idnits 2.17.1 draft-templin-intarea-6706bis-93.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- == The 'Obsoletes: ' line in the draft header should list only the _numbers_ of the RFCs which will be obsoleted by this document (if approved); it should not include the word 'RFC' in the list. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (February 9, 2021) is 1171 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Missing Reference: 'ULA' is mentioned on line 1008, but not defined == Missing Reference: 'ULA0' is mentioned on line 1008, but not defined == Missing Reference: 'ULA1' is mentioned on line 1008, but not defined == Missing Reference: 'ULA2' is mentioned on line 1008, but not defined == Unused Reference: 'RFC3971' is defined on line 3460, but no explicit reference was found in the text == Unused Reference: 'RFC3972' is defined on line 3465, but no explicit reference was found in the text == Unused Reference: 'RFC4193' is defined on line 3473, but no explicit reference was found in the text == Unused Reference: 'BGP' is defined on line 3518, but no explicit reference was found in the text == Unused Reference: 'I-D.bonica-6man-comp-rtg-hdr' is defined on line 3521, but no explicit reference was found in the text == Unused Reference: 'I-D.bonica-6man-crh-helper-opt' is defined on line 3527, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-intarea-frag-fragile' is defined on line 3532, but no explicit reference was found in the text == Unused Reference: 'I-D.templin-6man-dhcpv6-ndopt' is defined on line 3555, but no explicit reference was found in the text == Unused Reference: 'RFC4122' is defined on line 3633, but no explicit reference was found in the text == Unused Reference: 'RFC4982' is defined on line 3682, but no explicit reference was found in the text == Unused Reference: 'RFC6273' is defined on line 3738, but no explicit reference was found in the text == Unused Reference: 'RFC6355' is defined on line 3747, but no explicit reference was found in the text == Outdated reference: A later version (-99) exists of draft-templin-6man-omni-interface-69 == Outdated reference: A later version (-31) exists of draft-bonica-6man-comp-rtg-hdr-24 == Outdated reference: A later version (-04) exists of draft-bonica-6man-crh-helper-opt-02 == 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-19 == Outdated reference: A later version (-26) exists of draft-ietf-rtgwg-atn-bgp-10 -- 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 (~~), 24 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group F. Templin, Ed. 3 Internet-Draft Boeing Research & Technology 4 Obsoletes: rfc5320, rfc5558, rfc5720, February 9, 2021 5 rfc6179, rfc6706 (if 6 approved) 7 Intended status: Standards Track 8 Expires: August 13, 2021 10 Asymmetric Extended Route Optimization (AERO) 11 draft-templin-intarea-6706bis-93 13 Abstract 15 This document specifies an Asymmetric Extended Route Optimization 16 (AERO) service for IP internetworking over Overlay Multilink Network 17 (OMNI) interfaces. AERO/OMNI use an IPv6 link-local address format 18 that supports operation of the IPv6 Neighbor Discovery (ND) protocol 19 and links ND to IP forwarding. Prefix delegation/registration 20 services are employed for network admission and to manage the routing 21 system. Multilink operation, mobility management, multicast, quality 22 of service (QoS) signaling and route optimization are naturally 23 supported through dynamic neighbor cache updates. AERO is a widely- 24 applicable mobile internetworking service especially well-suited to 25 aviation services, intelligent transportation systems, mobile Virtual 26 Private Networks (VPNs) and many other applications. 28 Status of This Memo 30 This Internet-Draft is submitted in full conformance with the 31 provisions of BCP 78 and BCP 79. 33 Internet-Drafts are working documents of the Internet Engineering 34 Task Force (IETF). Note that other groups may also distribute 35 working documents as Internet-Drafts. The list of current Internet- 36 Drafts is at https://datatracker.ietf.org/drafts/current/. 38 Internet-Drafts are draft documents valid for a maximum of six months 39 and may be updated, replaced, or obsoleted by other documents at any 40 time. It is inappropriate to use Internet-Drafts as reference 41 material or to cite them other than as "work in progress." 43 This Internet-Draft will expire on August 13, 2021. 45 Copyright Notice 47 Copyright (c) 2021 IETF Trust and the persons identified as the 48 document authors. All rights reserved. 50 This document is subject to BCP 78 and the IETF Trust's Legal 51 Provisions Relating to IETF Documents 52 (https://trustee.ietf.org/license-info) in effect on the date of 53 publication of this document. Please review these documents 54 carefully, as they describe your rights and restrictions with respect 55 to this document. Code Components extracted from this document must 56 include Simplified BSD License text as described in Section 4.e of 57 the Trust Legal Provisions and are provided without warranty as 58 described in the Simplified BSD License. 60 Table of Contents 62 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4 63 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 5 64 3. Asymmetric Extended Route Optimization (AERO) . . . . . . . . 11 65 3.1. AERO Node Types . . . . . . . . . . . . . . . . . . . . . 11 66 3.2. The AERO Service over OMNI Links . . . . . . . . . . . . 12 67 3.2.1. AERO/OMNI Reference Model . . . . . . . . . . . . . . 12 68 3.2.2. Addressing and Node Identification . . . . . . . . . 15 69 3.2.3. AERO Routing System . . . . . . . . . . . . . . . . . 16 70 3.2.4. OMNI Link Segment Routing . . . . . . . . . . . . . . 18 71 3.2.5. Segment Routing Topologies (SRTs) . . . . . . . . . . 22 72 3.2.6. Segment Routing For OMNI Link Selection . . . . . . . 23 73 3.2.7. Segment Routing Within the OMNI Link . . . . . . . . 23 74 3.3. OMNI Interface Characteristics . . . . . . . . . . . . . 24 75 3.4. OMNI Interface Initialization . . . . . . . . . . . . . . 26 76 3.4.1. AERO Server/Relay Behavior . . . . . . . . . . . . . 26 77 3.4.2. AERO Proxy Behavior . . . . . . . . . . . . . . . . . 26 78 3.4.3. AERO Client Behavior . . . . . . . . . . . . . . . . 27 79 3.4.4. AERO Bridge Behavior . . . . . . . . . . . . . . . . 27 80 3.5. OMNI Interface Neighbor Cache Maintenance . . . . . . . . 27 81 3.5.1. OMNI Neighbor Interface Attributes . . . . . . . . . 29 82 3.5.2. OMNI Neighbor Advertisement Message Flags . . . . . . 30 83 3.6. OMNI Interface Encapsulation and Re-encapsulation . . . . 30 84 3.7. OMNI Interface Decapsulation . . . . . . . . . . . . . . 32 85 3.8. OMNI Interface Data Origin Authentication . . . . . . . . 32 86 3.9. OMNI Adaptation Layer and OMNI Interface MTU . . . . . . 32 87 3.10. OMNI Interface Forwarding Algorithm . . . . . . . . . . . 33 88 3.10.1. Client Forwarding Algorithm . . . . . . . . . . . . 34 89 3.10.2. Proxy Forwarding Algorithm . . . . . . . . . . . . . 35 90 3.10.3. Server/Relay Forwarding Algorithm . . . . . . . . . 36 91 3.10.4. Bridge Forwarding Algorithm . . . . . . . . . . . . 38 92 3.11. OMNI Interface Error Handling . . . . . . . . . . . . . . 39 93 3.12. AERO Router Discovery, Prefix Delegation and 94 Autoconfiguration . . . . . . . . . . . . . . . . . . . . 42 95 3.12.1. AERO Service Model . . . . . . . . . . . . . . . . . 42 96 3.12.2. AERO Client Behavior . . . . . . . . . . . . . . . . 42 97 3.12.3. AERO Server Behavior . . . . . . . . . . . . . . . . 44 98 3.13. The AERO Proxy . . . . . . . . . . . . . . . . . . . . . 47 99 3.13.1. Combined Proxy/Servers . . . . . . . . . . . . . . . 49 100 3.13.2. Detecting and Responding to Server Failures . . . . 50 101 3.13.3. Point-to-Multipoint Server Coordination . . . . . . 50 102 3.14. AERO Route Optimization / Address Resolution . . . . . . 51 103 3.14.1. Route Optimization Initiation . . . . . . . . . . . 52 104 3.14.2. Relaying the NS(AR) . . . . . . . . . . . . . . . . 52 105 3.14.3. Processing the NS(AR) and Sending the NA(AR) . . . . 53 106 3.14.4. Relaying the NA(AR) . . . . . . . . . . . . . . . . 54 107 3.14.5. Processing the NA(AR) . . . . . . . . . . . . . . . 54 108 3.14.6. Route Optimization Maintenance . . . . . . . . . . . 54 109 3.15. Neighbor Unreachability Detection (NUD) . . . . . . . . . 55 110 3.16. Mobility Management and Quality of Service (QoS) . . . . 57 111 3.16.1. Mobility Update Messaging . . . . . . . . . . . . . 57 112 3.16.2. Announcing Link-Layer Address and/or QoS Preference 113 Changes . . . . . . . . . . . . . . . . . . . . . . 59 114 3.16.3. Bringing New Links Into Service . . . . . . . . . . 59 115 3.16.4. Deactivating Existing Links . . . . . . . . . . . . 59 116 3.16.5. Moving Between Servers . . . . . . . . . . . . . . . 59 117 3.17. Multicast . . . . . . . . . . . . . . . . . . . . . . . . 61 118 3.17.1. Source-Specific Multicast (SSM) . . . . . . . . . . 61 119 3.17.2. Any-Source Multicast (ASM) . . . . . . . . . . . . . 63 120 3.17.3. Bi-Directional PIM (BIDIR-PIM) . . . . . . . . . . . 63 121 3.18. Operation over Multiple OMNI Links . . . . . . . . . . . 63 122 3.19. DNS Considerations . . . . . . . . . . . . . . . . . . . 64 123 3.20. Transition Considerations . . . . . . . . . . . . . . . . 65 124 3.21. Detecting and Reacting to Server and Bridge Failures . . 65 125 3.22. AERO Clients on the Open Internet . . . . . . . . . . . . 66 126 3.23. Time-Varying MNPs . . . . . . . . . . . . . . . . . . . . 69 127 4. Implementation Status . . . . . . . . . . . . . . . . . . . . 70 128 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 70 129 6. Security Considerations . . . . . . . . . . . . . . . . . . . 70 130 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 72 131 8. References . . . . . . . . . . . . . . . . . . . . . . . . . 73 132 8.1. Normative References . . . . . . . . . . . . . . . . . . 73 133 8.2. Informative References . . . . . . . . . . . . . . . . . 75 134 Appendix A. Non-Normative Considerations . . . . . . . . . . . . 81 135 A.1. Implementation Strategies for Route Optimization . . . . 82 136 A.2. Implicit Mobility Management . . . . . . . . . . . . . . 82 137 A.3. Direct Underlying Interfaces . . . . . . . . . . . . . . 83 138 A.4. AERO Critical Infrastructure Considerations . . . . . . . 83 139 A.5. AERO Server Failure Implications . . . . . . . . . . . . 84 140 A.6. AERO Client / Server Architecture . . . . . . . . . . . . 84 142 Appendix B. Change Log . . . . . . . . . . . . . . . . . . . . . 86 143 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 87 145 1. Introduction 147 Asymmetric Extended Route Optimization (AERO) fulfills the 148 requirements of Distributed Mobility Management (DMM) [RFC7333] and 149 route optimization [RFC5522] for aeronautical networking and other 150 network mobility use cases such as intelligent transportation 151 systems. AERO is an internetworking and mobility management service 152 that employs the Overlay Multilink Network Interface (OMNI) 153 [I-D.templin-6man-omni-interface] Non-Broadcast, Multiple Access 154 (NBMA) virtual link model. The OMNI link is a virtual overlay 155 configured over one or more underlying Internetworks, and nodes on 156 the link can exchange IP packets as single-hop neighbors via 157 encapsulation. The OMNI Adaptation Layer (OAL) supports multilink 158 operation for increased reliability, bandwidth optimization and 159 traffic path selection while accommodating Maximum Transmission Unit 160 (MTU) diversity. 162 The AERO service comprises Clients, Proxys, Servers and Relays that 163 are seen as OMNI link neighbors as well as Bridges that interconnect 164 OMNI link segments. Each node's OMNI interface uses an IPv6 link- 165 local address format that supports operation of the IPv6 Neighbor 166 Discovery (ND) protocol [RFC4861] and links ND to IP forwarding. A 167 node's OMNI interface can be configured over multiple underlying 168 interfaces, and therefore appears as a single interface with multiple 169 link-layer addresses. Each link-layer address is subject to change 170 due to mobility and/or QoS fluctuations, and link-layer address 171 changes are signaled by ND messaging the same as for any IPv6 link. 173 AERO provides a cloud-based service where mobile node Clients may use 174 any Server acting as a Mobility Anchor Point (MAP) and fixed nodes 175 may use any Relay on the link for efficient communications. Fixed 176 nodes forward packets destined to other AERO nodes via the nearest 177 Relay, which forwards them through the cloud. A mobile node's 178 initial packets are forwarded through the Server, while direct 179 routing is supported through asymmetric extended route optimization 180 while data packets are flowing. Both unicast and multicast 181 communications are supported, and mobile nodes may efficiently move 182 between locations while maintaining continuous communications with 183 correspondents and without changing their IP Address. 185 AERO Bridges are interconnected in a secured private BGP overlay 186 routing instance using encapsulation to provide a hybrid routing/ 187 bridging service that joins the underlying Internetworks of multiple 188 disjoint administrative domains into a single unified OMNI link. 189 Each OMNI link instance is characterized by the set of Mobility 190 Service Prefixes (MSPs) common to all mobile nodes. The link extends 191 to the point where a Relay/Server is on the optimal route from any 192 correspondent node on the link, and provides a conduit between the 193 underlying Internetwork and the OMNI link. To the underlying 194 Internetwork, the Relay/Server is the source of a route to the MSP, 195 and hence uplink traffic to the mobile node is naturally routed to 196 the nearest Relay/Server. 198 AERO can be used with OMNI links that span private-use Internetworks 199 and/or public Internetworks such as the global Internet. In the 200 latter case, some end systems may be located behind global Internet 201 Network Address Translators (NATs). A means for robust traversal of 202 NATs while avoiding "triangle routing" is therefore provided. 204 AERO assumes the use of PIM Sparse Mode in support of multicast 205 communication. In support of Source Specific Multicast (SSM) when a 206 Mobile Node is the source, AERO route optimization ensures that a 207 shortest-path multicast tree is established with provisions for 208 mobility and multilink operation. In all other multicast scenarios 209 there are no AERO dependencies. 211 AERO was designed for aeronautical networking for both manned and 212 unmanned aircraft, where the aircraft is treated as a mobile node 213 that can connect an Internet of Things (IoT). AERO is also 214 applicable to a wide variety of other use cases. For example, it can 215 be used to coordinate the links of mobile nodes (e.g., cellphones, 216 tablets, laptop computers, etc.) that connect into a home enterprise 217 network via public access networks using tunneling software such as 218 OpenVPN [OVPN] with VPN or non-VPN services enabled according to the 219 appropriate security model. AERO can also be used to facilitate 220 terrestrial vehicular and urban air mobility (as well as pedestrian 221 communication services) for future intelligent transportation systems 222 [I-D.ietf-ipwave-vehicular-networking][I-D.templin-ipwave-uam-its]. 223 Other applicable use cases are also in scope. 225 The following numbered sections present the AERO specification. The 226 appendices at the end of the document are non-normative. 228 2. Terminology 230 The terminology in the normative references applies; especially, the 231 terminology in the OMNI specification 232 [I-D.templin-6man-omni-interface] is used extensively throughout. 233 The following terms are defined within the scope of this document: 235 IPv6 Neighbor Discovery (ND) 236 an IPv6 control message service for coordinating neighbor 237 relationships between nodes connected to a common link. AERO uses 238 the ND service specified in [RFC4861]. 240 IPv6 Prefix Delegation 241 a networking service for delegating IPv6 prefixes to nodes on the 242 link. The nominal service is DHCPv6 [RFC8415], however alternate 243 services (e.g., based on ND messaging) are also in scope. Most 244 notably, a minimal form of prefix delegation known as "prefix 245 registration" can be used if the Client knows its prefix in 246 advance and can represent it in the IPv6 source address of an ND 247 message. 