 82 3.10. OMNI Interface Forwarding Algorithm . . . . . . . . . . . 37 83 3.10.1. Host Forwarding Algorithm . . . . . . . . . . . . . 39 84 3.10.2. Client Forwarding Algorithm . . . . . . . . . . . . 39 85 3.10.3. Proxy/Server and Relay Forwarding Algorithm . . . . 40 86 3.10.4. Gateway Forwarding Algorithm . . . . . . . . . . . . 43 87 3.11. OMNI Interface Error Handling . . . . . . . . . . . . . . 44 88 3.12. AERO Mobility Service Coordination . . . . . . . . . . . 47 89 3.12.1. AERO Service Model . . . . . . . . . . . . . . . . . 47 90 3.12.2. AERO Host and Client Behavior . . . . . . . . . . . 48 91 3.12.3. AERO Proxy/Server Behavior . . . . . . . . . . . . . 49 92 3.13. AERO Route Optimization . . . . . . . . . . . . . . . . . 56 93 3.13.1. Multilink Address Resolution . . . . . . . . . . . . 57 94 3.13.2. Multilink Route Optimization . . . . . . . . . . . . 61 95 3.13.3. Rapid Commit Route Optimization . . . . . . . . . . 73 96 3.13.4. Client/Gateway Route Optimization . . . . . . . . . 73 97 3.13.5. Client/Client Route Optimization . . . . . . . . . . 75 98 3.13.6. Client-to-Client OMNI Link Extension . . . . . . . . 77 99 3.13.7. Intra-ANET/ENET Route Optimization for AERO Peers . 78 100 3.14. Neighbor Unreachability Detection (NUD) . . . . . . . . . 78 101 3.15. Mobility Management and Quality of Service (QoS) . . . . 80 102 3.15.1. Mobility Update Messaging . . . . . . . . . . . . . 80 103 3.15.2. Announcing Link-Layer Information Changes . . . . . 81 104 3.15.3. Bringing New Links Into Service . . . . . . . . . . 82 105 3.15.4. Deactivating Existing Links . . . . . . . . . . . . 82 106 3.15.5. Moving Between Proxy/Servers . . . . . . . . . . . . 82 107 3.16. Multicast . . . . . . . . . . . . . . . . . . . . . . . . 83 108 3.16.1. Source-Specific Multicast (SSM) . . . . . . . . . . 84 109 3.16.2. Any-Source Multicast (ASM) . . . . . . . . . . . . . 85 110 3.16.3. Bi-Directional PIM (BIDIR-PIM) . . . . . . . . . . . 86 111 3.17. Operation over Multiple OMNI Links . . . . . . . . . . . 86 112 3.18. DNS Considerations . . . . . . . . . . . . . . . . . . . 87 113 3.19. Transition/Coexistence Considerations . . . . . . . . . . 87 114 3.20. Proxy/Server-Gateway Bidirectional Forwarding 115 Detection . . . . . . . . . . . . . . . . . . . . . . . 88 116 3.21. Time-Varying MNPs . . . . . . . . . . . . . . . . . . . . 88 117 4. Implementation Status . . . . . . . . . . . . . . . . . . . . 88 118 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 89 119 6. Security Considerations . . . . . . . . . . . . . . . . . . . 89 120 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 91 121 8. References . . . . . . . . . . . . . . . . . . . . . . . . . 93 122 8.1. Normative References . . . . . . . . . . . . . . . . . . 93 123 8.2. Informative References . . . . . . . . . . . . . . . . . 95 124 Appendix A. Non-Normative Considerations . . . . . . . . . . . . 102 125 A.1. Implementation Strategies for Route Optimization . . . . 103 126 A.2. Implicit Mobility Management . . . . . . . . . . . . . . 103 127 A.3. Direct Underlying Interfaces . . . . . . . . . . . . . . 104 128 A.4. AERO Critical Infrastructure Considerations . . . . . . . 104 129 A.5. AERO Server Failure Implications . . . . . . . . . . . . 105 130 A.6. AERO Client / Server Architecture . . . . . . . . . . . . 105 131 Appendix B. Change Log . . . . . . . . . . . . . . . . . . . . . 107 132 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 108 134 1. Introduction 136 Automatic Extended Route Optimization (AERO) fulfills the 137 requirements of Distributed Mobility Management (DMM) [RFC7333] and 138 route optimization [RFC5522] for aeronautical networking and other 139 network mobility use cases including intelligent transportation 140 systems and enterprise mobile device users. AERO is a secure 141 internetworking and mobility management service that employs the 142 Overlay Multilink Network Interface (OMNI) [I-D.templin-6man-omni] 143 Non-Broadcast, Multiple Access (NBMA) virtual link model. The OMNI 144 link is a virtual overlay configured over one or more concatenated 145 underlay Internetworks, and nodes on the link can exchange original 146 IP packets as single-hop neighbors. The OMNI Adaptation Layer (OAL) 147 supports multilink operation for increased reliability and path 148 optimization while providing fragmentation and reassembly services to 149 support improved performance and Maximum Transmission Unit (MTU) 150 diversity. This specification provides a mobility service 151 architecture companion to the OMNI specification. 153 The AERO service connects Hosts and Clients over Proxy/Servers and 154 Relays that are seen as OMNI link neighbors; AERO further includes 155 Gateways that interconnect diverse Internetworks as OMNI link 156 segments through OAL forwarding at a layer below IP. Each node's 157 OMNI interface uses an IPv6 link-local address format that supports 158 operation of the IPv6 Neighbor Discovery (IPv6 ND) protocol 159 [RFC4861]. A Client's OMNI interface can be configured over multiple 160 underlay interfaces, and therefore appears as a single interface with 161 multiple link-layer addresses. Each link-layer address is subject to 162 change due to mobility and/or multilink fluctuations, and link-layer 163 address changes are signaled by ND messaging the same as for any IPv6 164 link. 166 AERO provides a secure cloud-based service where mobile node Clients 167 may use Proxy/Servers acting as proxys and/or designated routers 168 while fixed nodes may use any Relay on the link for efficient 169 communications. Fixed nodes forward original IP packets destined to 170 other AERO nodes via the nearest Relay, which forwards them through 171 the cloud. Mobile node Clients discover shortest paths to OMNI link 172 neighbors through AERO route optimization. Both unicast and 173 multicast communications are supported, and Clients may efficiently 174 move between locations while maintaining continuous communications 175 with correspondents and without changing their IP Address. 177 AERO Gateways peer with Proxy/Servers in a secured private BGP 178 overlay routing instance to establish a Segment Routing Topology 179 (SRT) spanning tree over the underlay Internetworks of one or more 180 disjoint administrative domains concatenated as a single unified OMNI 181 link. Each OMNI link instance is characterized by the set of 182 Mobility Service Prefixes (MSPs) common to all mobile nodes. Relays 183 provide an optimal route from (fixed) correspondent nodes on underlay 184 Internetworks to (mobile or fixed) nodes on the OMNI link. To the 185 underlay Internetwork, the Relay is the source of a route to the MSP; 186 hence uplink traffic to mobile nodes is naturally routed to the 187 nearest Relay. 189 AERO can be used with OMNI links that span private-use Internetworks 190 and/or public Internetworks such as the global Internet. In both 191 cases, Clients may be located behind Network Address Translators 192 (NATs) on the path to their associated Proxy/Servers. A means for 193 robust traversal of NATs while avoiding "triangle routing" and 194 critical infrastructure traffic concentration is therefore provided. 196 AERO assumes the use of PIM Sparse Mode in support of multicast 197 communication. In support of Source Specific Multicast (SSM) when a 198 Mobile Node is the source, AERO route optimization ensures that a 199 shortest-path multicast tree is established with provisions for 200 mobility and multilink operation. In all other multicast scenarios 201 there are no AERO dependencies. 203 AERO provides a secure aeronautical internetworking service for both 204 manned and unmanned aircraft, where the aircraft is treated as a 205 mobile node that can connect an Internet of Things (IoT). AERO is 206 also applicable to a wide variety of other use cases. For example, 207 it can be used to coordinate the links of mobile nodes (e.g., 208 cellphones, tablets, laptop computers, etc.) that connect into a home 209 enterprise network via public access networks with VPN or non-VPN 210 services enabled according to the appropriate security model. AERO 211 can also be used to facilitate terrestrial vehicular and urban air 212 mobility (as well as pedestrian communication services) for future 213 intelligent transportation systems 214 [I-D.ietf-ipwave-vehicular-networking][I-D.templin-ipwave-uam-its]. 215 Other applicable use cases are also in scope. 217 Along with OMNI, AERO provides secured optimal routing support for 218 the "6M's" of modern Internetworking, including: 220 1. Multilink - a mobile node's ability to coordinate multiple 221 diverse underlay data links as a single logical unit (i.e., the 222 OMNI interface) to achieve the required communications 223 performance and reliability objectives. 225 2. Multinet - the ability to span the OMNI link over a segment 226 routing topology with multiple diverse administrative domain 227 network segments while maintaining seamless end-to-end 228 communications between mobile Clients and correspondents such as 229 air traffic controllers, fleet administrators, other mobile 230 Clients, etc. 232 3. Mobility - a mobile node's ability to change network points of 233 attachment (e.g., moving between wireless base stations) which 234 may result in an underlay interface address change, but without 235 disruptions to ongoing communication sessions with peers over the 236 OMNI link. 238 4. Multicast - the ability to send a single network transmission 239 that reaches multiple nodes belonging to the same interest group, 240 but without disturbing other nodes not subscribed to the interest 241 group. 243 5. Multihop - a mobile node vehicle-to-vehicle relaying capability 244 useful when multiple forwarding hops between vehicles may be 245 necessary to "reach back" to an infrastructure access point 246 connection to the OMNI link. 248 6. MTU assurance - the ability to deliver packets of various robust 249 sizes between peers without loss due to a link size restriction, 250 and to dynamically adjust packets sizes to achieve the optimal 251 performance for each independent traffic flow. 253 The following numbered sections present the AERO specification. The 254 appendices at the end of the document are non-normative. 256 2. Terminology 258 The terminology in the normative references applies; especially, the 259 terminology in the OMNI specification [I-D.templin-6man-omni] is used 260 extensively throughout. The following terms are defined within the 261 scope of this document: 263 IPv6 Neighbor Discovery (IPv6 ND) 264 a control message service for coordinating neighbor relationships 265 between nodes connected to a common link. AERO uses the IPv6 ND 266 messaging service specified in [RFC4861] in conjunction with the 267 OMNI extensions specified in [I-D.templin-6man-omni]. 269 IPv6 Prefix Delegation 270 a networking service for delegating IPv6 prefixes to nodes on the 271 link. The nominal service is DHCPv6 [RFC8415], however alternate 272 services (e.g., based on IPv6 ND messaging) are also in scope. A 273 minimal form of prefix delegation known as "prefix registration" 274 can be used if the Client knows its prefix in advance and can 275 represent it in the source address of an IPv6 ND message. 277 L2 278 The Data Link layer in the OSI network model. Also known as 279 "layer-2", "link-layer", "sub-IP layer", etc. 281 L3 282 The Network layer in the OSI network model. Also known as "layer- 283 3", "IP layer", etc. 285 Adaptation layer 286 A mid-layer that adapts L3 to a diverse collection of L2 underlay 287 interfaces and their encapsulations. (No layer number is 288 assigned, since numbering was an artifact of the legacy reference 289 model that need not carry forward in the modern architecture.) 290 The adaptation layer sees the upper layer as "L3" and sees all 291 lower layer encapsulations as "L2 encapsulations", which may 292 include UDP, IP and true link-layer (e.g., Ethernet, etc.) 293 headers. 295 Access Network (ANET) 296 a connected network region (e.g., an aviation radio access 297 network, satellite service provider network, cellular operator 298 network, WiFi network, etc.) that joins Clients to the Mobility 299 Service. Physical and/or data link level security is assumed, and 300 sometimes referred to as "protected spectrum". Private enterprise 301 networks and ground domain aviation service networks may provide 302 multiple secured IP hops between the Client's point of connection 303 and the nearest Proxy/Server. 305 Internetwork (INET) 306 a connected network region with a coherent IP addressing plan that 307 provides transit forwarding services between ANETs and OMNI nodes 308 that coordinate with the Mobility Service over unprotected media. 309 No physical and/or data link level security is assumed, therefore 310 security must be applied by upper layers. The global public 311 Internet itself is an example. 313 End-user Network (ENET) 314 a simple or complex "downstream" network that travels with the 315 Client as a single logical unit. The ENET could be as simple as a 316 single link connecting a single Host, or as complex as a large 317 network with many links, routers, bridges and Hosts. The ENET 318 could also provide an "upstream" link in a recursively-descending 319 chain of additional Clients and ENETs. In this way, an ENET of an 320 upstream Client is seen as the ANET of a downstream Client. 322 {A,I,E}NET interface 323 a node's attachment to a link in an {A,I,E}NET. 325 underlay network/interface 326 an ANET/INET/ENET network/interface over which an OMNI interface 327 is configured. The OMNI interface is seen as a L3 interface by 328 the IP layer, and the OMNI adaptation layer sees the underlay 329 interface as an L2 interface. The underlay interface either 330 connects directly to the physical communications media or 331 coordinates with another node where the physical media is hosted. 333 OMNI link 334 the same as defined in [I-D.templin-6man-omni]. The OMNI link 335 employs IPv6 encapsulation [RFC2473] to traverse intermediate 336 nodes in a spanning tree over underlay network segments the same 337 as a bridged campus LAN. AERO nodes on the OMNI link appear as 338 single-hop neighbors at the network layer even though they may be 339 separated by many underlay network hops; AERO nodes can employ 340 Segment Routing [RFC8402] to navigate between different OMNI 341 links, and/or to cause packets to visit selected waypoints within 342 the same OMNI link. 344 OMNI Adaptation Layer (OAL) 345 an OMNI interface sublayer service that encapsulates original IP 346 packets admitted into the interface in an IPv6 header and/or 347 subjects them to fragmentation and reassembly. The OAL is also 348 responsible for generating MTU-related control messages as 349 necessary, and for providing addressing context for spanning 350 multiple segments of an L2-extended OMNI link. 352 OMNI Interface 353 a node's attachment to an OMNI link (i.e., the same as defined in 354 [I-D.templin-6man-omni]). Since OMNI interface addresses are 355 managed for uniqueness, OMNI interfaces do not require Duplicate 356 Address Detection (DAD) and therefore set the administrative 357 variable 'DupAddrDetectTransmits' to zero [RFC4862]. 359 (network) partition 360 frequently, underlay networks such as large corporate enterprise 361 networks are sub-divided internally into separate isolated 362 partitions (a technique also known as "network segmentation"). 363 Each partition is fully connected internally but disconnected from 364 other partitions, and there is no requirement that separate 365 partitions maintain consistent Internet Protocol and/or addressing 366 plans. (Each partition is seen as a separate OMNI link segment as 367 discussed throughout this document.) 369 L2 encapsulation 370 the OAL encapsulation of a packet in an outer header or headers 371 that can be routed within the scope of the local {A,I,E}NET 372 partition. Common L2 encapsulation combinations include UDP/IP/ 373 Ethernet, etc. 375 L2 address(es) 376 the addresses that appear in the OAL L2 encapsulations for an 377 underlay interface. 379 INADDR 380 the UDP/IP addresses that appear in an L2 address. 382 original IP packet 383 a whole IP packet or fragment admitted into the OMNI interface by 384 the network layer prior to OAL encapsulation and fragmentation, or 385 an IP packet delivered to the network layer by the OMNI interface 386 following OAL decapsulation and reassembly. 388 OAL packet 389 an original IP packet encapsulated in an OAL IPv6 header before 390 OAL fragmentation, or following OAL reassembly. 392 OAL fragment 393 a portion of an OAL packet following fragmentation but prior to L2 394 encapsulation, or following L2 decapsulation but prior to OAL 395 reassembly. 397 (OAL) atomic fragment 398 an OAL packet that can be forwarded without fragmentation, but 399 still includes a Fragment Header with a valid Identification value 400 and with Fragment Offset and More Fragments both set to 0. 402 (OAL) carrier packet 403 an encapsulated OAL fragment following L2 encapsulation or prior 404 to L2 decapsulation. OAL sources and destinations exchange 405 carrier packets over underlay interfaces, and may be separated by 406 one or more OAL intermediate nodes. OAL intermediate nodes re- 407 encapsulate carrier packets during forwarding by removing the L2 408 headers of the previous hop underlay network and replacing them 409 with new L2 headers for the next hop underlay network. 411 OAL source 412 an OMNI interface acts as an OAL source when it encapsulates 413 original IP packets to form OAL packets, then performs OAL 414 fragmentation and L2 encapsulation to create carrier packets. 416 OAL destination 417 an OMNI interface acts as an OAL destination when it decapsulates 418 carrier packets, then performs OAL reassembly and decapsulation to 419 derive the original IP packet. 421 OAL intermediate node 422 an OMNI interface acts as an OAL intermediate node when it removes 423 the L2 headers of carrier packets received from a previous hop, 424 then re-encapsulates the carrier packets in new L2 headers and 425 forwards them to the next hop. OAL intermediate nodes decrement 426 the Hop Limit of the OAL IPv6 header during re-encapsulation, and 427 discard the packet if the Hop Limit reaches 0. OAL intermediate 428 nodes do not decrement the Hop Limit/TTL of the original IP 429 packet. 431 Mobility Service Prefix (MSP) 432 an aggregated IP Global Unicast Address (GUA) prefix (e.g., 433 2001:db8::/32, 192.0.2.0/24, etc.) assigned to the OMNI link and 434 from which more-specific Mobile Network Prefixes (MNPs) are 435 delegated. OMNI link administrators typically obtain MSPs from an 436 Internet address registry, however private-use prefixes can 437 alternatively be used subject to certain limitations (see: 438 [I-D.templin-6man-omni]). OMNI links that connect to the global 439 Internet advertise their MSPs to their interdomain routing peers. 441 Mobile Network Prefix (MNP) 442 a longer IP prefix delegated from an MSP (e.g., 443 2001:db8:1000:2000::/56, 192.0.2.8/30, etc.) and delegated to an 444 AERO Client or Relay. 446 Interface Identifier (IID) 447 the least significant 64 bits of an IPv6 address, as specified in 448 the IPv6 addressing architecture [RFC4291]. 450 Link Local Address (LLA) 451 an IPv6 address beginning with fe80::/64 per the IPv6 addressing 452 architecture [RFC4291] and with either a 64-bit MNP (LLA-MNP) or a 453 56-bit random value (LLA-RND) encoded in the IID as specified in 454 [I-D.templin-6man-omni]. 456 Unique Local Address (ULA) 457 an IPv6 address beginning with fd00::/8 followed by a 40-bit 458 Global ID followed by a 16-bit Subnet ID per [RFC4193] and with 459 either a 64-bit MNP (ULA-MNP) or a 56-bit random value (ULA-RND) 460 encoded in the IID as specified in [I-D.templin-6man-omni]. (Note 461 that [RFC4193] specifies a second form of ULAs based on the prefix 462 fc00::/8, which are referred to as "ULA-C" throughout this 463 document to distinguish them from the ULAs defined here.) 465 Temporary Local Address (TLA) 466 a ULA beginning with fd00::/16 followed by a 48-bit randomly- 467 initialized value followed by an MNP-based (TLA-MNP) or random 468 (TLA-RND) IID as specified in [I-D.templin-6man-omni]. Clients 469 use TLAs to bootstrap autoconfiguration in the presence of OMNI 470 link infrastructure or for sustained communications in the absence 471 of infrastructure. (Note that in some environments Clients can 472 instead use a (Hierarchical) Host Identity Tag ((H)HIT) instead of 473 a TLA - see: [I-D.templin-6man-omni].) 475 eXtended Local Address (XLA) 476 a TLA beginning with fd00::/64 followed by an MNP-based (XLA-MNP) 477 or random (XLA-RND) IID as specified in [I-D.templin-6man-omni]. 478 An XLA is simply a TLA with an all-0 48-bit value following 479 fd00::/16, and can be used to supply a "wildcard match" for IPv6 480 ND cache entries, a routing table entry for the OMNI link routing 481 system, etc. (Note that XLAs can also be statelessly formed from 482 LLAs (and vice-versa) simply by inverting prefix bits 7 and 8.) 484 AERO node 485 a node that is connected to an OMNI link and participates in the 486 AERO internetworking and mobility service. 488 AERO Host ("Host") 489 an AERO node that configures an OMNI interface over an ENET 490 underlying interface serviced by an upstream Client. The Host 491 does not assign an LLA or ULA to the OMNI interface, but instead 492 assigns the address taken from the ENET underlying interface. (As 493 an implementation matter, the Host may instead configure the "OMNI 494 interface" as a virtual sublayer of the underlay interface 495 itself.) When an AERO host forwards an original IP packet to 496 another AERO node on the same ENET, it uses simple IP-in-IP 497 encapsulation without including an OAL encapsulation header. The 498 Host is therefore an OMNI link termination endpoint. 500 AERO Client ("Client") 501 an AERO node that configures an OMNI interface over one or more 502 underlay interfaces and requests MNP delegation/registration 503 service from AERO Proxy/Servers. The Client assigns an XLA-MNP 504 (as well as Proxy/Server-specific ULA-MNPs) to the OMNI interface 505 for use in IPv6 ND exchanges with other AERO nodes and forwards 506 original IP packets to correspondents according to OMNI interface 507 neighbor cache state. The Client coordinates with Proxy/Servers 508 and/or other Clients over upstream ANET/INET interfaces and may 509 also provide Proxy/Server services for Hosts and/or other Clients 510 over downstream ENET interfaces. 512 AERO Proxy/Server ("Proxy/Server") 513 a node that provides a proxying service between AERO Clients and 514 external peers on its Client-facing ANET interfaces (i.e., in the 515 same fashion as for an enterprise network proxy) as well as 516 designated router services for coordination with correspondents on 517 its INET-facing interfaces. (Proxy/Servers in the open INET 518 instead configure only a single INET interface and no ANET 519 interfaces.) The Proxy/Server configures an OMNI interface and 520 assigns a ULA-RND to support the operation of IPv6 ND services, 521 while advertising any associated MNPs for which it is acting as a 522 hub via BGP peerings with AERO Gateways. 524 AERO Relay ("Relay") 525 a Proxy/Server that provides forwarding services between nodes 526 reached via the OMNI link and correspondents on other links/ 527 networks. AERO Relays configure an OMNI interface, assign a ULA- 528 RND and maintain BGP peerings with Gateways the same as Proxy/ 529 Servers and run a dynamic routing protocol to discover any non-MNP 530 IP GUA routes in service on other links/networks. The Relay 531 advertises the MSP(s) to its other links/networks, and 532 redistributes routes discovered on other links/networks into the 533 OMNI link BGP routing system the same as for Proxy/Servers. 534 (Relays that connect to major Internetworks such as the global 535 IPv6 or IPv4 Internet can also be configured to advertise 536 "default" routes into the OMNI link BGP routing system.) 538 AERO Gateway ("Gateway") 539 a BGP hub autonomous system node that also provides OAL forwarding 540 services for nodes on an OMNI link. Gateways forward carrier 541 packets between OMNI link segments as OAL intermediate nodes while 542 decrementing the OAL IPv6 header Hop Limit but without 543 decrementing the network layer IP TTL/Hop Limit. Gateways peer 544 with Proxy/Servers and other Gateways to form an IPv6-based OAL 545 spanning tree over all OMNI link segments and to discover the set 546 of all MNP and non-MNP prefixes in service. Gateways process 547 carrier packets received over the secured spanning tree that are 548 addressed to themselves, while forwarding all other carrier 549 packets to the next hop also via the secured spanning tree. 550 Gateways forward carrier packets received over the unsecured 551 spanning tree to the next hop either via the unsecured spanning 552 tree or via direct encapsulation if the next hop is on the same 553 OMNI link segment. 555 First-Hop Segment (FHS) Proxy/Server 556 a Proxy/Server for a source Client's underlay interface that 557 forwards the Client's packets into the segment routing topology. 558 FHS Proxy/Servers also act as intermediate forwarding nodes to 559 facilitate RS/RA exchanges between a Client and its Hub Proxy/ 560 Server. 562 Hub Proxy/Server 563 a single Proxy/Server selected by a Client that injects the 564 Client's MNP into the BGP routing system and provides a designated 565 router service for all of the Client's underlay interfaces. 566 Clients often select the first FHS Proxy/Server they coordinate 567 with to serve in the Hub role (as all FHS Proxy/Servers are 568 equally capable candidates to serve in that capacity), however the 569 Client can also select any available Proxy/Server for the OMNI 570 link (as there is no requirement that the Hub must also be one of 571 the Client's FHS Proxy/Servers). 573 Last-Hop Segment (LHS) Proxy/Server 574 a Proxy/Server for an underlay interface of the target Client that 575 forwards packets received from the segment routing topology to the 576 target Client over that interface. 578 Segment Routing Topology (SRT) 579 a Multinet OMNI link forwarding region between FHS and LHS Proxy/ 580 Servers. FHS/LHS Proxy/Servers and SRT Gateways span the OMNI 581 link on behalf of source/target Client pairs. The SRT maintains a 582 spanning tree established through BGP peerings between Gateways 583 and Proxy/Servers. Each SRT segment includes Gateways in a "hub" 584 and Proxy/Servers in "spokes", while adjacent segments are 585 interconnected by Gateway-Gateway peerings. The BGP peerings are 586 configured over both secured and unsecured underlay network paths 587 such that a secured spanning tree is available for critical 588 control messages while other messages can use the unsecured 589 spanning tree. 591 Mobile Node (MN) 592 an AERO Client and all of its downstream-attached networks that 593 move together as a single unit, i.e., an end system that connects 594 an Internet of Things. 596 Mobile Router (MR) 597 a MN's on-board router that forwards original IP packets between 598 any downstream-attached networks and the OMNI link. The MR is the 599 MN entity that hosts the AERO Client. 601 Route Optimization Source (ROS) 602 the AERO node nearest the source that initiates route 603 optimization. The ROS may be a FHS Proxy/Server or Relay for the 604 source, or may be the source Client itself. 606 Route Optimization responder (ROR) 607 the AERO node that responds to route optimization requests on 608 behalf of the target. The ROR may be either the target MNP Client 609 itself, the Client's current Hub Proxy/Server or a Relay for a 610 non-MNP target. 612 Potential Router List (PRL) 613 a geographically and/or topologically referenced list of addresses 614 of all Proxy/Servers within the same OMNI link. Each OMNI link 615 has its own PRL. 617 Distributed Mobility Management (DMM) 618 a BGP-based overlay routing service coordinated by Proxy/Servers 619 and Gateways that tracks all Proxy/Server-to-Client associations. 621 Mobility Service (MS) 622 the collective set of all Proxy/Servers, Gateways and Relays that 623 provide the AERO Service to Clients. 625 Multilink Forwarding Information Base (MFIB) 626 A forwarding table on each AERO/OMNI source, destination and 627 intermediate node that includes Multilink Forwarding Vectors (MFV) 628 with both next hop forwarding instructions and context for 629 reconstructing compressed headers for specific underlay interface 630 pairs used to communicate with peers. 632 Multilink Forwarding Vector (MFV) 633 An MFIB entry that includes soft state for each underlay interface 634 pairwise communication session between peer OMNI nodes. MFVs are 635 identified by both a next-hop and previous-hop MFV Index (MFVI), 636 with the next-hop established based on an IPv6 ND solicitation and 637 the previous hop established based on the solicited IPv6 ND 638 advertisement response. 640 Multilink Forwarding Vector Index (MVFI) 641 A 4 octet value selected by an AERO/OMNI node when it creates an 642 MFV, then advertises to either a next-hop or previous-hop. AERO/ 643 OMNI intermediate nodes assign two distinct local MFVIs for each 644 MFV and advertise one to the next-hop and the other to the 645 previous-hop. AERO/OMNI end systems assign and advertise a single 646 MFVI. AERO/OMNI nodes also discover the remote MFVIs advertised 647 by other nodes that indicate a value the remote node is willing to 648 accept. 650 Throughout the document, the simple terms "Host", "Client", "Proxy/ 651 Server", "Gateway" and "Relay" refer to "AERO Host", "AERO Client", 652 "AERO Proxy/Server", "AERO Gateway" and "AERO Relay", respectively. 653 Capitalization is used to distinguish these terms from other common 654 Internetworking uses in which they appear without capitalization. 656 The terminology of IPv6 ND [RFC4861], DHCPv6 [RFC8415] and OMNI 657 [I-D.templin-6man-omni] (including the names of node variables, 658 messages and protocol constants) is used throughout this document. 659 The terms "All-Routers multicast", "All-Nodes multicast", "Solicited- 660 Node multicast" and "Subnet-Router anycast" are defined in [RFC4291]. 661 Also, the term "IP" is used to generically refer to either Internet 662 Protocol version, i.e., IPv4 [RFC0791] or IPv6 [RFC8200]. 664 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 665 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 666 "OPTIONAL" in this document are to be interpreted as described in BCP 667 14 [RFC2119][RFC8174] when, and only when, they appear in all 668 capitals, as shown here. 670 3. Automatic Extended Route Optimization (AERO) 672 The following sections specify the operation of IP over OMNI links 673 using the AERO service: 675 3.1. AERO Node Types 677 AERO Hosts configure an OMNI interface over an underlay interface 678 connected to a Client's ENET and coordinate with both other AERO 679 Hosts and Clients over the ENET. As an implementation matter, the 680 Host either assigns the same (MNP-based) IP address from the underlay 681 interface to the OMNI interface, or configures the "OMNI interface" 682 as a virtual sublayer of the underlay interface itself. AERO Hosts 683 treat the ENET as an ANET, and treat the upstream Client for the ENET 684 as a Proxy/Server. AERO Hosts are seen as OMNI link termination 685 endpoints. 687 AERO Clients can be deployed as fixed infrastructure nodes close to 688 end systems, or as Mobile Nodes (MNs) that can change their network 689 attachment points dynamically. AERO Clients configure OMNI 690 interfaces over underlay interfaces with addresses that may change 691 due to mobility. AERO Clients register their Mobile Network Prefixes 692 (MNPs) with the AERO service, and distribute the MNPs to ENETs (which 693 may connect AERO Hosts and other Clients). AERO Clients provide 694 Proxy/Server-like services for Hosts and other Clients on downstream- 695 attached ENETs. 697 AERO Gateways, Proxy/Servers and Relays are critical infrastructure 698 elements in fixed (i.e., non-mobile) INET deployments and hence have 699 permanent and unchanging INET addresses. Together, they constitute 700 the AERO service which provides an OMNI link virtual overlay for 701 connecting AERO Clients and Hosts. AERO Gateways (together with 702 Proxy/Servers) provide the secured backbone supporting infrastructure 703 for a Segment Routing Topology (SRT) spanning tree for the OMNI link. 705 AERO Gateways forward carrier packets both within the same SRT 706 segment and between disjoint SRT segments based on an IPv6 707 encapsulation mid-layer known as the OMNI Adaptation Layer (OAL) 708 [I-D.templin-6man-omni]. The OMNI interface and OAL provide a 709 virtual bridging service, since the inner IP TTL/Hop Limit is not 710 decremented. Each Gateway also peers with Proxy/Servers and other 711 Gateways in a dynamic routing protocol instance to provide a 712 Distributed Mobility Management (DMM) service for the list of active 713 MNPs (see Section 3.2.3). Gateways assign one or more Mobility 714 Service Prefixes (MSPs) to the OMNI link and configure secured 715 tunnels with Proxy/Servers, Relays and other Gateways; they further 716 maintain forwarding table entries for each MNP or non-MNP prefix in 717 service on the OMNI link. 719 AERO Proxy/Servers distributed across one or more SRT segments 720 provide default forwarding and mobility/multilink services for AERO 721 Client mobile nodes. Each Proxy/Server also peers with Gateways in a 722 dynamic routing protocol instance to advertise its list of associated 723 MNPs (see Section 3.2.3). Hub Proxy/Servers provide prefix 724 delegation/registration services and track the mobility/multilink 725 profiles of each of their associated Clients, where each delegated 726 prefix becomes an MNP taken from an MSP. Proxy/Servers at ANET/INET 727 boundaries provide a forwarding service for ANET Clients and Hosts to 728 communicate with peers in external INETs, while Proxy/Servers in the 729 open INET provide an authentication service for INET Client IPv6 ND 730 messages but only a secondary forwarding service when the Client 731 cannot forward directly to a peer or Gateway. Source Clients 732 securely coordinate with target Clients by sending control messages 733 via a First-Hop Segment (FHS) Proxy/Server which forwards them over 734 the SRT spanning tree to a Last-Hop Segment (LHS) Proxy/Server which 735 finally forwards them to the target. 