249 Access Network (ANET) 250 a node's first-hop data link service network (e.g., a radio access 251 network, cellular service provider network, corporate enterprise 252 network, etc.) that often provides link-layer security services 253 such as IEEE 802.1X and physical-layer security prevent 254 unauthorized access internally and with border network-layer 255 security services such as firewalls and proxys that prevent 256 unauthorized outside access. 258 ANET interface 259 a node's attachment to a link in an ANET. 261 Internetwork (INET) 262 a connected IP network topology with a coherent routing and 263 addressing plan and that provides a transit backbone service for 264 ANET end systems. INETs also provide an underlay service over 265 which the AERO virtual link is configured. Example INETs include 266 corporate enterprise networks, aviation networks, and the public 267 Internet itself. When there is no administrative boundary between 268 an ANET and the INET, the ANET and INET are one and the same. 270 INET Partition 271 frequently, INETs such as large corporate enterprise networks are 272 sub-divided internally into separate isolated partitions. Each 273 partition is fully connected internally but disconnected from 274 other partitions, and there is no requirement that separate 275 partitions maintain consistent Internet Protocol and/or addressing 276 plans. (Each INET partition is seen as a separate OMNI link 277 segment as discussed below.) 279 INET interface 280 a node's attachment to a link in an INET. 282 INET address 283 an IP address assigned to a node's interface connection to an 284 INET. 286 INET encapsulation 287 the encapsulation of a packet in an outer header or headers that 288 can be routed within the scope of the local INET partition. 290 OMNI link 291 the same as defined in [I-D.templin-6man-omni-interface], and 292 manifested by IPv6 encapsulation [RFC2473]. The OMNI link spans 293 underlying INET segments joined by virtual bridges in a spanning 294 tree the same as a bridged campus LAN. AERO nodes on the OMNI 295 link appear as single-hop neighbors even though they may be 296 separated by multiple underlying INET hops, and can use Segment 297 Routing [RFC8402] to cause packets to visit selected waypoints on 298 the link. 300 OMNI Interface 301 a node's attachment to an OMNI link. Since the addresses assigned 302 to an OMNI interface are managed for uniqueness, OMNI interfaces 303 do not require Duplicate Address Detection (DAD) and therefore set 304 the administrative variable 'DupAddrDetectTransmits' to zero 305 [RFC4862]. 307 OMNI Adaptation Layer (OAL) 308 an OMNI interface process whereby packets admitted into the 309 interface are wrapped in a mid-layer IPv6 header and fragmented/ 310 reassembled if necessary to support the OMNI link Maximum 311 Transmission Unit (MTU) over a diversity of paths. The OAL is 312 also responsible for generating MTU-related control messages as 313 necessary, and for providing addressing context for spanning 314 multiple segments of a bridged OMNI link. 316 underlying interface 317 an ANET or INET interface over which an OMNI interface is 318 configured. 320 Mobility Service Prefix (MSP) 321 an aggregated IP Global Unicast Address (GUA) prefix (e.g., 322 2001:db8::/32, 192.0.2.0/24, etc.) assigned to the OMNI link and 323 from which more-specific Mobile Network Prefixes (MNPs) are 324 delegated. OMNI link administrators typically obtain MSPs from an 325 Internet address registry, however private-use prefixes can 326 alternatively be used subject to certain limitations (see: 327 [I-D.templin-6man-omni-interface]). OMNI links that connect to 328 the global Internet advertise their MSPs to their interdomain 329 routing peers. 331 Mobile Network Prefix (MNP) 332 a longer IP prefix delegated from an MSP (e.g., 333 2001:db8:1000:2000::/56, 192.0.2.8/30, etc.) and delegated to an 334 AERO Client or Relay. 336 Mobile Network Prefix Link Local Address (MNP-LLA) 337 an IPv6 Link Local Address that embeds the most significant 64 338 bits of an MNP in the lower 64 bits of fe80::/64, as specified in 339 [I-D.templin-6man-omni-interface]. 341 Mobile Network Prefix Unique Local Address (MNP-ULA) 342 an IPv6 Unique-Local Address derived from an MNP-LLA. 344 Administrative Link Local Address (ADM-LLA) 345 an IPv6 Link Local Address that embeds a 32-bit administratively- 346 assigned identification value in the lower 32 bits of fe80::/96, 347 as specified in [I-D.templin-6man-omni-interface]. 349 Administrative Unique Local Address (ADM-ULA) 350 an IPv6 Unique-Local Address derived from an ADM-LLA. 352 AERO node 353 a node that is connected to an OMNI link and participates in the 354 AERO internetworking and mobility service. 356 AERO Client ("Client") 357 an AERO node that connects over one or more underlying interfaces 358 and requests MNP delegation/registration service from AERO 359 Servers. The Client assigns an MNP-LLA to the OMNI interface for 360 use in ND exchanges with other AERO nodes and forwards packets to 361 correspondents according to OMNI interface neighbor cache state. 363 AERO Server ("Server") 364 an INET node that configures an OMNI interface to provide default 365 forwarding and Mobility Anchor Point (MAP) services for AERO 366 Clients. The Server assigns an ADM-LLA to its OMNI interface to 367 support the operation of the ND services, and advertises all of 368 its associated MNPs via BGP peerings with Bridges. 370 AERO Relay ("Relay") 371 an AERO Server that also provides forwarding services between 372 nodes reached via the OMNI link and correspondents on other links. 373 AERO Relays are provisioned with MNPs (i.e., the same as for an 374 AERO Client) and run a dynamic routing protocol to discover any 375 non-MNP IP GUA routes in service on its connected INET links. In 376 both cases, the Relay advertises the MSP(s) to its downstream 377 networks, and distributes all of its associated MNPs and non-MNP 378 IP GUA routes via BGP peerings with Bridges (i.e., the same as for 379 an AERO Server). 381 AERO Bridge ("Bridge") 382 a node that provides hybrid routing/bridging services (as well as 383 a security trust anchor) for nodes on an OMNI link. As a router, 384 the Bridge forwards packets using standard IP forwarding. As a 385 bridge, the Bridge forwards packets over the OMNI link without 386 decrementing the IPv6 Hop Limit. AERO Bridges peer with Servers 387 and other Bridges to discover the full set of MNPs for the link as 388 well as any non-MNP IP GUA routes that are reachable via Relays. 390 AERO Proxy ("Proxy") 391 a node that provides proxying services between Clients in an ANET 392 and Servers in external INETs. The AERO Proxy is a conduit 393 between the ANET and external INETs in the same manner as for 394 common web proxys, and behaves in a similar fashion as for ND 395 proxys [RFC4389]. A node may be configured to act as either a 396 Proxy and/or a Server, depending on Client Server selection 397 criteria. 399 ingress tunnel endpoint (ITE) 400 an OMNI interface endpoint that injects encapsulated packets into 401 an OMNI link. 403 egress tunnel endpoint (ETE) 404 an OMNI interface endpoint that receives encapsulated packets from 405 an OMNI link. 407 link-layer address 408 an IP address used as an encapsulation header source or 409 destination address from the perspective of the OMNI interface. 410 When an upper layer protocol (e.g., UDP) is used as part of the 411 encapsulation, the port number is also considered as part of the 412 link-layer address. 414 network layer address 415 the source or destination address of an encapsulated IP packet 416 presented to the OMNI interface. 418 end user network (EUN) 419 an internal virtual or external edge IP network that an AERO 420 Client or Relay connects to the rest of the network via the OMNI 421 interface. The Client/Relay sees each EUN as a "downstream" 422 network, and sees the OMNI interface as the point of attachment to 423 the "upstream" network. 425 Mobile Node (MN) 426 an AERO Client and all of its downstream-attached networks that 427 move together as a single unit, i.e., an end system that connects 428 an Internet of Things. 430 Mobile Router (MR) 431 a MN's on-board router that forwards packets between any 432 downstream-attached networks and the OMNI link. The MR is the MN 433 entity that hosts the AERO Client. 435 Route Optimization Source (ROS) 436 the AERO node nearest the source that initiates route 437 optimization. The ROS may be a Server or Proxy acting on behalf 438 of the source Client, or may be the Client itself if the Client is 439 connected to the INET either directly or through a NAT. 441 Route Optimization responder (ROR) 442 the AERO node nearest the target destination that responds to 443 route optimization requests. The ROR may be a Server acting on 444 behalf of a target MNP Client, or a Relay for a non-MNP 445 destination. 447 MAP List 448 a geographically and/or topologically referenced list of addresses 449 of all Servers within the same OMNI link. There is a single MAP 450 list for the entire OMNI link. 452 Distributed Mobility Management (DMM) 453 a BGP-based overlay routing service coordinated by Servers and 454 Bridges that tracks all Server-to-Client associations. 456 Mobility Service (MS) 457 the collective set of all Servers, Proxys, Bridges and Relays that 458 provide the AERO Service to Clients. 460 Mobility Service Endpoint MSE) 461 an individual Server, Proxy, Bridge or Relay in the Mobility 462 Service. 464 Throughout the document, the simple terms "Client", "Server", 465 "Bridge", "Proxy" and "Relay" refer to "AERO Client", "AERO Server", 466 "AERO Bridge", "AERO Proxy" and "AERO Relay", respectively. 467 Capitalization is used to distinguish these terms from other common 468 Internetworking uses in which they appear without capitalization. 470 The terminology of DHCPv6 [RFC8415] and IPv6 ND [RFC4861] (including 471 the names of node variables, messages and protocol constants) is used 472 throughout this document. The terms "All-Routers multicast", "All- 473 Nodes multicast", "Solicited-Node multicast" and "Subnet-Router 474 anycast" are defined in [RFC4291]. Also, the term "IP" is used to 475 generically refer to either Internet Protocol version, i.e., IPv4 476 [RFC0791] or IPv6 [RFC8200]. 478 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 479 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 480 "OPTIONAL" in this document are to be interpreted as described in BCP 481 14 [RFC2119][RFC8174] when, and only when, they appear in all 482 capitals, as shown here. 484 3. Asymmetric Extended Route Optimization (AERO) 486 The following sections specify the operation of IP over OMNI links 487 using the AERO service: 489 3.1. AERO Node Types 491 AERO Clients are Mobile Nodes (MNs) that configure OMNI interfaces 492 over underlying interfaces with addresses that may change when the 493 Client moves to a new network connection point. AERO Clients 494 register their Mobile Network Prefixes (MNPs) with the AERO service, 495 and distribute the MNPs to nodes on EUNs. AERO Bridges, Servers, 496 Proxys and Relays are critical infrastructure elements in fixed 497 (i.e., non-mobile) INET deployments and hence have permanent and 498 unchanging INET addresses. Together, they constitute the AERO 499 service which provides an OMNI link virtual overlay for connecting 500 AERO Clients. 502 AERO Bridges provide hybrid routing/bridging services (as well as a 503 security trust anchor) for nodes on an OMNI link. Bridges use 504 standard IPv6 routing to forward packets both within the same INET 505 partition and between disjoint INET partitions based on an IPv6 506 encapsulation mid-layer known as the OMNI Adaptation Layer (OAL) 507 [I-D.templin-6man-omni-interface]. During forwarding, the inner IP 508 layer experiences a virtual bridging service since the inner IP TTL/ 509 Hop Limit is not decremented. Each Bridge also peers with Servers 510 and other Bridges in a dynamic routing protocol instance to provide a 511 Distributed Mobility Management (DMM) service for the list of active 512 MNPs (see Section 3.2.3). Bridges present the OMNI link as a set of 513 one or more Mobility Service Prefixes (MSPs) and configure secured 514 tunnels with Servers, Relays, Proxys and other Bridges; they further 515 maintain IP forwarding table entries for each MNP and any other 516 reachable non-MNP prefixes. 518 AERO Servers in distributed INET locations provide default forwarding 519 and mobility/multilink services for AERO Client Mobile Nodes (MNs). 520 Each Server also peers with Bridges in a dynamic routing protocol 521 instance to advertise its list of associated MNPs (see 522 Section 3.2.3). Servers facilitate prefix delegation/registration 523 exchanges with Clients, where each delegated prefix becomes an MNP 524 taken from an MSP. Servers forward packets between OMNI interface 525 neighbors and track each Client's mobility profiles. Servers may 526 further act as Servers for some sets of Clients and as Proxys for 527 others. 529 AERO Proxys provide a conduit for ANET Clients to associate with 530 Servers in external INETs. Client and Servers exchange control plane 531 messages via the Proxy acting as a bridge between the ANET/INET 532 boundary. The Proxy forwards data packets between Clients and the 533 OMNI link according to forwarding information in the neighbor cache. 534 The Proxy function is specified in Section 3.13. Proxys may further 535 act as Proxys for some sets of Clients and as Servers for others. 537 AERO Relays are Servers that provide forwarding services between the 538 OMNI interface and INET/EUN interfaces. Relays are provisioned with 539 MNPs the same as for an AERO Client, and also run a dynamic routing 540 protocol to discover any non-MNP IP routes. The Relay advertises the 541 MSP(s) to its connected networks, and distributes all of its 542 associated MNPs and non-MNP IP GUA routes via BGP peerings with 543 Bridges 545 3.2. The AERO Service over OMNI Links 547 3.2.1. AERO/OMNI Reference Model 549 Figure 1 presents the basic OMNI link reference model: 551 +----------------+ 552 | AERO Bridge B1 | 553 | Nbr: S1, S2, P1| 554 |(X1->S1; X2->S2)| 555 | MSP M1 | 556 +-+---------+--+-+ 557 +--------------+ | Secured | | +--------------+ 558 |AERO Server S1| | tunnels | | |AERO Server S2| 559 | Nbr: C1, B1 +-----+ | +-----+ Nbr: C2, B1 | 560 | default->B1 | | | default->B1 | 561 | X1->C1 | | | X2->C2 | 562 +-------+------+ | +------+-------+ 563 | OMNI link | | 564 X===+===+===================+==)===============+===+===X 565 | | | | 566 +-----+--------+ +--------+--+-----+ +--------+-----+ 567 |AERO Client C1| | AERO Proxy P1 | |AERO Client C2| 568 | Nbr: S1 | |(Proxy Nbr Cache)| | Nbr: S2 | 569 | default->S1 | +--------+--------+ | default->S2 | 570 | MNP X1 | | | MNP X2 | 571 +------+-------+ .--------+------. +-----+--------+ 572 | (- Proxyed Clients -) | 573 .-. `---------------' .-. 574 ,-( _)-. ,-( _)-. 575 .-(_ IP )-. +-------+ +-------+ .-(_ IP )-. 576 (__ EUN )--|Host H1| |Host H2|--(__ EUN ) 577 `-(______)-' +-------+ +-------+ `-(______)-' 579 Figure 1: AERO/OMNI Reference Model 581 In this model: 583 o the OMNI link is an overlay network service configured over one or 584 more underlying INET partitions which may be managed by different 585 administrative authorities and have incompatible protocols and/or 586 addressing plans. 588 o AERO Bridge B1 aggregates Mobility Service Prefix (MSP) M1, 589 discovers Mobile Network Prefixes (MNPs) X* and advertises the MSP 590 via BGP peerings over secured tunnels to Servers (S1, S2). 591 Bridges connect the disjoint segments of a partitioned OMNI link. 593 o AERO Servers/Relays S1 and S2 configure secured tunnels with 594 Bridge B1 and also provide mobility, multilink and default router 595 services for their associated Clients C1 and C2. 597 o AERO Clients C1 and C2 associate with Servers S1 and S2, 598 respectively. They receive Mobile Network Prefix (MNP) 599 delegations X1 and X2, and also act as default routers for their 600 associated physical or internal virtual EUNs. Simple hosts H1 and 601 H2 attach to the EUNs served by Clients C1 and C2, respectively. 603 o AERO Proxy P1 configures a secured tunnel with Bridge B1 and 604 provides proxy services for AERO Clients in secured enclaves that 605 cannot associate directly with other OMNI link neighbors. 607 An OMNI link configured over a single INET appears as a single 608 unified link with a consistent underlying network addressing plan. 609 In that case, all nodes on the link can exchange packets via simple 610 INET encapsulation, since the underlying INET is connected. In 611 common practice, however, an OMNI link may be partitioned into 612 multiple "segments", where each segment is a distinct INET 613 potentially managed under a different administrative authority (e.g., 614 as for worldwide aviation service providers such as ARINC, SITA, 615 Inmarsat, etc.). Individual INETs may also themselves be partitioned 616 internally, in which case each internal partition is seen as a 617 separate segment. 619 The addressing plan of each segment is consistent internally but will 620 often bear no relation to the addressing plans of other segments. 621 Each segment is also likely to be separated from others by network 622 security devices (e.g., firewalls, proxys, packet filtering gateways, 623 etc.), and in many cases disjoint segments may not even have any 624 common physical link connections. Therefore, nodes can only be 625 assured of exchanging packets directly with correspondents in the 626 same segment, and not with those in other segments. The only means 627 for joining the segments therefore is through inter-domain peerings 628 between AERO Bridges. 630 The same as for traditional campus LANs, multiple OMNI link segments 631 can be joined into a single unified link via a virtual bridging 632 service using the OMNI Adaptation Layer (OAL) 633 [I-D.templin-6man-omni-interface] which inserts a mid-layer IPv6 634 encapsulation header that supports inter-segment forwarding (i.e., 635 bridging) without decrementing the network-layer TTL/Hop Limit. This 636 bridging of OMNI link segments is shown in Figure 2: 638 . . . . . . . . . . . . . . . . . . . . . . . 639 . . 640 . .-(::::::::) . 641 . .-(::::::::::::)-. +-+ . 642 . (:::: Segment A :::)--|B|---+ . 643 . `-(::::::::::::)-' +-+ | . 644 . `-(::::::)-' | . 645 . | . 646 . .-(::::::::) | . 647 . .-(::::::::::::)-. +-+ | . 648 . (:::: Segment B :::)--|B|---+ . 649 . `-(::::::::::::)-' +-+ | . 650 . `-(::::::)-' | . 651 . | . 652 . .-(::::::::) | . 653 . .-(::::::::::::)-. +-+ | . 654 . (:::: Segment C :::)--|B|---+ . 655 . `-(::::::::::::)-' +-+ | . 656 . `-(::::::)-' | . 657 . | . 658 . ..(etc).. x . 659 . . 660 . . 661 . <- OMNI link Bridged by encapsulation -> . 662 . . . . . . . . . . . . . .. . . . . . . . . 664 Figure 2: Bridging OMNI Link Segments 666 Bridges, Servers, Relays and Proxys connect via secured INET tunnels 667 over their respective segments in a spanning tree topology rooted at 668 the Bridges. The secured spanning tree supports strong 669 authentication for IPv6 ND control messages and may also be used to 670 convey the initial data packets in a flow. Route optimization can 671 then be employed to cause data packets to take more direct paths 672 between OMNI link neighbors without having to strictly follow the 673 spanning tree. 675 3.2.2. Addressing and Node Identification 677 AERO nodes on OMNI links use the Link-Local Address (LLA) prefix 678 fe80::/64 [RFC4291] to assign LLAs used for network-layer addresses 679 in link-scoped IPv6 ND and data messages. AERO Clients use LLAs 680 constructed from MNPs (i.e., "MNP-LLAs") while other AERO nodes use 681 LLAs constructed from administrative identification values ("ADM- 682 LLAs") as specified in [I-D.templin-6man-omni-interface]. 684 AERO nodes also use the Unique Local Address (ULA) prefix fd00::/8 685 followed by a pseudo-random 40-bit OMNI domain identifier to form the 686 prefix [ULA]::/48, then include a 16-bit OMNI link identifier '*' to 687 form the prefix [ULA*]::/64 [RFC4291]. The AERO node then uses the 688 prefix [ULA*]::/64 to form "MNP-ULAs" or "ADM-ULA"s as specified in 689 [I-D.templin-6man-omni-interface] to support OAL addressing. AERO 690 Clients also use Temporary ULAs constructed per 691 [I-D.templin-6man-omni-interface], where the addresses are typically 692 used only in initial control message exchanges until a stable MNP- 693 LLA/ULA is assigned. 695 AERO MSPs and MNPs are typically based on Global Unicast Addresses 696 (GUAs), but in some cases may be based on private-use addresses. See 697 [I-D.templin-6man-omni-interface] for a full specification of LLAs, 698 ULAs and GUAs used by AERO nodes on OMNI links. 700 Finally, AERO Clients and Servers configure node identification 701 values as specified in [I-D.templin-6man-omni-interface]. 703 3.2.3. AERO Routing System 705 The AERO routing system comprises a private instance of the Border 706 Gateway Protocol (BGP) [RFC4271] that is coordinated between Bridges 707 and Servers and does not interact with either the public Internet BGP 708 routing system or any underlying INET routing systems. 710 In a reference deployment, each Server is configured as an Autonomous 711 System Border Router (ASBR) for a stub Autonomous System (AS) using a 712 32-bit AS Number (ASN) [RFC4271] that is unique within the BGP 713 instance, and each Server further uses eBGP to peer with one or more 714 Bridges but does not peer with other Servers. Each INET of a multi- 715 segment OMNI link must include one or more Bridges, which peer with 716 the Servers and Proxys within that INET. All Bridges within the same 717 INET are members of the same hub AS, and use iBGP to maintain a 718 consistent view of all active MNPs currently in service. The Bridges 719 of different INETs peer with one another using eBGP. 721 Bridges advertise the OMNI link's MSPs and any non-MNP routes to each 722 of their Servers. This means that any aggregated non-MNPs (including 723 "default") are advertised to all Servers. Each Bridge configures a 724 black-hole route for each of its MSPs. By black-holing the MSPs, the 725 Bridge will maintain forwarding table entries only for the MNPs that 726 are currently active, and packets destined to all other MNPs will 727 correctly incur Destination Unreachable messages due to the black- 728 hole route. In this way, Servers have only partial topology 729 knowledge (i.e., they know only about the MNPs of their directly 730 associated Clients) and they forward all other packets to Bridges 731 which have full topology knowledge. 733 Each OMNI link segment assigns a unique ADM-ULA sub-prefix of 734 [ULA*]::/96. For example, a first segment could assign 735 [ULA*]::1000/116, a second could assign [ULA*]::2000/116, a third 736 could assign [ULA*]::3000/116, etc. Within each segment, each Proxy/ 737 Server configures an ADM-ULA within the segment's prefix, e.g., the 738 Proxy/Servers within [ULA*]::2000/116 could assign the ADM-ULAs 739 [ULA*]::2011/116, [ULA*]::2026/116, [ULA*]::2003/116, etc. 741 The administrative authorities for each segment must therefore 742 coordinate to assure mutually-exclusive ADM-ULA prefix assignments, 743 but internal provisioning of ADM-ULAs an independent local 744 consideration for each administrative authority. For each ADM-ULA 745 prefix, the Bridge(s) that connect that segment assign the all-zero's 746 address of the prefix as a Subnet Router Anycast address. For 747 example, the Subnet Router Anycast address for [ULA*]::1023/116 is 748 simply [ULA*]::1000. 750 ADM-ULA prefixes are statically represented in Bridge forwarding 751 tables. Bridges join multiple segments into a unified OMNI link over 752 multiple diverse administrative domains. They support a bridging 753 function by first establishing forwarding table entries for their 754 ADM-ULA prefixes either via standard BGP routing or static routes. 755 For example, if three Bridges ('A', 'B' and 'C') from different 756 segments serviced [ULA*]::1000/116, [ULA*]::2000/116 and 757 [ULA*]::3000/116 respectively, then the forwarding tables in each 758 Bridge are as follows: 760 A: [ULA*]::1000/116->local, [ULA*]::2000/116->B, [ULA*]::3000/116->C 762 B: [ULA*]::1000/116->A, [ULA*]::2000/116->local, [ULA*]::3000/116->C 764 C: [ULA*]::1000/116->A, [ULA*]::2000/116->B, [ULA*]::3000/116->local 766 These forwarding table entries are permanent and never change, since 767 they correspond to fixed infrastructure elements in their respective 768 segments. 770 MNP ULAs are instead dynamically advertised in the AERO routing 771 system by Servers and Relays that provide service for their 772 corresponding MNPs. For example, if three Servers ('D', 'E' and 'F') 773 service the MNPs 2001:db8:1000:2000::/56, 2001:db8:3000:4000::/56 and 774 2001:db8:5000:6000::/56 then the routing system would include: 776 D: [ULA*]:2001:db8:1000:2000/120 778 E: [ULA*]:2001:db8:3000:4000/120 780 F: [ULA*]:2001:db8:5000:6000/120 781 A full discussion of the BGP-based routing system used by AERO is 782 found in [I-D.ietf-rtgwg-atn-bgp]. 784 3.2.4. OMNI Link Segment Routing 786 With the Client and partition prefixes in place in Bridge forwarding 787 tables, the OMNI interface sends control and data packets toward AERO 788 destination nodes located in the same segment using simple INET 789 encapsulation without using the spanning tree. For destinations 790 located in different OMNI link segments, the OMNI interface instead 791 sends packets over the spanning tree via mid-layer encapsulation 792 using the IPv6 OMNI Adaptation Layer (OAL) header 793 [I-D.templin-6man-omni-interface]. When necessary, the OMNI 794 interface also includes an OMNI Routing Header (ORH) as an extension 795 to the OAL header if final segment forwarding information is 796 available, e.g., in the neighbor cache. The ORH is formatted as 797 shown in Figure 3: 799 0 1 2 3 800 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 801 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 802 | Next Header | Hdr Ext Len | Routing Type | SRT | FMT | 803 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 804 | Last Hop Segment-id (LHS) | 805 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 806 ~ Link Layer Address (L2ADDR) ~ 807 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 808 ~ Destination Suffix (if necessary) ~ 809 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 810 ~ Null Padding (if necessary) ~ 811 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 813 Figure 3: OMNI Routing Header (ORH) Format 815 In this format: 817 o Next Header identifies the type of header immediately following 818 the ORH. 820 o Hdr Ext Len is the length of the Routing header in 8-octet units 821 (not including the first 8 octets), with trailing padding added if 822 necessary to produce an integral number of 8-octet units. 824 o Routing Type is set to TBD (see IANA Considerations). 826 o Segments Left is omitted, and replaced by a 5-bit SRT and 3-bit 827 FMT field. 829 o SRT - a 5-bit Segment Routing Topology prefix length value that 830 (when added to 96) determines the prefix length to apply to the 831 ADM-ULA formed from concatenating [ULA*]::/96 with the 32 bit LHS 832 value that follows. For example, the value 16 corresponds to the 833 prefix length 112. 835 o FMT - a 3-bit "Framework/Mode/Type" code corresponding to the 836 included Link Layer Address as follows: 838 * When the most significant bit (i.e., "Framework") is set to 1, 839 L2ADDR is the INET encapsulation address for the target Client 840 itself; otherwise L2ADDR is the address of the Server/Proxy 841 named in the LHS. 843 * When the next most significant bit (i.e., "Mode") is set to 1, 844 the Framework node is (likely) located behind an INET Network 845 Address Translator (NAT); otherwise, it is on the open INET. 847 * When the least significant bit (i.e., "Type") is set to 0, 848 L2ADDR includes a UDP Port Number followed by an IPv4 address; 849 otherwise, it includes a UDP Port Number followed by an IPv6 850 address. 852 o LHS - the 32 bit ID of a node in the Last Hop Segment that 853 services the target. When SRT and LHS are both set to 0, the LHS 854 is considered unspecified. When SRT is set to 0 and LHS is non- 855 zero, the prefix length is set to 128. SRT and LHS provide 856 guidance to the OMNI interface forwarding algorithm. 857 Specifically, if SRT/LHS is located in the local OMNI link segment 858 then the OMNI interface can omit OAL/ORH encapsulation and 859 (following any necessary NAT traversal messaging) send directly to 860 the target using INET encapsulation according to FMT/L2ADDR. 861 Otherwise, it must perform INET/OAL/ORH encapsulation and forward 862 according to the OMNI link spanning tree. 864 o Link Layer Address (L2ADDR) - Formatted according to FMT, and 865 identifies the link-layer address (i.e., the encapsulation 866 address) of the target. The UDP Port Number appears in the first 867 two octets and the IP address appears in the next 4 octets for 868 IPv4 or 16 octets for IPv6. The Port Number and IP address are 869 recorded in network byte order, and in ones-compliment 870 "obfuscated" form per [RFC4380]. The OMNI interface forwarding 871 algorithm uses FMT/L2ADDR to determine the INET encapsulation 872 address for local forwarding when SRT/LHS is located in the same 873 OMNI link segment. Note that if the target is behind a NAT, 874 L2ADDR will contain the mapped INET address stored in the NAT; 875 otherwise, L2ADDR will contain the native INET information of the 876 target itself. 878 o Destination Suffix is a 64-bit field included only for OAL- 879 encapsulated packets that require OAL fragmentation. Present only 880 when Hdr Ext Len indicates that at least 8 bytes follow L2ADDR. 881 When present, encodes the 64-bit MNP-ULA suffix for the target 882 Client. For example, if the MNP-ULA is [ULA*]:2001:db8:1:2 then 883 the Destination suffix encodes the value 2001:db8:1:2. 885 o Null Padding contains zero-valued octets as necessary to pad the 886 ORH to an integral number of 8-octet units. 888 AERO nodes use OAL encapsulation when the target destination is 889 located in a different OMNI link segment, or when OAL fragmentation 890 is necessary (otherwise, OAL encapsulation is omitted and simple INET 891 encapsulation is used). When an AERO node uses OAL encapsulation for 892 a packet with source address 2001:db8:1:2::1 and destination address 893 2001:db8:1234:5678::1, it sets the OAL header source address to its 894 own ULA (e.g., [ULA*]::1000:2000) and sets the destination address to 895 the MNP-ULA corresponding to the IP destination address (e.g., 896 [ULA*]::2001:db8:1234:5678). 898 If the neighbor cache information indicates that the target is in a 899 different segment, the node next inserts an ORH immediately following 900 the OAL header while including the correct SRT, FMT, LHS, L2ADDR and 901 Destination Suffix if fragmentation is needed (in this case, the 902 Destination Suffix is 2001:db8:1234:5678). Next, the node overwrites 903 the OAL header destination address with the LHS Subnet Router Anycast 904 address (for example, for LHS 3000:4567 with SRT 16, the Subnet 905 Router Anycast address is [ULA*]::3000:0000). (Note: if ADM-ULA of 906 the last-hop Proxy/Server is known but the SRT, FMT, LHS and L2ADDR 907 are not (yet) known, the node instead overwrites the OAL header 908 destination address with the ADM-ULA.) 910 The node then fragments the OAL/ORH packet if necessary, with each 911 resulting fragment including the OAL/ORH headers while only the first 912 fragment includes the original IPv6 header. The node finally 913 encapsulates each resulting OAL/ORH packet/fragment in an INET header 914 with source address set to its own INET address (e.g., 192.0.2.100) 915 and destination set to the INET address of a Bridge (e.g., 916 192.0.2.1). 918 The encapsulation format in the above example is shown in Figure 4: 920 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 921 | INET Header | 922 | src = 192.0.2.100 | 923 | dst = 192.0.2.1 | 924 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 925 | OAL Header (if necessary) | 926 | src = [ULA*]::1000:2000 | 927 | dst= [ULA*]::3000:0000 | 928 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 929 | ORH Header (if necessary) | 930 | Destination Suffix: | 931 | 2001:db8:1234:5678 | 932 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 933 | Inner IP Header | 934 | src = 2001:db8:1:2::1 | 935 | dst = 2001:db8:1234:5678::1 | 936 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 937 | | 938 ~ ~ 939 ~ Inner Packet Body ~ 940 ~ ~ 941 | | 942 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 944 Figure 4: OAL/ORH Encapsulation 946 In this format, the inner IP header and packet body are the original 947 IP packet, the OAL header is an IPv6 header prepared according to 948 [RFC2473], the ORH is a Routing Header extension of the OAL header, 949 and the INET header is prepared as discussed in Section 3.6. 