737 AERO Relays are Proxy/Servers that provide forwarding services to 738 exchange original IP packets between the OMNI link and nodes on other 739 links/networks. Relays run a dynamic routing protocol to discover 740 any non-MNP prefixes in service on other links/networks, and Relays 741 that connect to larger Internetworks (such as the Internet) may 742 originate default routes. The Relay redistributes OMNI link MSP(s) 743 into other links/networks, and redistributes non-MNP prefixes via 744 OMNI link Gateway BGP peerings. 746 3.2. The AERO Service over OMNI Links 748 3.2.1. AERO/OMNI Reference Model 750 Figure 1 presents the basic OMNI link reference model: 752 +-----------------+ 753 | AERO Gateway G1 | 754 | Nbr: S1, S2, P1 | 755 |(X1->S1; X2->S2) | 756 | MSP M1 | 757 +--------+--------+ 758 +--------------+ | +--------------+ 759 | AERO P/S S1 | | | AERO P/S S2 | 760 | Nbr: C1, G1 | | | Nbr: C2, G1 | 761 | default->G1 | | | default->G1 | 762 | X1->C1 | | | X2->C2 | 763 +-------+------+ | +------+-------+ 764 | OMNI link | | 765 X===+===+==================+===================+===+===X 766 | | 767 +-----+--------+ +--------+-----+ 768 |AERO Client C1| |AERO Client C2| 769 | Nbr: S1 | | Nbr: S2 | 770 | default->S1 | | default->S2 | 771 | MNP X1 | | MNP X2 | 772 +------+-------+ +-----+--------+ 773 | | 774 .-. .-. 775 ,-( _)-. +-------+ +-------+ ,-( _)-. 776 .-(_ IP )-. | AERO | | AERO | .-(_ IP )-. 777 (__ ENET )--|Host H1| |Host H2|--(__ ENET ) 778 `-(______)-' +-------+ +-------+ `-(______)-' 780 Figure 1: AERO/OMNI Reference Model 782 In this model: 784 * the OMNI link is an overlay network service configured over one or 785 more underlay SRT segments which may be managed by different 786 administrative authorities and have incompatible protocols and/or 787 addressing plans. 789 * AERO Gateway G1 aggregates Mobility Service Prefix (MSP) M1, 790 discovers Mobile Network Prefixes (MNPs) X* and advertises the MSP 791 via BGP peerings over secured tunnels to Proxy/Servers (S1, S2). 792 Gateways provide the backbone for an SRT spanning tree for the 793 OMNI link. 795 * AERO Proxy/Servers S1 and S2 configure secured tunnels with 796 Gateway G1 and also provide mobility, multilink, multicast and 797 default router services for the MNPs of their associated Clients 798 C1 and C2. (Proxy/Servers that act as Relays can also advertise 799 non-MNP routes for non-mobile correspondent nodes the same as for 800 MNP Clients.) 802 * AERO Clients C1 and C2 associate with Proxy/Servers S1 and S2, 803 respectively. They receive MNP delegations X1 and X2, and also 804 act as default routers for their associated physical or internal 805 virtual ENETs. 807 * AERO Hosts H1 and H2 attach to the ENETs served by Clients C1 and 808 C2, respectively. 810 An OMNI link configured over a single underlay network appears as a 811 single unified link with a consistent addressing plan; all nodes on 812 the link can exchange carrier packets via simple L2 encapsulation 813 (i.e., following any necessary NAT traversal) since the underlay is 814 connected. In common practice, however, OMNI links are often 815 configured over an SRT spanning tree that bridges multiple distinct 816 underlay network segments managed under different administrative 817 authorities (e.g., as for worldwide aviation service providers such 818 as ARINC, SITA, Inmarsat, etc.). Individual underlay networks may 819 also be partitioned internally, in which case each internal partition 820 appears as a separate segment. 822 The addressing plan of each SRT segment is consistent internally but 823 will often bear no relation to the addressing plans of other 824 segments. Each segment is also likely to be separated from others by 825 network security devices (e.g., firewalls, proxys, packet filtering 826 gateways, etc.), and disjoint segments often have no common physical 827 link connections. Therefore, nodes can only be assured of exchanging 828 carrier packets directly with correspondents in the same segment, and 829 not with those in other segments. The only means for joining the 830 segments therefore is through inter-domain peerings between AERO 831 Gateways. 833 The OMNI link spans multiple SRT segments using the OMNI Adaptation 834 Layer (OAL) [I-D.templin-6man-omni] to provide the network layer with 835 a virtual abstraction similar to a bridged campus LAN. The OAL is an 836 OMNI interface sublayer that inserts a mid-layer IPv6 encapsulation 837 header for inter-segment forwarding (i.e., bridging) without 838 decrementing the network-layer TTL/Hop Limit of the original IP 839 packet. An example OMNI link SRT is shown in Figure 2: 841 . . . . . . . . . . . . . . . . . . . . . . . 842 . . 843 . .-(::::::::) . 844 . .-(::::::::::::)-. +-+ . 845 . (:::: Segment A :::)--|G|---+ . 846 . `-(::::::::::::)-' +-+ | . 847 . `-(::::::)-' | . 848 . | . 849 . .-(::::::::) | . 850 . .-(::::::::::::)-. +-+ | . 851 . (:::: Segment B :::)--|G|---+ . 852 . `-(::::::::::::)-' +-+ | . 853 . `-(::::::)-' | . 854 . | . 855 . .-(::::::::) | . 856 . .-(::::::::::::)-. +-+ | . 857 . (:::: Segment C :::)--|G|---+ . 858 . `-(::::::::::::)-' +-+ | . 859 . `-(::::::)-' | . 860 . | . 861 . ..(etc).. x . 862 . . 863 . . 864 . <- Segment Routing Topology (SRT) -> . 865 . (Spanned by OMNI Link) . 866 . . . . . . . . . . . . . .. . . . . . . . . 868 Figure 2: OMNI Link Segment Routing Topology (SRT) 870 Gateway, Proxy/Server and Relay OMNI interfaces are configured over 871 both secured tunnels and open INET underlay interfaces within their 872 respective SRT segments. Within each segment, Gateways configure 873 "hub-and-spokes" BGP peerings with Proxy/Servers and Relays as 874 "spokes". Adjacent SRT segments are joined by Gateway-to-Gateway 875 peerings to collectively form a spanning tree over the entire SRT. 876 The "secured" spanning tree supports authentication and integrity for 877 critical control plane messages. The "unsecured" spanning tree 878 conveys ordinary carrier packets without security codes and that must 879 be treated by destinations according to data origin authentication 880 procedures. AERO nodes can employ route optimization to cause 881 carrier packets to take more direct paths between OMNI link neighbors 882 without having to follow strict spanning tree paths. 884 The AERO Multinet service concatenates SRT segments to form larger 885 networks through Gateway-to-Gateway peerings as originally described 886 in the "Catenet Model for Internetworking" [IEN48]; especially 887 Figure 2 follows directly from the illustrations in [IEN48-2]. The 888 Catenet model inspired the global public Internet as it is known 889 today, while AERO applies the Catenet concepts to provide true 890 Multinet services for the future architecture. 892 3.2.2. Addressing and Node Identification 894 AERO nodes on OMNI links use the Link-Local Address (LLA) prefix 895 fe80::/64 [RFC4291] to assign LLAs to the OMNI interface in a latent 896 state and do not employ LLAs for any operational purposes (instead, 897 the LLAs are assigned solely to satisfy the requirements of 898 [RFC4861]). AERO Clients configure LLAs constructed from MNPs (i.e., 899 "LLA-MNPs") while AERO infrastructure nodes construct LLAs based on 900 56-bit random values ("LLA-RNDs") per [I-D.templin-6man-omni]. Non- 901 MNP routes are also represented the same as for MNPs, but may include 902 a prefix that is not properly covered by an MSP. 904 AERO nodes also use the Unique Local Address (ULA) prefix fd00::/8 905 followed by a pseudo-random 40-bit Global ID to form the prefix 906 {ULA}::/48, then include a 16-bit Subnet ID '*' to form the prefix 907 {ULA*}::/64 [RFC4291]. The AERO node then uses the prefix 908 {ULA*}::/64 to form "ULA-MNPs" or "ULA-RNDs" as specified in 909 [I-D.templin-6man-omni] to support OAL addressing. (The prefix 910 {ULA*}::/64 appearing alone and with no suffix represents "default" 911 for that prefix.) 913 AERO Clients also use Temporary Local Addresses (TLAs) and eXtended 914 Local Addresses (XLAs) constructed per [I-D.templin-6man-omni], where 915 TLAs are distinguished from ordinary ULAs based on the prefix 916 fd00::/16 and XLAs are distinguished from ULAs/TLAs based on the 917 prefix fd00::/64. Clients use TLA-RNDs only in initial control 918 message exchanges until a stable MNP is assigned, but may sometimes 919 also use them for sustained communications within a local routing 920 region. AERO nodes use XLA-MNPs to provide forwarding information 921 for the global routing table as well as IPv6 ND message and OAL 922 addressing information. 924 AERO MSPs, MNPs and non-MNP routes are typically based on Global 925 Unicast Addresses (GUAs), but in some cases may be based on IPv4 926 private addresses [RFC1918] or IPv6 ULA-D's [RFC4193]. A GUA block 927 is also reserved for OMNI link anycast purposes. See 928 [I-D.templin-6man-omni] for a full specification of LLAs, ULAs, TLAs, 929 XLAs, GUAs and anycast addresses used by AERO nodes on OMNI links. 931 Finally, AERO Clients and Proxy/Servers configure node identification 932 values as specified in [I-D.templin-6man-omni]. 934 3.2.3. AERO Routing System 936 The AERO routing system comprises a private Border Gateway Protocol 937 (BGP) [RFC4271] service coordinated between Gateways and Proxy/ 938 Servers (Relays also engage in the routing system as simplified 939 Proxy/Servers). The service supports carrier packet forwarding at a 940 layer below IP and does not interact with the public Internet BGP 941 routing system, but supports redistribution of information for other 942 links and networks connected by Relays. 944 In a reference deployment, each Proxy/Server is configured as an 945 Autonomous System Border Router (ASBR) for a stub Autonomous System 946 (AS) using a 32-bit AS Number (ASN) [RFC4271] that is unique within 947 the BGP instance, and each Proxy/Server further uses eBGP to peer 948 with one or more Gateways but does not peer with other Proxy/Servers. 949 Each SRT segment in the OMNI link must include one or more Gateways 950 in a "hub" AS, which peer with the Proxy/Servers within that segment 951 as "spoke" ASes. All Gateways within the same segment are members of 952 the same hub AS, and use iBGP to maintain a consistent view of all 953 active routes currently in service. The Gateways of different 954 segments peer with one another using eBGP. 956 Gateways maintain forwarding table entries only for ULA prefixes for 957 infrastrucutre elements and XLA-MNPs corresponding to MNP and non-MNP 958 routes that are currently active; Gateways also maintain black-hole 959 routes for the OMNI link MSPs so that carrier packets destined to 960 non-existent more-specific routes are dropped with a Destination 961 Unreachable message returned. In this way, Proxy/Servers and Relays 962 have only partial topology knowledge (i.e., they only maintain 963 routing information for their directly associated Clients and non- 964 AERO links) and they forward all other carrier packets to Gateways 965 which have full topology knowledge. 967 Each OMNI link segment assigns a unique sub-prefix of {ULA}::/48 968 known as the "SRT prefix". For example, a first segment could assign 969 {ULA}:1000::/56, a second could assign {ULA}:2000::/56, a third could 970 assign {ULA}:3000::/56, etc. Within each segment, each Proxy/Server 971 configures a ULA-RND within the segment's SRT prefix with a 56-bit 972 random value in the interface identifier as specified in 973 [I-D.templin-6man-omni]. 975 The administrative authorities for each segment must therefore 976 coordinate to assure mutually-exclusive ULA prefix assignments, but 977 internal provisioning of ULAs is an independent local consideration 978 for each administrative authority. For each ULA prefix, the 979 Gateway(s) that connect that segment assign the all-zero's address of 980 the prefix as a Subnet Router Anycast address. For example, the 981 Subnet Router Anycast address for {ULA}:1023::/64 is simply 982 {ULA}:1023::/64. 984 ULA prefixes are statically represented in Gateway forwarding tables. 985 Gateways join multiple SRT segments into a unified OMNI link over 986 multiple diverse network administrative domains. They support a 987 virtual bridging service by first establishing forwarding table 988 entries for their ULA prefixes either via standard BGP routing or 989 static routes. For example, if three Gateways ('A', 'B' and 'C') 990 from different segments serviced {ULA}:1000::/60, {ULA}:2000::/56 and 991 {ULA}:3000::/56 respectively, then the forwarding tables in each 992 Gateway appear as follows: 994 A: {ULA}:1000::/56->local, {ULA}:2000::/56->B, {ULA}:3000::/56->C 996 B: {ULA}:1000::/56->A, {ULA}:2000::/56->local, {ULA}:3000::/56->C 998 C: {ULA}:1000::/56->A, {ULA}:2000::/56->B, {ULA}:3000::/56->local 1000 These forwarding table entries rarely change, since they correspond 1001 to fixed infrastructure elements in their respective segments. 1003 MNP (and non-MNP) ULA routes are instead dynamically advertised in 1004 the AERO routing system by Proxy/Servers and Relays that provide 1005 service for their corresponding MNPs. The routes are advertised as 1006 XLA-MNP prefixes, i.e., as fd00::{MNP} (see: 1007 [I-D.templin-6man-omni]). For example, if three Proxy/Servers ('D', 1008 'E' and 'F') service the MNPs 2001:db8:1000:2000::/56, 1009 2001:db8:3000:4000::/56 and 2001:db8:5000:6000::/56 then the routing 1010 system would include: 1012 D: fd00::2001:db8:1000:2000/120 1014 E: fd00::2001:db8:3000:4000/120 1016 F: fd00::2001:db8:5000:6000/120 1018 A full discussion of the BGP-based routing system used by AERO is 1019 found in [I-D.ietf-rtgwg-atn-bgp]. 1021 3.2.4. Segment Routing Topologies (SRTs) 1023 The distinct {ULA}::/48 prefixes in an OMNI link domain identify 1024 distinct Segment Routing Topologies (SRTs). Each SRT is a mutually- 1025 exclusive OMNI link overlay instance using a distinct set of ULAs, 1026 and emulates a bridged campus LAN service for the OMNI link. In some 1027 cases (e.g., when redundant topologies are needed for fault tolerance 1028 and reliability) it may be beneficial to deploy multiple SRTs that 1029 act as independent overlay instances. A communication failure in one 1030 instance therefore will not affect communications in other instances. 1032 Each SRT is identified by a distinct value in the 40-bit ULA Global 1033 ID field and assigns an OMNI IPv6 anycast address used for OMNI 1034 interface determination in Safety-Based Multilink (SBM) as discussed 1035 in [I-D.templin-6man-omni]. Each OMNI interface further applies 1036 Performance-Based Multilink (PBM) internally. 1038 The Gateways and Proxy/Servers of each independent SRT engage in BGP 1039 peerings to form a spanning tree with the Gateways in non-leaf nodes 1040 and the Proxy/Servers in leaf nodes. The spanning tree is configured 1041 over both secured and unsecured underlay network paths. The secured 1042 spanning tree is used to convey secured control messages between 1043 Proxy/Servers and Gateways, while the unsecured spanning tree 1044 forwards data messages and/or unsecured control messages. 1046 Each SRT segment is identified by a unique ULA prefix used by all 1047 Proxy/Servers and Gateways in the segment. Each AERO node must 1048 therefore discover an SRT prefix that correspondents can use to 1049 determine the correct segment, and must publish the SRT prefix in 1050 IPv6 ND messages. 1052 Note: The distinct ULA prefixes in an OMNI link domain can be carried 1053 either in a common BGP routing protocol instance for all OMNI links 1054 or in distinct BGP routing protocol instances for different OMNI 1055 links. In some SBM environments, such separation may be necessary to 1056 ensure that distinct OMNI links do not include any common 1057 infrastructure elements as single points of failure. In other 1058 environments, carrying the ULAs of multiple OMNI links within a 1059 common routing system may be acceptable. 1061 3.2.5. Segment Routing For OMNI Link Selection 1063 Original IPv6 sources can direct IPv6 packets to an AERO node by 1064 including a standard IPv6 Segment Routing Header (SRH) [RFC8754] with 1065 the OMNI IPv6 anycast address for the selected OMNI link as either 1066 the IPv6 destination or as an intermediate hop within the SRH. This 1067 allows the original source to determine the specific OMNI link SRT an 1068 original IPv6 packet will traverse when there may be multiple 1069 alternatives. 1071 When an AERO node processes the SRH and forwards the original IPv6 1072 packet to the correct OMNI interface, the OMNI interface writes the 1073 next IPv6 address from the SRH into the IPv6 destination address and 1074 decrements Segments Left. If decrementing would cause Segments Left 1075 to become 0, the OMNI interface deletes the SRH before forwarding. 1076 This form of Segment Routing supports Safety-Based Multilink (SBM). 1078 3.3. OMNI Interface Characteristics 1080 OMNI interfaces are virtual interfaces configured over one or more 1081 underlay interfaces classified as follows: 1083 * ANET interfaces connect to a protected and secured ANET that is 1084 separated from the open INET by Proxy/Servers. The ANET interface 1085 may be either on the same L2 link segment as a Proxy/Server, or 1086 separated from a Proxy/Server by multiple IP hops. (Note that 1087 NATs may appear internally within an ANET and may require NAT 1088 traversal on the path to the Proxy/Server the same as for the INET 1089 case.) 1091 * INET interfaces connect to an INET either natively or through one 1092 or several IPv4 Network Address Translators (NATs). Native INET 1093 interfaces have global IP addresses that are reachable from any 1094 INET correspondent. NATed INET interfaces typically have private 1095 IP addresses and connect to a private network behind one or more 1096 NATs with the outermost NAT providing INET access. 1098 * ENET interfaces connect a Client's downstream-attached networks, 1099 where the Client provides forwarding services for ENET Host and 1100 Client communications to remote peers. An ENET be as simple as a 1101 small stub network that travels with a mobile Client (e.g., an 1102 Internet-of-Things) to as complex as a large private enterprise 1103 network that the Client connects to a larger ANET or INET. 1105 * VPNed interfaces use security encapsulation over an underlay 1106 network to a Client or Proxy/Server acting as a Virtual Private 1107 Network (VPN) gateway. Other than the link-layer encapsulation 1108 format, VPNed interfaces behave the same as for Direct interfaces. 1110 * Direct (aka "point-to-point") interfaces connect directly to a 1111 Client or Proxy/Server without crossing any networked paths. An 1112 example is a line-of-sight link between a remote pilot and an 1113 unmanned aircraft. 1115 OMNI interfaces use OAL encapsulation and fragmentation as discussed 1116 in Section 3.6. OMNI interfaces use L2 encapsulation (see: 1117 Section 3.6) to exchange carrier packets with OMNI link neighbors 1118 over INET or VPNed interfaces as well as over ANET interfaces for 1119 which the Client and FHS Proxy/Server may be multiple IP hops away. 1120 OMNI interfaces use link-layer encapsulation only (i.e., and no other 1121 L2 encapsulations) over Direct underlay interfaces or ANET interfaces 1122 when the Client and FHS Proxy/Server are known to be on the same 1123 underlay link. 1125 OMNI interfaces maintain a neighbor cache for tracking per-neighbor 1126 state the same as for any interface. OMNI interfaces use IPv6 ND 1127 messages including Router Solicitation (RS), Router Advertisement 1128 (RA), Neighbor Solicitation (NS), Neighbor Advertisement (NA) and 1129 Redirect for neighbor cache management. In environments where 1130 spoofing may be a threat, OMNI neighbors should invoke OAL 1131 Identification window synchronization in their IPv6 ND message 1132 exchanges. 1134 OMNI interfaces send IPv6 ND messages with an OMNI option formatted 1135 as specified in [I-D.templin-6man-omni]. The OMNI option includes 1136 prefix registration information, Interface Attributes and/or 1137 Multilink Forwarding Parameters containing link information 1138 parameters for the OMNI interface's underlay interfaces and any other 1139 per-neighbor information. 1141 A Host's OMNI interface is configured over an underlay interface 1142 connected to an ENET provided by an upstream Client. From the Host's 1143 perspective, the ENET appears as an ANET and the upstream Client 1144 appears as a Proxy/Server. The Host does not provide OMNI 1145 intermediate node services and is therefore a logical termination 1146 point for the OMNI link. 1148 A Client's OMNI interface may be configured over multiple ANET/INET 1149 underlay interfaces. For example, common mobile handheld devices 1150 have both wireless local area network ("WLAN") and cellular wireless 1151 links. These links are often used "one at a time" with low-cost WLAN 1152 preferred and highly-available cellular wireless as a standby, but a 1153 simultaneous-use capability could provide benefits. In a more 1154 complex example, aircraft frequently have many wireless data link 1155 types (e.g. satellite-based, cellular, terrestrial, air-to-air 1156 directional, etc.) with diverse performance and cost properties. 1158 If a Client's multiple ANET/INET underlay interfaces are used "one at 1159 a time" (i.e., all other interfaces are in standby mode while one 1160 interface is active), then successive IPv6 ND messages all include 1161 OMNI option Multilink Forwarding Parameters sub-options with the same 1162 underlay interface index. In that case, the Client would appear to 1163 have a single underlay interface but with a dynamically changing 1164 link-layer address. 1166 If the Client has multiple active ANET/INET underlay interfaces, then 1167 from the perspective of IPv6 ND it would appear to have multiple 1168 link-layer addresses. In that case, IPv6 ND message OMNI options MAY 1169 include Interface Attributes and/or Multilink Forwarding Parameters 1170 sub-options with different underlay interface indexes. 1172 Proxy/Servers on the open Internet include only a single INET 1173 underlay interface. INET Clients therefore discover only the INADDR 1174 information for the Proxy/Server's INET interface. Proxy/Servers on 1175 an ANET/INET boundary include both an ANET and INET underlay 1176 interface. ANET Clients therefore must discover both the ANET and 1177 INET INADDR information for the Proxy/Server. 1179 Gateway and Proxy/Server OMNI interfaces are configured over underlay 1180 interfaces that provide both secured tunnels for carrying IPv6 ND and 1181 BGP protocol control plane messages and open INET access for carrying 1182 unsecured messages. The OMNI interface configures a ULA-RND and acts 1183 as an OAL source to encapsulate and fragment original IP packets 1184 while presenting the resulting carrier packets over the secured or 1185 unsecured underlay paths. Note that Gateway and Proxy/Server end-to- 1186 end transport protocol sessions used by the BGP are run directly over 1187 the OMNI interface and use ULA-RND source and destination addresses. 1188 The OMNI interface employs the OAL to encapsulate the original IP 1189 packets for these sessions as carrier packets (i.e., even though the 1190 OAL header may use the same ULAs as the original IP header) and 1191 forwards them over the secured underlay path. 1193 3.4. OMNI Interface Initialization 1195 AERO Proxy/Servers, Clients and Hosts configure OMNI interfaces as 1196 their point of attachment to the OMNI link. AERO nodes assign the 1197 MSPs for the link to their OMNI interfaces (i.e., as a "route-to- 1198 interface") to ensure that original IP packets with destination 1199 addresses covered by an MNP not explicitly associated with another 1200 interface are directed to an OMNI interface. 1202 OMNI interface initialization procedures for Proxy/Servers, Clients 1203 Hosts and Gateways are discussed in the following sections. 1205 3.4.1. AERO Proxy/Server and Relay Behavior 1207 When a Proxy/Server enables an OMNI interface, it assigns a ULA-RND 1208 appropriate for the given OMNI link SRT segment. The Proxy/Server 1209 also configures secured tunnels and engages in BGP routing protocol 1210 sessions with one or more neighboring Gateways. 1212 The OMNI interface provides a single interface abstraction to the IP 1213 layer, but internally includes an NBMA nexus for sending carrier 1214 packets to OMNI interface neighbors over underlay INET interfaces and 1215 secured tunnels. The Proxy/Server further configures a service to 1216 facilitate IPv6 ND exchanges with AERO Clients and manages per-Client 1217 neighbor cache entries and IP forwarding table entries based on 1218 control message exchanges. 1220 Relays are simply Proxy/Servers that run a dynamic routing protocol 1221 to redistribute routes between the OMNI interface and INET/ENET 1222 interfaces (see: Section 3.2.3). The Relay provisions MNPs to 1223 networks on the INET/ENET interfaces (i.e., the same as a Client 1224 would do) and advertises the MSP(s) for the OMNI link over the INET/ 1225 ENET interfaces. The Relay further provides an attachment point of 1226 the OMNI link to a non-MNP-based global topology. 1228 3.4.2. AERO Client Behavior 1230 When a Client enables an OMNI interface, it assigns either an XLA-MNP 1231 or a TLA and sends OMNI-encapsulated RS messages over its ANET/INET 1232 underlay interfaces to an FHS Proxy/Server, which coordinates with a 1233 Hub Proxy/Server that returns an RA message with corresponding 1234 parameters. The RS/RA messages may pass through one or more NATs in 1235 the path between the Client and FHS Proxy/Server. (Note: if the 1236 Client used a TLA in its initial RS messages, it may discover ULA- 1237 MNPs in the corresponding RAs that it receives from FHS Proxy/Servers 1238 and begin using these new addresses. If the Client is operating 1239 outside the context of AERO infrastructure such as in a Mobile Ad-hoc 1240 Network (MANET), however, it may continue using TLAs for Client-to- 1241 Client communications until it encounters an infrastructure element 1242 that can delegate MNPs.) 1244 A Client can further extend the OMNI link over its (downstream) ENET 1245 interfaces where it provides a first-hop router for Hosts and other 1246 AERO Clients connected to the ENET. A downstream Client that 1247 connects via the ENET serviced by an upstream Client can in turn 1248 service further downstream ENETs that connect other Hosts and 1249 Clients. This OMNI link extension can be applied recursively over a 1250 "chain" of ENET Clients. 1252 3.4.3. AERO Host Behavior 1254 When a Host enables an OMNI interface, it assigns an address taken 1255 from the ENET underlay interface which may itself be a GUA delegated 1256 by the upstream Client. The Host does not assign a link-local 1257 address to the OMNI interface, since no autoconfiguration is 1258 necessary on that interface. (As an implementation matter, the Host 1259 could instead configure the "OMNI interface" as a virtual sublayer of 1260 the ENET underlay interface itself.) 1262 The Host sends OMNI-encapsulated RS messages over its ENET underlay 1263 interface to the upstream Client, which returns encapsulated RAs and 1264 provides routing services in the same fashion that Proxy/Servers 1265 provides services for Clients. Hosts represent the leaf end systems 1266 in recursively-nested chain of concatenated ENETs, i.e., they 1267 represent terminating endpoints for the OMNI link. 1269 3.4.4. AERO Gateway Behavior 1271 AERO Gateways configure an OMNI interface and assign a ULA-RND and 1272 corresponding Subnet Router Anycast address for each OMNI link SRT 1273 segment they connect to. Gateways configure secured tunnels with 1274 Proxy/Servers in the same SRT segment and other Gateways in the same 1275 (or an adjacent) SRT segment. Gateways then engage in a BGP routing 1276 protocol session with neighbors over the secured spanning tree (see: 1277 Section 3.2.3). 1279 3.5. OMNI Interface Neighbor Cache Maintenance 1281 Each Client, Proxy/Server and Gateway OMNI interface maintains a 1282 conceptual neighbor cache that includes a Neighbor Cache Entry (NCE) 1283 for each of its active neighbors on the OMNI link per [RFC4861]. 1284 Each NCE is indexed by the network layer address of the neighbor and 1285 determines the context for Identification verification. Clients and 1286 Proxy/Servers maintain NCEs through RS/RA exchanges, and also 1287 maintain NCEs for any active correspondent peers through NS/NA 1288 exchanges. 1290 Hosts also maintain NCEs for Clients and other Hosts through the 1291 exchange of RS/RA or NS/NA messages. Each NCE is indexed by the 1292 address assigned to the Host ENET interface, which is the same 1293 address used for OMNI L2 encapsulation (i.e., without the insertion 1294 of an OAL header). This encapsulation format identifies the NCE as a 1295 Host-based entry where the Host is a leaf end system in the 1296 recursively extended OMNI link. 1298 Gateways also maintain NCEs for Clients within their local segments 1299 based on NS/NA route optimization messaging (see: Section 3.13.4). 1300 When a Gateway creates/updates a NCE for a local segment Client based 1301 on NS/NA route optimization, it also maintains MFVI and INADDR state 1302 for messages destined to this local segment Client. 1304 Proxy/Servers add an additional state DEPARTED to the list of NCE 1305 states found in Section 7.3.2 of [RFC4861]. When a Client terminates 1306 its association, the Proxy/Server OMNI interface sets a "DepartTime" 1307 variable for the NCE to "DEPART_TIME" seconds. DepartTime is 1308 decremented unless a new IPv6 ND message causes the state to return 1309 to REACHABLE. While a NCE is in the DEPARTED state, the Proxy/Server 1310 forwards carrier packets destined to the target Client to the 1311 Client's new FHS/Hub Proxy/Server instead. It is RECOMMENDED that 1312 DEPART_TIME be set to the default constant value 10 seconds to accept 1313 any carrier packets that may be in flight. When DepartTime 1314 decrements to 0, the NCE is deleted. 1316 Clients determine the service profiles for their FHS and Hub Proxy/ 1317 Servers by setting the N/A/U flags in a Neighbor Coordination sub- 1318 option of the first OMNI option in RS messages. When the N/A/U flags 1319 are clear, Proxy/Servers forward all NS/NA messages to the Client, 1320 while the Client performs mobility update signaling through the 1321 transmission of uNA messages to all active neighbors following a 1322 mobility event. However, in some environments this may result in 1323 excessive NS/NA control message overhead especially for Clients 1324 connected to low-end data links. 1326 To minimize NS/NA message overhead, Clients can set the N/A/U flags 1327 in the OMNI option Neighbor Coordination header of RS messages they 1328 send. If the N flag is set, the FHS Proxy/Server that forwards the 1329 RS message assumes the role of responding to NS(NUD) messages and 1330 maintains peer NCEs associated with the NCE for this Client. If the 1331 A flag is set, the Hub Proxy/Server that processes the RS message 1332 assumes the role of responding to NS(AR) messages on behalf of this 1333 Client NCE. If the U flag is set, the Hub Proxy/Server that 1334 processes the RS message becomes responsible for maintaining a 1335 "Report List" of sources from which it has received an NS(AR) for 1336 this Client NCE. The Hub Proxy/Server maintains each Report List 1337 entry for REPORT_TIME seconds, and sends uNA messages to each member 1338 of the Report List when it receives a Client mobility update 1339 indication (e.g., through receipt of an RS with updated Interface 1340 Attributes, Traffic Selectors, etc.). 1342 Clients and their Hub Proxy/Servers have full knowledge of the 1343 Client's current underlay Interface Attributes, while FHS Proxy/ 1344 Servers acting in "proxy" mode have knowledge of only the individual 1345 Client underlay interfaces they service. Clients determine their FHS 1346 and Hub Proxy/Server service models by setting the N/A/U flags in the 1347 RS messages they send as discussed above. 1349 Clients act as RORs on their own behalf when they receive an NS(AR) 1350 from an ROS via their Hub Proxy/Server (Relays instead act as RORs on 1351 behalf of non-MNP targets specific to other links/networks that the 1352 Relay services and/or "default"). The ROR returns and NA(AR) 1353 response to the ROS, which creates or updates a NCE for the target 1354 network-layer and link-layer addresses. The ROS then (re)sets 1355 ReachableTime for the NCE to REACHABLE_TIME seconds and performs 1356 reachability tests over specific underlay interface pairs to 1357 determine paths for forwarding carrier packets directly to the 1358 target. The ROS otherwise decrements ReachableTime while no further 1359 solicited NA messages arrive. It is RECOMMENDED that REACHABLE_TIME 1360 be set to the default constant value 30 seconds as specified in 1361 [RFC4861]. 1363 AERO nodes also use the value MAX_UNICAST_SOLICIT to limit the number 1364 of NS messages sent when a correspondent may have gone unreachable, 1365 the value MAX_RTR_SOLICITATIONS to limit the number of RS messages 1366 sent without receiving an RA and the value MAX_NEIGHBOR_ADVERTISEMENT 1367 to limit the number of unsolicited NAs that can be sent based on a 1368 single event. It is RECOMMENDED that MAX_UNICAST_SOLICIT, 1369 MAX_RTR_SOLICITATIONS and MAX_NEIGHBOR_ADVERTISEMENT be set to 3 the 1370 same as specified in [RFC4861]. 1372 Different values for the above constants MAY be administratively set; 1373 however, if different values are chosen, all nodes on the link MUST 1374 consistently configure the same values. Most importantly, 1375 DEPART_TIME and REPORT_TIME SHOULD be set to a value that is 1376 sufficiently longer than REACHABLE_TIME to avoid packet loss due to 1377 stale route optimization state. 1379 3.5.1. OMNI ND Messages 1381 OMNI interfaces prepare IPv6 ND messages the same as for standard 1382 IPv6 ND, but also include a new option type termed the OMNI option 1383 [I-D.templin-6man-omni]. OMNI interfaces do not use LLAs as IPv6 ND 1384 message source and destination addresses, but instead use ULAs. This 1385 allows for multiple different OMNI links to be joined into a single 1386 link at some future time without requiring a global renumbering 1387 event. 1389 For each IPv6 ND message, OMNI interfaces include one or more OMNI 1390 options (and any other ND message options) then completely populate 1391 all option information. If the OMNI interface includes an 1392 authentication signature, it sets the IPv6 ND message Checksum field 1393 to 0 and calculates the authentication signature over the entire 1394 length of the OAL packet or super-packet (beginning with a pseudo- 1395 header of the IPv6 header) but does not then proceed to calculate the 1396 IPv6 ND message checksum itself. Otherwise, the OMNI interface 1397 calculates the standard IPv6 ND message checksum over the OAL packet 1398 or super-packet and writes the value in the Checksum field. OMNI 1399 interfaces verify authentication and integrity of each IPv6 ND 1400 message received according to the specific check(s) included, and 1401 process the message further only following verification. 1403 OMNI options include per-neighbor information that provides multilink 1404 forwarding, link-layer address and traffic selector information for 1405 the neighbor's underlay interfaces. This information is stored in 1406 the neighbor cache and provides the basis for the forwarding 1407 algorithm specified in Section 3.10. The information is cumulative 1408 and reflects the union of the OMNI information from the most recent 1409 IPv6 ND messages received from the neighbor; it is therefore not 1410 required that each IPv6 ND message contain all neighbor information. 1412 The OMNI option is distinct from any Source/Target Link-Layer Address 1413 Options (S/TLLAOs) that may appear in an IPv6 ND message according to 1414 the appropriate IPv6 over specific link layer specification (e.g., 1415 [RFC2464]). If both an OMNI option and S/TLLAO appear, the former 1416 pertains to encapsulation addresses while the latter pertains to the 1417 native L2 address format of the underlay media. 1419 OMNI interface IPv6 ND messages may also include other IPv6 ND 1420 options. In particular, solicitation messages may include a Nonce 1421 option if required for verification of advertisement replies. If an 1422 OMNI IPv6 ND solicitation message includes a Nonce option, the 1423 advertisement reply must echo the same Nonce. If an OMNI IPv6 ND 1424 advertisement message includes a Timestamp option, the recipient 1425 should check the Timestamp to determine if the message is current. 1427 AERO Clients send RS messages to the link-scoped All-Routers 1428 multicast address or a ULA-RND while using unicast or anycast L2 1429 addresses. AERO Proxy/Servers respond by returning unicast RA 1430 messages. During the RS/RA exchange, AERO Clients and Proxy/Servers 1431 include state synchronization parameters to establish Identification 1432 windows and other state. 1434 AERO Hosts and Clients on ENET underlay networks send RS messages to 1435 the link-scoped All-Routers multicast address, a ULA-RND of a remote 1436 Hub Proxy/Server or the ULA-MNP of an upstream Client while using 1437 unicast or anycast L2 addresses. The upstream AERO Client responds 1438 by returning a unicast RA message. 1440 AERO nodes use NS/NA messages for the following purposes: 1442 * NS/NA(AR) messages are used for address resolution and optionally 1443 to establish sequence number windows. The ROS sends an NS(AR) to 1444 the solicited-node multicast address of the target, and an ROR 1445 with addressing information for the target returns a unicast 1446 NA(AR) that contains current, consistent and authentic target 1447 address resolution information. NS/NA(AR) messages must be 1448 secured. 1450 * NS/NA(NUD) messages are used to establish multilink forwarding 1451 state and determine target reachability. The source sends an 1452 NS(NUD) to the unicast address of the target while naming a 1453 specific underlay interface pair, and the target returns a unicast 1454 NA(NUD). NS/NA(NUD) messages that use an in-window sequence 1455 number and do not update any other state need not include an 1456 authentication signature but instead must include an IPv6 ND 1457 message checksum. NS/NA(NUD) messages may also be used to 1458 establish window synchronization and/or MFIB state, in which case 1459 the messages must be secured. 1461 * Unsolicited NA (uNA) messages are used to signal addressing and/or 1462 other neighbor state changes (e.g., address changes due to 1463 mobility, signal degradation, traffic selector updates, etc.). uNA 1464 messages that update state information must be secured. 1466 * NS/NA(DAD) messages are not used in AERO, since Duplicate Address 1467 Detection is not required. 1469 Additionally, nodes may set the OMNI option PNG flag in NA/RA 1470 messages to receive a uNA response from the neighbor. The uNA 1471 response MUST set the ACK flag (without also setting the SYN or PNG 1472 flags) with the Acknowledgement field set to the Identification used 1473 in the PNG message. 1475 3.5.2. OMNI Neighbor Advertisement Message Flags 1477 As discussed in Section 4.4 of [RFC4861] NA messages include three 1478 flag bits R, S and O. OMNI interface NA messages treat the flags as 1479 follows: 1481 * R: The R ("Router") flag is set to 1 in the NA messages sent by 1482 all AERO/OMNI node types. Simple hosts that would set R to 0 do 1483 not occur on the OMNI link itself, but may occur on the downstream 1484 links of Clients and Relays. 1486 * S: The S ("Solicited") flag is set exactly as specified in 1487 Section 4.4. of [RFC4861], i.e., it is set to 1 for Solicited NAs 1488 and set to 0 for uNAs (both unicast and multicast). 1490 * O: The O ("Override") flag is set to 0 for solicited NAs returned 1491 by a Proxy/Server ROR and set to 1 for all other solicited and 1492 unsolicited NAs. For further study is whether solicited NAs for 1493 anycast targets apply for OMNI links. Since XLA-MNPs must be 1494 uniquely assigned to Clients to support correct IPv6 ND protocol 1495 operation, however, no role is currently seen for assigning the 1496 same XLA-MNP to multiple Clients. 1498 3.5.3. OMNI Neighbor Window Synchronization 1500 In secured environments (e.g., between secured spanning tree 1501 neighbors, between neighbors on the same secured ANET, etc.), OMNI 1502 interface neighbors can exchange OAL packets using randomly- 1503 initialized and monotonically-increasing Identification values 1504 (modulo 2**32) without window synchronization. In environments where 1505 spoofing is considered a threat, OMNI interface neighbors instead 1506 invoke window synchronization in NS/NA message exchanges to maintain 1507 send/receive window state in their respective neighbor cache entries 1508 as specified in [I-D.templin-6man-omni]. 1510 3.6. OMNI Interface Encapsulation and Fragmentation 1512 When the network layer forwards an original IP packet into an OMNI 1513 interface, the interface locates or creates a Neighbor Cache Entry 1514 (NCE) that matches the destination. The OMNI interface then invokes 1515 the OMNI Adaptation Layer (OAL) as discussed in 1516 [I-D.templin-6man-omni] which encapsulates the packet in an IPv6 1517 header to produce an OAL packet. For example, an original IP packet 1518 with source address 2001:db8:1:2::1 and destination address 1519 2001:db8:1234:5678::1 might cause the OAL encapsulation header to 1520 include source address {XLA*}::2001:db8:1:2 (i.e., an XLA-MNP) and 1521 destination address {ULA*}::0012:3456:789a:bcde (i.e., a ULA-RND). 1523 Following encapsulation, the OAL source then calculates a 2-octet 1524 checksum and fragments the OAL packet while including an identical 1525 Identification value for each fragment that must be within the window 1526 for the LHS Proxy/Server or the target Client itself. The OAL source 1527 finally includes the checksum as the final 2 octets of the final 1528 fragment, i.e., as a "trailer". 1530 The OAL source next includes an identical Compressed Routing Header 1531 with 32-bit ID fields (CRH-32) [I-D.bonica-6man-comp-rtg-hdr] with 1532 each fragment containing one or more Multilink Forwarding Vector 1533 Indices (MFVIs) if necessary as discussed in Section 3.13. The OAL 1534 source can instead invoke OAL header compression by replacing the OAL 1535 IPv6 header, CRH-32 and Fragment Header with an OAL Compressed Header 1536 (OCH). 1538 The OAL source finally encapsulates each resulting OAL fragment in L2 1539 headers to form an OAL carrier packet, with source address set to its 1540 own L2 address (e.g., 192.0.2.100) and destination set to the L2 1541 address of the next hop OAL intermediate node or destination (e.g., 1542 192.0.2.1). The carrier packet encapsulation format in the above 1543 example is shown in Figure 3: 1545 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1546 | L2 Headers | 1547 | src = 192.0.2.100 | 1548 | dst = 192.0.2.1 | 1549 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1550 | OAL IPv6 Header | 1551 | src = {XLA*}::2001:db8:1:2 | 1552 |dst={ULA*}::0012:3456:789a:bcde| 1553 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1554 | CRH-32 (if necessary) | 1555 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1556 | OAL Fragment Header | 1557 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1558 | Original IP Header | 1559 | (first-fragment only) | 1560 | src = 2001:db8:1:2::1 | 1561 | dst = 2001:db8:1234:5678::1 | 1562 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1563 | | 1564 ~ ~ 1565 ~ Original Packet Body/Fragment ~ 1566 ~ ~ 1567 | | 1568 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1569 | OAL Trailing Checksum | 1570 | (final-fragment only) | 1571 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1573 Figure 3: Carrier Packet Format 1575 Note: the carrier packets exchanged by Hosts on ENETs do not include 1576 the OAL IPv6 or CRH-32 headers, i.e., the OAL encapsulation is NULL 1577 and only the Fragment Header and L2 encapsulations are included. 1579 In this format, the OAL source encapsulates the original IP header 1580 and packet body/fragment in an OAL IPv6 header prepared according to 1581 [RFC2473], the CRH-32 is a Routing Header extension of the OAL 1582 header, the Fragment Header identifies each fragment, and the L2 1583 headers are prepared as discussed in [I-D.templin-6man-omni]. The 1584 OAL source transmits each such carrier packet into the SRT spanning 1585 tree, where they are forwarded over possibly multiple OAL 1586 intermediate nodes until they arrive at the OAL destination. 1588 The OMNI link control plane service distributes Client XLA-MNP prefix 1589 information that may change dynamically due to regional node mobility 1590 as well as XLA-MNP prefix information for Relay non-MNPs and per- 1591 segment ULA prefix information that rarely changes. OMNI link 1592 Gateways and Proxy/Servers use the information to establish and 1593 maintain a forwarding plane spanning tree that connects all nodes on 1594 the link. The spanning tree supports a carrier packet virtual 1595 bridging service according to link-layer (instead of network-layer) 1596 information, but may often include longer paths than necessary. 1598 Each OMNI interface therefore also includes a Multilink Forwarding 1599 Information Base (MFIB) with Multilink Forwarding Vectors (MFVs) that 1600 can often provide more direct forwarding "shortcuts" that avoid 1601 strict spanning tree paths. As a result, the spanning tree is always 1602 available but OMNI interfaces can often use the MFIB to greatly 1603 improve performance and reduce load on critical infrastructure 1604 elements. 1606 For carrier packets undergoing re-encapsulation at an OAL 1607 intermediate node, the OMNI interface decrements the OAL IPv6 header 1608 Hop Limit and discards the carrier packet if the Hop Limit reaches 0. 1609 The intermediate node next removes the L2 encapsulation headers from 1610 the first segment and re-encapsulates the packet in new L2 1611 encapsulation headers for the next segment. 1613 When an FHS Gateway receives a carrier packet with an OCH header that 1614 must be forwarded to an LHS Gateway over the unsecured spanning tree, 1615 it reconstructs the headers based on MFV state, inserts a CRH-32 1616 immediately following the OAL header and adjusts the OAL payload 1617 length and destination address field. The FHS Gateway includes a 1618 single MFVI in the CRH-32 that will be meaningful to the LHS Gateway. 1619 When the LHS Gateway receives the carrier packet, it locates the MFV 1620 for the next hop based on the CRH-32 MFVI then re-applies header 1621 compression (resulting in the removal of the CRH-32) and forwards the 1622 carrier packet to the next hop. 1624 3.7. OMNI Interface Decapsulation 1626 OMNI interfaces (acting as OAL destinations) decapsulate and 1627 reassemble OAL packets into original IP packets destined either to 1628 the AERO node itself or to a destination reached via an interface 1629 other than the OMNI interface the original IP packet was received on. 1630 When carrier packets containing OAL fragments addressed to itself 1631 arrive, this OAL destination discards the NET encapsulation headers 1632 and reassembles to obtain the OAL packet or super-packet (see: 1633 [I-D.templin-6man-omni]). The OAL destination then verifies the OAL 1634 checksum, discards the OAL encapsulations to obtain the original IP 1635 packet(s) and finally forwards them to either the network layer or a 1636 next-hop on the OMNI link. 1638 3.8. OMNI Interface Data Origin Authentication 1640 AERO nodes employ simple data origin authentication procedures. In 1641 particular: 1643 * AERO Gateways and Proxy/Servers accept carrier packets received 1644 from the secured spanning tree. 1646 * AERO Proxy/Servers and Clients accept carrier packets and original 1647 IP packets that originate from within the same secured ANET. 1649 * AERO Clients and Relays accept original IP packets from downstream 1650 network correspondents based on ingress filtering. 1652 * AERO Hosts, Clients, Relays, Proxy/Servers and Gateways verify 1653 carrier packet L2 encapsulation addresses according to 1654 [I-D.templin-6man-omni]. 1656 * AERO nodes accept carrier packets addressed to themselves with 1657 Identification values within the current window for the OAL source 1658 neighbor and drop any carrier packets with out-of-window 1659 Identification values. (AERO nodes may forward carrier packets 1660 not addressed to themselves without verifying the Identification 1661 value.) 1663 AERO nodes silently drop any packets that do not satisfy the above 1664 data origin authentication procedures. Further security 1665 considerations are discussed in Section 6. 1667 3.9. OMNI Interface MTU 1669 The OMNI interface observes the link nature of tunnels, including the 1670 Maximum Transmission Unit (MTU), Maximum Reassembly Unit (MRU) and 1671 the role of fragmentation and reassembly [I-D.ietf-intarea-tunnels]. 1672 The OMNI interface employs an OMNI Adaptation Layer (OAL) that 1673 accommodates multiple underlay links with diverse MTUs while 1674 observing both a minimum and per-path Maximum Payload Size (MPS). 1675 The functions of the OAL and OMNI interface MTU/MRU/MPS 1676 considerations are specified in [I-D.templin-6man-omni]. (Note that 1677 the OMNI interface accommodates an assured MTU of 65535 octets due to 1678 the use of fragmentation, and can optionally expose larger MTUs to 1679 upper layers for best-effort Jumbogram services.) 1681 When the network layer presents an original IP packet to the OMNI 1682 interface, the OAL source encapsulates and fragments the original IP 1683 packet if necessary. When the network layer presents the OMNI 1684 interface with multiple original IP packets bound to the same OAL 1685 destination, the OAL source can concatenate them as a single OAL 1686 super-packet as discussed in [I-D.templin-6man-omni] before applying 1687 fragmentation. The OAL source then encapsulates each OAL fragment in 1688 L2 headers for transmission as carrier packets over an underlay 1689 interface connected to either a physical link (e.g., Ethernet, WiFi, 1690 Cellular, etc.) or a virtual link such as an Internet or higher-layer 1691 tunnel (see the definition of link in [RFC8200]). 1693 Note: Although a CRH-32 may be inserted or removed by a Gateway in 1694 the path (see: Section 3.10.4), this does not interfere with the 1695 destination's ability to reassemble since the CRH-32 is not included 1696 in the fragmentable part and its removal/transformation does not 1697 invalidate fragment header information. 1699 3.10. OMNI Interface Forwarding Algorithm 1701 Original IP packets enter a node's OMNI interface either from the 1702 network layer (i.e., from a local application or the IP forwarding 1703 system) while carrier packets enter from the link layer (i.e., from 1704 an OMNI interface neighbor). All original IP packets and carrier 1705 packets entering a node's OMNI interface first undergo data origin 1706 authentication as discussed in Section 3.8. Those that satisfy data 1707 origin authentication are processed further, while all others are 1708 dropped silently. 1710 Original IP packets that enter the OMNI interface from the network 1711 layer are forwarded to an OMNI interface neighbor using OAL 1712 encapsulation and fragmentation to produce carrier packets for 1713 transmission over underlay interfaces. (If routing indicates that 1714 the original IP packet should instead be forwarded back to the 1715 network layer, the packet is dropped to avoid looping). Carrier 1716 packets that enter the OMNI interface from the link layer are either 1717 re-encapsulated and re-admitted into the OMNI link, or reassembled 1718 and forwarded to the network layer where they are subject to either 1719 local delivery or IP forwarding. In all cases, the OAL MUST NOT 1720 decrement the original IP packet TTL/Hop-count since its forwarding 1721 actions occur below the network layer. 1723 OMNI interfaces may have multiple underlay interfaces and/or neighbor 1724 cache entries for neighbors with multiple underlay interfaces (see 1725 Section 3.3). The OAL uses Interface Attributes and/or Traffic 1726 Selectors (e.g., port numbers, flow specifications, etc.) to select 1727 an outbound underlay interface for each OAL packet and also to select 1728 segment routing and/or link-layer destination addresses based on the 1729 neighbor's underlay interfaces. AERO implementations SHOULD permit 1730 network management to dynamically adjust Traffic Selector values at 1731 runtime. 1733 If an OAL packet matches the Traffic Selectors of multiple outgoing 1734 interfaces and/or neighbor interfaces, the OMNI interface replicates 1735 the packet and sends one copy via each of the (outgoing / neighbor) 1736 interface pairs; otherwise, it sends a single copy of the OAL packet 1737 via an interface with the best matching Traffic Selector. (While not 1738 strictly required, the likelihood of successful reassembly may 1739 improve when the OMNI interface sends all fragments of the same 1740 fragmented OAL packet consecutively over the same underlay interface 1741 pair to avoid complicating factors such as delay variance and 1742 reordering.) AERO nodes keep track of which underlay interfaces are 1743 currently "reachable" or "unreachable", and only use "reachable" 1744 interfaces for forwarding purposes. 1746 The Subnet ID value in ULAs is used only for subnet coordination 1747 within a local OMNI link segment. When a node forwards a packet with 1748 a ULA with a foreign Global and/or Subnet ID value it forwards the 1749 packet based solely on the OMNI link routing information. For this 1750 reason, OMNI link routing and forwarding table entries always include 1751 both ULAs with their associated prefix lengths and XLA-MNPs which 1752 encode an MNP while leaving the Global and Subnet ID values set to 0. 1754 The following sections discuss the OMNI interface forwarding 1755 algorithms for Hosts, Clients, Proxy/Servers and Gateways. In the 1756 following discussion, an original IP packet's destination address is 1757 said to "match" if it is the same as a cached address, or if it is 1758 covered by a cached prefix (which may be encoded in a *LA-MNP). 1760 3.10.1. Host Forwarding Algorithm 1762 When an original IP packet enters a Host's OMNI interface from the 1763 network layer the Host searches for a NCE that matches the 1764 destination. If there is a matching NCE, the Host performs L2 1765 encapsulation, fragments the encapsulated packet if necessary and 1766 forwards the packets into the ENET addressed to the L2 address of the 1767 neighbor. 1769 After sending a packet, the Host may receive a Redirect message from 1770 its upstream Client to inform it of another AERO node on the same 1771 ENET that would provide a better first hop. The Host authenticates 1772 the Redirect message, then updates its neighbor cache accordingly. 1774 3.10.2. Client Forwarding Algorithm 1776 When an original IP packet enters a Client's OMNI interface from the 1777 network layer the Client searches for a NCE that matches the 1778 destination. If there is a matching NCE on an ANET/INET interface 1779 (i.e., an upstream interface), the Client selects one or more 1780 "reachable" neighbor interfaces in the entry for forwarding purposes. 1781 Otherwise, the Client invokes route optimization per Section 3.13 and 1782 follows the multilink forwarding procedures outlined there. If there 1783 is a matching NCE on an ENET interface (i.e., a downstream 1784 interface), the Client instead performs OAL and/or L2 encapsulation 1785 and forwards the packet to the downstream Host or Client. 1787 When a carrier packet enters a Client's OMNI interface from the link- 1788 layer, if the OAL destination matches one of the Client's *LAs the 1789 Client (acting as an OAL destination) verifies that the 1790 Identification is in-window for this OAL source, then reassembles and 1791 decapsulates as necessary and delivers the original IP packet to the 1792 network layer. If the OAL destination matches a NCE for a Host or 1793 Client on an ENET interface, the Client instead forwards the carrier 1794 packet to the Host/Client. If the OAL destination does not match, 1795 the Client drops the original IP packet and MAY return a network- 1796 layer ICMP Destination Unreachable message subject to rate limiting 1797 (see: Section 3.11). 1799 When a Client forwards a carrier packet from an ENET Host to a 1800 neighbor connected to the same ENET, it also returns a Redirect 1801 message to inform the source that it can reach the neighbor directly 1802 as an ENET peer. 1804 Note: Clients and their FHS Proxy/Server (and other Client) peers can 1805 exchange original IP packets over ANET underlay interfaces without 1806 invoking the OAL, since the ANET is secured at the link and physical 1807 layers. By forwarding original IP packets without invoking the OAL, 1808 however, the ANET peers can engage only in classical path MTU 1809 discovery since the packets are subject to loss and/or corruption due 1810 to the various per-link MTU limitations that may occur within the 1811 ANET. Moreover, the original IP packets do not include either the 1812 OAL integrity check or per-packet Identification values that can be 1813 used for data origin authentication and link-layer retransmissions. 1814 The tradeoff therefore involves an assessment of the per-packet 1815 encapsulation overhead saved by bypassing the OAL vs. inheritance of 1816 classical network "brittleness". (Note however that ANET peers can 1817 send small original IP packets without invoking the OAL, while 1818 invoking the OAL for larger packets. This presents the beneficial 1819 aspects of both small packet efficiency and large packet robustness, 1820 with delay variance and reordering as possible side effects.) 1822 Note: The forwarding table entries established in peer Clients of a 1823 multihop forwarding region are based on ULA-MNPs and/or TLAs used to 1824 seed the multihop routing protocols. When ULA-MNPs are used, the ULA 1825 /64 prefix provides topological relevance for the multihop forwarding 1826 region, while the 64-bit Interface Identifier encodes the Client MNP. 1827 Therefore, Clients can forward atomic fragments with compressed OAL 1828 headers that do not include ULA or MFVI information by examining the 1829 MNP-based addresses in the actual IP packet header. In other words, 1830 each forwarding table entry contains two pieces of forwarding 1831 information - the ULA information in the prefix and the MNP 1832 information in the interface identifier. 1834 3.10.3. Proxy/Server and Relay Forwarding Algorithm 1836 When a Proxy/Server receives an original IP packet from the network 1837 layer, it drops the packet if routing indicates that it should be 1838 forwarded back to the network layer to avoid looping. Otherwise, the 1839 Proxy/Server regards the original IP packet the same as if it had 1840 arrived as carrier packets with OAL destination set to its own ULA. 1841 When the Proxy/Server receives carrier packets on underlay interfaces 1842 with OAL destination set to its own ULA, it performs OAL reassembly 1843 if necessary to obtain the original IP packet. The Proxy/Server then 1844 supports multilink forwarding procedures as specified in 1845 Section 3.13.2 and/or acts as an ROS to initiate route optimization 1846 as specified in Section 3.13. 1848 When the Proxy/Server receives a carrier packet with OAL destination 1849 set to a *LA-MNP that does not match the MSP, it accepts the carrier 1850 packet only if data origin authentication succeeds and if there is a 1851 network layer routing table entry for a GUA route that matches the 1852 *LA-MNP. If there is no route, the Proxy/Server drops the carrier 1853 packet; otherwise, it reassembles and decapsulates to obtain the 1854 original IP packet then acts as a Relay to present it to the network 1855 layer where it will be delivered according to standard IP forwarding. 1857 When a Proxy/Server receives a carrier packet from one of its Client 1858 neighbors with OAL destination set to another node, it forwards the 1859 packets via a matching NCE or via the spanning tree if there is no 1860 matching entry. When the Proxy/Server receives a carrier packet with 1861 OAL destination set to a *LA-MNP of one of its Client neighbors 1862 established through RS/RA exchanges, it accepts the carrier packet 1863 only if data origin authentication succeeds. If the NCE state is 1864 DEPARTED, the Proxy/Server changes the OAL destination address to the 1865 ULA of the new Proxy/Server, then re-encapsulates the carrier packet 1866 and forwards it to a Gateway which will eventually deliver it to the 1867 new Proxy/Server. If the neighbor cache state for the Client is 1868 REACHABLE, the Proxy/Server forwards the carrier packets to the 1869 Client which then must reassemble. (Note that the Proxy/Server does 1870 not reassemble carrier packets not explicitly addressed to its own 1871 ULA, since some of the carrier packets of the same original IP packet 1872 could be forwarded through a different Proxy/Server.) In that case, 1873 the Client may receive fragments that are smaller than its link MTU 1874 but that can still be reassembled. 1876 Proxy/Servers process carrier packets with OAL destinations that do 1877 not match their ULA in the same manner as for traditional IP 1878 forwarding within the OAL, i.e., nodes use IP forwarding to forward 1879 packets not explicitly addressed to themselves. (Proxy/Servers 1880 include a special case that accepts and reassembles carrier packets 1881 destined to a *LA-MNP of one of their Clients received over the 1882 secured spanning tree.) Proxy/Servers process carrier packets with 1883 their ULA as the destination by first examining the packet for a 1884 CRH-32 header or an OCH header. In that case, the Proxy/Server 1885 examines the next MFVI in the carrier packet to locate the MFV entry 1886 in the MFIB for next hop forwarding (i.e., without examining IP 1887 addresses). When the Proxy/Server forwards the carrier packet, it 1888 changes the destination address according to the MFVI value for the 1889 next hop found either in the CRH-32 header or in the node's own MFIB. 1890 Proxy/Servers must verify that the L2 addresses of carrier packets 1891 not received from the secured spanning tree are "trusted" before 1892 forwarding according to an MFV (otherwise, the carrier packet must be 1893 dropped). 1895 Note: Proxy/Servers may receive carrier packets addressed to their 1896 own ULA with CRH-32s that include additional forwarding information. 1897 Proxy/Servers use the forwarding information to determine the correct 1898 NCE and underlay interface for forwarding to the target Client, then 1899 remove the CRH-32 and forward the carrier packet. If necessary, the 1900 Proxy/Server reassembles first before re-encapsulating (and possibly 1901 also re-fragmenting) then forwards to the target Client. 1903 Note: Clients and their FHS Proxy/Server peers can exchange original 1904 IP packets over ANET underlay interfaces without invoking the OAL, 1905 since the ANET is secured at the link and physical layers. By 1906 forwarding original IP packets without invoking the OAL, however, the 1907 Client and Proxy/Server can engage only in classical path MTU 1908 discovery since the packets are subject to loss and/or corruption due 1909 to the various per-link MTU limitations that may occur within the 1910 ANET. Moreover, the original IP packets do not include either the 1911 OAL integrity check or per-packet Identification values that can be 1912 used for data origin authentication and link-layer retransmissions. 1913 The tradeoff therefore involves an assessment of the per-packet 1914 encapsulation overhead saved by bypassing the OAL vs. inheritance of 1915 classical network "brittleness". (Note however that ANET peers can 1916 send small original IP packets without invoking the OAL, while 1917 invoking the OAL for larger packets. This presents the beneficial 1918 aspects of both small packet efficiency and large packet robustness.) 1920 Note: When a Proxy/Server receives a (non-OAL) original IP packet 1921 from an ANET Client, or a carrier packet with OAL destination set to 1922 its own ULA from any Client, the Proxy/Server reassembles if 1923 necessary then performs ROS functions on behalf of the Client. The 1924 Client may at some later time begin sending carrier packets to the 1925 OAL address of the actual target instead of the Proxy/Server, at 1926 which point it may begin functioning as an ROS on its own behalf and 1927 thereby "override" the Proxy/Server's ROS role. 1929 Note; Proxy/Servers drop any original IP packets (received either 1930 directly from an ANET Client or following reassembly of carrier 1931 packets received from an ANET/INET Client) with a destination that 1932 corresponds to the Client's delegated MNP. Similarly, Proxy/Servers 1933 drop any carrier packet received with both a source and destination 1934 that correspond to the Client's delegated MNP regardless of their 1935 OMNI link point of origin. These checks are necessary to prevent 1936 Clients from either accidentally or intentionally establishing 1937 endless loops that could congest Proxy/Servers and/or ANET/INET 1938 links. 