951 This gives rise to a routing system that contains both Client prefix 952 routes that may change dynamically due to regional node mobility and 953 partition prefix routes that rarely if ever change. The Bridges can 954 therefore provide link-layer bridging by sending packets over the 955 spanning tree instead of network-layer routing according to MNP 956 routes. As a result, opportunities for packet loss due to node 957 mobility between different segments are mitigated. 959 In normal operations, IPv6 ND messages are conveyed over secured 960 paths between OMNI link neighbors so that specific Proxys, Servers or 961 Relays can be addressed without being subject to mobility events. 962 Conversely, only the first few packets destined to Clients need to 963 traverse secured paths until route optimization can determine a more 964 direct path. 966 Note: When the source OAL has multiple inner IP packets to send to 967 the same neighbor, it can perform "packing" to generate a "super- 968 packet" as discussed in Section 5.2 of 969 [I-D.templin-6man-omni-interface]. In that case, the OAL/ORH super- 970 packet may include up to N inner packets as long as the total length 971 of the super-packet does not exceed the OMNI interface MTU. 973 Note: Use of an IPv6 "minimal encapsulation" format (i.e., an IPv6 974 variant of [RFC2004]) based on extensions to the ORH was considered 975 and abandoned. In the approach, the ORH would be inserted as an 976 extension header to the original IPv6 packet header. The IPv6 977 destination address would then be written into the ORH, and the ULA 978 corresponding to the destination would be overwritten in the IPv6 979 destination address. This would seemingly convey enough forwarding 980 information so that OAL encapsulation could be avoided. However, 981 this "minimal encapsulation" IPv6 packet would then have a non-ULA 982 source address and ULA destination address, an incorrect value in 983 upper layer protocol checksums, and a Hop Limit that is decremented 984 within the spanning tree when it should not be. The insertion and 985 removal of the ORH would also entail rewriting the Payload Length and 986 Next Header fields - again, invalidating upper layer checksums. 987 These irregularities would result in implementation challenges and 988 the potential for operational issues, e.g., since actionable ICMPv6 989 error reports could not be delivered to the original source. In 990 order to address the issues, still more information such as the 991 original IPv6 source address could be written into the ORH. However, 992 with the additional information the benefit of the "minimal 993 encapsulation" savings quickly diminishes, and becomes overshadowed 994 by the implementation and operational irregularities. 996 3.2.5. Segment Routing Topologies (SRTs) 998 The 64-bit sub-prefixes of [ULA]::/48 identify up to 2^16 distinct 999 Segment Routing Topologies (SRTs). Each SRT is a mutually-exclusive 1000 OMNI link overlay instance using a distinct set of ULAs, and emulates 1001 a Virtual LAN (VLAN) service for the OMNI link. In some cases (e.g., 1002 when redundant topologies are needed for fault tolerance and 1003 reliability) it may be beneficial to deploy multiple SRTs that act as 1004 independent overlay instances. A communication failure in one 1005 instance therefore will not affect communications in other instances. 1007 Each SRT is identified by a distinct value in bits 48-63 of 1008 [ULA]::/48, i.e., as [ULA0]::/64, [ULA1]::/64, [ULA2]::/64, etc. 1009 Each OMNI interface is identified by a unique interface name (e.g., 1010 omni0, omni1, omni2, etc.) and assigns an anycast ADM-ULA 1011 corresponding to its SRT prefix length. The anycast ADM-ULA is used 1012 for OMNI interface determination in Safety-Based Multilink (SBM) as 1013 discussed in [I-D.templin-6man-omni-interface]. Each OMNI interface 1014 further applies Performance-Based Multilink (PBM) internally. 1016 3.2.6. Segment Routing For OMNI Link Selection 1018 An original IPv6 source can direct an IPv6 packet to an AERO node by 1019 including a standard IPv6 Segment Routing Header (SRH) [RFC8754] with 1020 the anycast ADM-ULA for the selected SRT as either the IPv6 1021 destination or as an intermediate hop within the SRH. This allows 1022 the original source to determine the specific OMNI link topology a 1023 packet will traverse when there may be multiple alternatives. 1025 When the AERO node processes the SRH and forwards the packet to the 1026 correct OMNI interface, the OMNI interface writes the next IPv6 1027 address from the SRH into the IPv6 destination address and decrements 1028 Segments Left. If decrementing would cause Segments Left to become 1029 0, the OMNI interface deletes the SRH before forwarding. This form 1030 of Segment Routing supports Safety-Based Multilink (SBM). 1032 3.2.7. Segment Routing Within the OMNI Link 1034 AERO node OMNI interfaces can insert OAL/ORH headers for Segment 1035 Routing within the OMNI link to influence the paths of packets 1036 destined to targets in remote segments without requiring all packets 1037 to traverse strict spanning tree paths. 1039 When an AERO node's OMNI interface has a packet to send to a target 1040 discovered through route optimization located in the same OMNI link 1041 segment, it encapsulates the packet in OAL/ORH headers if necessary 1042 as discussed above. The node then uses the target's Link Layer 1043 Address (L2ADDR) information for INET encapsulation. 1045 When an AERO node's OMNI interface has a packet to send to a route 1046 optimization target located in a remote OMNI link segment, it 1047 encapsulates the packet in OAL/ORH headers as discussed above while 1048 forwarding the packet to a Bridge with destination set to the Subnet 1049 Router Anycast address for the final OMNI link segment. 1051 When a Bridge receives a packet destined to its Subnet Router Anycast 1052 address with an OAL/ORH with SRT/LHS values corresponding to the 1053 local segment, it examines the L2ADDR according to FMT and removes 1054 the ORH from the packet; if the packet is not a fragment, the Bridge 1055 also removes the OAL header. If the packet was a fragment, the 1056 Bridge instead writes the MNP-ULA corresponding to the ORH 1057 Destination Suffix into the OAL destination address. The Bridge then 1058 encapsulates the packet/fragment in an INET header according to 1059 L2ADDR and forwards the packet within the INET either to the LHS 1060 Server/Proxy or directly to the target Client itself. In this way, 1061 the Bridge participates in route optimization to reduce traffic load 1062 and suboptimal routing through strict spanning tree paths. 1064 3.3. OMNI Interface Characteristics 1066 OMNI interfaces are virtual interfaces configured over one or more 1067 underlying interfaces classified as follows: 1069 o INET interfaces connect to an INET either natively or through one 1070 or several NATs. Native INET interfaces have global IP addresses 1071 that are reachable from any INET correspondent. All Server, Relay 1072 and Bridge interfaces are native interfaces, as are INET-facing 1073 interfaces of Proxys. NATed INET interfaces connect to a private 1074 network behind one or more NATs that provide INET access. Clients 1075 that are behind a NAT are required to send periodic keepalive 1076 messages to keep NAT state alive when there are no data packets 1077 flowing. 1079 o ANET interfaces connect to an ANET that is separated from the open 1080 INET by a Proxy. Proxys can actively issue control messages over 1081 the INET on behalf of the Client to reduce ANET congestion. 1083 o VPNed interfaces use security encapsulation over the INET to a 1084 Virtual Private Network (VPN) server that also acts as a Server or 1085 Proxy. Other than the link-layer encapsulation format, VPNed 1086 interfaces behave the same as Direct interfaces. 1088 o Direct (i.e., single-hop point-to-point) interfaces connect a 1089 Client directly to a Server or Proxy without crossing any ANET/ 1090 INET paths. An example is a line-of-sight link between a remote 1091 pilot and an unmanned aircraft. The same Client considerations 1092 apply as for VPNed interfaces. 1094 OMNI interfaces use OAL/ORH encapsulation as necessary as discussed 1095 in Section 3.2.4. OMNI interfaces use link-layer encapsulation (see: 1096 Section 3.6) to exchange packets with OMNI link neighbors over INET 1097 or VPNed interfaces as well as over ANET interfaces for which the 1098 Client and Proxy may be multiple IP hops away. OMNI interfaces do 1099 not use link-layer encapsulation over Direct underlying interfaces or 1100 ANET interfaces when the Client and Proxy are known to be on the same 1101 underlying link. 1103 OMNI interfaces maintain a neighbor cache for tracking per-neighbor 1104 state the same as for any interface. OMNI interfaces use ND messages 1105 including Router Solicitation (RS), Router Advertisement (RA), 1106 Neighbor Solicitation (NS) and Neighbor Advertisement (NA) for 1107 neighbor cache management. 1109 OMNI interfaces send ND messages with an OMNI option formatted as 1110 specified in [I-D.templin-6man-omni-interface]. The OMNI option 1111 includes prefix registration information and Interface Attributes 1112 containing link information parameters for the OMNI interface's 1113 underlying interfaces. Each OMNI option may include multiple 1114 Interface Attributes sub-options, each identified by an ifIndex 1115 value. 1117 A Client's OMNI interface may be configured over multiple underlying 1118 interface connections. For example, common mobile handheld devices 1119 have both wireless local area network ("WLAN") and cellular wireless 1120 links. These links are often used "one at a time" with low-cost WLAN 1121 preferred and highly-available cellular wireless as a standby, but a 1122 simultaneous-use capability could provide benefits. In a more 1123 complex example, aircraft frequently have many wireless data link 1124 types (e.g. satellite-based, cellular, terrestrial, air-to-air 1125 directional, etc.) with diverse performance and cost properties. 1127 If a Client's multiple underlying interfaces are used "one at a time" 1128 (i.e., all other interfaces are in standby mode while one interface 1129 is active), then ND message OMNI options include only a single 1130 Interface Attributes sub-option set to constant values. In that 1131 case, the Client would appear to have a single interface but with a 1132 dynamically changing link-layer address. 1134 If the Client has multiple active underlying interfaces, then from 1135 the perspective of ND it would appear to have multiple link-layer 1136 addresses. In that case, ND message OMNI options MAY include 1137 multiple Interface Attributes sub-options - each with values that 1138 correspond to a specific interface. Every ND message need not 1139 include Interface Attributes for all underlying interfaces; for any 1140 attributes not included, the neighbor considers the status as 1141 unchanged. 1143 Bridge, Server and Proxy OMNI interfaces may be configured over one 1144 or more secured tunnel interfaces. The OMNI interface configures 1145 both an ADM-LLA and its corresponding ADM-ULA, while the underlying 1146 secured tunnel interfaces are either unnumbered or configure the same 1147 ULA. The OMNI interface encapsulates each IP packet in OAL/ORH 1148 headers and presents the packet to the underlying secured tunnel 1149 interface. Routing protocols such as BGP that run over the OMNI 1150 interface do not employ OAL/ORH encapsulation, but rather present the 1151 routing protocol messages directly to the underlying secured tunnels 1152 while using the ULA as the source address. This distinction must be 1153 honored consistently according to each node's configuration so that 1154 the IP forwarding table will associate discovered IP routes with the 1155 correct interface. 1157 3.4. OMNI Interface Initialization 1159 AERO Servers, Proxys and Clients configure OMNI interfaces as their 1160 point of attachment to the OMNI link. AERO nodes assign the MSPs for 1161 the link to their OMNI interfaces (i.e., as a "route-to-interface") 1162 to ensure that packets with destination addresses covered by an MNP 1163 not explicitly assigned to a non-OMNI interface are directed to the 1164 OMNI interface. 1166 OMNI interface initialization procedures for Servers, Proxys, Clients 1167 and Bridges are discussed in the following sections. 1169 3.4.1. AERO Server/Relay Behavior 1171 When a Server enables an OMNI interface, it assigns an ADM-{LLA,ULA} 1172 appropriate for the given OMNI link segment. The Server also 1173 configures secured tunnels with one or more neighboring Bridges and 1174 engages in a BGP routing protocol session with each Bridge. 1176 The OMNI interface provides a single interface abstraction to the IP 1177 layer, but internally comprises multiple secured tunnels as well as 1178 an NBMA nexus for sending encapsulated data packets to OMNI interface 1179 neighbors. The Server further configures a service to facilitate ND 1180 exchanges with AERO Clients and manages per-Client neighbor cache 1181 entries and IP forwarding table entries based on control message 1182 exchanges. 1184 Relays are simply Servers that run a dynamic routing protocol to 1185 redistribute routes between the OMNI interface and INET/EUN 1186 interfaces (see: Section 3.2.3). The Relay provisions MNPs to 1187 networks on the INET/EUN interfaces (i.e., the same as a Client would 1188 do) and advertises the MSP(s) for the OMNI link over the INET/EUN 1189 interfaces. The Relay further provides an attachment point of the 1190 OMNI link to a non-MNP-based global topology. 1192 3.4.2. AERO Proxy Behavior 1194 When a Proxy enables an OMNI interface, it assigns an ADM-{LLA, ULA} 1195 and configures permanent neighbor cache entries the same as for 1196 Servers. The Proxy also configures secured tunnels with one or more 1197 neighboring Bridges and maintains per-Client neighbor cache entries 1198 based on control message exchanges. Importantly Proxys are often 1199 configured to act as Servers, and vice-versa. 1201 3.4.3. AERO Client Behavior 1203 When a Client enables an OMNI interface, it assigns either an 1204 MNP-{LLA, ULA} or a Temporary ULA and sends RS messages with ND 1205 parameters over its underlying interfaces to a Server, which returns 1206 an RA message with corresponding parameters. The RS/RA messages may 1207 pass through a Proxy in the case of a Client's ANET interface, or 1208 through one or more NATs in the case of a Client's INET interface. 1209 (Note: if the Client used a Temporary ULA in its initial RS message, 1210 it will discover an MNP-{LLA, ULA} in the corresponding RA that it 1211 receives from the Server and begin using these new addresses while 1212 deprecating the Temporary addresses. If the Client is operating 1213 outside the context of AERO infrastructure, however, it may continue 1214 using Temporary ULAs for Client-to-Client communications until it 1215 encounters an infrastructure element that can provide an MNP.) 1217 3.4.4. AERO Bridge Behavior 1219 AERO Bridges configure an OMNI interface and assign the ADM-ULA 1220 Subnet Router Anycast address for each OMNI link segment they connect 1221 to. Bridges configure secured tunnels with Servers, Proxys and other 1222 Bridges, and engage in a BGP routing protocol session with neighbors 1223 on the spanning tree (see: Section 3.2.3). 1225 3.5. OMNI Interface Neighbor Cache Maintenance 1227 Each OMNI interface maintains a conceptual neighbor cache that 1228 includes an entry for each neighbor it communicates with on the OMNI 1229 link per [RFC4861]. OMNI interface neighbor cache entries are said 1230 to be one of "permanent", "symmetric", "asymmetric" or "proxy". 1232 Permanent neighbor cache entries are created through explicit 1233 administrative action; they have no timeout values and remain in 1234 place until explicitly deleted. AERO Bridges maintain permanent 1235 neighbor cache entries for their associated Proxys/Servers (and vice- 1236 versa). Each entry maintains the mapping between the neighbor's 1237 network-layer LLA and corresponding INET address. 1239 Symmetric neighbor cache entries are created and maintained through 1240 RS/RA exchanges as specified in Section 3.12, and remain in place for 1241 durations bounded by prefix lifetimes. AERO Servers maintain 1242 symmetric neighbor cache entries for each of their associated 1243 Clients, and AERO Clients maintain symmetric neighbor cache entries 1244 for each of their associated Servers. 