1940 Note: Proxy/Servers forward secure control plane carrier packets via 1941 the SRT secured spanning tree and forward other carrier packets via 1942 the unsecured spanning tree. When a Proxy/Server receives a carrier 1943 packet from the secured spanning tree, it considers the message as 1944 authentic without having to verify upper layer authentication 1945 signatures. When a Proxy/Server receives a carrier packet from the 1946 unsecured spanning tree, it applies data origin authentication itself 1947 and/or forwards the unsecured message toward the destination which 1948 must apply data origin authentication on its own behalf. 1950 Note: If the Proxy/Server has multiple original IP packets to send to 1951 the same neighbor, it can concatenate them in a single OAL super- 1952 packet [I-D.templin-6man-omni]. 1954 3.10.4. Gateway Forwarding Algorithm 1956 Gateways forward spanning tree carrier packets while decrementing the 1957 OAL header Hop Count but not the original IP header Hop Count/TTL. 1958 Gateways convey carrier packets that encapsulate critical IPv6 ND 1959 control messages or routing protocol control messages via the SRT 1960 secured spanning tree, and may convey other carrier packets via the 1961 secured/unsecured spanning tree or via more direct paths according to 1962 MFIB information. When the Gateway receives a carrier packet, it 1963 removes the L2 headers and searches for an MFIB entry that matches an 1964 MFVI or an IP forwarding table entry that matches the OAL destination 1965 address. 1967 Gateways process carrier packets with OAL destinations that do not 1968 match their ULA or the SRT Subnet Router Anycast address in the same 1969 manner as for traditional IP forwarding within the OAL, i.e., nodes 1970 use IP forwarding to forward packets not explicitly addressed to 1971 themselves. Gateways process carrier packets with their ULA or the 1972 SRT Subnet Router Anycast address as the destination by first 1973 examining the packet for a full OAL header with a CRH-32 extension or 1974 an OCH header. In that case, the Gateway examines the next MFVI in 1975 the carrier packet to locate the MFV entry in the MFIB for next hop 1976 forwarding (i.e., without examining IP addresses). When the Gateway 1977 forwards the carrier packet, it changes the destination address 1978 according to the MFVI value for the next hop found either in the 1979 CRH-32 header or in the node's own MFIB. If the Gateway has a NCE 1980 for the target Client with an entry for the target underlay interface 1981 and current L2 addresses, the Gateway instead forwards directly to 1982 the target Client while using the final hop MFVI instead of the next 1983 hop (see: Section 3.13.4). 1985 Gateways forward carrier packets received from a first segment via 1986 the secured spanning tree to the next segment also via the secured 1987 spanning tree. Gateways forward carrier packets received from a 1988 first segment via the unsecured spanning tree to the next segment 1989 also via the unsecured spanning tree. Gateways use a single IPv6 1990 routing table that always determines the same next hop for a given 1991 OAL destination, where the secured/unsecured spanning tree is 1992 determined through the selection of the underlay interface to be used 1993 for transmission (i.e., a secured tunnel or an open INET interface). 1995 As for Proxy/Servers, Gateways must verify that the L2 addresses of 1996 carrier packets not received from the secured spanning tree are 1997 "trusted" before forwarding according to an MFV (otherwise, the 1998 carrier packet must be dropped). 2000 3.11. OMNI Interface Error Handling 2002 When an AERO node admits an original IP packet into the OMNI 2003 interface, it may receive link-layer or network-layer error 2004 indications. The AERO node may also receive OMNI link error 2005 indications in OAL-encapsulated uNA messages that include 2006 authentication signatures. 2008 A link-layer error indication is an ICMP error message generated by a 2009 router in an underlay network on the path to the neighbor or by the 2010 neighbor itself. The message includes an IP header with the address 2011 of the node that generated the error as the source address and with 2012 the link-layer address of the AERO node as the destination address. 2014 The IP header is followed by an ICMP header that includes an error 2015 Type, Code and Checksum. Valid type values include "Destination 2016 Unreachable", "Time Exceeded" and "Parameter Problem" 2017 [RFC0792][RFC4443]. (OMNI interfaces ignore link-layer IPv4 2018 "Fragmentation Needed" and IPv6 "Packet Too Big" messages for carrier 2019 packets that are no larger than the minimum/path MPS as discussed in 2020 Section 3.9, however these messages may provide useful hints of probe 2021 failures during path MPS probing.) 2023 The ICMP header is followed by the leading portion of the carrier 2024 packet that generated the error, also known as the "packet-in-error". 2025 For ICMPv6, [RFC4443] specifies that the packet-in-error includes: 2026 "As much of invoking packet as possible without the ICMPv6 packet 2027 exceeding the minimum IPv6 MTU" (i.e., no more than 1280 bytes). For 2028 ICMPv4, [RFC0792] specifies that the packet-in-error includes: 2029 "Internet Header + 64 bits of Original Data Datagram", however 2030 [RFC1812] Section 4.3.2.3 updates this specification by stating: "the 2031 ICMP datagram SHOULD contain as much of the original datagram as 2032 possible without the length of the ICMP datagram exceeding 576 2033 bytes". 2035 The link-layer error message format is shown in Figure 4: 2037 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2038 ~ ~ 2039 | IP Header of link layer | 2040 | error message | 2041 ~ ~ 2042 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2043 | ICMP Header | 2044 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 2045 ~ ~ P 2046 | carrier packet L2 and OAL | a 2047 | encapsulation headers | c 2048 ~ ~ k 2049 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ e 2050 ~ ~ t 2051 | original IP packet headers | 2052 | (first-fragment only) | i 2053 ~ ~ n 2054 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2055 ~ ~ e 2056 | Portion of the body of | r 2057 | the original IP packet | r 2058 | (all fragments) | o 2059 ~ ~ r 2060 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 2062 Figure 4: OMNI Interface Link-Layer Error Message Format 2064 The AERO node rules for processing these link-layer error messages 2065 are as follows: 2067 * When an AERO node receives a link-layer Parameter Problem message, 2068 it processes the message the same as described as for ordinary 2069 ICMP errors in the normative references [RFC0792][RFC4443]. 2071 * When an AERO node receives persistent link-layer Time Exceeded 2072 messages, the IP ID field may be wrapping before earlier fragments 2073 awaiting reassembly have been processed. In that case, the node 2074 should begin including integrity checks and/or institute rate 2075 limits for subsequent packets. 2077 * When an AERO node receives persistent link-layer Destination 2078 Unreachable messages in response to carrier packets that it sends 2079 to one of its neighbor correspondents, the node should process the 2080 message as an indication that a path may be failing, and 2081 optionally initiate NUD over that path. If it receives 2082 Destination Unreachable messages over multiple paths, the node 2083 should allow future carrier packets destined to the correspondent 2084 to flow through a default route and re-initiate route 2085 optimization. 2087 * When an AERO Client receives persistent link-layer Destination 2088 Unreachable messages in response to carrier packets that it sends 2089 to one of its neighbor Proxy/Servers, the Client should mark the 2090 path as unusable and use another path. If it receives Destination 2091 Unreachable messages on many or all paths, the Client should 2092 associate with a new Proxy/Server and release its association with 2093 the old Proxy/Server as specified in Section 3.15.5. 2095 * When an AERO Proxy/Server receives persistent link-layer 2096 Destination Unreachable messages in response to carrier packets 2097 that it sends to one of its neighbor Clients, the Proxy/Server 2098 should mark the underlay path as unusable and use another underlay 2099 path. 2101 * When an AERO Proxy/Server receives link-layer Destination 2102 Unreachable messages in response to a carrier packet that it sends 2103 to one of its permanent neighbors, it treats the messages as an 2104 indication that the path to the neighbor may be failing. However, 2105 the dynamic routing protocol should soon reconverge and correct 2106 the temporary outage. 2108 When an AERO Gateway receives a carrier packet for which the network- 2109 layer destination address is covered by an MSP assigned to a black- 2110 hole route, the Gateway drops the packet if there is no more-specific 2111 routing information for the destination and returns an OMNI interface 2112 Destination Unreachable message subject to rate limiting. 2114 When an AERO node receives a carrier packet for which reassembly is 2115 currently congested, it returns an OMNI interface Packet Too Big 2116 (PTB) message as discussed in [I-D.templin-6man-omni] (note that the 2117 PTB messages could indicate either "hard" or "soft" errors). 2119 AERO nodes include ICMPv6 error messages intended for an OAL source 2120 as sub-options in the OMNI option of secured uNA messages. When the 2121 OAL source receives the uNA message, it can extract the ICMPv6 error 2122 message enclosed in the OMNI option and either process it locally or 2123 translate it into a network-layer error to return to the original 2124 source. 2126 3.12. AERO Mobility Service Coordination 2128 AERO nodes observes the Router Discovery and Prefix Registration 2129 specifications found in Section 15 of [I-D.templin-6man-omni]. AERO 2130 nodes further coordinate their autoconfiguration actions with the 2131 mobility service as discussed in the following sections. 2133 3.12.1. AERO Service Model 2135 Each AERO Proxy/Server on the OMNI link is configured to facilitate 2136 Client prefix delegation/registration requests. Each Proxy/Server is 2137 provisioned with a database of MNP-to-Client ID mappings for all 2138 Clients enrolled in the AERO service, as well as any information 2139 necessary to authenticate each Client. The Client database is 2140 maintained by a central administrative authority for the OMNI link 2141 and securely distributed to all Proxy/Servers, e.g., via the 2142 Lightweight Directory Access Protocol (LDAP) [RFC4511], via static 2143 configuration, etc. Clients receive the same service regardless of 2144 the Proxy/Servers they select. 2146 Clients associate each of their ANET/INET underlay interfaces with a 2147 FHS Proxy/Server. Each FHS Proxy/Server locally services one or more 2148 of the Client's underlay interfaces, and the Client typically selects 2149 one among them to serve as the Hub Proxy/Server (the Client may 2150 instead select a "third-party" Hub Proxy/Server that does not 2151 directly service any of its underlay interfaces). All of the 2152 Client's other FHS Proxy/Servers forward proxyed copies of RS/RA 2153 messages between the Hub Proxy/Server and Client without assuming the 2154 Hub role functions themselves. 2156 Each Client associates with a single Hub Proxy/Server at a time, 2157 while all other Proxy/Servers are candidates for providing the Hub 2158 role for other Clients. An FHS Proxy/Server assumes the Hub role 2159 when it receives an RS message with its own ULA or link-scoped All- 2160 Routers multicast as the destination. An FHS Proxy/Server assumes 2161 the proxy role when it receives an RS message with the ULA of another 2162 Proxy/Server as the destination. (An FHS Proxy/Server can also 2163 assume the proxy role when it receives an RS message addressed to 2164 link-scoped All-Routers multicast if it can determine the ULA of 2165 another Proxy/Server to serve as a Hub.) 2167 Hosts and Clients on ENET interfaces associate with an upstream 2168 Client on the ENET the same as a Client would associate with an ANET 2169 Proxy/Server. In particular, the Host/Client sends an RS message via 2170 the ENET which directs the message to the upstream Client. The 2171 upstream Client returns an RA message. In this way, the downstream 2172 nodes see the ENET as an ANET and see the upstream Client as a Proxy/ 2173 Server for that ANET. 2175 AERO Hosts, Clients and Proxy/Servers use IPv6 ND messages to 2176 maintain neighbor cache entries. AERO Proxy/Servers configure their 2177 OMNI interfaces as advertising NBMA interfaces, and therefore send 2178 unicast RA messages with a short Router Lifetime value (e.g., 2179 ReachableTime seconds) in response to a Client's RS message. 2180 Thereafter, Clients send additional RS messages to keep Proxy/Server 2181 state alive. 2183 AERO Clients and Hub Proxy/Servers include prefix delegation and/or 2184 registration parameters in RS/RA messages. The IPv6 ND messages are 2185 exchanged between the Client and Hub Proxy/Server (via any FHS Proxy/ 2186 Servers acting as proxys) according to the prefix management schedule 2187 required by the service. If the Client knows its MNP in advance, it 2188 can employ prefix registration by including its XLA-MNP as the source 2189 address of an RS message and with an OMNI option with valid prefix 2190 registration information for the MNP. If the Hub Proxy/Server 2191 accepts the Client's MNP assertion, it injects the MNP into the 2192 routing system and establishes the necessary neighbor cache state. 2193 If the Client does not have a pre-assigned MNP, it can instead employ 2194 prefix delegation by including a TLA as the source address of an RS 2195 message and with an OMNI option with prefix delegation parameters to 2196 request an MNP. 2198 The following sections outlines Host, Client and Proxy/Server 2199 behaviors based on the Router Discovery and Prefix Registration 2200 specifications found in Section 15 of [I-D.templin-6man-omni]. These 2201 sections observe all of the OMNI specifications, and include 2202 additional specifications of the interactions of Client-Proxy/Server 2203 RS/RA exchanges with the AERO mobility service. 2205 3.12.2. AERO Host and Client Behavior 2207 AERO Hosts and Clients discover the addresses of candidate Proxy/ 2208 Servers by resolving the Potential Router List (PRL) in a similar 2209 manner as described in [RFC5214]. Discovery methods include static 2210 configuration (e.g., a flat-file map of Proxy/Server addresses and 2211 locations), or through an automated means such as Domain Name System 2212 (DNS) name resolution [RFC1035]. Alternatively, the Host/Client can 2213 discover Proxy/Server addresses through a layer 2 data link login 2214 exchange, or through an RA response to a multicast/anycast RS as 2215 described below. In the absence of other information, the Host/ 2216 Client can resolve the DNS Fully-Qualified Domain Name (FQDN) 2217 "linkupnetworks.[domainname]" where "linkupnetworks" is a constant 2218 text string and "[domainname]" is a DNS suffix for the OMNI link 2219 (e.g., "example.com"). The name resolution returns a set of resource 2220 records with Proxy/Server address information. 2222 The Host/Client then performs RS/RA exchanges over each of its 2223 underlay interfaces to associate with (possibly multiple) FHS Proxy/ 2224 Serves and a single Hub Proxy/Server as specified in Section 15 of 2225 [I-D.templin-6man-omni]. The Host/Client then sends each RS (either 2226 directly via Direct interfaces, via a VPN for VPNed interfaces, via 2227 an access router for ANET interfaces or via INET encapsulation for 2228 INET interfaces) and waits up to RetransTimer milliseconds for an RA 2229 message reply (see Section 3.12.3) while retrying up to 2230 MAX_RTR_SOLICITATIONS if necessary. If the Host/Client receives no 2231 RAs, or if it receives an RA with Router Lifetime set to 0, the 2232 Client SHOULD abandon attempts through the first candidate Proxy/ 2233 Server and try another Proxy/Server. 2235 After the Host/Client registers its underlay interfaces, it may wish 2236 to change one or more registrations, e.g., if an interface changes 2237 address or becomes unavailable, if traffic selectors change, etc. To 2238 do so, the Host/Client prepares an RS message to send over any 2239 available underlay interface as above. The RS includes an OMNI 2240 option with prefix registration/delegation information and with an 2241 Interface Attributes sub-option specific to the selected underlay 2242 interface. When the Host/Client receives the Hub Proxy/Server's RA 2243 response, it has assurance that both the Hub and FHS Proxy/Servers 2244 have been updated with the new information. 2246 If the Host/Client wishes to discontinue use of a Hub Proxy/Server it 2247 issues an RS message over any underlay interface with an OMNI option 2248 with a prefix release indication (i.e., by setting the OMNI Neighbor 2249 Coordination header Preflen to 0). When the Hub Proxy/Server 2250 processes the message, it releases the MNP, sets the NCE state for 2251 the Host/Client to DEPARTED and returns an RA reply with Router 2252 Lifetime set to 0. After a short delay (e.g., 2 seconds), the Hub 2253 Proxy/Server withdraws the MNP from the routing system. 2254 (Alternatively, when the Host/Client associates with a new FHS/Hub 2255 Proxy/Server it can include an OMNI "Proxy/Server Departure" sub- 2256 option in RS messages with the ULAs of the Old FHS/Hub Proxy/ 2257 Servers.) 2259 3.12.3. AERO Proxy/Server Behavior 2261 AERO Proxy/Servers act as both IP routers and IPv6 ND proxys, and 2262 support a prefix delegation/registration service for Clients. Proxy/ 2263 Servers arrange to add their ULAs to the PRL maintained in a static 2264 map of Proxy/Server addresses for the link, the DNS resource records 2265 for the FQDN "linkupnetworks.[domainname]", etc. before entering 2266 service. The PRL should be arranged such that Clients can discover 2267 the addresses of Proxy/Servers that are geographically and/or 2268 topologically "close" to their underlay network connections. 2270 When a FHS/Hub Proxy/Server receives a prospective Client's RS 2271 message, it SHOULD return an immediate RA reply with Router Lifetime 2272 set to 0 if it is currently too busy or otherwise unable to service 2273 the Client; otherwise, it processes the RS as specified in Section 15 2274 of [I-D.templin-6man-omni]. When the Hub Proxy/Server receives the 2275 RS, it determines the correct MNPs to provide to the Client by 2276 processing the XLA-MNP prefix parameters and/or the DHCPv6 OMNI sub- 2277 option. When the Hub Proxy/Server returns the MNPs, it also creates 2278 an XLA-MNP forwarding table entry for the MNP resulting in a BGP 2279 update (see: Section 3.2.3). The Hub Proxy/Server then returns an RA 2280 to the Client with destination set to the source of the RS (if an FHS 2281 Proxy/Server on the return path proxys the RA, it changes the 2282 destination to the Client's ULA-MNP). 2284 After the initial RS/RA exchange, the Hub Proxy/Server maintains a 2285 ReachableTime timer for each of the Client's underlay interfaces 2286 individually (and for the Client's NCE collectively) set to expire 2287 after ReachableTime seconds. If the Client (or an FHS Proxy/Server) 2288 issues additional RS messages, the Hub Proxy/Server sends an RA 2289 response and resets ReachableTime. If the Hub Proxy/Server receives 2290 an IPv6 ND message with a prefix release indication it sets the 2291 Client's NCE to the DEPARTED state and withdraws the MNP route from 2292 the routing system after a short delay (e.g., 2 seconds). If 2293 ReachableTime expires before a new RS is received on an individual 2294 underlay interface, the Hub Proxy/Server marks the interface as DOWN. 2295 If ReachableTime expires before any new RS is received on any 2296 individual underlay interface, the Hub Proxy/Server sets the NCE 2297 state to STALE and sets a 10 second timer. If the Hub Proxy/Server 2298 has not received a new RS or uNA message with a prefix release 2299 indication before the 10 second timer expires, it deletes the NCE and 2300 withdraws the XLA-MNP from the routing system. 2302 The Hub Proxy/Server processes any IPv6 ND messages pertaining to the 2303 Client while forwarding to the Client or responding on the Client's 2304 behalf as necessary. The Hub Proxy/Server may also issue unsolicited 2305 RA messages, e.g., with reconfigure parameters to cause the Client to 2306 renegotiate its prefix delegation/registrations, with Router Lifetime 2307 set to 0 if it can no longer service this Client, etc. The Hub 2308 Proxy/Server may also receive carrier packets via the secured 2309 spanning tree that contain initial data packets sent while route 2310 optimization is in progress. The Hub Proxy/Server reassembles, then 2311 re-encapsulates/re-fragments and forwards the packets to the target 2312 Client via an FHS Proxy/Server if necessary. Finally, If the NCE is 2313 in the DEPARTED state, the old Hub Proxy/Server forwards any carrier 2314 packets it receives from the secured spanning tree and destined to 2315 the Client to the new Hub Proxy/Server, then deletes the entry after 2316 DepartTime expires. 2318 Note: Clients SHOULD arrange to notify former Hub Proxy/Servers of 2319 their departures, but Hub Proxy/Servers are responsible for expiring 2320 neighbor cache entries and withdrawing XLA-MNP routes even if no 2321 departure notification is received (e.g., if the Client leaves the 2322 network unexpectedly). Hub Proxy/Servers SHOULD therefore set Router 2323 Lifetime to ReachableTime seconds in solicited RA messages to 2324 minimize persistent stale cache information in the absence of Client 2325 departure notifications. A short Router Lifetime also ensures that 2326 proactive RS/RA messaging between Clients and FHS Proxy/Servers will 2327 keep any NAT state alive (see above). 2329 Note: All Proxy/Servers on an OMNI link MUST advertise consistent 2330 values in the RA Cur Hop Limit, M and O flags, Reachable Time and 2331 Retrans Timer fields the same as for any link, since unpredictable 2332 behavior could result if different Proxy/Servers on the same link 2333 advertised different values. 2335 3.12.3.1. Additional Proxy/Server Considerations 2337 AERO Clients register with FHS Proxy/Servers for each underlay 2338 interface. Each of the Client's FHS Proxy/Servers must inform a 2339 single Hub Proxy/Server of the Client's underlay interface(s) that it 2340 services. For Clients on Direct and VPNed underlay interfaces, the 2341 FHS Proxy/Server for each interface is directly connected, for 2342 Clients on ANET underlay interfaces the FHS Proxy/Server is located 2343 on the ANET/INET boundary, and for Clients on INET underlay 2344 interfaces the FHS Proxy/Server is located somewhere in the connected 2345 Internetwork. When FHS Proxy/Server "B" processes a Client 2346 registration, it must either assume the Hub role or forward a proxyed 2347 registration to another Proxy/Server "A" acting as the Hub. Proxy/ 2348 Servers satisfy these requirements as follows: 2350 * when FHS Proxy/Server "B" receives a Client RS message, it first 2351 verifies that the OAL Identification is within the window for the 2352 NCE that matches the *LA-MNP for this Client neighbor and 2353 authenticates the message. If no NCE was found, Proxy/Server "B" 2354 instead creates one in the STALE state and caches the Client- 2355 supplied Interface Attributes, Origin Indication and OMNI Neighbor 2356 Coordination header window synchronization parameters as well as 2357 the Client's observed L2 addresses (noting that they may differ 2358 from the Origin addresses if there were NATs on the path). Proxy/ 2359 Server "B" then examines the network-layer destination address. 2360 If the destination address is the ULA of a different Proxy/Server 2361 "A", Proxy/Server "B" prepares a separate proxyed version of the 2362 RS message with an OAL header with source set to its own ULA and 2363 destination set to Proxy/Server B's ULA. Proxy/Server "B" also 2364 writes its own information over the Interface Attributes sub- 2365 option supplied by the Client, omits or zeros the Origin 2366 Indication sub-option then forwards the message into the OMNI link 2367 secured spanning tree. 2369 * when Hub Proxy/Server "A" receives the RS, it assume the Hub role 2370 and creates/updates a NCE for the Client with FHS Proxy/Server 2371 "B"'s Interface Attributes as the link-layer address information 2372 for this FHS omIndex. Hub Proxy/Server "A" then prepares an RA 2373 message with source set to its own ULA and destination set to the 2374 source of the RS message, then encapsulates the RA in an OAL 2375 header with source set to its own ULA and destination set to the 2376 ULA of FHS Proxy/Server "B". Hub Proxy/Server "A" then performs 2377 fragmentation if necessary and sends the resulting carrier packets 2378 into the secured spanning tree. 2380 * when FHS Proxy/Server "B" reassembles the RA, it locates the 2381 Client NCE based on the RA destination. If the RA message 2382 includes an OMNI "Proxy/Server Departure" sub-option, Proxy/Server 2383 "B" first sends a uNA to the old FHS/Hub Proxy/Servers named in 2384 the sub-option. Proxy/Server "B" then changes the RA destination 2385 address to the ULA-MNP of the Client, then re-encapsulates the 2386 message with OAL source set to its own ULA and OAL destination set 2387 to ULA that appeared in the Client's RS source, with an 2388 appropriate Identification value, with an authentication signature 2389 if necessary, with the Client's Interface Attributes sub-option 2390 echoed and with the cached observed L2 addresses written into an 2391 Origin Indication sub-option. Proxy/Server "B" sets the P flag in 2392 the RA flags field to indicate that the message has passed through 2393 a proxy [RFC4389], includes responsive window synchronization 2394 parameters, then fragments the RA if necessary and returns the 2395 fragments to the Client. 2397 * The Client repeats this process over each of its additional 2398 underlay interfaces while treating each additional FHS Proxy/ 2399 Server "C", "D", "E", etc. as a proxy to facilitate RS/RA 2400 exchanges between the Hub and the Client. The Client creates/ 2401 updates NCEs for each such FHS Proxy/Server as well as the Hub 2402 Proxy/Server in the process. 2404 After the initial RS/RA exchanges each FHS Proxy/Server forwards any 2405 of the Client's carrier packets with OAL destinations for which there 2406 is no matching NCE to a Gateway using OAL encapsulation with its own 2407 ULA as the source and with destination determined by the Client. The 2408 Proxy/Server instead forwards any carrier packets destined to a 2409 neighbor cache target directly to the target according to the OAL/ 2410 link-layer information - the process of establishing neighbor cache 2411 entries is specified in Section 3.13. 2413 While the Client is still associated with FHS Proxy/Servers "B", "C", 2414 "D", etc., each FHS Proxy/Server can send NS, RS and/or unsolicited 2415 NA messages to update the neighbor cache entries of other AERO nodes 2416 on behalf of the Client based on changes in Interface Attributes, 2417 Traffic Selectors, etc. This allows for higher-frequency Proxy- 2418 initiated RS/RA messaging over well-connected INET infrastructure 2419 supplemented by lower-frequency Client-initiated RS/RA messaging over 2420 constrained ANET data links. 2422 If the Hub Proxy/Server "A" ceases to send solicited RAs, FHS Proxy/ 2423 Servers "B", "C", "D" can send unsolicited RAs over the Client's 2424 underlay interface with destination set to (link-local) All-Nodes 2425 multicast and with Router Lifetime set to zero to inform Clients that 2426 the Hub Proxy/Server has failed. Although FHS Proxy/Servers "B", "C" 2427 and "D" can engage in IPv6 ND exchanges on behalf of the Client, the 2428 Client can also send IPv6 ND messages on its own behalf, e.g., if it 2429 is in a better position to convey state changes. The IPv6 ND 2430 messages sent by the Client include the Client's XLA-MNP as the 2431 source in order to differentiate them from the IPv6 ND messages sent 2432 by a FHS Proxy/Server. 2434 If the Client becomes unreachable over all underlay interface it 2435 serves, the Hub Proxy/Server sets the NCE state to DEPARTED and 2436 retains the entry for DepartTime seconds. While the state is 2437 DEPARTED, the Hub Proxy/Server forwards any carrier packets destined 2438 to the Client to a Gateway via OAL encapsulation. When DepartTime 2439 expires, the Hub Proxy/Server deletes the NCE, withdraws the XLA-MNP 2440 route and discards any further carrier packets destined to the former 2441 Client. 2443 In some ANETs that employ a Proxy/Server, the Client's MNP can be 2444 injected into the ANET routing system. In that case, the Client can 2445 send original IP packets without invoking the OAL so that the ANET 2446 routing system transports the original IP packets to the Proxy/ 2447 Server. This can be beneficial, e.g., if the Client connects to the 2448 ANET via low-end data links such as some aviation wireless links. 2450 If the ANET first-hop access router is on the same underlay link as 2451 the Client and recognizes the AERO/OMNI protocol, the Client can 2452 avoid OAL encapsulation for both its control and data messages. When 2453 the Client connects to the link, it can send an unencapsulated RS 2454 message with source address set to its own XLA-MNP (or to a TLA), and 2455 with destination address set to the ULA of the Client's selected 2456 Proxy/Server or to link-scoped All-Routers multicast. The Client 2457 includes an OMNI option formatted as specified in 2458 [I-D.templin-6man-omni]. The Client then sends the unencapsulated RS 2459 message, which will be intercepted by the AERO-aware ANET access 2460 router. 2462 The ANET access router then performs OAL encapsulation on the RS 2463 message and forwards it to a Proxy/Server at the ANET/INET boundary. 2464 When the access router and Proxy/Server are one and the same node, 2465 the Proxy/Server would share an underlay link with the Client but its 2466 message exchanges with outside correspondents would need to pass 2467 through a security gateway at the ANET/INET border. The method for 2468 deploying access routers and Proxys (i.e. as a single node or 2469 multiple nodes) is an ANET-local administrative consideration. 2471 Note: When a Proxy/Server alters the IPv6 ND message contents before 2472 forwarding (e.g., such as altering the OMNI option contents), the 2473 original IPv6 ND message checksum or authentication signature is 2474 invalidated, and a new checksum or authentication signature must be 2475 calculated and included. 2477 Note: When a Proxy/Server receives a secured Client NS message, it 2478 performs the same proxying procedures as for described for RS 2479 messages above. The proxying procedures for NS/NA message exchanges 2480 is specified in Section 3.13. 2482 3.12.3.2. Detecting and Responding to Proxy/Server Failures 2484 In environments where fast recovery from Proxy/Server failure is 2485 required, FHS Proxy/Servers SHOULD use proactive Neighbor 2486 Unreachability Detection (NUD) to track Hub Proxy/Server reachability 2487 in a similar fashion as for Bidirectional Forwarding Detection (BFD) 2488 [RFC5880]. Each FHS Proxy/Server can then quickly detect and react 2489 to failures so that cached information is re-established through 2490 alternate paths. The NS/NA(NUD) control messaging is carried only 2491 over well-connected ground domain networks (i.e., and not low-end 2492 aeronautical radio links) and can therefore be tuned for rapid 2493 response. 2495 FHS Proxy/Servers perform continuous NS/NA(NUD) exchanges with the 2496 Hub Proxy/Server, e.g., one exchange per second. The FHS Proxy/ 2497 Server sends the NS(NUD) message via the spanning tree with its own 2498 ULA as the source and the ULA of the Hub Proxy/Server as the 2499 destination, and the Hub Proxy/Server responds with an NA(NUD). When 2500 the FHS Proxy/Server is also sending RS messages to a Hub Proxy/ 2501 Server on behalf of Clients, the resulting RA responses can be 2502 considered as equivalent hints of forward progress. This means that 2503 the FHS Proxy/Server need not also send a periodic NS(NUD) if it has 2504 already sent an RS within the same period. If the Hub Proxy/Server 2505 fails (i.e., if the FHS Proxy/Server ceases to receive 2506 advertisements), the FHS Proxy/Server can quickly inform Clients by 2507 sending unsolicited RA messages 2509 The FHS Proxy/Server sends unsolicited RA messages with source 2510 address set to the Hub Proxy/Server's address, destination address 2511 set to (link-local) All-Nodes multicast, and Router Lifetime set to 2512 0. The FHS Proxy/Server SHOULD send MAX_FINAL_RTR_ADVERTISEMENTS RA 2513 messages separated by small delays [RFC4861]. Any Clients that had 2514 been using the failed Hub Proxy/Server will receive the RA messages 2515 and select one of its other FHS Proxy/Servers to assume the Hub role 2516 (i.e., by sending an RS with destination set to the ULA of the new 2517 Hub). 2519 3.12.3.3. DHCPv6-Based Prefix Registration 2521 When a Client is not pre-provisioned with an MNP, it will need for 2522 the Hub Proxy/Server to select one or more MNPs on its behalf and set 2523 up the correct state in the AERO routing service. (A Client with a 2524 pre-provisioned MNP may also request the Hub Proxy/Server to select 2525 additional MNPs.) The DHCPv6 service [RFC8415] is used to support 2526 this requirement. 2528 When a Client needs to have the Hub Proxy/Server select MNPs, it 2529 sends an RS message with source address set to a TLA and with an OMNI 2530 option that includes a DHCPv6 message sub-option with DHCPv6 Prefix 2531 Delegation (DHCPv6-PD) parameters. When the Hub Proxy/Server 2532 receives the RS message, it extracts the DHCPv6-PD message from the 2533 OMNI option. 2535 The Hub Proxy/Server then acts as a "Proxy DHCPv6 Client" in a 2536 message exchange with the locally-resident DHCPv6 server, which 2537 delegates MNPs and returns a DHCPv6-PD Reply message. (If the Hub 2538 Proxy/Server wishes to defer creation of MN state until the DHCPv6-PD 2539 Reply is received, it can instead act as a Lightweight DHCPv6 Relay 2540 Agent per [RFC6221] by encapsulating the DHCPv6-PD message in a 2541 Relay-forward/reply exchange with Relay Message and Interface ID 2542 options.) 2544 When the Hub Proxy/Server receives the DHCPv6-PD Reply, it creates an 2545 XLA based on the delegated MNP adds an XLA-MNP route to the routing 2546 system. The Hub Proxy/Server then sends an RA back to the Client 2547 either directly or via an FHS Proxy/Server acting as a proxy. The 2548 Proxy/Server that returns the RA directly to the Client sets the 2549 (newly-created) ULA-MNP as the destination address and with the 2550 DHCPv6-PD Reply message and OMNI Neighbor Coordination header Preflen 2551 coded in the OMNI option. When the Client receives the RA, it 2552 creates a default route, assigns the Subnet Router Anycast address 2553 and sets its {ULA,XLA}-MNP based on the delegated MNP. 2555 Note: Further details of the DHCPv6-PD based MNP registration (as 2556 well as a minimal MNP delegation alternative that avoids including a 2557 DHCPv6 message sub-option in the RS) are found in 2558 [I-D.templin-6man-omni]. 2560 Note: when the Hub Proxy/Server forwards an RA to the Client via a 2561 different node acting as a FHS Proxy/Server, the Hub sets the RA 2562 destination to the same address that appeared in the RS source. The 2563 FHS Proxy/Server then subsequently sets the RA destination to the 2564 ULA-MNP when it forwards the Proxyed version of the RA to the Client 2565 - see [I-D.templin-6man-omni] for further details. 2567 3.13. AERO Route Optimization 2569 AERO nodes invoke route optimization when they need to forward 2570 initial packets to new target destinations over ANET/INET interfaces 2571 and for ongoing multilink forwarding for current destinations. Route 2572 optimization is based on IPv6 ND Address Resolution messaging between 2573 a Route Optimization Source (ROS) and a Relay or the target Client 2574 itself (reached via the current Hub Proxy/Server) acting as a Route 2575 Optimization Responder (ROR). Route optimization is initiated by the 2576 first eligible ROS closest to the source as follows: 2578 * For Clients on VPNed and Direct interfaces, the Client's FHS 2579 Proxy/Server is the ROS. 2581 * For Clients on ANET interfaces, either the Client or the FHS 2582 Proxy/Server may be the ROS. 2584 * For Clients on INET interfaces, the Client itself is the ROS. 2586 * For correspondent nodes on INET/ENET interfaces serviced by a 2587 Relay, the Relay is the ROS. 2589 * For Clients that engage the Hub Proxy/Server in "mobility anchor" 2590 mode, the Hub Proxy/Server is the ROS. 2592 * For peers within the same ANET/ENET, route optimization is through 2593 receipt of Redirect messages from a Proxy/Server. 2595 The AERO routing system directs a route optimization request sent by 2596 the ROS to the ROR, which returns a route optimization reply which 2597 must include information that is current, consistent and authentic. 2598 The ROS is responsible for periodically refreshing the route 2599 optimization, and the ROR is responsible for quickly informing the 2600 ROS of any changes. Following address resolution, the ROS and ROR 2601 perform ongoing multilink route optimizations to maintain optimal 2602 forwarding profiles. 2604 The route optimization procedures are specified in the following 2605 sections. 2607 3.13.1. Multilink Address Resolution 2609 When one or more original IP packets from a source node destined to a 2610 target node arrives, the ROS checks for a NCE with an XLA that 2611 matches the target destination. If there is a NCE in the REACHABLE 2612 state, the ROS invokes the OAL and forwards the resulting carrier 2613 packets according to the cached state then returns from processing. 2614 Otherwise, if there is no NCE the ROS creates one in the INCOMPLETE 2615 state. 2617 The ROS next prepares an NS message for Address Resolution (NS(AR)) 2618 to send toward an ROR while including the original IP packet(s) as 2619 trailing data following the NS(AR) in an OAL super-packet 2620 [I-D.templin-6man-omni]. The resulting NS(AR) message must be sent 2621 securely, and includes: 2623 * the ULA of the ROS as the source address. 2625 * the XLA corresponding to the original IP packet's destination as 2626 the Target Address, e.g., for 2001:db8:1:2::10:2000 the Target 2627 Address is fd00::2001:db8:1:2. 2629 * the Solicited-Node multicast address [RFC4291] formed from the 2630 lower 24 bits of the original IP packet's destination as the 2631 destination address, e.g., for 2001:db8:1:2::10:2000 the NS(AR) 2632 destination address is ff02:0:0:0:0:1:ff10:2000. 2634 The NS(AR) message also includes an OMNI option with an 2635 authentication sub-option if necessary and with OMNI extension header 2636 Preflen set to the prefix length associated with the NS(AR) source. 2637 The ROS also includes Interface Attributes and Traffic Selectors for 2638 all of the source Client's underlay interfaces, calculates the 2639 authentication signature or checksum, then selects an Identification 2640 value and submits the NS(AR) message for OAL encapsulation with OAL 2641 source set to its own ULA and OAL destination set to the XLA 2642 corresponding to the target and with window synchronization 2643 parameters. The ROS then inserts a fragment header, performs 2644 fragmentation and L2 encapsulation, then sends the resulting carrier 2645 packets into the SRT secured spanning tree without decrementing the 2646 network-layer TTL/Hop Limit field. 2648 When the ROS is a Client, it must instead use the ULA of one of its 2649 FHS Proxy/Servers as the OAL destination. The ROS Client then 2650 fragments, performs L2 encapsulation and forwards the carrier packets 2651 to the FHS Proxy/Server. The FHS Proxy/Server then reassembles, 2652 verifies the NS(AR) authentication signature or checksum, changes the 2653 OAL source to its own ULA, changes the OAL destination to the XLA 2654 corresponding to the target, selects an appropriate Identification, 2655 then re-fragments and forwards the resulting carrier packets into the 2656 secured spanning tree on behalf of the Client. 2658 Note: both the target Client and its Hub Proxy/Server include current 2659 and accurate information for the Client's multilink Interface 2660 Attributes profile. The Hub Proxy/Server can be trusted to provide 2661 an authoritative response on behalf of the Client should the need 2662 arise. While the Client has no such trust basis, any attempt by the 2663 Client to mount an attack by providing false Interface Attributes 2664 information would only result in black-holing of return traffic, 2665 i.e., the "attack" could only result in denial of service to the 2666 Client itself. Therefore, the Client's asserted Interface Attributes 2667 need not be validated by the Hub Proxy/Server. 2669 3.13.1.1. Relaying the NS(AR) *NET Packet(s) 2671 When the Gateway receives carrier packets containing the NS(AR), it 2672 discards the L2 headers and determines the next hop by consulting its 2673 standard IPv6 forwarding table for the OAL header destination 2674 address. The Gateway then decrements the OAL header Hop-Limit, then 2675 re-encapsulates and forwards the carrier packet(s) via the secured 2676 spanning tree the same as for any IPv6 router, where they may 2677 traverse multiple OMNI link segments. The final-hop Gateway will 2678 deliver the carrier packet via the secured spanning tree to the Hub 2679 Proxy/Server (or Relay) that services the target. 2681 3.13.1.2. Processing and Responding to the NS(AR) 2683 When the Hub Proxy/Server for the target receives the NS(AR) secured 2684 carrier packets with the XLA of the target as the OAL destination, it 2685 reassembles then forwards the message to the target Client (while 2686 including an authentication signature and encapsulation if necessary) 2687 or processes the NS(AR) locally if it is acting as a Relay/IP router 2688 or the Client's designated ROR. The Hub Proxy/Server processes the 2689 message as follows: 2691 * if the NS(AR) target matches a Client NCE in the DEPARTED state, 2692 the (old) Hub Proxy/Server re-encapsulates by setting the OAL 2693 destination address to the ULA of the Client's new Hub Proxy/ 2694 Server. The old Hub Proxy/Server then re-fragments and re- 2695 encapsulates, then forwards the resulting carrier packets over the 2696 secured spanning tree. 2698 * If the NS(AR) target matches a Client NCE in the REACHABLE state, 2699 the Hub Proxy/Server notes whether the NS(AR) arrived from the 2700 secured spanning tree then sets the OAL destination address to the 2701 ULA-MNP of the Client or the ULA of the selected FHS Proxy/Server 2702 for the Client (i.e., if the Hub is not also the FHS Proxy/Server 2703 for the selected underlay interface). If the message arrived via 2704 the secured spanning tree the Hub Proxy/Server verifies the 2705 checksum; otherwise, it must verify the message authentication 2706 signature before forwarding. When the Hub Proxy/Server determines 2707 the underlay interface for the target Client, it then changes the 2708 OAL destination to the ULA of the target Client's FHS Proxy/ 2709 Server, re-fragments and forwards the resulting carrier packets 2710 into the secured spanning tree. When the FHS Proxy/Server 2711 receives the carrier packets, it reassembles and verifies the 2712 checksum, then includes an authentication signature if necessary, 2713 changes the OAL source to its own ULA and destination to the ULA- 2714 MNP of the target Client, includes an Identification value within 2715 the current window, then re-fragments and forwards the resulting 2716 carrier packets to the target Client ROR. (Note that if the Hub 2717 and FHS Proxy/Server are one and the same the Hub itself will 2718 perform the FHS procedures.) 2720 * If the NS(AR) target matches one of its non-MNP routes, the Hub 2721 Proxy/Server serves as both a Relay and a ROR, since the Relay 2722 forwards IP packets toward the (fixed network) target at the 2723 network layer. 2725 The ROR then creates a NCE for the NS(AR) source address if 2726 necessary, processes the window synchronization parameters, caches 2727 all Interface Attributes and Traffic Selector information, and 2728 prepares a (solicited) NA(AR) message to return to the ROS with the 2729 source address set to its own ULA-MNP, the destination address set to 2730 the NS(AR) ULA source address and the Target Address set to the same 2731 value that appeared in the NS(AR) Target Address. The ROR includes 2732 an OMNI option with OMNI Neighbor Coordination header Preflen set to 2733 the prefix length associated with the NA(AR) source address. 2735 The ROR then sets the NA(AR) message R flag to 1 (as a router) and S 2736 flag to 1 (as a response to a solicitation) and sets the O flag to 1 2737 (as an authoritative responder). The ROR finally submits the NA(AR) 2738 for OAL encapsulation with source set to its own ULA and destination 2739 set to either the ULA corresponding to the NS(AR) source or the ULA 2740 of its FHS Proxy/Server, selects an appropriate Identification, and 2741 includes window synchronization parameters and authentication 2742 signature or checksum. The ROR then includes Interface Attributes 2743 and Traffic Selector sub-options for all of the target's underlay 2744 interfaces with current information for each interface, fragments and 2745 encapsulates each fragment in appropriate L2 headers, then forwards 2746 the resulting (L2-encapsulated) carrier packets to the FHS Proxy/ 2747 Server. 2749 When the FHS Proxy/Server receives the carrier packets, it 2750 reassembles if necessary and verifies the authentication signature or 2751 checksum. The FHS Proxy/Server then changes the OAL source address 2752 to its own ULA, changes the destination to the ULA or XLA 2753 corresponding to the NA(AR) destination, includes an appropriate 2754 Identification, then fragments and forwards the carrier packets into 2755 the secured spanning tree. 2757 Note: If the Hub Proxy/Server is acting as the Client's ROR but not 2758 as a Relay/IP router (i.e., by virtue of receipt of an RS message 2759 with the A flag set), it prepares the NS(AR) with the R flag set to 0 2760 but without setting the SYN flag in the OMNI Neighbor Coordination 2761 header window synchronization parameters. This informs the ROS that 2762 it must initiate multilink route optimization to synchronize with the 2763 Client either directly or via a FHS Proxy/Server (see: 2764 Section 3.13.2). 2766 3.13.1.3. Relaying the NA(AR) 2768 When the Gateway receives NA(AR) carrier packets, it discards the L2 2769 headers and determines the next hop by consulting its standard IPv6 2770 forwarding table for the OAL header destination address. The Gateway 2771 then decrements the OAL header Hop-Limit, re-encapsulates the carrier 2772 packet and forwards it via the SRT secured spanning tree, where it 2773 may traverse multiple OMNI link segments. The final-hop Gateway will 2774 deliver the carrier packet via the secured spanning tree to a Proxy/ 2775 Server for the ROS. 2777 3.13.1.4. Processing the NA(AR) 2779 When the ROS receives the NA(AR) message, it first searches for a NCE 2780 that matches the NA(AR) target address. The ROS then processes the 2781 message the same as for standard IPv6 Address Resolution [RFC4861]. 2782 In the process, it caches all OMNI option information in the target 2783 NCE (including all Interface Attributes), and caches the NA(AR) XLA 2784 source address as the address of the target Client. 2786 When the ROS is a Client, the SRT secured spanning tree will first 2787 deliver the solicited NA(AR) message to the FHS Proxy/Server, which 2788 re-encapsulates and forwards the message to the Client. If the 2789 Client is on a well-managed ANET, physical security and protected 2790 spectrum ensures security for the NA(AR) without needing an 2791 additional authentication signature; if the Client is on the open 2792 INET the Proxy/Server must instead include an authentication 2793 signature (while adjusting the OMNI option size, if necessary). The 2794 Proxy/Server uses its own ULA as the OAL source and the ULA-MNP of 2795 the Client as the OAL destination. 2797 3.13.2. Multilink Route Optimization 2799 Following address resolution, the ROS and ROR can assert multilink 2800 paths through underlay interface pairs serviced by the same source/ 2801 destination ULAs by sending unicast NS/NA messages with Multilink 2802 Forwarding Parameters and OMNI Neighbor Coordination window 2803 synchronization parameters when necessary. The unicast NS/NA 2804 messages establish multilink forwarding state in intermediate nodes 2805 in the path between the ROS and ROR. 2807 To support multilink route optimization, OMNI interfaces include an 2808 additional forwarding table termed the Multilink Forwarding 2809 Information Base (MFIB) that supports carrier packet forwarding based 2810 on OMNI neighbor underlay interface pairs. The MFIB contains 2811 Multilink Forwarding Vectors (MFVs) indexed by 4-octet values known 2812 as MFV Indexes (MFVIs). 2814 OAL source, intermediate and destination nodes create MFVs/MFVIs when 2815 they process an NS message with a Multilink Forwarding Parameters 2816 sub-option with Job code '00' (Initialize; Build B) or a solicited NA 2817 with Job code '01' (Follow B; Build A) (see: 2818 [I-D.templin-6man-omni]). The OAL source of the NS (and OAL 2819 destination of the solicited NA) are considered to reside in the 2820 "First Hop Segment (FHS)", while the OAL destination of the NS (and 2821 OAL source of the solicited NA) are considered to reside in the "Last 2822 Hop Segment (LHS)". 2824 When an OAL node processes an NS with Job code '00', it creates an 2825 MFV, records the NS source and destination ULAs and assigns a "B" 2826 MFVI. When the "B" MVFI is referenced, the MVF retains the ULAs in 2827 (dst,src) order the opposite of how they appeared in the original NS 2828 to support full header reconstruction. (If the NS message included a 2829 nested OAL encapsulation, the ULAs of both OAL headers are retained.) 2831 When an OAL node processes a solicited NA with Job code '01', it 2832 locates the MFV created by the NS and assigns an "A" MFVI. When the 2833 "A" MFVI is referenced, the MFV retains the ULAs in (src,dst) order 2834 the same as they appeared in the original NS to support full header 2835 reconstruction. (If the NS message included a nested OAL 2836 encapsulation, the ULAs of both OAL headers are retained.) 2838 OAL nodes generate random 32-bit values as candidate A/B MFVIs which 2839 must first be tested for local uniqueness. If a candidate MFVI s 2840 already in use (or if the value is 0), the OAL node repeats the 2841 random generation process until it obtains a unique non-zero value. 2842 (Since the number of MFVs in service at each OAL node is likely to be 2843 much smaller than 2**32, the process will generate a unique value 2844 after a small number of tries; also, an MFVI generated by a first OAL 2845 node is never tested for uniqueness on other OAL nodes, since the 2846 uniqueness property is node-local only.) 2848 OAL nodes maintain A/B MFVIs as follows: 2850 * "B1" - a locally-unique MFVI maintained independently by each OAL 2851 node on the path from the FHS OAL source to the last OAL 2852 intermediate node before the LHS OAL destination. The OAL node 2853 generates and assigns a "B1" MFVI to a newly-created MFV when it 2854 processes an NS message with Job code '00'. When the OAL node 2855 receives future carrier packets that include this value, it can 2856 unambiguously locate the correct MFV and determine directionality 2857 without examining addresses. 2859 * "A1" - a locally unique MFVI maintained independently by each OAL 2860 node on the path from the LHS OAL source to the last OAL 2861 intermediate node before the FHS OAL destination. The OAL node 2862 generates and assigns an "A1" MFVI to the MVF that configures the 2863 corresponding "B1" MFVI when it processes a solicited NA message 2864 with Job code '01'. When the OAL node receives future carrier 2865 packets that include this value, it can unambiguously locate the 2866 correct MFV and determine directionality without examining 2867 addresses. 2869 * "A2" - the A1 MFVI of a remote OAL node discovered by an FHS OAL 2870 source or OAL intermediate node when it processes an NA message 2871 with Job code '01' that originated from an LHS OAL source. A2 2872 values MUST NOT be tested for uniqueness within the OAL node's 2873 local context. 2875 * "B2" - the B1 MFVI of a remote OAL node discovered by an LHS OAL 2876 source or OAL intermediate node when it processes an NS message 2877 with Job code '00' that originated from an FHS OAL source. B2 2878 values MUST NOT be tested for uniqueness within the OAL node's 2879 local context. 2881 When an FHS OAL source has an original IP packet to send to an LHS 2882 OAL destination discovered via multilink address resolution, it first 2883 selects a source and target underlay interface pair. The OAL source 2884 uses its cached information for the target underlay interface as LHS 2885 information then prepares an NS message with an OMNI Multilink 2886 Forwarding Parameters sub-option with Job code '00' and with source 2887 set to its own ULA or XLA. If the LHS FMT-Forward and FMT-Mode bits 2888 are both clear, the OAL source sets the destination to the ULA of the 2889 LHS Proxy/Server; otherwise, it sets the destination to the XLA of 2890 the target Client. The OAL source then sets window synchronization 2891 information in the OMNI Neighbor Coordination header and creates a 2892 NCE for the selected destination ULA or XLA in the INCOMPLETE state. 2893 The OAL source next creates an MFV based on the NS source and 2894 destination ULAs, then generates a "B1" MFVI and assigns it to the 2895 MFV while also including it as the first B entry in the MFVI List. 2896 The OAL source then populates the NS Multilink Forwarding Parameters 2897 based on any FHS/LHS information it knows locally. OAL intermediate 2898 nodes on the path to the OAL destination may populate additional FHS/ 2899 LHS information on a hop-by-hop basis. 2901 If the OAL source is the FHS Proxy/Server, it then performs OAL 2902 encapsulation/fragmentation while setting the source to its own ULA 2903 and setting the destination to the FHS Subnet Router Anycast ULA 2904 determined by applying the FHS SRT prefix length to its ULA. The FHS 2905 Proxy/Server next examines the LHS FMT code. If FMT-Forward is clear 2906 and FMT-Mode is set, the FHS Proxy/Server checks for a NCE for the 2907 ULA of the LHS Proxy/Server. If there is no NCE, the FHS Proxy/ 2908 Server creates one in the INCOMPLETE state. If a new NCE was created 2909 (or if the existing NCE requires fresh window synchronization), the 2910 FHS Proxy/Server then writes window synchronization parameters into 2911 the OMNI Multilink Forwarding Parameters Tunnel Window 2912 Synchronization fields. The FHS Proxy/Server then selects an 2913 appropriate Identification value and L2 headers and forwards the 2914 resulting carrier packets into the secured spanning tree which will 2915 deliver them to a Gateway interface that assigns the FHS Subnet 2916 Router Anycast ULA. 2918 If the OAL source is the FHS Client, it instead includes an 2919 authentication signature if necessary, performs OAL encapsulation, 2920 sets the source to its own ULA-MNP, sets the destination to the ULA 2921 of the FHS Proxy/Server and selects an appropriate Identification 2922 value for the FHS Proxy/Server. If FHS FMT-Forward is set and LHS 2923 FMT-Forward is clear, the FHS Client creates/updates a NCE for the 2924 ULA of the LHS Proxy/Server as above and includes Tunnel Window 2925 Synchronization parameters. The FHS Client then fragments and 2926 encapsulates in appropriate L2 headers then forwards the carrier 2927 packets to the FHS Proxy/Server. When the FHS Proxy/Server receives 2928 the carrier packets, it verifies the Identification, reassembles/ 2929 decapsulates to obtain the NS then verifies the authentication 2930 signature or checksum. The FHS Proxy/Server then creates an MFV 2931 (i.e., the same as the FHS Client had done) while assigning the 2932 current B entry in the MFVI List (i.e., the one included by the FHS 2933 Client) as the "B2" MFVI for this MVF. The FHS Proxy/Server next 2934 generates a new unique "B1" MFVI, then both assigns it to the MFV and 2935 writes it as the next B entry in the OMNI Multilink Forwarding 2936 Parameters MFVI List (while also writing any FHS Client and Proxy/ 2937 Server addressing information). The FHS Proxy/Server then checks 2938 FHS/LHS FMT-Forward/Mode to determine whether to create a NCE for the 2939 LHS Proxy/Server ULA and include Tunnel Window Synchronization 2940 parameters the same as above. The FHS Proxy/Server then calculates 2941 the checksum, re-fragments while setting the OAL source address to 2942 its own ULA and destination address to the FHS Subnet Router Anycast 2943 ULA, and includes an Identification appropriate for the secured 2944 spanning tree. The FHS Proxy/Server finally includes appropriate L2 2945 headers and forwards the carrier packets into the secured spanning 2946 tree the same as above. 2948 Gateways in the spanning tree forward carrier packets not explicitly 2949 addressed to themselves, while forwarding those that arrived via the 2950 secured spanning tree to the next hop also via the secured spanning 2951 tree and forwarding all others via the unsecured spanning tree. When 2952 an FHS Gateway receives a carrier packet over the secured spanning 2953 tree addressed to its ULA or the FHS Subnet Router Anycast ULA, it 2954 instead reassembles/decapsulates to obtain the NS then verifies the 2955 checksum. The FHS Gateway next creates an MFV (i.e., the same as the 2956 FHS Proxy/Server had done) while assigning the current B entry in the 2957 MFVI List as the MFV "B2" index. The FHS Gateway also caches the NS 2958 Multilink Forwarding Parameters FHS information in the MFV, and also 2959 caches the first B entry in the MFVI List as "FHS-Client" when FHS 2960 FMT-Forward/Mode are both set to enable future direct forwarding to 2961 this FHS Client. The FHS Gateway then generates a "B1" MFVI for the 2962 MFV and also writes it as the next B entry in the NS's MFVI List. 2964 The FHS Gateway then examines the SRT prefixes corresponding to both 2965 FHS and LHS. If the FHS Gateway has a local interface connection to 2966 both the FHS and LHS (whether they are the same or different 2967 segments), the FHS/LHS Gateway caches the NS LHS information, writes 2968 its ULA suffix and LHS INADDR into the NS OMNI Multilink Forwarding 2969 Parameters LHS fields, then sets its own ULA as the source and the 2970 ULA of the LHS Proxy/Server as the destination while selecting an 2971 appropriate identification. If the FHS and LHS prefixes are 2972 different, the FHS Gateway instead sets the LHS Subnet Router Anycast 2973 ULA as the destination. The FHS Gateway then recalculates the NS 2974 checksum, selects an appropriate Identification and L2 headers as 2975 above then forwards the carrier packets into the secured spanning 2976 tree. 2978 When the FHS and LHS Gateways are different, the LHS Gateway will 2979 receive carrier packets over the secured spanning tree from the FHS 2980 Gateway. The LHS Gateway reassembles/decapsulates to obtain the NS 2981 then verifies the checksum and creates an MFV (i.e., the same as the 2982 FHS Gateway had done) while assigning the current B entry in the MFVI 2983 List as the MFV "B2" index. The LHS Gateway also caches the ULA of 2984 the FHS Gateway found in the Multilink Forwarding Parameters as the 2985 spanning tree address for "B2", caches the NS Multilink Forwarding 2986 Parameters LHS information then generates a "B1" MFVI for the MFV 2987 while also writing it as the next B entry in the MFVI List. The LHS 2988 Gateway also writes its own ULA suffix and LHS INADDR into the OMNI 2989 Multilink Forwarding Parameters. The LHS Gateway then sets the its 2990 own ULA as the source and the ULA of the LHS Proxy/Server as the OAL 2991 destination, recalculates the checksum, selects an appropriate 2992 Identification, then fragments while including appropriate L2 headers 2993 and forwards the carrier packets into the secured spanning tree. 2995 When the LHS Proxy/Server receives the carrier packets from the 2996 secured spanning tree, it reassembles/decapsulates to obtain the NS, 2997 verifies the checksum then verifies that the LHS information supplied 2998 by the FHS source is consistent with its own cached information. If 2999 the information is consistent, the LHS Proxy/Server then creates an 3000 MFV and assigns the current B entry in the MFVI List as the "B2" MFVI 3001 the same as for the prior hop. If the NS destination is the XLA of 3002 the target Client, the LHS Proxy/Server also generates a "B1" MFVI 3003 and assigns it both to the MFVI and as the next B entry in the MFVI 3004 List. The LHS Proxy/Server then examines FHS FMT; if FMT-Forward is 3005 clear and FMT-Mode is set, the LHS Proxy/Server creates a NCE for the 3006 ULA of the FHS Proxy/Server (if necessary) and sets the state to 3007 STALE, then caches any Tunnel Window Synchronization parameters. 3009 If the NS destination is its own ULA, the LHS Proxy/Server next 3010 prepares to return a solicited NA with Job code '01'. If the NS 3011 source was the XLA of the FHS Client, the LHS Proxy/Server first 3012 creates or updates an NCE for the XLA with state set to STALE. The 3013 LHS Proxy/Server next caches the NS OMNI Neighbor Coordination header 3014 window synchronization parameters and Multilink Forwarding Parameters 3015 information (including the MFVI List) in the NCE corresponding to the 3016 ULA source. When the LHS Proxy/Server forwards future carrier 3017 packets based on the NCE, it can populate reverse-path forwarding 3018 information in a CRH-32 routing header to enable forwarding based on 3019 the cached MFVI List B entries instead of ULA addresses. 3021 The LHS Proxy/Server then creates an NA with Job code '01' while 3022 copying the NS OMNI Multilink Forwarding Parameters FHS/LHS 3023 information into the corresponding fields in the NA. The LHS Proxy/ 3024 Server then generates an "A1" MFVI and both assigns it to the MFV and 3025 includes it as the first A entry in NA's MFVI List (see: 3026 [I-D.templin-6man-omni] for details on MFVI List A/B processing). 3027 The LHS Proxy/Server then includes end-to-end window synchronization 3028 parameters in the OMNI Neighbor Coordination header (if necessary) 3029 and also tunnel window synchronization parameters in the Multilink 3030 Forwarding Parameters (if necessary). The LHS Proxy/Server then 3031 encapsulates the NA, calculates the checksum, sets the source to its 3032 own ULA, sets the destination to the ULA of the LHS Gateway, selects 3033 an appropriate Identification value and L2 headers then forwards the 3034 carrier packets into the secured spanning tree. 3036 If the NS destination was the XLA of the LHS Client, the LHS Proxy/ 3037 Server instead includes an authentication signature in the NS if 3038 necessary (otherwise recalculates the checksum), then changes the OAL 3039 source to its own ULA and changes the destination to the ULA-MNP of 3040 the LHS Client. The LHS Proxy/Server then selects an appropriate 3041 Identification value, fragments if necessary, includes appropriate L2 3042 headers and forwards the carrier packets to the LHS Client. When the 3043 LHS Client receives the carrier packets, it verifies the 3044 Identification and reassembles/decapsulates to obtain the NS then 3045 verifies the authentication signature or checksum. The LHS Client 3046 then creates a NCE for the NS ULA source address in the STALE state. 3047 If LHS FMT-Forward is set, FHS FMT-Forward is clear and the NS source 3048 was an XLA, the Client also creates a NCE for the ULA of the FHS 3049 Proxy/Server in the STALE state and caches any Tunnel Window 3050 Synchronization parameters. The Client then caches the NS OMNI 3051 Neighbor Coordination header window synchronization parameters and 3052 Multilink Forwarding Parameters in the NCE corresponding to the NS 3053 ULA source, then creates an MFV and assigns both the current MFVI 3054 List B entry as "B2" and a locally generated "A1" MFVI the same as 3055 for previous hops (the LHS Client also includes the "A1" value in the 3056 solicited NA - see above and below). The LHS Client also caches the 3057 previous MFVI List B entry as "LHS-Gateway" since it can include this 3058 value when it sends future carrier packets directly to the Gateway 3059 (following appropriate neighbor coordination). 3061 The LHS Client then prepares an NA using exactly the same procedures 3062 as for the LHS Proxy/Server above, except that it uses its XLA as the 3063 source and the ULA or XLA of the FHS correspondent as the 3064 destination. The LHS Client also includes an authentication 3065 signature if necessary (otherwise calculates the checksum), then 3066 encapsulates the NA with OAL source set to its own ULA-MNP and 3067 destination set to the ULA of the LHS Proxy/Server, includes an 3068 appropriate Identification and L2 headers and forwards the carrier 3069 packets to the LHS Proxy/Server. When the LHS Proxy/Server receives 3070 the carrier packets, it verifies the Identifications, reassembles/ 3071 decapsulates to obtain the NA, verifies the authentication signature 3072 or checksum, then uses the current MVFI List B entry to locate the 3073 MFV. The LHS Proxy/Server then writes the current MFVI List A entry 3074 as the "A2" value for the MVF, generates an "A1" MFVI and both 3075 assigns it to the MFV and writes it as the next MFVI List A entry. 