1246 Asymmetric neighbor cache entries are created or updated based on 1247 route optimization messaging as specified in Section 3.14, and are 1248 garbage-collected when keepalive timers expire. AERO ROSs maintain 1249 asymmetric neighbor cache entries for active targets with lifetimes 1250 based on ND messaging constants. Asymmetric neighbor cache entries 1251 are unidirectional since only the ROS (and not the ROR) creates an 1252 entry. 1254 Proxy neighbor cache entries are created and maintained by AERO 1255 Proxys when they process Client/Server ND exchanges, and remain in 1256 place for durations bounded by ND and prefix lifetimes. AERO Proxys 1257 maintain proxy neighbor cache entries for each of their associated 1258 Clients. Proxy neighbor cache entries track the Client state and the 1259 address of the Client's associated Server(s). 1261 To the list of neighbor cache entry states in Section 7.3.2 of 1262 [RFC4861], Proxy and Server OMNI interfaces add an additional state 1263 DEPARTED that applies to symmetric and proxy neighbor cache entries 1264 for Clients that have recently departed. The interface sets a 1265 "DepartTime" variable for the neighbor cache entry to "DEPART_TIME" 1266 seconds. DepartTime is decremented unless a new ND message causes 1267 the state to return to REACHABLE. While a neighbor cache entry is in 1268 the DEPARTED state, packets destined to the target Client are 1269 forwarded to the Client's new location instead of being dropped. 1270 When DepartTime decrements to 0, the neighbor cache entry is deleted. 1271 It is RECOMMENDED that DEPART_TIME be set to the default constant 1272 value REACHABLE_TIME plus 10 seconds (40 seconds by default) to allow 1273 a window for packets in flight to be delivered while stale route 1274 optimization state may be present. 1276 When an ROR receives an authentic NS message used for route 1277 optimization, it searches for a symmetric neighbor cache entry for 1278 the target Client. The ROR then returns a solicited NA message 1279 without creating a neighbor cache entry for the ROS, but creates or 1280 updates a target Client "Report List" entry for the ROS and sets a 1281 "ReportTime" variable for the entry to REPORT_TIME seconds. The ROR 1282 resets ReportTime when it receives a new authentic NS message, and 1283 otherwise decrements ReportTime while no authentic NS messages have 1284 been received. It is RECOMMENDED that REPORT_TIME be set to the 1285 default constant value REACHABLE_TIME plus 10 seconds (40 seconds by 1286 default) to allow a window for route optimization to converge before 1287 ReportTime decrements below REACHABLE_TIME. 1289 When the ROS receives a solicited NA message response to its NS 1290 message used for route optimization, it creates or updates an 1291 asymmetric neighbor cache entry for the target network-layer and 1292 link-layer addresses. The ROS then (re)sets ReachableTime for the 1293 neighbor cache entry to REACHABLE_TIME seconds and uses this value to 1294 determine whether packets can be forwarded directly to the target, 1295 i.e., instead of via a default route. The ROS otherwise decrements 1296 ReachableTime while no further solicited NA messages arrive. It is 1297 RECOMMENDED that REACHABLE_TIME be set to the default constant value 1298 30 seconds as specified in [RFC4861]. 1300 AERO nodes also use the value MAX_UNICAST_SOLICIT to limit the number 1301 of NS keepalives sent when a correspondent may have gone unreachable, 1302 the value MAX_RTR_SOLICITATIONS to limit the number of RS messages 1303 sent without receiving an RA and the value MAX_NEIGHBOR_ADVERTISEMENT 1304 to limit the number of unsolicited NAs that can be sent based on a 1305 single event. It is RECOMMENDED that MAX_UNICAST_SOLICIT, 1306 MAX_RTR_SOLICITATIONS and MAX_NEIGHBOR_ADVERTISEMENT be set to 3 the 1307 same as specified in [RFC4861]. 1309 Different values for DEPART_TIME, REPORT_TIME, REACHABLE_TIME, 1310 MAX_UNICAST_SOLICIT, MAX_RTR_SOLCITATIONS and 1311 MAX_NEIGHBOR_ADVERTISEMENT MAY be administratively set; however, if 1312 different values are chosen, all nodes on the link MUST consistently 1313 configure the same values. Most importantly, DEPART_TIME and 1314 REPORT_TIME SHOULD be set to a value that is sufficiently longer than 1315 REACHABLE_TIME to avoid packet loss due to stale route optimization 1316 state. 1318 3.5.1. OMNI Neighbor Interface Attributes 1320 OMNI interface IPv6 ND messages include OMNI options 1321 [I-D.templin-6man-omni-interface] with Interface Attributes that 1322 provide Link-Layer Address and QoS Preference information for the 1323 neighbor's underlying interfaces. This information is stored in the 1324 neighbor cache and provides the basis for the forwarding algorithm 1325 specified in Section 3.10. The information is cumulative and 1326 reflects the union of the OMNI information from the most recent ND 1327 messages received from the neighbor; it is therefore not required 1328 that each ND message contain all neighbor information. 1330 The OMNI option Interface Attributes for each underlying interface 1331 includes a two-part "Link-Layer Address" consisting of a simple IP 1332 encapsulation address determined by the FMT and L2ADDR fields and an 1333 ADM-ULA determined by the SRT and LHS fields. If the neighbor is 1334 located in the local OMNI link segment (and, if any necessary NAT 1335 state has been established) forwarding via simple IP encapsulation 1336 can be used; otherwise, OAL encapsulation must be used. Underlying 1337 interfaces are further selected based on their associated preference 1338 values "high", "medium", "low" or "disabled". 1340 Note: the OMNI option is distinct from any Source/Target Link-Layer 1341 Address Options (S/TLLAOs) that may appear in an ND message according 1342 to the appropriate IPv6 over specific link layer specification (e.g., 1343 [RFC2464]). If both an OMNI option and S/TLLAO appear, the former 1344 pertains to encapsulation addresses while the latter pertains to the 1345 native L2 address format of the underlying media. 1347 3.5.2. OMNI Neighbor Advertisement Message Flags 1349 As discussed in Section 4.4 of [RFC4861] NA messages include three 1350 flag bits R, S and O. OMNI interface NA messages treat the flags as 1351 follows: 1353 o R: The R ("Router") flag is set to 1 in the NA messages sent by 1354 all AERO/OMNI node types. Simple hosts that would set R to 0 do 1355 not occur on the OMNI link itself, but may occur on the downstream 1356 links of Clients and Relays. 1358 o S: The S ("Solicited") flag is set exactly as specified in 1359 Section 4.4. of [RFC4861], i.e., it is set to 1 for Solicited NAs 1360 and set to 0 for Unsolicited NAs (both unicast and multicast). 1362 o O: The O ("Override") flag is set to 0 for solicited proxy NAs and 1363 set to 1 for all other solicited and unsolicited NAs. For further 1364 study is whether solicited NAs for anycast targets apply for OMNI 1365 links. Since MNP-LLAs must be uniquely assigned to Clients to 1366 support correct ND protocol operation, however, no role is 1367 currently seen for assigning the same MNP-LLA to multiple Clients. 1369 3.6. OMNI Interface Encapsulation and Re-encapsulation 1371 The OMNI Adaptation Layer (OAL) inserts mid-layer IPv6 headers known 1372 as the OAL/ORH headers when necessary as discussed in the following 1373 sections. After either inserting or omitting the OAL/ORH headers, 1374 the OMNI interface also inserts or omits an outer ANET/INET 1375 encapsulation header as discussed below. 1377 OMNI interfaces avoid outer encapsulation over Direct underlying 1378 interfaces and ANET underlying interfaces for which the Client and 1379 Proxy are connected to the same underlying link. Otherwise, OMNI 1380 interfaces encapsulate packets according to whether they are entering 1381 the OMNI interface from the network layer or if they are being re- 1382 admitted into the same OMNI link they arrived on. This latter form 1383 of encapsulation is known as "re-encapsulation". 1385 For packets entering the OMNI interface from the network layer, the 1386 OMNI interface copies the "TTL/Hop Limit", "Type of Service/Traffic 1387 Class" [RFC2983], "Flow Label"[RFC6438] (for IPv6) and "Congestion 1388 Experienced" [RFC3168] values in the inner packet's IP header into 1389 the corresponding fields in the OAL and outer encapsulation 1390 header(s). 1392 For packets undergoing re-encapsulation, the OMNI interface instead 1393 copies these values from the original encapsulation header into the 1394 new encapsulation header, i.e., the values are transferred between 1395 encapsulation headers and *not* copied from the encapsulated packet's 1396 network-layer header. (Note especially that by copying the TTL/Hop 1397 Limit between encapsulation headers the value will eventually 1398 decrement to 0 if there is a (temporary) routing loop.) 1400 OMNI interfaces configured over ANET underlying interfaces which 1401 employ a different IP protocol version (and/or when the Client and 1402 Proxy may be separated by multiple ANET IP hops) use IP-in-IP 1403 encapsulation so that the inner packet can traverse the ANET without 1404 decrementing the TTL/Hop-Limit. IPv6 underlying ANET interfaces use 1405 [RFC2473] encapsulation, while IPv4 interfaces use the appropriate 1406 encapsulation per one of [RFC5214][RFC2003]. 1408 OMNI interfaces configured over INET underlying interfaces 1409 encapsulate packets in INET headers according to the next hop 1410 determined in the forwarding algorithm in Section 3.10. If the next 1411 hop is reached via a secured tunnel, the OMNI interface uses an 1412 encapsulation format specific to the secured tunnel type (see: 1413 Section 6). If the next hop is reached via an unsecured INET 1414 interface, the OMNI interface instead uses UDP/IP encapsulation per 1415 [RFC6081][RFC4380]. 1417 When UDP/IP encapsulation is used, the OMNI interface next sets the 1418 UDP source port to a constant value that it will use in each 1419 successive packet it sends, and sets the UDP length field to the 1420 length of the encapsulated packet plus 8 bytes for the UDP header 1421 itself plus the length of any included extension headers or trailers. 1422 The encapsulated packet may be either IPv6 or IPv4, as distinguished 1423 by the version number found in the first four bits following the UDP 1424 header. 1426 For UDP/IP-encapsulated packets sent to a Server, Relay or Bridge, 1427 the OMNI interface sets the UDP destination port to 8060, i.e., the 1428 IANA-registered port number for AERO. For packets sent to a Client, 1429 the OMNI interface sets the UDP destination port to the port value 1430 stored in the neighbor cache entry for this Client. The OMNI 1431 interface finally includes/omits the UDP checksum according to 1432 [RFC6935][RFC6936]. 1434 When a Proxy, Relay or Server re-encapsulates a packet received from 1435 a Client that includes an OAL but no ORH, it inserts an ORH 1436 immediately following the OAL header and adjusts the OAL payload 1437 length and destination address field. The inserted ORH will be 1438 removed by the final-hop Bridge, but its insertion and removal will 1439 not interfere with reassembly at the final destination. For this 1440 reason, Clients must reserve 40 bytes for a maximum-length ORH when 1441 they perform OAL encapsulation (see: Section 3.9). 1443 3.7. OMNI Interface Decapsulation 1445 OMNI interfaces decapsulate packets destined either to the AERO node 1446 itself or to a destination reached via an interface other than the 1447 OMNI interface the packet was received on. When the encapsulated 1448 packet arrives in multiple OAL fragments, the OMNI interface 1449 reassembles as discussed in Section 3.9. Further decapsulation steps 1450 are performed according to the appropriate encapsulation format 1451 specification. 1453 3.8. OMNI Interface Data Origin Authentication 1455 AERO nodes employ simple data origin authentication procedures. In 1456 particular: 1458 o AERO Bridges, Servers and Proxys accept encapsulated data packets 1459 and control messages received from the (secured) spanning tree. 1461 o AERO Proxys and Clients accept packets that originate from within 1462 the same secured ANET. 1464 o AERO Clients and Relays accept packets from downstream network 1465 correspondents based on ingress filtering. 1467 o AERO Clients, Relays and Servers verify the outer UDP/IP 1468 encapsulation addresses according to [RFC4380]. 1470 AERO nodes silently drop any packets that do not satisfy the above 1471 data origin authentication procedures. Further security 1472 considerations are discussed in Section 6. 1474 3.9. OMNI Adaptation Layer and OMNI Interface MTU 1476 The OMNI interface observes the link nature of tunnels, including the 1477 Maximum Transmission Unit (MTU), Maximum Reassembly Unit (MRU) and 1478 the role of fragmentation and reassembly [I-D.ietf-intarea-tunnels]. 1479 The OMNI interface employs an OMNI Adaptation Layer (OAL) for 1480 accommodating multiple underlying links with diverse MTUs. The 1481 functions of the OAL and the OMNI interface MTU/MRU are specified in 1482 Section 5 of [I-D.templin-6man-omni-interface], with MTU/MRU both set 1483 to the constant value 9180 bytes. 1485 When the network layer presents an IP packet to the OMNI interface, 1486 the OAL encapsulates the packet in OAL and ORH headers if necessary. 1487 When the network layer presents the OMNI interface with multiple IP 1488 packets bound to the same neighbor, the OAL can concatenate multiple 1489 IP packets together into a single OAL super-packet as discussed in 1490 Section 5.2 of [I-D.templin-6man-omni-interface]. The OAL then 1491 fragments the encapsulated packet if necessary such that the OAL/ORH 1492 headers appear in each fragment while the original IP packet header 1493 appears only in the first fragment. The OAL then transmits each OAL/ 1494 ORH packet/fragment over an underlying interface connected to either 1495 a physical link such as Ethernet, WiFi and the like or a virtual link 1496 such as an Internet or higher-layer tunnel (see the definition of 1497 link in [RFC8200]). 1499 Note: A Client that does not (yet) have neighbor cache state for a 1500 target may omit the ORH in packets with the understanding that a 1501 Proxy/Server may insert an ORH on its behalf. For this reason, 1502 Clients reserve 40 bytes for the largest possible ORH in their OAL 1503 fragment size calculations. 1505 Note: Although the ORH may be removed by a Bridge on the path (see: 1506 Section 3.10.4), this does not interfere with the destination's 1507 ability to reassemble in the event that the packet was fragmented. 1508 This is due to the fact that the ORH is not included in the 1509 fragmentable part; therefore, its removal does not invalidate the 1510 offset values in any fragment headers. 1512 3.10. OMNI Interface Forwarding Algorithm 1514 IP packets enter a node's OMNI interface either from the network 1515 layer (i.e., from a local application or the IP forwarding system) or 1516 from the link layer (i.e., from an OMNI interface neighbor). All 1517 packets entering a node's OMNI interface first undergo data origin 1518 authentication as discussed in Section 3.8. Packets that satisfy 1519 data origin authentication are processed further, while all others 1520 are dropped silently. The OMNI interface OAL wraps accepted packets 1521 in OAL/ORH headers if necessary as discussed above. 1523 Packets that enter the OMNI interface from the network layer are 1524 forwarded to an OMNI interface neighbor. Packets that enter the OMNI 1525 interface from the link layer are either re-admitted into the OMNI 1526 link or forwarded to the network layer where they are subject to 1527 either local delivery or IP forwarding. In all cases, the OMNI 1528 interface itself MUST NOT decrement the network layer TTL/Hop-count 1529 since its forwarding actions occur below the network layer. 1531 OMNI interfaces may have multiple underlying interfaces and/or 1532 neighbor cache entries for neighbors with multiple underlying 1533 interfaces (see Section 3.3). The OMNI interface uses interface 1534 attributes and/or traffic classifiers (e.g., DSCP value, port number, 1535 flow specification, etc.) to select an outgoing underlying interface 1536 for each packet based on the node's own QoS preferences, and also to 1537 select a destination link-layer address based on the neighbor's 1538 underlying interface with the highest preference. AERO 1539 implementations SHOULD allow for QoS preference values to be modified 1540 at runtime through network management. 