3076 The LHS Proxy/Server then examines the FHS/LHS FMT codes to determine 3077 if it needs to include Tunnel Window Synchronization parameters. The 3078 LHS Proxy/Server then recalculates the checksum, re-fragments the NA 3079 while setting the OAL source to its own ULA and destination to the 3080 ULA of the LHS Gateway, includes an appropriate Identification and L2 3081 headers and forwards the carrier packets into the secured spanning 3082 tree. 3084 When the LHS Gateway receives the carrier packets, it reassembles/ 3085 decapsulates to obtain the NA while verifying the checksum then uses 3086 the current MFVI List B entry to locate the MFV. The LHS Gateway 3087 then writes the current MFVI List A entry as the MFV "A2" index and 3088 generates a new "A1" value which it both assigns the MFV and writes 3089 as the next MFVI List A entry. (The LHS Gateway also caches the 3090 first A entry in the MFVI List as "LHS-Client" when LHS FMT-Forward/ 3091 Mode are both set to enable future direct forwarding to this LHS 3092 Client.) If the LHS Gateway is connected directly to both the FHS 3093 and LHS segments (whether the segments are the same or different), 3094 the FHS/LHS Gateway will have already cached the FHS/LHS information 3095 based on the original NS. The FHS/LHS Gateway recalculates the 3096 checksum then re-fragments the NA while setting the OAL source to its 3097 own ULA and destination to the ULA of the FHS Proxy/Server. If the 3098 FHS and LHS prefixes are different, the FHS Gateway instead re- 3099 fragments while setting the destination to the ULA of the FHS 3100 Gateway. The LHS Gateway selects an appropriate Identification and 3101 L2 headers then forwards the carrier packets into the secured 3102 spanning tree. 3104 When the FHS and LHS Gateways are different, the FHS Gateway will 3105 receive the carrier packets from the LHS Gateway over the secured 3106 spanning tree. The FHS Gateway reassembles/decapsulates to obtain 3107 the NA while verifying the checksum, then locates the MFV based on 3108 the current MFVI List B entry. The FHS Gateway then assigns the 3109 current MFVI List A entry as the MFV "A2" index and caches the ULA of 3110 the LHS Gateway as the spanning tree address for "A2". The FHS 3111 Gateway then generates an "A1" MVFI and both assigns it to the MVF 3112 and writes it as the next MFVI List A entry while also writing its 3113 ULA and INADDR in the NA FHS Gateway fields. The FHS Gateway then 3114 recalculates the checksum, re-encapsulates/re-fragments with its own 3115 ULA as the source, with the ULA of the FHS Proxy/Server as the 3116 destination, then selects an appropriate Identification value and L2 3117 headers and forwards the carrier packets into the secured spanning 3118 tree. 3120 When the FHS Proxy/Server receives the carrier packets from the 3121 secured spanning tree, it reassembles/decapsulates to obtain the NA 3122 while verifying the checksum then locates the MFV based on the 3123 current MFVI List B entry. The FHS Proxy/Server then assigns the 3124 current MFVI List A entry as the "A2" MFVI the same as for the prior 3125 hop. If the NA destination is its own ULA, the FHS Proxy/Server then 3126 caches the NA Multilink Forwarding Parameters with the MFV and 3127 examines LHS FMT. If FMT-Forward is clear, the FHS Proxy/Server 3128 locates the NCE for the ULA of the LHS Proxy/Server and sets the 3129 state to REACHABLE then caches any Tunnel Window Synchronization 3130 parameters. If the NA source is the XLA of the LHS Client, the FHS 3131 Proxy/Server then locates the LHS Client NCE and sets the state to 3132 REACHABLE then caches the OMNI Neighbor Coordination header window 3133 synchronization parameters and prepares to return an NA 3134 acknowledgement, if necessary. 3136 If the NA destination is the XLA of the FHS Client, the FHS Proxy/ 3137 Server also searches for and updates the NCE for the ULA of the LHS 3138 Proxy/Server if necessary the same as above. The FHS Proxy/Server 3139 then generates an "A1" MFVI and assigns it both to the MFVI and as 3140 the next MFVI List A entry, then includes an authentication signature 3141 or checksum in the NA message. The FHS Proxy/Server then sets the 3142 OAL source to its own ULA and sets the destination to the ULA-MNP of 3143 the FHS Client, then selects an appropriate Identification value and 3144 L2 headers and forwards the carrier packets to the FHS Client. 3146 When the FHS Client receives the carrier packets, it verifies the 3147 Identification, reassembles/decapsulates to obtain the NA, verifies 3148 the authentication signature or checksum, then locates the MFV based 3149 on the current MFVI List B entry. The FHS Client then assigns the 3150 current MFVI List A entry as the "A2" MFVI the same as for the prior 3151 hop. The FHS Client then caches the NA Multilink Forwarding 3152 Parameters (including the MFVI List) with the MFV and examines LHS 3153 FMT. If FMT-Forward is clear, the FHS Client locates the NCE for the 3154 ULA of the LHS Proxy/Server and sets the state to REACHABLE then 3155 caches any Tunnel Window Synchronization parameters. If the NA 3156 source is the XLA of the LHS Client, the FHS Proxy/Server then 3157 locates the LHS Client NCE and sets the state to REACHABLE then 3158 caches the OMNI Neighbor Coordination header window synchronization 3159 parameters and prepares to return an NA acknowledgement, if 3160 necessary. The FHS Client also caches the previous MFVI List A entry 3161 as "FHS-Gateway" since it can include this value when it sends future 3162 carrier packets directly to the Gateway (following appropriate 3163 neighbor coordination). 3165 If either the FHS Client or FHS Proxy/Server needs to return an 3166 acknowledgement to complete window synchronization, it prepares a uNA 3167 message with an OMNI Multilink Forwarding Parameters sub-option with 3168 Job code set to '10' (Follow A; Record B) (note that this step is 3169 unnecessary when Rapid Commit route optimization is used per 3170 Section 3.13.3). The FHS node sets the source to its own ULA or XLA, 3171 sets the destination to the ULA or XLA of the LHS node then includes 3172 Tunnel Window Synchronization parameters if necessary. The FHS node 3173 next sets the MFVI List to the cached list of A entries received in 3174 the Job code '01' NA, but need not set any other FHS/LHS information. 3175 The FHS node then encapsulates the uNA message in an OAL header with 3176 its own ULA as the source. If the FHS node is the Client, it next 3177 sets the ULA of the FHS Proxy/Server as the OAL destination, includes 3178 an authentication signature or checksum, selects an appropriate 3179 Identification value and L2 headers and forwards the carrier packets 3180 to the FHS Proxy/Server. The FHS Proxy/Server then verifies the 3181 Identification, reassembles/decapsulates, verifies the authentication 3182 signature or checksum, then uses the current MFVI List A entry to 3183 locate the MFV. The FHS Proxy/Server then writes its "B1" MFVI as 3184 the next MFVI List B entry and determines whether it needs to include 3185 Tunnel Window Synchronization parameters the same as it had done when 3186 it forwarded the original NS. 3188 The FHS Proxy/Server recalculates the uNA checksum then re-fragments 3189 while setting its own ULA as the source and the ULA of the FHS 3190 Gateway as the destination, then selects an appropriate 3191 Identification and L2 headers and forwards the carrier packets into 3192 the secured spanning tree. When the FHS Gateway receives the carrier 3193 packets, it reassembles/decapsulates to obtain the uNA while 3194 verifying the checksum then uses the current MFVI List A entry to 3195 locate the MFV. The FHS Gateway then writes its "B1" MFVI as the 3196 next MFVI List B entry, then re-fragments while setting the OAL 3197 source and destination. If the FHS Gateway is also the LHS Gateway, 3198 it sets the ULA of the LHS Proxy/Server as the destination; otherwise 3199 it sets the ULA of the LHS Gateway. The FHS Gateway recalculates the 3200 checksum then selects an appropriate Identification and L2 headers, 3201 re-fragments/forwards the carrier packets into the secured spanning 3202 tree. If an LHS Gateway receives the carrier packets, it processes 3203 them exactly the same as the FHS Gateway had done while setting the 3204 carrier packet destination to the ULA of the LHS Proxy/Server. 3206 When the LHS Proxy/Server receives the carrier packets, it 3207 reassembles/decapsulates to obtain the uNA message while verifying 3208 the checksum. The LHS Proxy/Server then locates the MFV based on the 3209 current MFVI List A entry then determines whether it is a tunnel 3210 ingress the same as for the original NS. If it is a tunnel ingress, 3211 the LHS Proxy/Server updates the NCE for the tunnel far-end based on 3212 the Tunnel Window Synchronization parameters. If the uNA destination 3213 is its own ULA, the LHS Proxy/Server next updates the NCE for the 3214 source ULA based on the OMNI Neighbor Coordination header window 3215 synchronization parameters and MAY compare the MVFI List to the 3216 version it had cached in the MFV based on the original NS. 3218 If the uNA destination is the XLA of the LHS Client, the LHS Proxy/ 3219 Server instead writes its "B1" MFV as the next MFVI List B entry, 3220 includes an authentication signature or checksum, writes its own ULA 3221 as the OAL source and the ULA-MNP of the Client as the OAL 3222 destination then selects an appropriate Identification and L2 headers 3223 and forwards the resulting carrier packets to the LHS Client. When 3224 the LHS Client receives the carrier packets, it verifies the 3225 Identification, reassembles/decapsulates to obtain the uNA, verifies 3226 the authentication signature or checksum then processes the message 3227 exactly the same as for the LHS Proxy/Server case above. 3229 Following the NS/NA exchange with Multilink Forwarding Parameters, 3230 OAL end systems and tunnel endpoints can begin exchanging ordinary 3231 carrier packets with Identification values within their respective 3232 send/receive windows without requiring security signatures and/or 3233 secured spanning tree traversal. Either peer can refresh window 3234 synchronization parameters and/or send other carrier packets 3235 requiring security at any time using the same secured procedures 3236 described above. OAL end systems and intermediate nodes can also use 3237 their own A1/B1 MFVIs when they receive carrier packets to 3238 unambiguously locate the correct MFV and determine directionality and 3239 can use any discovered A2/B2 MFVIs to forward carrier packets to 3240 other OAL nodes that configure the corresponding A1/B1 MFVIs. When 3241 an OAL node uses an MFVI included in a carrier packet to locate an 3242 MFV, it need not also examine the carrier packet addresses. 3244 OAL sources can also begin including CRH-32s in carrier packets with 3245 a list of A/B MFVIs that OAL intermediate nodes can use for shortest- 3246 path carrier packet forwarding based on MFVIs instead of spanning 3247 tree addresses. OAL sources and intermediate nodes can also begin 3248 forwarding carrier packets with OAL compressed headers termed "OCH" 3249 (see: [I-D.templin-6man-omni]) that include only a single A/B MFVI 3250 meaningful to the next hop, since all nodes in the path up to (and 3251 sometimes including) the OAL destination have already established MFV 3252 forwarding information. Note that when an FHS OAL source receives a 3253 solicited NA with Job code '01', the message will contain an MFVI 3254 List with A entries populated in the reverse order needed for 3255 populating a CRH-32 routing header. The FHS OAL source must 3256 therefore write the MFVI List A entries last-to-first when it 3257 populates a CRH-32, or must select the correct A entry to include in 3258 an OCH header based on the intended OAL intermediate node or 3259 destination. 3261 When a Gateway receives unsecured carrier packets destined to a local 3262 segment Client that has asserted direct reachability, the Gateway 3263 performs direct carrier packet forwarding while bypassing the local 3264 Proxy/Server based on the Client's advertised MFVIs and discovered 3265 NATed INADDR information (see: Section 3.13.4). If the Client cannot 3266 be reached directly (or if NAT traversal has not yet converged), the 3267 Gateway instead forwards carrier packets directly to the local Proxy/ 3268 Server. 3270 When a Proxy/Server receives carrier packets destined to a local 3271 Client or forwards carrier packets received from a local Client, it 3272 first locates the correct MFV. If the carrier packets include a 3273 secured IPv6 ND message, the Proxy/Server uses the Client's NCE 3274 established through RS/RA exchanges to re-encapsulate/re-fragment 3275 while forwarding outbound secured carrier packets via the secured 3276 spanning tree and forwarding inbound secured carrier packets while 3277 including an authentication signature or checksum. For ordinary 3278 carrier packets, the Proxy/Server uses the same MFV if directed by 3279 MFVI and/or OAL addressing. Otherwise it locates an MFV established 3280 through an NS/NA exchange between the Client and the remote peer, and 3281 forwards the carrier packets without first reassembling/ 3282 decapsulating. 3284 When a Proxy/Server or Client configured as a tunnel ingress receives 3285 a carrier packet with a full OAL header with a ULA-MNP source and 3286 CRH-32 routing header, or an OCH header with an MFVI that matches an 3287 MFV, the ingress encapsulates the carrier packet in a new full OAL 3288 header or an OCH header containing the next hop MVFI and an 3289 Identification value appropriate for the end-to-end window and the 3290 outer header containing an Identification value appropriate for the 3291 tunnel endpoints. When a Proxy/Server or Client configured as a 3292 tunnel egress receives an encapsulated carrier packet, it verifies 3293 the Identification in the outer header, then discards the outer 3294 header and forwards the inner carrier packet to the final 3295 destination. 3297 When a Proxy/Server with FMT-Forward/Mode set to 0/1 for a source 3298 Client receives carrier packets from the source Client, it first 3299 reassembles to obtain the original OAL packet then re-fragments if 3300 necessary to cause the Client's packets to match the MPS on the path 3301 from the Proxy/Server as a tunnel ingress to the tunnel egress. The 3302 Proxy/Server then performs OAL-in-OAL encapsulation and forwards the 3303 resulting carrier packets to the tunnel egress. When a Proxy/Server 3304 with FMT-Forward/Mode set to 0/1 for a target Client receives carrier 3305 packets from a tunnel ingress, it first decapsulates to obtain the 3306 original fragments then reassembles to obtain the original OAL 3307 packet. The Proxy/Server then re-fragments if necessary to cause the 3308 fragments to match the target Client's underlay interface (Path) MTU 3309 and forwards the resulting carrier packets to the target Client. 3311 When a source Client forwards carrier packets it can employ header 3312 compression according to the MFVIs established through an NS/NA 3313 exchange with a remote or local peer. When the source Client 3314 forwards to a remote peer, it can forward carrier packets to a local 3315 SRT Gateway (following the establishment of INADDR information) while 3316 bypassing the Proxy/Server (see: Section 3.13.4). When a target 3317 Client receives carrier packets that match a local MFV, the Client 3318 first verifies the Identification then decompresses the headers if 3319 necessary, reassembles if necessary to obtain the OAL packet then 3320 decapsulates and delivers the IP packet to upper layers. 3322 When synchronized peer Clients in the same SRT segment with FMT- 3323 Forward and FMT-Mode set discover each other's NATed INADDR 3324 addresses, they can exchange carrier packets directly with header 3325 compression using MFVIs discovered as above (see: Section 3.13.5). 3326 The FHS Client will have cached the A MFVI for the LHS Client, which 3327 will have cached the B MVFI for the FHS Client. 3329 After window synchronization state has been established, the ROS and 3330 ROR can begin forwarding carrier packets while performing additional 3331 NS/NA exchanges as above to update window state, register new 3332 interface pairs for optimized multilink forwarding and/or confirm 3333 reachability. The ROS sends carrier packets to the FHS Gateway 3334 discovered through the NS/NA exchange. The FHS Gateway then forwards 3335 the carrier packets over the unsecured spanning tree to the LHS 3336 Gateway, which forwards them via LHS encapsulation to the LHS Proxy/ 3337 Server or directly to the target Client itself. The target Client in 3338 turn sends packets to the ROS in the reverse direction while 3339 forwarding through the Gateways to minimize Proxy/Server load 3340 whenever possible. 3342 While the ROS continues to actively forward packets to the target 3343 Client, it is responsible for updating window synchronization state 3344 and per-interface reachability before expiration. Window 3345 synchronization state is shared by all underlay interfaces in the 3346 ROS' NCE that use the same destination ULA so that a single NS/NA 3347 exchange applies for all interfaces regardless of the specific 3348 interface used to conduct the exchange. However, the window 3349 synchronization exchange only confirms target Client reachability 3350 over the specific underlay interface pair. Reachability for other 3351 underlay interfaces that share the same window synchronization state 3352 must be determined individually using additional NS/NA messages. 3354 3.13.3. Rapid Commit Route Optimization 3356 When the ROR receives an NS(AR) with a set of Interface Attributes 3357 for the source Client, it can perform "rapid commit" by immediately 3358 invoking multilink route optimization as above instead of returning 3359 an NA(AR). In order to perform rapid commit, the ROR prepares a 3360 unicast NS message with an OMNI option with window synchronization 3361 information responsive to the NS(AR), with a Multilink Forwarding 3362 Parameters sub-option selected for a specific underlay interface pair 3363 and with Interface Attributes for all of the ROR's other underlay 3364 interfaces. The ROR can also include ordinary IP packets as OAL 3365 super-packet extensions to the NS message if it has immediate data to 3366 send to the ROS. The ROR then returns the NS to the ROS the same as 3367 for the NA(AR) case. 3369 When the NS message traverses the return path to the ROR, all 3370 intermediate nodes in the path establish state exactly the same as 3371 for an ordinary NS/NA multilink route optimization exchange. When 3372 the NS message arrives at the ROS, the window synchronization 3373 parameters confirm that the NS is taking the place of the NA(AR), 3374 thereby eliminating an extraneous message transmission and associated 3375 delay. The ROS then completes the route optimization by returning a 3376 responsive NA. 3378 Note: The ROS must accept unicast NS messages with an ACK matching 3379 the SYN included in the NS(AR) as an equivalent message replacement 3380 for the NA(AR). Address resolution and multilink forwarding 3381 coordination can therefore be coordinated in a single three-way 3382 handshake connection with minimal messaging and delay (i.e., as 3383 opposed to a four-message exchange). 3385 3.13.4. Client/Gateway Route Optimization 3387 Following multilink route optimization for specific underlay 3388 interface pairs, ROS/ROR Clients located on open INETs can invoke 3389 Client/Gateway route optimization to improve performance and reduce 3390 load and congestion on their respective FHS/LHS Proxy/Servers. To 3391 initiate Client/Gateway route optimization, the Client prepares an NS 3392 message with its own XLA address as the source and the ULA of its 3393 Gateway as the destination while creating a NCE for the Gateway if 3394 necessary. The NS message must be no larger than the minimum MPS and 3395 encapsulated as an atomic fragment. 3397 The Client then includes an Interface Attributes sub-option for its 3398 underlay interface as well as an authentication signature but does 3399 not include window synchronization parameters. The Client then 3400 performs OAL encapsulation with its own ULA-MNP as the source and the 3401 ULA of the Gateway as the destination while including a randomly- 3402 chosen Identification value, then performs L2 encapsulation on the 3403 atomic fragment and sends the resulting carrier packet directly to 3404 the Gateway. 3406 When the Gateway receives the carrier packet, it verifies the 3407 authentication signature then creates a NCE for the Client. The 3408 Gateway then caches the L2 encapsulation addresses (which may have 3409 been altered by one or more NATs on the path) as well as the 3410 Interface Attributes for this Client omIndex, and marks this Client 3411 underlay interface as "trusted". The Gateway then prepares an NA 3412 reply with its own ULA as the source and the XLA of the Client as the 3413 destination where the NA again must be no larger than the minimum 3414 MPS. 3416 The Gateway then echoes the Client's Interface Attributes, includes 3417 an Origin Indication with the Client's observed L2 addresses and 3418 includes an authentication signature. The Gateway then performs OAL 3419 encapsulation with its own ULA as the source and the ULA-MNP of the 3420 Client as the destination while using the same Identification value 3421 that appeared in the NS, then performs L2 encapsulation on the atomic 3422 fragment and sends the resulting carrier packet directly to the 3423 Client. 3425 When the Client receives the NA reply, it caches the carrier packet 3426 L2 source address information as the Gateway target address via this 3427 underlay interface while marking the interface as "trusted". The 3428 Client also caches the Origin Indication L2 address information as 3429 its own (external) source address for this underlay interface. 3431 After the Client and Gateway have established NCEs as well as 3432 "trusted" status for a particular underlay interface pair, each node 3433 can begin forwarding ordinary carrier packets intended for this 3434 multilink route optimization directly to one another while omitting 3435 the Proxy/Server from the forwarding path while the status is 3436 "trusted". The NS/NA messaging will have established the correct 3437 state in any NATs in the path so that NAT traversal is naturally 3438 supported. The Client and Gateway must maintain a timer that watches 3439 for activity on the path; if no carrier packets and/or NS/NA messages 3440 are sent or received over the path before NAT state is likely to have 3441 expired, the underlay interface pair status becomes "untrusted". 3443 Thereafter, when the Client forwards a carrier packet with an MFVI 3444 toward the Gateway as the next hop, the Client uses the MFVI for the 3445 Gateway (discovered during multilink route optimization) instead of 3446 the MFVI for its Proxy/Server; the Gateway will accept the packet 3447 from the Client if and only if the underlay interface status is 3448 trusted and if the MFVI is correct for the next hop toward the final 3449 destination. (The same is true in the reverse direction when the 3450 Gateway sends carrier packets directly to the Client.) 3452 Note that the Client and Gateway each maintain a single NCE, but that 3453 the NCE may aggregate multiple underlay interface pairs. Each 3454 underlay interface pair may use differing source and target L2 3455 addresses according to NAT mappings, and the "trusted/untrusted" 3456 status of each pair must be tested independently. When no "trusted" 3457 pairs remain, the NCE is deleted. 3459 Note that the above method requires Gateways to participate in NS/NA 3460 message authentication signature application and verification. In an 3461 alternate approach, the Client could instead exchange NS/NA messages 3462 with authentication signatures via its Proxy/Server but addressed to 3463 the ULA of the Gateway, and the Proxy/Server and Gateway could relay 3464 the messages over the secured spanning tree. However, this would 3465 still require the Client to send additional messages toward the L2 3466 address of the Gateway to populate NAT state; hence the savings in 3467 complexity for Gateways would result in increased message overhead 3468 for Clients. 3470 3.13.5. Client/Client Route Optimization 3472 When the ROS/ROR Clients are both located on the same SRT segment, 3473 Client-to-Client route optimization is possible following the 3474 establishment of any necessary state in NATs in the path. Both 3475 Clients will have already established state via their respective 3476 shared segment Proxy/Servers (and possibly also the shared segment 3477 Gateway) and can begin forwarding packets directly via NAT traversal 3478 while avoiding any Proxy/Server and/or Gateway hops. 3480 When the ROR/ROS Clients on the same SRT segment perform the initial 3481 NS/NA exchange to establish Multilink Forwarding state, they also 3482 include an Origin Indication (i.e., in addition to Multilink 3483 Forwarding Parameters) with the mapped addresses discovered during 3484 the RS/RA exchanges with their respective Proxy/Servers. After the 3485 MFV paths have been established, both Clients can begin sending 3486 packets via strict MFV paths while establishing a direct path for 3487 Client-to-Client route optimization. 3489 To establish the direct path, either Client (acting as the source) 3490 transmits a bubble to the mapped L2 address for the target Client 3491 which primes its local chain of NATs for reception of future packets 3492 from that L2 address (see: [RFC4380] and [I-D.templin-6man-omni]). 3493 The source Client then prepares an NS message with its own XLA as the 3494 source, with the XLA of the target as the destination and with an 3495 OMNI option with an Interface Attributes sub-option. The source 3496 Client then encapsulates the NS in an OAL header with its own ULA-MNP 3497 as the source, with the ULA-MNP of the target Client as the 3498 destination and with an in-window Identification for the target. The 3499 source Client then fragments and encapsulates in L2 headers addressed 3500 to its FHS Proxy/Server then forwards the resulting carrier packets 3501 to the Proxy/Server. 3503 When the FHS Proxy/Server receives the carrier packets, it re- 3504 encapsulates and forwards them as unsecured carrier packets according 3505 to MFV state where they will eventually arrive at the target Client 3506 which can verify that the identifications are within the acceptable 3507 window and reassemble if necessary. Following reassembly, the target 3508 Client prepares an NA message with its own XLA as the source, with 3509 the XLA of the source Client as the destination and with an OMNI 3510 option with an Interface Attributes sub-option. The target Client 3511 then encapsulates the NA in an OAL header with its own ULA-MNP as the 3512 source, with the ULA-MNP of the source Client as the destination and 3513 with an in-window Identification for the source Client. The target 3514 Client then fragments and encapsulates in L2 headers addressed to the 3515 source Client's Origin addresses then forwards the resulting carrier 3516 packets directly to the source Client. 3518 Following the initial NS/NA exchange, both Clients mark their 3519 respective (source, target) underlay interface pairs as "trusted" for 3520 no more than ReachableTime seconds. While the Clients continue to 3521 exchange carrier packets via the direct path avoiding all Proxy/ 3522 Servers and Gateways, they should perform additional NS/NA exchanges 3523 via their local Proxy/Servers to refresh NCE state as well as send 3524 additional bubbles to the peer's Origin address information if 3525 necessary to refresh NAT state. 3527 Note that these procedures are suitable for a widely-deployed but 3528 basic class of NATs. Procedures for advanced NAT classes are 3529 outlined in [RFC6081], which provides mechanisms that can be employed 3530 equally for AERO using the corresponding sub-options specified by 3531 OMNI. 3533 Note also that each communicating pair of Clients may need to 3534 maintain NAT state for peer to peer communications via multiple 3535 underlay interface pairs. It is therefore important that Origin 3536 Indications are maintained with the correct peer interface and that 3537 the NCE may cache information for multiple peer interfaces. 3539 Note that the source and target Client exchange Origin information 3540 during the secured NS/NA multilink route optimization exchange. This 3541 allows for subsequent NS/NA exchanges to proceed using only the 3542 Identification value as a data origin confirmation. However, Client- 3543 to-Client peerings that require stronger security may also include 3544 authentication signatures for mutual authentication. 3546 3.13.6. Client-to-Client OMNI Link Extension 3548 Clients may be recursively nested within the ENETs of other Clients. 3549 When a Client is the downstream-attached ENET neighbor of an upstream 3550 Client, it still supports the route optimization functions discussed 3551 above by maintaining an MFIB and assigning MFVI values. When the 3552 Client processes an IPv6 ND NS/NA message that includes a Multilink 3553 Forwarding Parameters sub-option, it writes its MFVI information as 3554 the first/last MFVI list entry the same as for the single Client case 3555 discussed above. 3557 The Client then forwards the NS/NA message to the next Client in the 3558 extended OMNI link toward the FHS/LHS Proxy/Server, which records the 3559 MVFI value then overwrites the MFVI list entry with its own MFVI 3560 value. This process iteratively continues until the Client that will 3561 forward the NS/NA message to the FHS/LHS Proxy/Server is reached, at 3562 which point the NS/NA MFVI list entries are populated by the 3563 intermediate nodes on the path to the LHS/FHS the same as discussed 3564 above. 3566 In this way, each Client in the extended OMNI link discovers the A/B 3567 MVFIs of the next/previous Client without intruding into the 3568 Multilink Forwarding Parameters MFVI list. Therefore the list can 3569 remain fixed at 5 entries even though the Client-to-Client OMNI link 3570 extension can be arbitrarily long. Therefore, route optimization is 3571 not possible between consecutive Client members of the extended OMNI 3572 link but becomes possible at the Internetworking border that 3573 separates the FHS and LHS elements. 3575 3.13.7. Intra-ANET/ENET Route Optimization for AERO Peers 3577 When a Client forwards a packet from a Host or another Client 3578 connected to one of its downstream ENETs to a peer within the same 3579 downstream ENET, the Client returns an IPv6 ND Redirect message to 3580 inform the source that that target can be reached directly. The 3581 contents of the Redirect message are the same as specified in 3582 [RFC4861]. 3584 In the same fashion, when a Proxy/Server forwards a packet from a 3585 Host or Client connected to one of its downstream ANETs to a peer 3586 within the same downstream ANET, the Proxy/Server returns an IPv6 ND 3587 Redirect message. 3589 All other route optimization functions are conducted per the NS/NA 3590 messaging discussed in the previous sections. 3592 3.14. Neighbor Unreachability Detection (NUD) 3594 AERO nodes perform Neighbor Unreachability Detection (NUD) per 3595 [RFC4861] either reactively in response to persistent link-layer 3596 errors (see Section 3.11) or proactively to confirm reachability. 3597 The NUD algorithm is based on periodic control message exchanges and 3598 may further be seeded by IPv6 ND hints of forward progress, but care 3599 must be taken to avoid inferring reachability based on spoofed 3600 information. For example, IPv6 ND message exchanges that include 3601 authentication codes and/or in-window Identifications may be 3602 considered as acceptable hints of forward progress, while spurious 3603 random carrier packets should be ignored. 3605 AERO nodes can perform NS/NA(NUD) exchanges over the OMNI link 3606 secured spanning tree (i.e. the same as described above) to test 3607 reachability without risk of DoS attacks from nodes pretending to be 3608 a neighbor. These NS/NA(NUD) messages use the unicast XLAs/ULAs of 3609 the parties involved in the NUD test. When only reachability 3610 information is required without updating any other NCE state, AERO 3611 nodes can instead perform NS/NA(NUD) exchanges directly between 3612 neighbors without employing the secured spanning tree as long as they 3613 include in-window Identifications and either an authentication 3614 signature or checksum. 3616 After an ROR directs an ROS to a target neighbor with one or more 3617 link-layer addresses, either node may invoke multilink forwarding 3618 state initialization to establish authentic intermediate node state 3619 between specific underlay interface pairs which also tests their 3620 reachability. Thereafter, either node acting as the source may 3621 perform additional reachability probing through NS(NUD) messages over 3622 the SRT secured or unsecured spanning tree, or through NS(NUD) 3623 messages sent directly to an underlay interface of the target itself. 3624 While testing a target underlay interface, the source can optionally 3625 continue to forward carrier packets via alternate interfaces, 3626 maintain a small queue of carrier packets until target reachability 3627 is confirmed or include them as trailing data with the NS(NUD) in an 3628 OAL super-packet [I-D.templin-6man-omni]. 