1542 If multiple outgoing interfaces and/or neighbor interfaces have a 1543 preference of "high", the AERO node replicates the packet and sends 1544 one copy via each of the (outgoing / neighbor) interface pairs; 1545 otherwise, the node sends a single copy of the packet via an 1546 interface with the highest preference. AERO nodes keep track of 1547 which underlying interfaces are currently "reachable" or 1548 "unreachable", and only use "reachable" interfaces for forwarding 1549 purposes. 1551 The following sections discuss the OMNI interface forwarding 1552 algorithms for Clients, Proxys, Servers and Bridges. In the 1553 following discussion, a packet's destination address is said to 1554 "match" if it is the same as a cached address, or if it is covered by 1555 a cached prefix (which may be encoded in an MNP-LLA). 1557 3.10.1. Client Forwarding Algorithm 1559 When an IP packet enters a Client's OMNI interface from the network 1560 layer the Client searches for an asymmetric neighbor cache entry that 1561 matches the destination. If there is a match, the Client uses one or 1562 more "reachable" neighbor interfaces in the entry for packet 1563 forwarding. If there is no asymmetric neighbor cache entry, the 1564 Client instead forwards the packet toward a Server (the packet is 1565 intercepted by a Proxy if there is a Proxy on the path). The Client 1566 encapsulates the packet in OAL/ORH headers if necessary and fragments 1567 according to MTU requirements (see: Section 3.9). The Client sets 1568 the OAL destination address to the MNP-ULA of the neighbor if it has 1569 an asymmetric neighbor cache entry; otherwise, it sets the OAL 1570 destination to the ADM-ULA of the Proxy/Server. If the Client has 1571 multiple IP packets to send to the same neighbor, it can concatenate 1572 them in a single super-packet as discussed in Section 5.2 of 1573 [I-D.templin-6man-omni-interface]. 1575 If there is an asymmetric neighbor cache entry and the neighbor 1576 interface selected for forwarding is located on the same OMNI link 1577 segment and not behind a NAT, the Client forwards the packet/ 1578 fragments directly according to the L2ADDR information for the 1579 neighbor. If the neighbor interface is behind a NAT on the same OMNI 1580 link segment, the Client instead forwards the initial packet/ 1581 fragments to its Server and initiates NAT traversal procedures. If 1582 the Client's intended source underlying interface is also behind a 1583 NAT and located on the same OMNI link segment, it sends a "direct 1584 bubble" over the interface per [RFC6081][RFC4380] to the L2ADDR found 1585 in the neighbor cache in order to establish state in its own NAT by 1586 generating traffic toward the neighbor (note that no response to the 1587 bubble is expected). 1589 The Client next sends an NS(NUD) message toward the MNP-ULA of the 1590 neighbor via its Server as discussed in Section 3.15. If the Client 1591 receives an NA(NUD) from the neighbor over the underlying interface, 1592 it marks the neighbor interface as "trusted" and sends future 1593 packets/fragments directly to the L2ADDR information for the neighbor 1594 instead of indirectly via the Server. The Client must honor the 1595 neighbor cache maintenance procedure by sending additional direct 1596 bubbles and/or NS/NA(NUD) messages as discussed in [RFC6081][RFC4380] 1597 in order to keep NAT state alive as long as packets are still 1598 flowing. 1600 When an IP packet enters a Client's OMNI interface from the link- 1601 layer, if the destination matches one of the Client's MNPs or link- 1602 local addresses the Client reassembles and decapsulates as necessary 1603 and delivers the inner packet to the network layer. Otherwise, the 1604 Client drops the packet and MAY return a network-layer ICMP 1605 Destination Unreachable message subject to rate limiting (see: 1606 Section 3.11). 1608 3.10.2. Proxy Forwarding Algorithm 1610 For control messages originating from or destined to a Client, the 1611 Proxy intercepts the message and updates its proxy neighbor cache 1612 entry for the Client. The Proxy then forwards a (proxyed) copy of 1613 the control message. (For example, the Proxy forwards a proxyed 1614 version of a Client's NS/RS message to the target neighbor, and 1615 forwards a proxyed version of the neighbor's NA/RA reply to the 1616 Client.) 1618 When the Proxy receives a data packet from a Client within the ANET, 1619 the Proxy reassembles and re-fragments if necessary then searches for 1620 an asymmetric neighbor cache entry that matches the destination and 1621 forwards as follows: 1623 o if the destination matches an asymmetric neighbor cache entry, the 1624 Proxy uses one or more "reachable" neighbor interfaces in the 1625 entry for packet forwarding using OAL/ORH encapsulation if 1626 necessary according to the cached link-layer address information. 1627 If the neighbor interface is in a different OMNI link segment, the 1628 Proxy forwards the packet to a Bridge; otherwise, it forwards the 1629 packet directly to the neighbor. If the neighbor is behind a NAT, 1630 the Proxy instead forwards initial packets via a Bridge while 1631 sending an NS(NUD) to the neighbor. When the Proxy receives the 1632 NA(NUD), it can begin forwarding packets directly to the neighbor 1633 the same as discussed in Section 3.10.1 while sending additional 1634 NUD messages as necessary to maintain NAT state. Note that no 1635 direct bubbles are necessary since the Proxy is by definition not 1636 located behind a NAT. 1638 o else, the Proxy uses OAL/ORH encapsulation and forwards the packet 1639 to a Bridge while using the MNP-ULA corresponding to the packet's 1640 destination as the destination address. 1642 When the Proxy receives an encapsulated data packet from an INET 1643 neighbor or from a secured tunnel from a Bridge, it accepts the 1644 packet only if data origin authentication succeeds and if there is a 1645 proxy neighbor cache entry that matches the inner destination. Next, 1646 the Proxy reassembles the packet (if necessary) and continues 1647 processing. If the reassembly is complete and the neighbor cache 1648 state is REACHABLE, the Proxy then returns a PTB if necessary (see: 1649 Section 3.9) then either drops or forwards the packet to the Client 1650 while performing OAL/ORH re-encapsulation and re-fragmentation if 1651 necessary. If the neighbor cache entry state is DEPARTED, the Proxy 1652 instead changes the destination address to the address of the new 1653 Server and forwards it to a Bridge while performing OAL/ORH re- 1654 encapsulation if necessary. 1656 Note: When the Proxy receives fragmented OAL packets with destination 1657 address set to the MNP-ULA of the Client, it can either reassemble 1658 first and then re-encapsulate/re-fragment before forwarding to the 1659 Client or forward the raw fragments on to the Client which then must 1660 reassemble. In the former case, the Proxy can re-fragment to a size 1661 that better matches the link MTU for the Client, which may be 1662 important for low-end links with large MTUs. In the latter case, the 1663 Client may receive fragments that are smaller than its link MTU but 1664 can still be reassembled; this case may provide an important 1665 performance benefit to Proxys by permitting them to avoid excessive 1666 reassembly and re-fragmentation overhead.) 1668 Note: If the Proxy has multiple IP packets to send to the same 1669 neighbor, it can concatenate them in a single super-packet as 1670 discussed in Section 5.2 of [I-D.templin-6man-omni-interface]. 1672 3.10.3. Server/Relay Forwarding Algorithm 1674 For control messages destined to a target Client's MNP-LLA that are 1675 received from a secured tunnel, the Server intercepts the message and 1676 returns a Proxyed response on behalf of the Client. (For example, 1677 the Server sends a Proxyed NA message reply in response to an NS 1678 message directed to one of its associated Clients.) If the Client's 1679 neighbor cache entry is in the DEPARTED state, however, the Server 1680 instead forwards the packet to the Client's new Server using OAL/ORH 1681 re-encapsulation the same as discussed for Proxys in Section 3.10.2. 1683 When the Server receives an encapsulated data packet from an INET 1684 neighbor or from a secured tunnel, it accepts the packet only if data 1685 origin authentication succeeds. If the packet was an OAL fragment 1686 with destination set to the Server's ADM-ULA, the Server submits the 1687 packet for reassembly (if the destination was the MNP-ULA of one of 1688 its Clients, the Server instead either submits the packet for 1689 reassembly or forwards the fragment directly to the target Client). 1690 If reassembly is complete (or if no reassembly was required), the 1691 Server then (re)fragments if necessary and forwards as follows: 1693 o if the network layer destination matches a symmetric neighbor 1694 cache entry in the REACHABLE state the Server prepares the packet 1695 for forwarding to the destination Client. The Server then 1696 forwards the packet as specified in Section 3.9. 1698 o else, if the destination matches a symmetric neighbor cache entry 1699 in the DEPARETED state the Server re-encapsulates the packet and 1700 forwards it using the ADM-ULA of the Client's new Server as the 1701 destination. 1703 o else, if the destination matches an asymmetric neighbor cache 1704 entry, the Server proceeds the same as specified for a Proxy in 1705 Section 3.10.2. Again, the Server need not send direct bubbles 1706 since it is not located behind a NAT. 1708 o else, if the destination matches a non-MNP route in the IP 1709 forwarding table or an ADM-LLA assigned to the Server's OMNI 1710 interface, the Server decapsulates the packet and releases it to 1711 the network layer for local delivery or IP forwarding. 1713 o else, the Server drops the packet. 1715 When the Server's OMNI interface receives a data packet from the 1716 network layer or from a VPNed or Direct Client, it performs OAL/ORH 1717 encapsulation and fragmentation if necessary, then processes the 1718 packet according to the network-layer destination address as follows: 1720 o if the destination matches a symmetric or asymmetric neighbor 1721 cache entry the Server processes the packet as above. 1723 o else, the Server encapsulates the packet in OLA/ORH headers and 1724 forwards it to a Bridge using its own ADM-ULA as the source and 1725 the MNP-ULA corresponding to the destination as the destination. 1727 Note: The same as for Proxys, when the Server receives fragmented OAL 1728 packets with destination address set to the MNP-ULA of the target, it 1729 can either reassemble first and then re-encapsulate/re-fragment 1730 before sending or forward the raw fragments on to the target which 1731 then must reassemble. In the former case, the Server can re-fragment 1732 to a size that better matches the link MTU for the Client, which may 1733 be important for low-end links with large MTUs. In the latter case, 1734 the target may receive fragments that are smaller than its link MTU 1735 but can still be reassembled; this case may provide an important 1736 performance benefit to Servers by permitting them to avoid excessive 1737 reassembly and re-fragmentation overhead.) 1739 Note: If the Server has multiple IP packets to send to the same 1740 neighbor, it can concatenate them in a single super-packet as 1741 discussed in Section 5.2 of [I-D.templin-6man-omni-interface]. 1743 3.10.4. Bridge Forwarding Algorithm 1745 Bridges forward OAL/ORH-encapsulated packets over secured tunnels the 1746 same as any IP router. When the Bridge receives an OAL/ORH- 1747 encapsulated packet via a secured tunnel, it removes the outer INET 1748 header and searches for a forwarding table entry that matches the OAL 1749 destination address. The Bridge then processes the packet as 1750 follows: 1752 o if the destination matches its ADM-ULA Subnet Router Anycast 1753 address the Bridge processes the packet locally before forwarding. 1754 The Bridge drops the packet if the next header is not an ORH; 1755 otherwise, for NA(NUD) messages the Bridge replaces the OMNI 1756 option Interface Attributes sub-option with information for its 1757 own interface while retaining the ifIndex value supplied by the 1758 NA(NUD) message source. For all packet types, the Bridge next 1759 examines the ORH FMT code. If the code indicates the destination 1760 is a Client on the open INET (or, a Client behind a NAT for which 1761 NAT traversal procedures have already converged) the Bridge 1762 removes the ORH then writes the MNP-ULA formed from the ORH 1763 Destination Suffix into the OAL destination if the packet is a 1764 fragment (otherwise, the Bridge also removes the OAL header). The 1765 Bridge then forwards the packet/fragment via INET encapsulation to 1766 the ORH L2ADDR. For all other destination cases, the Bridge 1767 instead writes the ADM-ULA formed from the ORH SRT/LHS into the 1768 OAL destination address and forwards the OAL/ORH encapsulated 1769 packet to the ADM-ULA Server while invoking NAT traversal 1770 procedures the same as for Proxys and Servers if necessary, noting 1771 that no direct bubbles are necessary since only the Client and not 1772 the Bridge is behind a NAT. 1774 o else, if the destination matches one of the Bridge's own 1775 addresses, the Bridge submits the packet for local delivery. 1777 o else, if the destination matches a forwarding table entry the 1778 Bridge forwards the packet via a secured tunnel to the next hop. 1779 If the destination matches an MSP without matching an MNP, 1780 however, the Bridge instead drops the packet and returns an ICMP 1781 Destination Unreachable message subject to rate limiting (see: 1782 Section 3.11). 1784 o else, the Bridge drops the packet and returns an ICMP Destination 1785 Unreachable as above. 1787 As for any IP router, the Bridge decrements the TTL/Hop Limit when it 1788 forwards the packet. Therefore, when an OAL header is present only 1789 the Hop Limit in the OAL header is decremented and not the TTL/Hop 1790 Limit in the inner packet header. Bridges do not insert OAL/ORH 1791 headers themselves; instead, they act as IPv6 routers and forward 1792 packets based on the destination address found in the headers of 1793 packets they receive. 1795 3.11. OMNI Interface Error Handling 1797 When an AERO node admits a packet into the OMNI interface, it may 1798 receive link-layer or network-layer error indications. 1800 A link-layer error indication is an ICMP error message generated by a 1801 router in the INET on the path to the neighbor or by the neighbor 1802 itself. The message includes an IP header with the address of the 1803 node that generated the error as the source address and with the 1804 link-layer address of the AERO node as the destination address. 1806 The IP header is followed by an ICMP header that includes an error 1807 Type, Code and Checksum. Valid type values include "Destination 1808 Unreachable", "Time Exceeded" and "Parameter Problem" 1809 [RFC0792][RFC4443]. (OMNI interfaces ignore all link-layer IPv4 1810 "Fragmentation Needed" and IPv6 "Packet Too Big" messages since they 1811 only emit packets that are guaranteed to be no larger than the IP 1812 minimum link MTU as discussed in Section 3.9.) 1814 The ICMP header is followed by the leading portion of the packet that 1815 generated the error, also known as the "packet-in-error". For 1816 ICMPv6, [RFC4443] specifies that the packet-in-error includes: "As 1817 much of invoking packet as possible without the ICMPv6 packet 1818 exceeding the minimum IPv6 MTU" (i.e., no more than 1280 bytes). For 1819 ICMPv4, [RFC0792] specifies that the packet-in-error includes: 1820 "Internet Header + 64 bits of Original Data Datagram", however 1821 [RFC1812] Section 4.3.2.3 updates this specification by stating: "the 1822 ICMP datagram SHOULD contain as much of the original datagram as 1823 possible without the length of the ICMP datagram exceeding 576 1824 bytes". 1826 The link-layer error message format is shown in Figure 5 (where, "L2" 1827 and "L3" refer to link-layer and network-layer, respectively): 1829 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1830 ~ ~ 1831 | L2 IP Header of | 1832 | error message | 1833 ~ ~ 1834 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1835 | L2 ICMP Header | 1836 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1837 ~ ~ P 1838 | IP and other encapsulation | a 1839 | headers of original L3 packet | c 1840 ~ ~ k 1841 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ e 1842 ~ ~ t 1843 | IP header of | 1844 | original L3 packet | i 1845 ~ ~ n 1846 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1847 ~ ~ e 1848 | Upper layer headers and | r 1849 | leading portion of body | r 1850 | of the original L3 packet | o 1851 ~ ~ r 1852 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1854 Figure 5: OMNI Interface Link-Layer Error Message Format 1856 The AERO node rules for processing these link-layer error messages 1857 are as follows: 1859 o When an AERO node receives a link-layer Parameter Problem message, 1860 it processes the message the same as described as for ordinary 1861 ICMP errors in the normative references [RFC0792][RFC4443]. 1863 o When an AERO node receives persistent link-layer Time Exceeded 1864 messages, the IP ID field may be wrapping before earlier fragments 1865 awaiting reassembly have been processed. In that case, the node 1866 should begin including integrity checks and/or institute rate 1867 limits for subsequent packets. 1869 o When an AERO node receives persistent link-layer Destination 1870 Unreachable messages in response to encapsulated packets that it 1871 sends to one of its asymmetric neighbor correspondents, the node 1872 should process the message as an indication that a path may be 1873 failing, and optionally initiate NUD over that path. If it 1874 receives Destination Unreachable messages over multiple paths, the 1875 node should allow future packets destined to the correspondent to 1876 flow through a default route and re-initiate route optimization. 1878 o When an AERO Client receives persistent link-layer Destination 1879 Unreachable messages in response to encapsulated packets that it 1880 sends to one of its symmetric neighbor Servers, the Client should 1881 mark the path as unusable and use another path. If it receives 1882 Destination Unreachable messages on many or all paths, the Client 1883 should associate with a new Server and release its association 1884 with the old Server as specified in Section 3.16.5. 1886 o When an AERO Server receives persistent link-layer Destination 1887 Unreachable messages in response to encapsulated packets that it 1888 sends to one of its symmetric neighbor Clients, the Server should 1889 mark the underlying path as unusable and use another underlying 1890 path. 1892 o When an AERO Server or Proxy receives link-layer Destination 1893 Unreachable messages in response to an encapsulated packet that it 1894 sends to one of its permanent neighbors, it treats the messages as 1895 an indication that the path to the neighbor may be failing. 1896 However, the dynamic routing protocol should soon reconverge and 1897 correct the temporary outage. 1899 When an AERO Bridge receives a packet for which the network-layer 1900 destination address is covered by an MSP, the Bridge drops the packet 1901 if there is no more-specific routing information for the destination 1902 and returns a network-layer Destination Unreachable message subject 1903 to rate limiting. The Bridge writes the network-layer source address 1904 of the original packet as the destination address and uses one of its 1905 non link-local addresses as the source address of the message. 1907 When an AERO node receives an encapsulated packet for which the 1908 reassembly buffer it too small, it drops the packet and returns a 1909 network-layer Packet Too Big (PTB) message. The node first writes 1910 the MRU value into the PTB message MTU field, writes the network- 1911 layer source address of the original packet as the destination 1912 address and writes one of its non link-local addresses as the source 1913 address. 1915 3.12. AERO Router Discovery, Prefix Delegation and Autoconfiguration 1917 AERO Router Discovery, Prefix Delegation and Autoconfiguration are 1918 coordinated as discussed in the following Sections. 1920 3.12.1. AERO Service Model 1922 Each AERO Server on the OMNI link is configured to facilitate Client 1923 prefix delegation/registration requests. Each Server is provisioned 1924 with a database of MNP-to-Client ID mappings for all Clients enrolled 1925 in the AERO service, as well as any information necessary to 1926 authenticate each Client. The Client database is maintained by a 1927 central administrative authority for the OMNI link and securely 1928 distributed to all Servers, e.g., via the Lightweight Directory 1929 Access Protocol (LDAP) [RFC4511], via static configuration, etc. 1930 Clients receive the same service regardless of the Servers they 1931 select. 1933 AERO Clients and Servers use ND messages to maintain neighbor cache 1934 entries. AERO Servers configure their OMNI interfaces as advertising 1935 NBMA interfaces, and therefore send unicast RA messages with a short 1936 Router Lifetime value (e.g., ReachableTime seconds) in response to a 1937 Client's RS message. Thereafter, Clients send additional RS messages 1938 to keep Server state alive. 1940 AERO Clients and Servers include prefix delegation and/or 1941 registration parameters in RS/RA messages (see 1942 [I-D.templin-6man-omni-interface]). The ND messages are exchanged 1943 between Client and Server according to the prefix management schedule 1944 required by the service. If the Client knows its MNP in advance, it 1945 can employ prefix registration by including its MNP-LLA as the source 1946 address of an RS message and with an OMNI option with valid prefix 1947 registration information for the MNP. If the Server (and Proxy) 1948 accept the Client's MNP assertion, they inject the prefix into the 1949 routing system and establish the necessary neighbor cache state. If 1950 the Client does not have a pre-assigned MNP, it can instead employ 1951 prefix delegation by including the unspecified address (::) as the 1952 source address of an RS message and with an OMNI option with a DHCPv6 1953 message option to request an MNP. 1955 The following sections specify the Client and Server behavior. 1957 3.12.2. AERO Client Behavior 1959 AERO Clients discover the addresses of Servers in a similar manner as 1960 described in [RFC5214]. Discovery methods include static 1961 configuration (e.g., from a flat-file map of Server addresses and 1962 locations), or through an automated means such as Domain Name System 1963 (DNS) name resolution [RFC1035]. Alternatively, the Client can 1964 discover Server addresses through a layer 2 data link login exchange, 1965 or through a unicast RA response to a multicast/anycast RS as 1966 described below. In the absence of other information, the Client can 1967 resolve the DNS Fully-Qualified Domain Name (FQDN) 1968 "linkupnetworks.[domainname]" where "linkupnetworks" is a constant 1969 text string and "[domainname]" is a DNS suffix for the OMNI link 1970 (e.g., "example.com"). 1972 To associate with a Server, the Client acts as a requesting router to 1973 request MNPs. The Client prepares an RS message with prefix 1974 management parameters and includes a Nonce and Timestamp option if 1975 the Client needs to correlate RA replies. If the Client already 1976 knows the Server's ADM-LLA, it includes the LLA as the network-layer 1977 destination address; otherwise, the Client includes the (link-local) 1978 All-Routers multicast as the network-layer destination. If the 1979 Client already knows its own MNP-LLA, it can use the MNP-LLA as the 1980 network-layer source address and include an OMNI option with prefix 1981 registration information. Otherwise, the Client uses the unspecified 1982 address (::) as the network-layer source address and includes a 1983 DHCPv6 sub-option in the OMNI option (see: 1984 [I-D.templin-6man-omni-interface]). The Client includes Interface 1985 Attributes corresponding to the underlying interface over which it 1986 will send the RS message, and MAY include additional Interface 1987 Attributes specific to other underlying interfaces. 1989 The Client then sends the RS message (either directly via Direct 1990 interfaces, via a VPN for VPNed interfaces, via a Proxy for ANET 1991 interfaces or via INET encapsulation for INET interfaces) while using 1992 OAL encapsulation/fragmentation if necessary, then waits for an RA 1993 message reply (see Section 3.12.3). The Client retries up to 1994 MAX_RTR_SOLICITATIONS times until an RA is received. If the Client 1995 receives no RAs, or if it receives an RA with Router Lifetime set to 1996 0, the Client SHOULD abandon this Server and try another Server. 1997 Otherwise, the Client processes the prefix information found in the 1998 RA message. 2000 When the Client processes an RA, it first performs OAL reassembly if 2001 necessary then creates a symmetric neighbor cache entry with the 2002 Server's ADM-LLA as the network-layer address and the Server's 2003 encapsulation and/or link-layer addresses as the link-layer address. 2004 The Client next records the RA Router Lifetime field value in the 2005 neighbor cache entry as the time for which the Server has committed 2006 to maintaining the MNP in the routing system via this underlying 2007 interface, and caches the other RA configuration information 2008 including Cur Hop Limit, M and O flags, Reachable Time and Retrans 2009 Timer. The Client then autoconfigures MNP-LLAs for any delegated 2010 MNPs and assigns them to the OMNI interface. The Client also caches 2011 any MSPs included in Route Information Options (RIOs) [RFC4191] as 2012 MSPs to associate with the OMNI link, and assigns the MTU value in 2013 the MTU option to the underlying interface. 2015 The Client then registers additional underlying interfaces with the 2016 Server by sending RS messages via each additional interface as 2017 described above. The RS messages include the same parameters as for 2018 the initial RS/RA exchange, but with destination address set to the 2019 Server's ADM-LLA. The Client finally sub-delegates the MNPs to its 2020 attached EUNs and/or the Client's own internal virtual interfaces as 2021 described in [I-D.templin-v6ops-pdhost] to support the Client's 2022 downstream attached "Internet of Things (IoT)". The Client then 2023 sends additional RS messages over each underlying interface before 2024 the Router Lifetime received for that interface expires. 2026 After the Client registers its underlying interfaces, it may wish to 2027 change one or more registrations, e.g., if an interface changes 2028 address or becomes unavailable, if QoS preferences change, etc. To 2029 do so, the Client prepares an RS message to send over any available 2030 underlying interface as above. The RS includes an OMNI option with 2031 prefix registration/delegation information, with Interface Attributes 2032 specific to the selected underlying interface, and with any 2033 additional Interface Attributes specific to other underlying 2034 interfaces. When the Client receives the Server's RA response, it 2035 has assurance that the Server has been updated with the new 2036 information. 2038 If the Client wishes to discontinue use of a Server it issues an RS 2039 message over any underlying interface with an OMNI option with a 2040 prefix release indication. When the Server processes the message, it 2041 releases the MNP, sets the symmetric neighbor cache entry state for 2042 the Client to DEPARTED and returns an RA reply with Router Lifetime 2043 set to 0. After a short delay (e.g., 2 seconds), the Server 2044 withdraws the MNP from the routing system. 2046 3.12.3. AERO Server Behavior 2048 AERO Servers act as IP routers and support a prefix delegation/ 2049 registration service for Clients. Servers arrange to add their ADM- 2050 LLAs to a static map of Server addresses for the link and/or the DNS 2051 resource records for the FQDN "linkupnetworks.[domainname]" before 2052 entering service. Server addresses should be geographically and/or 2053 topologically referenced, and made available for discovery by Clients 2054 on the OMNI link. 2056 When a Server receives a prospective Client's RS message on its OMNI 2057 interface, it SHOULD return an immediate RA reply with Router 2058 Lifetime set to 0 if it is currently too busy or otherwise unable to 2059 service the Client. Otherwise, the Server performs OAL reassembly on 2060 the RS message if necessary, then authenticates the message and 2061 processes the prefix delegation/registration parameters. The Server 2062 first determines the correct MNPs to provide to the Client by 2063 processing the MNP-LLA prefix parameters and/or the DHCPv6 OMNI sub- 2064 option. When the Server returns the MNPs, it also creates a 2065 forwarding table entry for the MNP-ULA corresponding to each MNP so 2066 that the MNPs are propagated into the routing system (see: 2067 Section 3.2.3). For IPv6, the Server creates an IPv6 forwarding 2068 table entry for each MNP. For IPv4, the Server creates an IPv6 2069 forwarding table entry with the IPv4-compatibility MNP-ULA prefix 2070 corresponding to the IPv4 address. 2072 The Server next creates a symmetric neighbor cache entry for the 2073 Client using the base MNP-LLA as the network-layer address and with 2074 lifetime set to no more than the smallest prefix lifetime. Next, the 2075 Server updates the neighbor cache entry by recording the information 2076 in each Interface Attributes sub-option in the RS OMNI option. The 2077 Server also records the actual OAL/INET addresses in the neighbor 2078 cache entry. 2080 Next, the Server prepares an RA message using its ADM-LLA as the 2081 network-layer source address and the network-layer source address of 2082 the RS message as the network-layer destination address. The Server 2083 sets the Router Lifetime to the time for which it will maintain both 2084 this underlying interface individually and the symmetric neighbor 2085 cache entry as a whole. The Server also sets Cur Hop Limit, M and O 2086 flags, Reachable Time and Retrans Timer to values appropriate for the 2087 OMNI link. The Server includes the MNPs, any other prefix management 2088 parameters and an OMNI option with no Interface Attributes. The 2089 Server then includes one or more RIOs that encode the MSPs for the 2090 OMNI link, plus an MTU option (see Section 3.9). The Server finally 2091 forwards the message to the Client using OAL/INET, INET, or NULL 2092 encapsulation and OAL fragmentation as necessary. 2094 After the initial RS/RA exchange, the Server maintains a 2095 ReachableTime timer for each of the Client's underlying interfaces 2096 individually (and for the Client's symmetric neighbor cache entry 2097 collectively) set to expire after ReachableTime seconds. If the 2098 Client (or Proxy) issues additional RS messages, the Server sends an 2099 RA response and resets ReachableTime. If the Server receives an ND 2100 message with a prefix release indication it sets the Client's 2101 symmetric neighbor cache entry to the DEPARTED state and withdraws 2102 the MNP from the routing system after a short delay (e.g., 2 2103 seconds). If ReachableTime expires before a new RS is received on an 2104 individual underlying interface, the Server marks the interface as 2105 DOWN. If ReachableTime expires before any new RS is received on any 2106 individual underlying interface, the Server sets the symmetric 2107 neighbor cache entry state to STALE and sets a 10 second timer. If 2108 the Server has not received a new RS or ND message with a prefix 2109 release indication before the 10 second timer expires, it deletes the 2110 neighbor cache entry and withdraws the MNP from the routing system. 2112 The Server processes any ND messages pertaining to the Client and 2113 returns an NA/RA reply in response to solicitations. The Server may 2114 also issue unsolicited RA messages, e.g., with reconfigure parameters 2115 to cause the Client to renegotiate its prefix delegation/ 2116 registrations, with Router Lifetime set to 0 if it can no longer 2117 service this Client, etc. Finally, If the symmetric neighbor cache 2118 entry is in the DEPARTED state, the Server deletes the entry after 2119 DepartTime expires. 