3630 NS(NUD) messages are encapsulated, fragmented and transmitted as 3631 carrier packets the same as for ordinary original IP data packets, 3632 however the encapsulated destinations are either the ULA or XLA of 3633 the source and either the ULA of the LHS Proxy/Server or the XLA of 3634 the target itself. The source encapsulates the NS(NUD) message the 3635 same as described in Section 3.13.2 and includes an Interface 3636 Attributes sub-option with omIndex set to identify its underlay 3637 interface used for forwarding. The source then includes an in-window 3638 Identification, fragments the OAL packet and forwards the resulting 3639 carrier packets into the unsecured spanning tree, directly to the 3640 target if it is in the local segment or directly to a Gateway in the 3641 local segment. 3643 When the target receives the NS(NUD) carrier packets, it verifies 3644 that it has a NCE for this source and that the Identification is in- 3645 window, then submits the carrier packets for reassembly. The target 3646 then verifies the authentication signature or checksum, then searches 3647 for Interface Attributes in its NCE for the source that match the 3648 NS(NUD) for the NA(NUD) reply. The target then prepares the NA(NUD) 3649 with the source and destination addresses reversed, encapsulates and 3650 sets the OAL source and destination, includes an Interface Attributes 3651 sub-option in the NA(NUD) to identify the omIndex of the underlay 3652 interface the NS(NUD) arrived on and sets the Target Address to the 3653 same value included in the NS(NUD). The target next sets the R flag 3654 to 1, the S flag to 1 and the O flag to 1, then selects an in-window 3655 Identification for the source and performs fragmentation. The node 3656 then forwards the carrier packets into the unsecured spanning tree, 3657 directly to the source if it is in the local segment or directly to a 3658 Gateway in the local segment. 3660 When the source receives the NA(NUD), it marks the target underlay 3661 interface tested as "trusted". Note that underlay interface states 3662 are maintained independently of the overall NCE REACHABLE state, and 3663 that a single NCE may have multiple target underlay interfaces in 3664 various "trusted/untrusted" states while the NCE state as a whole 3665 remains REACHABLE. 3667 3.15. Mobility Management and Quality of Service (QoS) 3669 AERO is a fully Distributed Mobility Management (DMM) service in 3670 which each Proxy/Server is responsible for only a small subset of the 3671 Clients on the OMNI link. This is in contrast to a Centralized 3672 Mobility Management (CMM) service where there are only one or a few 3673 network mobility collective entities for large Client populations. 3674 Clients coordinate with their associated FHS and Hub Proxy/Servers 3675 via RS/RA exchanges to maintain the DMM profile, and the AERO routing 3676 system tracks all current Client/Proxy/Server peering relationships. 3678 Hub Proxy/Servers provide a designated router service for their 3679 dependent Clients, while FHS Proxy/Servers provide a proxy conduit 3680 between the Client and both the Hub and OMNI link in general. 3681 Clients are responsible for maintaining neighbor relationships with 3682 their Proxy/Servers through periodic RS/RA exchanges, which also 3683 serves to confirm neighbor reachability. When a Client's underlay 3684 interface attributes change, the Client is responsible for updating 3685 the Hub Proxy/Server through new RS/RA exchanges using the FHS Proxy/ 3686 Server as a first-hop conduit. The FHS Proxy/Server can also act as 3687 a proxy to perform some IPv6 ND exchanges on the Client's behalf 3688 without consuming bandwidth on the Client underlay interface. 3690 Mobility management considerations are specified in the following 3691 sections. 3693 3.15.1. Mobility Update Messaging 3695 RORs and ROSs accommodate Client mobility and/or multilink change 3696 events by sending secured uNA messages to each active neighbor. When 3697 an ROR/ROS sends a uNA message, it sets the IPv6 source address to 3698 the its own ULA or XLA, sets the destination address to the 3699 neighbor's ULA or XLA and sets the Target Address to the Client's 3700 XLA. The ROR/ROS also includes an OMNI option with OMNI Neighbor 3701 Coordination header Preflen set to the prefix length associated with 3702 the Client's XLA, includes Interface Attributes and Traffic Selectors 3703 for the Client's underlay interfaces and includes an authentication 3704 signature if necessary. The ROR then sets the uNA R flag to 1, S 3705 flag to 0 and O flag to 1, then encapsulates the message in an OAL 3706 header with source set to its own ULA and destination set to its FHS 3707 Proxy/Server's ULA. When the FHS Proxy/Server receives the uNA, it 3708 reassembles, verifies the authentication signature, then changes the 3709 destination to the ULA corresponding to the uNA destination and 3710 forwards the uNA into the secured spanning tree. 3712 As discussed in Section 7.2.6 of [RFC4861], the transmission and 3713 reception of uNA messages is unreliable but provides a useful 3714 optimization. In well-connected Internetworks with robust data links 3715 uNA messages will be delivered with high probability, but in any case 3716 the ROR/ROS can optionally send up to MAX_NEIGHBOR_ADVERTISEMENT uNAs 3717 to each neighbor to increase the likelihood that at least one will be 3718 received. Alternatively, the ROR/ROS can set the PNG flag in the uNA 3719 OMNI option header to request a uNA acknowledgement as specified in 3720 [I-D.templin-6man-omni]. 3722 When the ROR/ROS Proxy/Server receives a uNA message prepared as 3723 above, if the uNA destination was its own ULA the Proxy/Server uses 3724 the included OMNI option information to update its NCE for the target 3725 but does not reset ReachableTime since the receipt of a uNA message 3726 does not provide confirmation that any forward paths to the target 3727 Client are working. If the destination was the XLA of the ROR/ROS 3728 Client, the Proxy/Server instead changes the OAL source to its own 3729 ULA, includes an authentication signature if necessary, and includes 3730 an in-window Identification for this Client. Finally, if the uNA 3731 message PNG flag was set, the node that processes the uNA returns a 3732 uNA acknowledgement as specified in [I-D.templin-6man-omni]. 3734 3.15.2. Announcing Link-Layer Information Changes 3736 When a Client needs to change its underlay Interface Attributes and/ 3737 or Traffic Selectors (e.g., due to a mobility event), the Client 3738 sends an RS message to its Hub Proxy/Server via a first-hop FHS 3739 Proxy/Server, if necessary. The RS includes an OMNI option with an 3740 Interface Attributes sub-option with the omIndex and with new link 3741 quality and any other information. 3743 Note that the first FHS Proxy/Server may change due to the underlay 3744 interface change. If the Client supplies the address of the former 3745 FHS Proxy/Server, the new FHS Proxy/Server can send a departure 3746 indication (see below); otherwise, any stale state in the former FHS 3747 Proxy/Server will simply expire after ReachableTime expires with no 3748 effect on the Hub Proxy/Server. 3750 Up to MAX_RTR_SOLICITATIONS RS messages MAY be sent in parallel with 3751 sending carrier packets containing user data in case one or more RAs 3752 are lost. If all RAs are lost, the Client SHOULD re-associate with a 3753 new Proxy/Server. 3755 After performing the RS/RA exchange, the Client sends uNA messages to 3756 all neighbors the same as described in the previous section. 3758 3.15.3. Bringing New Links Into Service 3760 When a Client needs to bring new underlay interfaces into service 3761 (e.g., when it activates a new data link), it sends an RS message to 3762 the Hub Proxy/Server via a FHS Proxy/Server for the underlay 3763 interface (if necessary) with an OMNI option that includes an 3764 Interface Attributes sub-option with appropriate link quality values 3765 and with link-layer address information for the new link. The Client 3766 then again sends uNA messages to all neighbors the same as described 3767 above. 3769 3.15.4. Deactivating Existing Links 3771 When a Client needs to deactivate an existing underlay interface, it 3772 sends a uNA message toward the Hub Proxy/Server via an FHS Proxy/ 3773 Server with an OMNI option with appropriate Interface Attributes 3774 values for the deactivated link - in particular, the link quality 3775 value 0 assures that neighbors will cease to use the link. 3777 If the Client needs to send uNA messages over an underlay interface 3778 other than the one being deactivated, it MUST include Interface 3779 Attributes with appropriate link quality values for any underlay 3780 interfaces being deactivated. The Client then again sends uNA 3781 messages to all neighbors the same as described above. 3783 Note that when a Client deactivates an underlay interface, neighbors 3784 that receive the ensuing uNA messages need not purge all references 3785 for the underlay interface from their neighbor cache entries. The 3786 Client may reactivate or reuse the underlay interface and/or its 3787 omIndex at a later point in time, when it will send new RS messages 3788 to an FHS Proxy/Server with fresh interface parameters to update any 3789 neighbors. 3791 3.15.5. Moving Between Proxy/Servers 3793 The Client performs the procedures specified in Section 3.12.2 when 3794 it first associates with a new Hub Proxy/Server or renews its 3795 association with an existing Hub Proxy/Server. 3797 When a Client associates with a new Hub Proxy/Server, it sends RS 3798 messages to register its underlay interfaces with the new Hub while 3799 including the old Hub's ULA in the "Old Hub Proxy/Server ULA" field 3800 of a Proxy/Server Departure OMNI sub-option. When the new Hub Proxy/ 3801 Server returns the RA message via the FHS Proxy/Server (acting as a 3802 Proxy), the FHS Proxy/Server sends a uNA to the old Hub Proxy/Server 3803 (i.e., if the ULA is non-zero and different from its own). The uNA 3804 has the XLA of the Client as the source and the ULA of the old hub as 3805 the destination and with OMNI Neighbor Coordination header Preflen 3806 set to 0. The FHS Proxy/Server encapsulates the uNA in an OAL header 3807 with the ULA of the new Hub as the source and the ULA of the old Hub 3808 as the destination, the fragments and sends the carrier packets via 3809 the secured spanning tree. 3811 When the old Hub Proxy/Server receives the uNA, it changes the 3812 Client's NCE state to DEPARTED, resets DepartTime and caches the new 3813 Hub Proxy/Server ULA. After a short delay (e.g., 2 seconds) the old 3814 Hub Proxy/Server withdraws the Client's MNP from the routing system. 3815 While in the DEPARTED state, the old Hub Proxy/Server forwards any 3816 carrier packets received via the secured spanning tree destined to 3817 the Client's ULA-MNP to the new Hub Proxy/Server's ULA. After 3818 DepartTime expires, the old Hub Proxy/Server deletes the Client's 3819 NCE. 3821 Mobility events may also cause a Client to change to a new FHS Proxy/ 3822 Server over a specific underlay interface at any time such that a 3823 Client RS/RA exchange over the underlay interface will engage the new 3824 FHS Proxy/Server instead of the old. The Client can arrange to 3825 inform the old FHS Proxy/Server of the departure by including a 3826 Proxy/Server Departure sub-option with a ULA for the "Old FHS Proxy/ 3827 Server ULA", and the new FHS Proxy/Server will issue a uNA using the 3828 same procedures as outlined for the Hub above while using its own ULA 3829 as the source address. This can often result in successful delivery 3830 of packets that would otherwise be lost due to the mobility event. 3832 Clients SHOULD NOT move rapidly between Hub Proxy/Servers in order to 3833 avoid causing excessive oscillations in the AERO routing system. 3834 Examples of when a Client might wish to change to a different Hub 3835 Proxy/Server include a Hub Proxy/Server that has gone unreachable, 3836 topological movements of significant distance, movement to a new 3837 geographic region, movement to a new OMNI link segment, etc. 3839 3.16. Multicast 3841 Clients provide an IGMP (IPv4) [RFC2236] or MLD (IPv6) [RFC3810] 3842 proxy service for its ENETs and/or hosted applications [RFC4605] and 3843 act as a Protocol Independent Multicast - Sparse-Mode (PIM-SM, or 3844 simply "PIM") Designated Router (DR) [RFC7761] on the OMNI link. 3845 Proxy/Servers act as OMNI link PIM routers for Clients on ANET, VPNed 3846 or Direct interfaces, and Relays also act as OMNI link PIM routers on 3847 behalf of nodes on other links/networks. 3849 Clients on VPNed, Direct or ANET underlay interfaces for which the 3850 ANET has deployed native multicast services forward IGMP/MLD messages 3851 into the ANET. The IGMP/MLD messages may be further forwarded by a 3852 first-hop ANET access router acting as an IGMP/MLD-snooping switch 3853 [RFC4541], then ultimately delivered to an ANET Proxy/Server. The 3854 FHS Proxy/Server then acts as an ROS to send NS(AR) messages to an 3855 ROR for the multicast source. Clients on INET and ANET underlay 3856 interfaces without native multicast services instead send NS(AR) 3857 messages as an ROS to cause their FHS Proxy/Server forward the 3858 message to an ROR. When the ROR receives an NA(AR) response, it 3859 initiates PIM protocol messaging according to the Source-Specific 3860 Multicast (SSM) and Any-Source Multicast (ASM) operational modes as 3861 discussed in the following sections. 3863 3.16.1. Source-Specific Multicast (SSM) 3865 When an ROS "X" (i.e., either a Client or Proxy/Server) acting as PIM 3866 router receives a Join/Prune message from a node on its downstream 3867 interfaces containing one or more ((S)ource, (G)roup) pairs, it 3868 updates its Multicast Routing Information Base (MRIB) accordingly. 3869 For each S belonging to a prefix reachable via X's non-OMNI 3870 interfaces, X then forwards the (S, G) Join/Prune to any PIM routers 3871 on those interfaces per [RFC7761]. 3873 For each S belonging to a prefix reachable via X's OMNI interface, X 3874 sends an NS(AR) message (see: Section 3.13) using its own ULA or XLA 3875 as the source address, the solicited node multicast address 3876 corresponding to S as the destination and the XLA of S as the target 3877 address. X then encapsulates the NS(AR) in an OAL header with source 3878 address set to its own ULA and destination address set to the ULA for 3879 S, then forwards the message into the secured spanning tree which 3880 delivers it to ROR "Y" that services S. The resulting NA(AR) will 3881 return an OMNI option with Interface Attributes for any underlay 3882 interfaces that are currently servicing S. 3884 When X processes the NA(AR) it selects one or more underlay 3885 interfaces for S and performs an NS/NA multilink route optimization 3886 exchange over the secured spanning tree while including a PIM Join/ 3887 Prune message for each multicast group of interest in the OMNI 3888 option. If S is located behind any Proxys "Z"*, each Z* then updates 3889 its MRIB accordingly and maintains the XLA of X as the next hop in 3890 the reverse path. Since Gateways forward messages not addressed to 3891 themselves without examining them, this means that the (reverse) 3892 multicast tree path is simply from each Z* (and/or S) to X with no 3893 other multicast-aware routers in the path. 3895 Following the initial combined Join/Prune and NS/NA messaging, X 3896 maintains a NCE for each S the same as if X was sending unicast data 3897 traffic to S. In particular, X performs additional NS/NA exchanges 3898 to keep the NCE alive for up to t_periodic seconds [RFC7761]. If no 3899 new Joins are received within t_periodic seconds, X allows the NCE to 3900 expire. Finally, if X receives any additional Join/Prune messages 3901 for (S,G) it forwards the messages over the secured spanning tree. 3903 Client C that holds an MNP for source S may later depart from a first 3904 Proxy/Server Z1 and/or connect via a new Proxy/Server Z2. In that 3905 case, Y sends a uNA message to X the same as specified for unicast 3906 mobility in Section 3.15. When X receives the uNA message, it 3907 updates its NCE for the XLA for source S and sends new Join messages 3908 in NS/NA exchanges addressed to the new target Client underlay 3909 interface connection for S. There is no requirement to send any 3910 Prune messages to old Proxy/Server Z1 since source S will no longer 3911 source any multicast data traffic via Z1. Instead, the multicast 3912 state for (S,G) in Proxy/Server Z1 will soon expire since no new 3913 Joins will arrive. 3915 3.16.2. Any-Source Multicast (ASM) 3917 When an ROS X acting as a PIM router receives Join/Prune messages 3918 from a node on its downstream interfaces containing one or more (*,G) 3919 pairs, it updates its Multicast Routing Information Base (MRIB) 3920 accordingly. X first performs an NS/NA(AR) exchange to receive route 3921 optimization information for Rendezvous Point (RP) R for each G. X 3922 then includes a copy of each Join/Prune message in the OMNI option of 3923 an NS message with its own ULA or XLA as the source address and the 3924 ULA or XLA for R as the destination address, then encapsulates the NS 3925 message in an OAL header with its own ULA as the source and the ULA 3926 of R's Proxy/Server as the destination then sends the message into 3927 the secured spanning tree. 3929 For each source S that sends multicast traffic to group G via R, 3930 Client S* that aggregates S (or its Proxy/Server) encapsulates the 3931 original IP packets in PIM Register messages, includes the PIM 3932 Register messages in the OMNI options of uNA messages, performs OAL 3933 encapsulation and fragmentation then forwards the resulting carrier 3934 packets with Identification values within the receive window for 3935 Client R* that aggregates R. Client R* may then elect to send a PIM 3936 Join to S* in the OMNI option of a uNA over the secured spanning 3937 tree. This will result in an (S,G) tree rooted at S* with R as the 3938 next hop so that R will begin to receive two copies of the original 3939 IP packet; one native copy from the (S, G) tree and a second copy 3940 from the pre-existing (*, G) tree that still uses uNA PIM Register 3941 encapsulation. R can then issue a uNA PIM Register-stop message over 3942 the secured spanning tree to suppress the Register-encapsulated 3943 stream. At some later time, if Client S* moves to a new Proxy/ 3944 Server, it resumes sending original IP packets via uNA PIM Register 3945 encapsulation via the new Proxy/Server. 3947 At the same time, as multicast listeners discover individual S's for 3948 a given G, they can initiate an (S,G) Join for each S under the same 3949 procedures discussed in Section 3.16.1. Once the (S,G) tree is 3950 established, the listeners can send (S, G) Prune messages to R so 3951 that multicast original IP packets for group G sourced by S will only 3952 be delivered via the (S, G) tree and not from the (*, G) tree rooted 3953 at R. All mobility considerations discussed for SSM apply. 3955 3.16.3. Bi-Directional PIM (BIDIR-PIM) 3957 Bi-Directional PIM (BIDIR-PIM) [RFC5015] provides an alternate 3958 approach to ASM that treats the Rendezvous Point (RP) as a Designated 3959 Forwarder (DF). Further considerations for BIDIR-PIM are out of 3960 scope. 3962 3.17. Operation over Multiple OMNI Links 3964 An AERO Client can connect to multiple OMNI links the same as for any 3965 data link service. In that case, the Client maintains a distinct 3966 OMNI interface for each link, e.g., 'omni0' for the first link, 3967 'omni1' for the second, 'omni2' for the third, etc. Each OMNI link 3968 would include its own distinct set of Gateways and Proxy/Servers, 3969 thereby providing redundancy in case of failures. 3971 Each OMNI link could utilize the same or different ANET connections. 3972 The links can be distinguished at the link-layer via the SRT prefix 3973 in a similar fashion as for Virtual Local Area Network (VLAN) tagging 3974 (e.g., IEEE 802.1Q) and/or through assignment of distinct sets of 3975 MSPs on each link. This gives rise to the opportunity for supporting 3976 multiple redundant networked paths (see: Section 3.2.4). 3978 The Client's IP layer can select the outgoing OMNI interface 3979 appropriate for a given traffic profile while (in the reverse 3980 direction) correspondent nodes must have some way of steering their 3981 original IP packets destined to a target via the correct OMNI link. 3983 In a first alternative, if each OMNI link services different MSPs the 3984 Client can receive a distinct MNP from each of the links. IP routing 3985 will therefore assure that the correct OMNI link is used for both 3986 outbound and inbound traffic. This can be accomplished using 3987 existing technologies and approaches, and without requiring any 3988 special supporting code in correspondent nodes or Gateways. 3990 In a second alternative, if each OMNI link services the same MSP(s) 3991 then each link could assign a distinct "OMNI link Anycast" address 3992 that is configured by all Gateways on the link. Correspondent nodes 3993 can then perform Segment Routing to select the correct SRT, which 3994 will then direct the original IP packet over multiple hops to the 3995 target. 3997 3.18. DNS Considerations 3999 AERO Client MNs and INET correspondent nodes consult the Domain Name 4000 System (DNS) the same as for any Internetworking node. When 4001 correspondent nodes and Client MNs use different IP protocol versions 4002 (e.g., IPv4 correspondents and IPv6 MNs), the INET DNS must maintain 4003 A records for IPv4 address mappings to MNs which must then be 4004 populated in Relay NAT64 mapping caches. In that way, an IPv4 4005 correspondent node can send original IPv4 packets to the IPv4 address 4006 mapping of the target MN, and the Relay will translate the IPv4 4007 header and destination address into an IPv6 header and IPv6 4008 destination address of the MN. 4010 When an AERO Client registers with an AERO Proxy/Server, the Proxy/ 4011 Server can return the address(es) of DNS servers in RDNSS options 4012 [RFC6106]. The DNS server provides the IP addresses of other MNs and 4013 correspondent nodes in AAAA records for IPv6 or A records for IPv4. 4015 3.19. Transition/Coexistence Considerations 4017 OAL encapsulation ensures that dissimilar INET partitions can be 4018 joined into a single unified OMNI link, even though the partitions 4019 themselves may have differing protocol versions and/or incompatible 4020 addressing plans. However, a commonality can be achieved by 4021 incrementally distributing globally routable (i.e., native) IP 4022 prefixes to eventually reach all nodes (both mobile and fixed) in all 4023 OMNI link segments. This can be accomplished by incrementally 4024 deploying AERO Gateways on each INET partition, with each Gateway 4025 distributing its MNPs and/or discovering non-MNP IP GUA prefixes on 4026 its INET links. 4028 This gives rise to the opportunity to eventually distribute native IP 4029 addresses to all nodes, and to present a unified OMNI link view even 4030 if the INET partitions remain in their current protocol and 4031 addressing plans. In that way, the OMNI link can serve the dual 4032 purpose of providing a mobility/multilink service and a transition/ 4033 coexistence service. Or, if an INET partition is transitioned to a 4034 native IP protocol version and addressing scheme that is compatible 4035 with the OMNI link MNP-based addressing scheme, the partition and 4036 OMNI link can be joined by Gateways. 4038 Relays that connect INETs/ENETs with dissimilar IP protocol versions 4039 may need to employ a network address and protocol translation 4040 function such as NAT64 [RFC6146]. 4042 3.20. Proxy/Server-Gateway Bidirectional Forwarding Detection 4044 In environments where rapid failure recovery is required, Proxy/ 4045 Servers and Gateways SHOULD use Bidirectional Forwarding Detection 4046 (BFD) [RFC5880]. Nodes that use BFD can quickly detect and react to 4047 failures so that cached information is re-established through 4048 alternate nodes. BFD control messaging is carried only over well- 4049 connected ground domain networks (i.e., and not low-end radio links) 4050 and can therefore be tuned for rapid response. 4052 Proxy/Servers and Gateways maintain BFD sessions in parallel with 4053 their BGP peerings. If a Proxy/Server or Gateway fails, BGP peers 4054 will quickly re-establish routes through alternate paths the same as 4055 for common BGP deployments. Similarly, Proxys maintain BFD sessions 4056 with their associated Gateways even though they do not establish BGP 4057 peerings with them. 4059 3.21. Time-Varying MNPs 4061 In some use cases, it is desirable, beneficial and efficient for the 4062 Client to receive a constant MNP that travels with the Client 4063 wherever it moves. For example, this would allow air traffic 4064 controllers to easily track aircraft, etc. In other cases, however 4065 (e.g., intelligent transportation systems), the MN may be willing to 4066 sacrifice a modicum of efficiency in order to have time-varying MNPs 4067 that can be changed every so often to defeat adversarial tracking. 4069 The DHCPv6 service offers a way for Clients that desire time-varying 4070 MNPs to obtain short-lived prefixes (e.g., on the order of a small 4071 number of minutes). In that case, the identity of the Client would 4072 not be bound to the MNP but rather to a Node Identification value 4073 (see: [I-D.templin-6man-omni]) to be used as the Client ID seed for 4074 MNP prefix delegation. The Client would then be obligated to 4075 renumber its internal networks whenever its MNP (and therefore also 4076 its XLA) changes. This should not present a challenge for Clients 4077 with automated network renumbering services, however presents limits 4078 for the durations of ongoing sessions that would prefer to use a 4079 constant address. 4081 4. Implementation Status 4083 An early AERO implementation based on OpenVPN (https://openvpn.net/) 4084 was announced on the v6ops mailing list on January 10, 2018 and an 4085 initial public release of the AERO proof-of-concept source code was 4086 announced on the intarea mailing list on August 21, 2015. 4088 Many AERO/OMNI functions are implemented and undergoing final 4089 integration. OAL fragmentation/reassembly buffer management code has 4090 been cleared for public release. 4092 5. IANA Considerations 4094 The IANA has assigned the UDP port number "8060" for an earlier 4095 experimental first version of AERO [RFC6706]. This document together 4096 with [I-D.templin-6man-omni] reclaims UDP port number "8060" as the 4097 service port for UDP/IP encapsulation. This document makes no 4098 request of IANA, since [I-D.templin-6man-omni] already provides 4099 instructions. (Note: although [RFC6706] was not widely implemented 4100 or deployed, it need not be obsoleted since its messages use the 4101 invalid ICMPv6 message type number '0' which implementations of this 4102 specification can easily distinguish and ignore.) 4104 No further IANA actions are required. 4106 6. Security Considerations 4108 AERO Gateways configure secured tunnels with AERO Proxy/Servers and 4109 Relays within their local OMNI link segments. Applicable secured 4110 tunnel alternatives include IPsec [RFC4301], TLS/SSL [RFC8446], DTLS 4111 [RFC6347], WireGuard [WG], etc. The AERO Gateways of all OMNI link 4112 segments in turn configure secured tunnels for their neighboring AERO 4113 Gateways in a secured spanning tree topology. Therefore, control 4114 messages exchanged between any pair of OMNI link neighbors over the 4115 secured spanning tree are already protected. 4117 To prevent spoofing vectors, Proxy/Servers MUST discard without 4118 responding to any unsecured NS/NA(AR) messages. Also, Proxy/Servers 4119 MUST discard without forwarding any original IP packets received from 4120 one of their own Clients (whether directly or following OAL 4121 reassembly) with a source address that does not match the Client's 4122 MNP and/or a destination address that does match the Client's MNP. 4123 Finally, Proxy/Servers MUST discard without forwarding any carrier 4124 packets with an OAL source and destination that both match the same 4125 MNP. 4127 For INET partitions that require strong security in the data plane, 4128 two options for securing communications include 1) disable route 4129 optimization so that all traffic is conveyed over secured tunnels, or 4130 2) enable on-demand secure tunnel creation between Client neighbors. 4131 Option 1) would result in longer routes than necessary and impose 4132 traffic concentration on critical infrastructure elements. Option 2) 4133 could be coordinated between Clients using NS/NA messages with OMNI 4134 Host Identity Protocol (HIP) "Initiator/Responder" message sub- 4135 options [RFC7401][I-D.templin-6man-omni] to create a secured tunnel 4136 on-demand, or to use the QUIC-TLS protocol to establish a secured 4137 connection [RFC9000][RFC9001][RFC9002]. 4139 AERO Clients that connect to secured ANETs need not apply security to 4140 their IPv6 ND messages, since the messages will be authenticated and 4141 forwarded by a perimeter Proxy/Server that applies security on its 4142 INET-facing interface as part of the secured spanning tree (see 4143 above). AERO Clients connected to the open INET can use network and/ 4144 or transport layer security services such as VPNs or can by some 4145 other means establish a direct link to a Proxy/Server. When a VPN or 4146 direct link may be impractical, however, INET Clients and Proxy/ 4147 Servers SHOULD include and verify authentication signatures for their 4148 IPv6 ND messages as specified in [I-D.templin-6man-omni]. 4150 Application endpoints SHOULD use transport-layer (or higher-layer) 4151 security services such as QUIC-TLS, TLS/SSL, DTLS or SSH [RFC4251] to 4152 assure the same level of protection as for critical secured Internet 4153 services. AERO Clients that require host-based VPN services SHOULD 4154 use network and/or transport layer security services such as IPsec, 4155 TLS/SSL, DTLS, etc. AERO Proxys and Proxy/Servers can also provide a 4156 network-based VPN service on behalf of the Client, e.g., if the 4157 Client is located within a secured enclave and cannot establish a VPN 4158 on its own behalf. 4160 AERO Proxy/Servers and Gateways present targets for traffic 4161 amplification Denial of Service (DoS) attacks. This concern is no 4162 different than for widely-deployed VPN security gateways in the 4163 Internet, where attackers could send spoofed packets to the gateways 4164 at high data rates. This can be mitigated through the AERO/OMNI data 4165 origin authentication procedures, as well as connecting Proxy/Servers 4166 and Gateways over dedicated links with no connections to the Internet 4167 and/or when connections to the Internet are only permitted through 4168 well-managed firewalls. Traffic amplification DoS attacks can also 4169 target an AERO Client's low data rate links. This is a concern not 4170 only for Clients located on the open Internet but also for Clients in 4171 secured enclaves. AERO Proxy/Servers and Proxys can institute rate 4172 limits that protect Clients from receiving packet floods that could 4173 DoS low data rate links. 4175 AERO Relays must implement ingress filtering to avoid a spoofing 4176 attack in which spurious messages with ULA addresses are injected 4177 into an OMNI link from an outside attacker. AERO Clients MUST ensure 4178 that their connectivity is not used by unauthorized nodes on their 4179 ENETs to gain access to a protected network, i.e., AERO Clients that 4180 act as routers MUST NOT provide routing services for unauthorized 4181 nodes. (This concern is no different than for ordinary hosts that 4182 receive an IP address delegation but then "share" the address with 4183 other nodes via some form of Internet connection sharing such as 4184 tethering.) 4186 The PRL MUST be well-managed and secured from unauthorized tampering, 4187 even though the list contains only public information. The PRL can 4188 be conveyed to the Client in a similar fashion as in [RFC5214] (e.g., 4189 through layer 2 data link login messaging, secure upload of a static 4190 file, DNS lookups, etc.). 4192 The AERO service for open INET Clients depends on a public key 4193 distribution service in which Client public keys and identities are 4194 maintained in a shared database accessible to all open INET Proxy/ 4195 Servers. Similarly, each Client must be able to determine the public 4196 key of each Proxy/Server, e.g. by consulting an online database. 4197 When AERO nodes register their public keys indexed by a unique Host 4198 Identity Tag (HIT) [RFC7401] in a distributed database such as the 4199 DNS, and use the HIT as an identity for applying IPv6 ND message 4200 authentication signatures, a means for determining public key 4201 attestation is available. 4203 Security considerations for IPv6 fragmentation and reassembly are 4204 discussed in [I-D.templin-6man-omni]. In environments where spoofing 4205 is considered a threat, OMNI nodes SHOULD employ Identification 4206 window synchronization and OAL destinations SHOULD configure an (end- 4207 system-based) firewall. 4209 SRH authentication facilities are specified in [RFC8754]. Security 4210 considerations for accepting link-layer ICMP messages and reflected 4211 packets are discussed throughout the document. 4213 7. Acknowledgements 4215 Discussions in the IETF, aviation standards communities and private 4216 exchanges helped shape some of the concepts in this work. 4217 Individuals who contributed insights include Mikael Abrahamsson, Mark 4218 Andrews, Fred Baker, Bob Braden, Stewart Bryant, Scott Burleigh, 4219 Brian Carpenter, Wojciech Dec, Pavel Drasil, Ralph Droms, Adrian 4220 Farrel, Nick Green, Sri Gundavelli, Brian Haberman, Bernhard Haindl, 4221 Joel Halpern, Tom Herbert, Bob Hinden, Sascha Hlusiak, Lee Howard, 4222 Christian Huitema, Zdenek Jaron, Andre Kostur, Hubert Kuenig, Eliot 4223 Lear, Ted Lemon, Andy Malis, Satoru Matsushima, Tomek Mrugalski, 4224 Thomas Narten, Madhu Niraula, Alexandru Petrescu, Behcet Saikaya, 4225 Michal Skorepa, Dave Thaler, Joe Touch, Bernie Volz, Ryuji Wakikawa, 4226 Tony Whyman, Lloyd Wood and James Woodyatt. Members of the IESG also 4227 provided valuable input during their review process that greatly 4228 improved the document. Special thanks go to Stewart Bryant, Joel 4229 Halpern and Brian Haberman for their shepherding guidance during the 4230 publication of the AERO first edition. 4232 This work has further been encouraged and supported by Boeing 4233 colleagues including Akash Agarwal, Kyle Bae, M. Wayne Benson, Dave 4234 Bernhardt, Cam Brodie, John Bush, Balaguruna Chidambaram, Irene Chin, 4235 Bruce Cornish, Claudiu Danilov, Don Dillenburg, Joe Dudkowski, Wen 4236 Fang, Samad Farooqui, Anthony Gregory, Jeff Holland, Seth Jahne, 4237 Brian Jaury, Greg Kimberly, Ed King, Madhuri Madhava Badgandi, Laurel 4238 Matthew, Gene MacLean III, Kyle Mikos, Rob Muszkiewicz, Sean 4239 O'Sullivan, Satish Raghavendran, Vijay Rajagopalan, Greg Saccone, 4240 Bhargava Raman Sai Prakash, Rod Santiago, Madhanmohan Savadamuthu, 4241 Kent Shuey, Brian Skeen, Mike Slane, Carrie Spiker, Katie Tran, 4242 Brendan Williams, Amelia Wilson, Julie Wulff, Yueli Yang, Eric Yeh 4243 and other members of the Boeing mobility, networking and autonomy 4244 teams. Akash Agarwal, Kyle Bae, Wayne Benson, Madhuri Madhava 4245 Badgandi, Vijayasarathy Rajagopalan, Bhargava Raman Sai Prakash, 4246 Katie Tran and Eric Yeh are especially acknowledged for their work on 4247 the AERO implementation. Chuck Klabunde is honored and remembered 4248 for his early leadership, and we mourn his untimely loss. 4250 This work was inspired by the support and encouragement of countless 4251 outstanding colleagues, managers and program directors over the span 4252 of many decades. Beginning in the late 1980s,' the Digital Equipment 4253 Corporation (DEC) Ultrix Engineering and DECnet Architects groups 4254 identified early issues with fragmentation and bridging links with 4255 diverse MTUs. In the early 1990s, engagements at DEC Project Sequoia 4256 at UC Berkeley and the DEC Western Research Lab in Palo Alto included 4257 investigations into large-scale networked filesystems, ATM vs 4258 Internet and network security proxys. In the mid-1990s to early 4259 2000s employment at the NASA Ames Research Center (Sterling Software) 4260 and SRI International supported early investigations of IPv6, ONR UAV 4261 Communications and the IETF. An employment at Nokia where important 4262 IETF documents were published gave way to a present-day engagement 4263 with The Boeing Company. The work matured at Boeing through major 4264 programs including Future Combat Systems, Advanced Airplane Program, 4265 DTN for the International Space Station, Mobility Vision Lab, CAST, 4266 Caravan, Airplane Internet of Things, the NASA UAS/CNS program, the 4267 FAA/ICAO ATN/IPS program and many others. An attempt to name all who 4268 gave support and encouragement would double the current document size 4269 and result in many unintentional omissions - but to all a humble 4270 thanks. 4272 Earlier works on NBMA tunneling approaches are found in 4273 [RFC2529][RFC5214][RFC5569]. 4275 Many of the constructs presented in this second edition of AERO are 4276 based on the author's earlier works, including: 4278 * The Internet Routing Overlay Network (IRON) 4279 [RFC6179][I-D.templin-ironbis] 4281 * Virtual Enterprise Traversal (VET) 4282 [RFC5558][I-D.templin-intarea-vet] 4284 * The Subnetwork Encapsulation and Adaptation Layer (SEAL) 4285 [RFC5320][I-D.templin-intarea-seal] 4287 * AERO, First Edition [RFC6706] 4289 Note that these works cite numerous earlier efforts that are not also 4290 cited here due to space limitations. The authors of those earlier 4291 works are acknowledged for their insights. 4293 This work is aligned with the NASA Safe Autonomous Systems Operation 4294 (SASO) program under NASA contract number NNA16BD84C. 4296 This work is aligned with the FAA as per the SE2025 contract number 4297 DTFAWA-15-D-00030. 4299 This work is aligned with the Boeing Commercial Airplanes (BCA) 4300 Internet of Things (IoT) and autonomy programs. 4302 This work is aligned with the Boeing Information Technology (BIT) 4303 MobileNet program. 4305 8. References 4307 8.1. Normative References 4309 [I-D.templin-6man-omni] 4310 Templin, F. L., "Transmission of IP Packets over Overlay 4311 Multilink Network (OMNI) Interfaces", Work in Progress, 4312 Internet-Draft, draft-templin-6man-omni-60, 22 April 2022, 4313 . 4316 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 4317 DOI 10.17487/RFC0791, September 1981, 4318 . 4320 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 4321 RFC 792, DOI 10.17487/RFC0792, September 1981, 4322 . 4324 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 4325 Requirement Levels", BCP 14, RFC 2119, 4326 DOI 10.17487/RFC2119, March 1997, 4327 . 4329 [RFC2473] Conta, A. and S. Deering, "Generic Packet Tunneling in 4330 IPv6 Specification", RFC 2473, DOI 10.17487/RFC2473, 4331 December 1998, . 4333 [RFC3971] Arkko, J., Ed., Kempf, J., Zill, B., and P. Nikander, 4334 "SEcure Neighbor Discovery (SEND)", RFC 3971, 4335 DOI 10.17487/RFC3971, March 2005, 4336 . 4338 [RFC3972] Aura, T., "Cryptographically Generated Addresses (CGA)", 4339 RFC 3972, DOI 10.17487/RFC3972, March 2005, 4340 . 4342 [RFC4191] Draves, R. and D. Thaler, "Default Router Preferences and 4343 More-Specific Routes", RFC 4191, DOI 10.17487/RFC4191, 4344 November 2005, . 4346 [RFC4193] Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast 4347 Addresses", RFC 4193, DOI 10.17487/RFC4193, October 2005, 4348 . 4350 [RFC4380] Huitema, C., "Teredo: Tunneling IPv6 over UDP through 4351 Network Address Translations (NATs)", RFC 4380, 4352 DOI 10.17487/RFC4380, February 2006, 4353 . 4355 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 4356 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 4357 DOI 10.17487/RFC4861, September 2007, 4358 . 4360 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 4361 Address Autoconfiguration", RFC 4862, 4362 DOI 10.17487/RFC4862, September 2007, 4363 . 4365 [RFC6081] Thaler, D., "Teredo Extensions", RFC 6081, 4366 DOI 10.17487/RFC6081, January 2011, 4367 . 4369 [RFC7401] Moskowitz, R., Ed., Heer, T., Jokela, P., and T. 4370 Henderson, "Host Identity Protocol Version 2 (HIPv2)", 4371 RFC 7401, DOI 10.17487/RFC7401, April 2015, 4372 . 4374 [RFC7739] Gont, F., "Security Implications of Predictable Fragment 4375 Identification Values", RFC 7739, DOI 10.17487/RFC7739, 4376 February 2016, . 4378 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 4379 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 4380 May 2017, . 4382 [RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6 4383 (IPv6) Specification", STD 86, RFC 8200, 4384 DOI 10.17487/RFC8200, July 2017, 4385 . 4387 [RFC8415] Mrugalski, T., Siodelski, M., Volz, B., Yourtchenko, A., 4388 Richardson, M., Jiang, S., Lemon, T., and T. Winters, 4389 "Dynamic Host Configuration Protocol for IPv6 (DHCPv6)", 4390 RFC 8415, DOI 10.17487/RFC8415, November 2018, 4391 . 4393 8.2. Informative References 4395 [BGP] Huston, G., "BGP in 2015, http://potaroo.net", January 4396 2016. 4398 [I-D.bonica-6man-comp-rtg-hdr] 4399 Bonica, R., Kamite, Y., Alston, A., Henriques, D., and L. 4400 Jalil, "The IPv6 Compact Routing Header (CRH)", Work in 4401 Progress, Internet-Draft, draft-bonica-6man-comp-rtg-hdr- 4402 27, 15 November 2021, . 4405 [I-D.bonica-6man-crh-helper-opt] 4406 Li, X., Bao, C., Ruan, E., and R. Bonica, "Compressed 4407 Routing Header (CRH) Helper Option", Work in Progress, 4408 Internet-Draft, draft-bonica-6man-crh-helper-opt-04, 11 4409 October 2021, . 4412 [I-D.ietf-intarea-frag-fragile] 4413 Bonica, R., Baker, F., Huston, G., Hinden, R. M., Troan, 4414 O., and F. Gont, "IP Fragmentation Considered Fragile", 4415 Work in Progress, Internet-Draft, draft-ietf-intarea-frag- 4416 fragile-17, 30 September 2019, 4417 . 4420 [I-D.ietf-intarea-tunnels] 4421 Touch, J. and M. Townsley, "IP Tunnels in the Internet 4422 Architecture", Work in Progress, Internet-Draft, draft- 4423 ietf-intarea-tunnels-10, 12 September 2019, 4424 . 4427 [I-D.ietf-ipwave-vehicular-networking] 4428 Jeong, J. (., "IPv6 Wireless Access in Vehicular 4429 Environments (IPWAVE): Problem Statement and Use Cases", 4430 Work in Progress, Internet-Draft, draft-ietf-ipwave- 4431 vehicular-networking-28, 30 March 2022, 4432 . 4435 [I-D.ietf-rtgwg-atn-bgp] 4436 Templin, F. L., Saccone, G., Dawra, G., Lindem, A., and V. 4437 Moreno, "A Simple BGP-based Mobile Routing System for the 4438 Aeronautical Telecommunications Network", Work in 4439 Progress, Internet-Draft, draft-ietf-rtgwg-atn-bgp-17, 19 4440 April 2022, . 4443 [I-D.templin-6man-dhcpv6-ndopt] 4444 Templin, F. L., "A Unified Stateful/Stateless 4445 Configuration Service for IPv6", Work in Progress, 4446 Internet-Draft, draft-templin-6man-dhcpv6-ndopt-11, 1 4447 January 2021, . 4450 [I-D.templin-intarea-seal] 4451 Templin, F. L., "The Subnetwork Encapsulation and 4452 Adaptation Layer (SEAL)", Work in Progress, Internet- 4453 Draft, draft-templin-intarea-seal-68, 3 January 2014, 4454 . 4457 [I-D.templin-intarea-vet] 4458 Templin, F. L., "Virtual Enterprise Traversal (VET)", Work 4459 in Progress, Internet-Draft, draft-templin-intarea-vet-40, 4460 3 May 2013, . 4463 [I-D.templin-ipwave-uam-its] 4464 Templin, F. L., "Urban Air Mobility Implications for 4465 Intelligent Transportation Systems", Work in Progress, 4466 Internet-Draft, draft-templin-ipwave-uam-its-04, 4 January 4467 2021, . 4470 [I-D.templin-ironbis] 4471 Templin, F. L., "The Interior Routing Overlay Network 4472 (IRON)", Work in Progress, Internet-Draft, draft-templin- 4473 ironbis-16, 28 March 2014, 4474 . 4477 [I-D.templin-v6ops-pdhost] 4478 Templin, F. L., "IPv6 Prefix Delegation and Multi- 4479 Addressing Models", Work in Progress, Internet-Draft, 4480 draft-templin-v6ops-pdhost-27, 1 January 2021, 4481 . 4484 [IEN48] Cerf, V., "The Catenet Model For Internetworking, 4485 https://www.rfc-editor.org/ien/ien48.txt", July 1978. 4487 [IEN48-2] Cerf, V., "The Catenet Model For Internetworking (with 4488 figures), http://www.postel.org/ien/pdf/ien048.pdf", July 4489 1978. 4491 [OVPN] OpenVPN, O., "http://openvpn.net", October 2016. 4493 [RFC1035] Mockapetris, P., "Domain names - implementation and 4494 specification", STD 13, RFC 1035, DOI 10.17487/RFC1035, 4495 November 1987, . 4497 [RFC1812] Baker, F., Ed., "Requirements for IP Version 4 Routers", 4498 RFC 1812, DOI 10.17487/RFC1812, June 1995, 4499 . 4501 [RFC1918] Rekhter, Y., Moskowitz, B., Karrenberg, D., de Groot, G. 4502 J., and E. Lear, "Address Allocation for Private 4503 Internets", BCP 5, RFC 1918, DOI 10.17487/RFC1918, 4504 February 1996, . 4506 [RFC2003] Perkins, C., "IP Encapsulation within IP", RFC 2003, 4507 DOI 10.17487/RFC2003, October 1996, 4508 . 4510 [RFC2004] Perkins, C., "Minimal Encapsulation within IP", RFC 2004, 4511 DOI 10.17487/RFC2004, October 1996, 4512 . 4514 [RFC2236] Fenner, W., "Internet Group Management Protocol, Version 4515 2", RFC 2236, DOI 10.17487/RFC2236, November 1997, 4516 . 4518 [RFC2464] Crawford, M., "Transmission of IPv6 Packets over Ethernet 4519 Networks", RFC 2464, DOI 10.17487/RFC2464, December 1998, 4520 . 4522 [RFC2529] Carpenter, B. and C. Jung, "Transmission of IPv6 over IPv4 4523 Domains without Explicit Tunnels", RFC 2529, 4524 DOI 10.17487/RFC2529, March 1999, 4525 . 4527 [RFC2983] Black, D., "Differentiated Services and Tunnels", 4528 RFC 2983, DOI 10.17487/RFC2983, October 2000, 4529 . 4531 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 4532 of Explicit Congestion Notification (ECN) to IP", 4533 RFC 3168, DOI 10.17487/RFC3168, September 2001, 4534 . 4536 [RFC3330] IANA, "Special-Use IPv4 Addresses", RFC 3330, 4537 DOI 10.17487/RFC3330, September 2002, 4538 . 4540 [RFC3810] Vida, R., Ed. and L. Costa, Ed., "Multicast Listener 4541 Discovery Version 2 (MLDv2) for IPv6", RFC 3810, 4542 DOI 10.17487/RFC3810, June 2004, 4543 . 4545 [RFC4122] Leach, P., Mealling, M., and R. Salz, "A Universally 4546 Unique IDentifier (UUID) URN Namespace", RFC 4122, 4547 DOI 10.17487/RFC4122, July 2005, 4548 . 4550 [RFC4251] Ylonen, T. and C. Lonvick, Ed., "The Secure Shell (SSH) 4551 Protocol Architecture", RFC 4251, DOI 10.17487/RFC4251, 4552 January 2006, . 4554 [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A 4555 Border Gateway Protocol 4 (BGP-4)", RFC 4271, 4556 DOI 10.17487/RFC4271, January 2006, 4557 . 4559 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 4560 Architecture", RFC 4291, DOI 10.17487/RFC4291, February 4561 2006, . 4563 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 4564 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 4565 December 2005, . 4567 [RFC4389] Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery 4568 Proxies (ND Proxy)", RFC 4389, DOI 10.17487/RFC4389, April 4569 2006, . 4571 [RFC4443] Conta, A., Deering, S., and M. Gupta, Ed., "Internet 4572 Control Message Protocol (ICMPv6) for the Internet 4573 Protocol Version 6 (IPv6) Specification", STD 89, 4574 RFC 4443, DOI 10.17487/RFC4443, March 2006, 4575 . 4577 [RFC4511] Sermersheim, J., Ed., "Lightweight Directory Access 4578 Protocol (LDAP): The Protocol", RFC 4511, 4579 DOI 10.17487/RFC4511, June 2006, 4580 . 4582 [RFC4541] Christensen, M., Kimball, K., and F. Solensky, 4583 "Considerations for Internet Group Management Protocol 4584 (IGMP) and Multicast Listener Discovery (MLD) Snooping 4585 Switches", RFC 4541, DOI 10.17487/RFC4541, May 2006, 4586 . 4588 [RFC4605] Fenner, B., He, H., Haberman, B., and H. Sandick, 4589 "Internet Group Management Protocol (IGMP) / Multicast 4590 Listener Discovery (MLD)-Based Multicast Forwarding 4591 ("IGMP/MLD Proxying")", RFC 4605, DOI 10.17487/RFC4605, 4592 August 2006, . 4594 [RFC4982] Bagnulo, M. and J. Arkko, "Support for Multiple Hash 4595 Algorithms in Cryptographically Generated Addresses 4596 (CGAs)", RFC 4982, DOI 10.17487/RFC4982, July 2007, 4597 . 4599 [RFC5015] Handley, M., Kouvelas, I., Speakman, T., and L. Vicisano, 4600 "Bidirectional Protocol Independent Multicast (BIDIR- 4601 PIM)", RFC 5015, DOI 10.17487/RFC5015, October 2007, 4602 . 4604 [RFC5214] Templin, F., Gleeson, T., and D. Thaler, "Intra-Site 4605 Automatic Tunnel Addressing Protocol (ISATAP)", RFC 5214, 4606 DOI 10.17487/RFC5214, March 2008, 4607 . 4609 [RFC5320] Templin, F., Ed., "The Subnetwork Encapsulation and 4610 Adaptation Layer (SEAL)", RFC 5320, DOI 10.17487/RFC5320, 4611 February 2010, . 4613 [RFC5522] Eddy, W., Ivancic, W., and T. Davis, "Network Mobility 4614 Route Optimization Requirements for Operational Use in 4615 Aeronautics and Space Exploration Mobile Networks", 4616 RFC 5522, DOI 10.17487/RFC5522, October 2009, 4617 . 4619 [RFC5558] Templin, F., Ed., "Virtual Enterprise Traversal (VET)", 4620 RFC 5558, DOI 10.17487/RFC5558, February 2010, 4621 . 4623 [RFC5569] Despres, R., "IPv6 Rapid Deployment on IPv4 4624 Infrastructures (6rd)", RFC 5569, DOI 10.17487/RFC5569, 4625 January 2010, . 4627 [RFC5880] Katz, D. and D. Ward, "Bidirectional Forwarding Detection 4628 (BFD)", RFC 5880, DOI 10.17487/RFC5880, June 2010, 4629 . 4631 [RFC6106] Jeong, J., Park, S., Beloeil, L., and S. Madanapalli, 4632 "IPv6 Router Advertisement Options for DNS Configuration", 4633 RFC 6106, DOI 10.17487/RFC6106, November 2010, 4634 . 4636 [RFC6139] Russert, S., Ed., Fleischman, E., Ed., and F. Templin, 4637 Ed., "Routing and Addressing in Networks with Global 4638 Enterprise Recursion (RANGER) Scenarios", RFC 6139, 4639 DOI 10.17487/RFC6139, February 2011, 4640 . 4642 [RFC6146] Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful 4643 NAT64: Network Address and Protocol Translation from IPv6 4644 Clients to IPv4 Servers", RFC 6146, DOI 10.17487/RFC6146, 4645 April 2011, . 4647 [RFC6179] Templin, F., Ed., "The Internet Routing Overlay Network 4648 (IRON)", RFC 6179, DOI 10.17487/RFC6179, March 2011, 4649 . 4651 [RFC6221] Miles, D., Ed., Ooghe, S., Dec, W., Krishnan, S., and A. 4652 Kavanagh, "Lightweight DHCPv6 Relay Agent", RFC 6221, 4653 DOI 10.17487/RFC6221, May 2011, 4654 . 4656 [RFC6273] Kukec, A., Krishnan, S., and S. Jiang, "The Secure 4657 Neighbor Discovery (SEND) Hash Threat Analysis", RFC 6273, 4658 DOI 10.17487/RFC6273, June 2011, 4659 . 4661 [RFC6347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer 4662 Security Version 1.2", RFC 6347, DOI 10.17487/RFC6347, 4663 January 2012, . 4665 [RFC6355] Narten, T. and J. Johnson, "Definition of the UUID-Based 4666 DHCPv6 Unique Identifier (DUID-UUID)", RFC 6355, 4667 DOI 10.17487/RFC6355, August 2011, 4668 . 4670 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 4671 for Equal Cost Multipath Routing and Link Aggregation in 4672 Tunnels", RFC 6438, DOI 10.17487/RFC6438, November 2011, 4673 . 4675 [RFC6706] Templin, F., Ed., "Asymmetric Extended Route Optimization 4676 (AERO)", RFC 6706, DOI 10.17487/RFC6706, August 2012, 4677 . 4679 [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and 4680 UDP Checksums for Tunneled Packets", RFC 6935, 4681 DOI 10.17487/RFC6935, April 2013, 4682 . 4684 [RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement 4685 for the Use of IPv6 UDP Datagrams with Zero Checksums", 4686 RFC 6936, DOI 10.17487/RFC6936, April 2013, 4687 . 4689 [RFC7333] Chan, H., Ed., Liu, D., Seite, P., Yokota, H., and J. 4690 Korhonen, "Requirements for Distributed Mobility 4691 Management", RFC 7333, DOI 10.17487/RFC7333, August 2014, 4692 . 4694 [RFC7761] Fenner, B., Handley, M., Holbrook, H., Kouvelas, I., 4695 Parekh, R., Zhang, Z., and L. Zheng, "Protocol Independent 4696 Multicast - Sparse Mode (PIM-SM): Protocol Specification 4697 (Revised)", STD 83, RFC 7761, DOI 10.17487/RFC7761, March 4698 2016, . 4700 [RFC8402] Filsfils, C., Ed., Previdi, S., Ed., Ginsberg, L., 4701 Decraene, B., Litkowski, S., and R. Shakir, "Segment 4702 Routing Architecture", RFC 8402, DOI 10.17487/RFC8402, 4703 July 2018, . 4705 [RFC8446] Rescorla, E., "The Transport Layer Security (TLS) Protocol 4706 Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018, 4707 . 4709 [RFC8754] Filsfils, C., Ed., Dukes, D., Ed., Previdi, S., Leddy, J., 4710 Matsushima, S., and D. Voyer, "IPv6 Segment Routing Header 4711 (SRH)", RFC 8754, DOI 10.17487/RFC8754, March 2020, 4712 . 4714 [RFC9000] Iyengar, J., Ed. and M. Thomson, Ed., "QUIC: A UDP-Based 4715 Multiplexed and Secure Transport", RFC 9000, 4716 DOI 10.17487/RFC9000, May 2021, 4717 . 4719 [RFC9001] Thomson, M., Ed. and S. Turner, Ed., "Using TLS to Secure 4720 QUIC", RFC 9001, DOI 10.17487/RFC9001, May 2021, 4721 . 4723 [RFC9002] Iyengar, J., Ed. and I. Swett, Ed., "QUIC Loss Detection 4724 and Congestion Control", RFC 9002, DOI 10.17487/RFC9002, 4725 May 2021, . 4727 [WG] Wireguard, "WireGuard, https://www.wireguard.com", August 4728 2020. 4730 Appendix A. Non-Normative Considerations 4732 AERO can be applied to a multitude of Internetworking scenarios, with 4733 each having its own adaptations. The following considerations are 4734 provided as non-normative guidance: 4736 A.1. Implementation Strategies for Route Optimization 4738 Route optimization as discussed in Section 3.13 results in the 4739 creation of NCEs. The NCE state is set to REACHABLE for at most 4740 ReachableTime seconds. In order to refresh the NCE lifetime before 4741 the ReachableTime timer expires, the specification requires 4742 implementations to issue a new NS/NA(AR) exchange to reset 4743 ReachableTime while data packets are still flowing. However, the 4744 decision of when to initiate a new NS/NA(AR) exchange and to 4745 perpetuate the process is left as an implementation detail. 4747 One possible strategy may be to monitor the NCE watching for data 4748 packets for (ReachableTime - 5) seconds. If any data packets have 4749 been sent to the neighbor within this timeframe, then send an NS(AR) 4750 to receive a new NA(AR). If no data packets have been sent, wait for 4751 5 additional seconds and send an immediate NS(AR) if any data packets 4752 are sent within this "expiration pending" 5 second window. If no 4753 additional data packets are sent within the 5 second window, reset 4754 the NCE state to STALE. 4756 The monitoring of the neighbor data packet traffic therefore becomes 4757 an ongoing process during the NCE lifetime. If the NCE expires, 4758 future data packets will trigger a new NS/NA(AR) exchange while the 4759 packets themselves are delivered over a longer path until route 4760 optimization state is re-established. 4762 A.2. Implicit Mobility Management 4764 OMNI interface neighbors MAY provide a configuration option that 4765 allows them to perform implicit mobility management in which no IPv6 4766 ND messaging is used. In that case, the Client only transmits 4767 packets over a single interface at a time, and the neighbor always 4768 observes packets arriving from the Client from the same link-layer 4769 source address. 4771 If the Client's underlay interface address changes (either due to a 4772 readdressing of the original interface or switching to a new 4773 interface) the neighbor immediately updates the NCE for the Client 4774 and begins accepting and sending packets according to the Client's 4775 new address. This implicit mobility method applies to use cases such 4776 as cellphones with both WiFi and Cellular interfaces where only one 4777 of the interfaces is active at a given time, and the Client 4778 automatically switches over to the backup interface if the primary 4779 interface fails. 4781 A.3. Direct Underlying Interfaces 4783 When a Client's OMNI interface is configured over a Direct interface, 4784 the neighbor at the other end of the Direct link can receive packets 4785 without any encapsulation. In that case, the Client sends packets 4786 over the Direct link according to traffic selectors. If the Direct 4787 interface is selected, then the Client's IP packets are transmitted 4788 directly to the peer without going through an ANET/INET. If other 4789 interfaces are selected, then the Client's IP packets are transmitted 4790 via a different interface, which may result in the inclusion of 4791 Proxy/Servers and Gateways in the communications path. Direct 4792 interfaces must be tested periodically for reachability, e.g., via 4793 NUD. 4795 A.4. AERO Critical Infrastructure Considerations 4797 AERO Gateways can be either Commercial off-the Shelf (COTS) standard 4798 IP routers or virtual machines in the cloud. Gateways must be 4799 provisioned, supported and managed by the INET administrative 4800 authority, and connected to the Gateways of other INETs via inter- 4801 domain peerings. Cost for purchasing, configuring and managing 4802 Gateways is nominal even for very large OMNI links. 4804 AERO INET Proxy/Servers can be standard dedicated server platforms, 4805 but most often will be deployed as virtual machines in the cloud. 4806 The only requirements for INET Proxy/Servers are that they can run 4807 the AERO/OMNI code and have at least one network interface connection 4808 to the INET. INET Proxy/Servers must be provisioned, supported and 4809 managed by the INET administrative authority. Cost for purchasing, 4810 configuring and managing cloud Proxy/Servers is nominal especially 4811 for virtual machines. 4813 AERO ANET Proxy/Servers are most often standard dedicated server 4814 platforms with one underlay interface connected to the ANET and a 4815 second interface connected to an INET. As with INET Proxy/Servers, 4816 the only requirements are that they can run the AERO/OMNI code and 4817 have at least one interface connection to the INET. ANET Proxy/ 4818 Servers must be provisioned, supported and managed by the ANET 4819 administrative authority. Cost for purchasing, configuring and 4820 managing Proxys is nominal, and borne by the ANET administrative 4821 authority. 4823 AERO Relays are simply Proxy/Servers connected to INETs and/or ENETs 4824 that provide forwarding services for non-MNP destinations. The Relay 4825 connects to the OMNI link and engages in eBGP peering with one or 4826 more Gateways as a stub AS. The Relay then injects its MNPs and/or 4827 non-MNP prefixes into the BGP routing system, and provisions the 4828 prefixes to its downstream-attached networks. The Relay can perform 4829 ROS/ROR services the same as for any Proxy/Server, and can route 4830 between the MNP and non-MNP address spaces. 4832 A.5. AERO Server Failure Implications 4834 AERO Proxy/Servers may appear as a single point of failure in the 4835 architecture, but such is not the case since all Proxy/Servers on the 4836 link provide identical services and loss of a Proxy/Server does not 4837 imply immediate and/or comprehensive communication failures. Proxy/ 4838 Server failure is quickly detected and conveyed by Bidirectional 4839 Forward Detection (BFD) and/or proactive NUD allowing Clients to 4840 migrate to new Proxy/Servers. 4842 If a Proxy/Server fails, ongoing packet forwarding to Clients will 4843 continue by virtue of the neighbor cache entries that have already 4844 been established in route optimization sources (ROSs). If a Client 4845 also experiences mobility events at roughly the same time the Proxy/ 4846 Server fails, uNA messages may be lost but neighbor cache entries in 4847 the DEPARTED state will ensure that packet forwarding to the Client's 4848 new locations will continue for up to DepartTime seconds. 4850 If a Client is left without a Proxy/Server for a considerable length 4851 of time (e.g., greater than ReachableTime seconds) then existing 4852 neighbor cache entries will eventually expire and both ongoing and 4853 new communications will fail. The original source will continue to 4854 retransmit until the Client has established a new Proxy/Server 4855 relationship, after which time continuous communications will resume. 4857 Therefore, providing many Proxy/Servers on the link with high 4858 availability profiles provides resilience against loss of individual 4859 Proxy/Servers and assurance that Clients can establish new Proxy/ 4860 Server relationships quickly in event of a Proxy/Server failure. 4862 A.6. AERO Client / Server Architecture 4864 The AERO architectural model is client / server in the control plane, 4865 with route optimization in the data plane. The same as for common 4866 Internet services, the AERO Client discovers the addresses of AERO 4867 Proxy/Servers and connects to one or more of them. The AERO service 4868 is analogous to common Internet services such as google.com, 4869 yahoo.com, cnn.com, etc. However, there is only one AERO service for 4870 the link and all Proxy/Servers provide identical services. 4872 Common Internet services provide differing strategies for advertising 4873 server addresses to clients. The strategy is conveyed through the 4874 DNS resource records returned in response to name resolution queries. 4875 As of January 2020 Internet-based 'nslookup' services were used to 4876 determine the following: 4878 * When a client resolves the domainname "google.com", the DNS always 4879 returns one A record (i.e., an IPv4 address) and one AAAA record 4880 (i.e., an IPv6 address). The client receives the same addresses 4881 each time it resolves the domainname via the same DNS resolver, 4882 but may receive different addresses when it resolves the 4883 domainname via different DNS resolvers. But, in each case, 4884 exactly one A and one AAAA record are returned. 4886 * When a client resolves the domainname "ietf.org", the DNS always 4887 returns one A record and one AAAA record with the same addresses 4888 regardless of which DNS resolver is used. 4890 * When a client resolves the domainname "yahoo.com", the DNS always 4891 returns a list of 4 A records and 4 AAAA records. Each time the 4892 client resolves the domainname via the same DNS resolver, the same 4893 list of addresses are returned but in randomized order (i.e., 4894 consistent with a DNS round-robin strategy). But, interestingly, 4895 the same addresses are returned (albeit in randomized order) when 4896 the domainname is resolved via different DNS resolvers. 4898 * When a client resolves the domainname "amazon.com", the DNS always 4899 returns a list of 3 A records and no AAAA records. As with 4900 "yahoo.com", the same three A records are returned from any 4901 worldwide Internet connection point in randomized order. 4903 The above example strategies show differing approaches to Internet 4904 resilience and service distribution offered by major Internet 4905 services. The Google approach exposes only a single IPv4 and a 4906 single IPv6 address to clients. Clients can then select whichever IP 4907 protocol version offers the best response, but will always use the 4908 same IP address according to the current Internet connection point. 4909 This means that the IP address offered by the network must lead to a 4910 highly-available server and/or service distribution point. In other 4911 words, resilience is predicated on high availability within the 4912 network and with no client-initiated failovers expected (i.e., it is 4913 all-or-nothing from the client's perspective). However, Google does 4914 provide for worldwide distributed service distribution by virtue of 4915 the fact that each Internet connection point responds with a 4916 different IPv6 and IPv4 address. The IETF approach is like google 4917 (all-or-nothing from the client's perspective), but provides only a 4918 single IPv4 or IPv6 address on a worldwide basis. This means that 4919 the addresses must be made highly-available at the network level with 4920 no client failover possibility, and if there is any worldwide service 4921 distribution it would need to be conducted by a network element that 4922 is reached via the IP address acting as a service distribution point. 4924 In contrast to the Google and IETF philosophies, Yahoo and Amazon 4925 both provide clients with a (short) list of IP addresses with Yahoo 4926 providing both IP protocol versions and Amazon as IPv4-only. The 4927 order of the list is randomized with each name service query 4928 response, with the effect of round-robin load balancing for service 4929 distribution. With a short list of addresses, there is still 4930 expectation that the network will implement high availability for 4931 each address but in case any single address fails the client can 4932 switch over to using a different address. The balance then becomes 4933 one of function in the network vs function in the end system. 4935 The same implications observed for common highly-available services 4936 in the Internet apply also to the AERO client/server architecture. 4937 When an AERO Client connects to one or more ANETs, it discovers one 4938 or more AERO Proxy/Server addresses through the mechanisms discussed 4939 in earlier sections. Each Proxy/Server address presumably leads to a 4940 fault-tolerant clustering arrangement such as supported by Linux-HA, 4941 Extended Virtual Synchrony or Paxos. Such an arrangement has 4942 precedence in common Internet service deployments in lightweight 4943 virtual machines without requiring expensive hardware deployment. 4944 Similarly, common Internet service deployments set service IP 4945 addresses on service distribution points that may relay requests to 4946 many different servers. 4948 For AERO, the expectation is that a combination of the Google/IETF 4949 and Yahoo/Amazon philosophies would be employed. The AERO Client 4950 connects to different ANET access points and can receive 1-2 Proxy/ 4951 Server ULAs at each point. It then selects one AERO Proxy/Server 4952 address, and engages in RS/RA exchanges with the same Proxy/Server 4953 from all ANET connections. The Client remains with this Proxy/Server 4954 unless or until the Proxy/Server fails, in which case it can switch 4955 over to an alternate Proxy/Server. The Client can likewise switch 4956 over to a different Proxy/Server at any time if there is some reason 4957 for it to do so. So, the AERO expectation is for a balance of 4958 function in the network and end system, with fault tolerance and 4959 resilience at both levels. 4961 Appendix B. Change Log 4963 << RFC Editor - remove prior to publication >> 4965 Changes from earlier versions: 4967 * Submit for RFC publication. 4969 Author's Address 4971 Fred L. Templin (editor) 4972 Boeing Research & Technology 4973 P.O. Box 3707 4974 Seattle, WA 98124 4975 United States of America 4976 Email: fltemplin@acm.org