2121 Note: Clients SHOULD notify former Servers of their departures, but 2122 Servers are responsible for expiring neighbor cache entries and 2123 withdrawing routes even if no departure notification is received 2124 (e.g., if the Client leaves the network unexpectedly). Servers 2125 SHOULD therefore set Router Lifetime to ReachableTime seconds in 2126 solicited RA messages to minimize persistent stale cache information 2127 in the absence of Client departure notifications. A short Router 2128 Lifetime also ensures that proactive Client/Server RS/RA messaging 2129 will keep any NAT state alive (see above). 2131 Note: All Servers on an OMNI link MUST advertise consistent values in 2132 the RA Cur Hop Limit, M and O flags, Reachable Time and Retrans Timer 2133 fields the same as for any link, since unpredictable behavior could 2134 result if different Servers on the same link advertised different 2135 values. 2137 3.12.3.1. DHCPv6-Based Prefix Registration 2139 When a Client is not pre-provisioned with an MNP-LLA, it will need 2140 for the Server to select one or more MNPs on its behalf and set up 2141 the correct state in the AERO routing service. (A Client with a pre- 2142 provisioned MNP may also request the Server to select additional 2143 MNPs.) The DHCPv6 service [RFC8415] is used to support this 2144 requirement. 2146 When a Client needs to have the Server select MNPs, it sends an RS 2147 message with source address set to the unspecified address (::) and 2148 with an OMNI option that includes a DHCPv6 message sub-option with 2149 DHCPv6 Prefix Delegation (DHCPv6-PD) parameters. When the Server 2150 receives the RS message, it extracts the DHCPv6-PD message from the 2151 OMNI option. 2153 The Server then acts as a "Proxy DHCPv6 Client" in a message exchange 2154 with the locally-resident DHCPv6 server, which delegates MNPs and 2155 returns a DHCPv6-PD Reply message. (If the Server wishes to defer 2156 creation of MN state until the DHCPv6-PD Reply is received, it can 2157 instead act as a Lightweight DHCPv6 Relay Agent per [RFC6221] by 2158 encapsulating the DHCPv6-PD message in a Relay-forward/reply exchange 2159 with Relay Message and Interface ID options.) 2161 When the Server receives the DHCPv6-PD Reply, it adds a route to the 2162 routing system and creates an MNP-LLA based on the delegated MNP. 2163 The Server then sends an RA back to the Client with the (newly- 2164 created) MNP-LLA as the destination address and with the DHCPv6-PD 2165 Reply message coded in the OMNI option. When the Client receives the 2166 RA, it creates a default route, assigns the Subnet Router Anycast 2167 address and sets its MNP-LLA based on the delegated MNP. 2169 3.13. The AERO Proxy 2171 Clients may connect to protected-spectrum ANETs that employ physical 2172 and/or link-layer security services to facilitate communications to 2173 Servers in outside INETs. In that case, the ANET can employ an AERO 2174 Proxy. The Proxy is located at the ANET/INET border and listens for 2175 RS messages originating from or RA messages destined to ANET Clients. 2176 The Proxy acts on these control messages as follows: 2178 o when the Proxy receives an RS message from a new ANET Client, it 2179 first authenticates the message then examines the network-layer 2180 destination address. If the destination address is a Server's 2181 ADM-LLA, the Proxy proceeds to the next step. Otherwise, if the 2182 destination is (link-local) All-Routers multicast, the Proxy 2183 selects a "nearby" Server that is likely to be a good candidate to 2184 serve the Client and replaces the destination address with the 2185 Server's ADM-LLA. Next, the Proxy creates a proxy neighbor cache 2186 entry and caches the Client and Server link-layer addresses along 2187 with the OMNI option information and any other identifying 2188 information including Transaction IDs, Client Identifiers, Nonce 2189 values, etc. The Proxy finally encapsulates the (proxyed) RS 2190 message in an OAL header with source set to the Proxy's ADM-ULA 2191 and destination set to the Server's ADM-ULA. The Proxy also 2192 includes an OMNI header with an Interface Attributes option that 2193 includes its own INET address plus a unique Port Number for this 2194 Client, then forwards the message into the OMNI link spanning 2195 tree. (Note: including a unique Port Number allows the Server to 2196 distinguish different Clients located behind the same Proxy at the 2197 link-layer, whereas the link-layer addresses would otherwise be 2198 indistinguishable) 2200 o when the Server receives the RS, it authenticates the message then 2201 creates or updates a symmetric neighbor cache entry for the Client 2202 with the Proxy's ADM-ULA, INET address and Port Number as the 2203 link-layer address information. The Server then sends an RA 2204 message back to the Proxy via the spanning tree. 2206 o when the Proxy receives the RA, it authenticates the message and 2207 matches it with the proxy neighbor cache entry created by the RS. 2208 The Proxy then caches the prefix information as a mapping from the 2209 Client's MNPs to the Client's link-layer address, caches the 2210 Server's advertised Router Lifetime and sets the neighbor cache 2211 entry state to REACHABLE. The Proxy then optionally rewrites the 2212 Router Lifetime and forwards the (proxyed) message to the Client. 2213 The Proxy finally includes an MTU option (if necessary) with an 2214 MTU to use for the underlying ANET interface. 2216 After the initial RS/RA exchange, the Proxy forwards any Client data 2217 packets for which there is no matching asymmetric neighbor cache 2218 entry to a Bridge using OAL encapsulation with its own ADM-ULA as the 2219 source and the MNP-ULA corresponding to the Client as the 2220 destination. The Proxy instead forwards any Client data destined to 2221 an asymmetric neighbor cache target directly to the target according 2222 to the OAL/link-layer information - the process of establishing 2223 asymmetric neighbor cache entries is specified in Section 3.14. 2225 While the Client is still attached to the ANET, the Proxy sends NS, 2226 RS and/or unsolicited NA messages to update the Server's symmetric 2227 neighbor cache entries on behalf of the Client and/or to convey QoS 2228 updates. This allows for higher-frequency Proxy-initiated RS/RA 2229 messaging over well-connected INET infrastructure supplemented by 2230 lower-frequency Client-initiated RS/RA messaging over constrained 2231 ANET data links. 2233 If the Server ceases to send solicited advertisements, the Proxy 2234 sends unsolicited RAs on the ANET interface with destination set to 2235 (link-local) All-Nodes multicast and with Router Lifetime set to zero 2236 to inform Clients that the Server has failed. Although the Proxy 2237 engages in ND exchanges on behalf of the Client, the Client can also 2238 send ND messages on its own behalf, e.g., if it is in a better 2239 position than the Proxy to convey QoS changes, etc. For this reason, 2240 the Proxy marks any Client-originated solicitation messages (e.g. by 2241 inserting a Nonce option) so that it can return the solicited 2242 advertisement to the Client instead of processing it locally. 2244 If the Client becomes unreachable, the Proxy sets the neighbor cache 2245 entry state to DEPARTED and retains the entry for DepartTime seconds. 2246 While the state is DEPARTED, the Proxy forwards any packets destined 2247 to the Client to a Bridge via OAL/ORH encapsulation. When DepartTime 2248 expires, the Proxy deletes the neighbor cache entry and discards any 2249 further packets destined to this (now forgotten) Client. 2251 In some ANETs that employ a Proxy, the Client's MNP can be injected 2252 into the ANET routing system. In that case, the Client can send data 2253 messages without encapsulation so that the ANET routing system 2254 transports the unencapsulated packets to the Proxy. This can be very 2255 beneficial, e.g., if the Client connects to the ANET via low-end data 2256 links such as some aviation wireless links. 2258 If the first-hop ANET access router is on the same underlying link 2259 and recognizes the AERO/OMNI protocol, the Client can avoid 2260 encapsulation for both its control and data messages. When the 2261 Client connects to the link, it can send an unencapsulated RS message 2262 with source address set to its own MNP-LLA (or to a Temporary LLA), 2263 and with destination address set to the ADM-LLA of the Client's 2264 selected Server or to (link-local) All-Routers multicast. The Client 2265 includes an OMNI option formatted as specified in 2266 [I-D.templin-6man-omni-interface]. 2268 The Client then sends the unencapsulated RS message, which will be 2269 intercepted by the AERO-Aware access router. The access router then 2270 encapsulates the RS message in an ANET header with its own address as 2271 the source address and the address of a Proxy as the destination 2272 address. The access router further remembers the address of the 2273 Proxy so that it can encapsulate future data packets from the Client 2274 via the same Proxy. If the access router needs to change to a new 2275 Proxy, it simply sends another RS message toward the Server via the 2276 new Proxy on behalf of the Client. 2278 In some cases, the access router and Proxy may be one and the same 2279 node. In that case, the node would be located on the same physical 2280 link as the Client, but its message exchanges with the Server would 2281 need to pass through a security gateway at the ANET/INET border. The 2282 method for deploying access routers and Proxys (i.e. as a single node 2283 or multiple nodes) is an ANET-local administrative consideration. 2285 Note: The Proxy can apply packing as discussed in Section 5.2 of 2286 [I-D.templin-6man-omni-interface] if an opportunity arises to 2287 concatenate multiple payload packets that will be destined to the 2288 same neighbor. 2290 3.13.1. Combined Proxy/Servers 2292 Clients may need to connect directly to Servers via INET, Direct and 2293 VPNed interfaces (i.e., non-ANET interfaces). If the Client's 2294 underlying interfaces all connect via the same INET partition, then 2295 it can connect to a single controlling Server via all interfaces. 2297 If some Client interfaces connect via different INET partitions, 2298 however, the Client still selects a set of controlling Servers and 2299 sends RS messages via their directly-connected Servers while using 2300 the ADM-LLA of the controlling Server as the destination. 2302 When a Server receives an RS with destination set to the ADM-LLA of a 2303 controlling Server, it acts as a Proxy to forward the message to the 2304 controlling Server while forwarding the corresponding RA reply to the 2305 Client. 2307 3.13.2. Detecting and Responding to Server Failures 2309 In environments where fast recovery from Server failure is required, 2310 Proxys SHOULD use proactive Neighbor Unreachability Detection (NUD) 2311 to track Server reachability in a similar fashion as for 2312 Bidirectional Forwarding Detection (BFD) [RFC5880]. Proxys can then 2313 quickly detect and react to failures so that cached information is 2314 re-established through alternate paths. The NUD control messaging is 2315 carried only over well-connected ground domain networks (i.e., and 2316 not low-end aeronautical radio links) and can therefore be tuned for 2317 rapid response. 2319 Proxys perform proactive NUD with Servers for which there are 2320 currently active ANET Clients by sending continuous NS messages in 2321 rapid succession, e.g., one message per second. The Proxy sends the 2322 NS message via the spanning tree with the Proxy's ADM-LLA as the 2323 source and the ADM-LLA of the Server as the destination. When the 2324 Proxy is also sending RS messages to the Server on behalf of ANET 2325 Clients, the resulting RA responses can be considered as equivalent 2326 hints of forward progress. This means that the Proxy need not also 2327 send a periodic NS if it has already sent an RS within the same 2328 period. If the Server fails (i.e., if the Proxy ceases to receive 2329 advertisements), the Proxy can quickly inform Clients by sending 2330 multicast RA messages on the ANET interface. 2332 The Proxy sends RA messages on the ANET interface with source address 2333 set to the Server's address, destination address set to (link-local) 2334 All-Nodes multicast, and Router Lifetime set to 0. The Proxy SHOULD 2335 send MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small 2336 delays [RFC4861]. Any Clients on the ANET that had been using the 2337 failed Server will receive the RA messages and associate with a new 2338 Server. 2340 3.13.3. Point-to-Multipoint Server Coordination 2342 In environments where Client messaging over ANETs is bandwidth- 2343 limited and/or expensive, Clients can enlist the services of the 2344 Proxy to coordinate with multiple Servers in a single RS/RA message 2345 exchange. The Client can send a single RS message to (link-local) 2346 All-Routers multicast that includes the ID's of multiple Servers in 2347 MS-Register sub-options of the OMNI option. 2349 When the Proxy receives the RS and processes the OMNI option, it 2350 sends a separate RS to each MS-Register Server ID. When the Proxy 2351 receives an RA, it can optionally return an immediate "singleton" RA 2352 to the Client or record the Server's ID for inclusion in a pending 2353 "aggregate" RA message. The Proxy can then return aggregate RA 2354 messages to the Client including multiple Server IDs in order to 2355 conserve bandwidth. Each RA includes a proper subset of the Server 2356 IDs from the original RS message, and the Proxy must ensure that the 2357 message contents of each RA are consistent with the information 2358 received from the (aggregated) Servers. 2360 Clients can thereafter employ efficient point-to-multipoint Server 2361 coordination under the assistance of the Proxy to reduce the number 2362 of messages sent over the ANET while enlisting the support of 2363 multiple Servers for fault tolerance. Clients can further include 2364 MS-Release sub-options in IPv6 ND messages to request the Proxy to 2365 release from former Servers via the procedures discussed in 2366 Section 3.16.5. 2368 The OMNI interface specification [I-D.templin-6man-omni-interface] 2369 provides further discussion of the Client/Proxy RS/RA messaging 2370 involved in point-to-multipoint coordination. 2372 3.14. AERO Route Optimization / Address Resolution 2374 While data packets are flowing between a source and target node, 2375 route optimization SHOULD be used. Route optimization is initiated 2376 by the first eligible Route Optimization Source (ROS) closest to the 2377 source as follows: 2379 o For Clients on VPNed and Direct interfaces, the Server is the ROS. 2381 o For Clients on ANET interfaces, the Proxy is the ROS. 2383 o For Clients on INET interfaces, the Client itself is the ROS. 2385 o For correspondent nodes on INET/EUN interfaces serviced by a 2386 Relay, the Relay is the ROS. 2388 The route optimization procedure is conducted between the ROS and the 2389 target Server/Relay acting as a Route Optimization Responder (ROR) in 2390 the same manner as for IPv6 ND Address Resolution and using the same 2391 NS/NA messaging. The target may either be a MNP Client serviced by a 2392 Server, or a non-MNP correspondent reachable via a Relay. 2394 The procedures are specified in the following sections. 2396 3.14.1. Route Optimization Initiation 2398 While data packets are flowing from the source node toward a target 2399 node, the ROS performs address resolution by sending an NS message 2400 for Address Resolution (NS(AR)) to receive a solicited NA(AR) message 2401 from the ROR. When the ROS sends an NS(AR), it includes: 2403 o the LLA of the ROS as the source address. 2405 o the data packet's destination as the Target Address. 2407 o the Solicited-Node multicast address [RFC4291] formed from the 2408 lower 24 bits of the data packet's destination as the destination 2409 address, e.g., for 2001:db8:1:2::10:2000 the NS destination 2410 address is ff02:0:0:0:0:1:ff10:2000. 2412 The NS(AR) message includes an OMNI option with no Interface 2413 Attributes, such that the target will not create a neighbor cache 2414 entry. The Prefix Length in the OMNI option is set to the Prefix 2415 Length associated with the ROS's LLA. 2417 The ROS then encapsulates the NS(AR) message in an OAL header with 2418 source set to its own ULA and destination set to the MNP-ULA 2419 corresponding to the target, then sends the message into the spanning 2420 tree without decrementing the network-layer TTL/Hop Limit field. 2421 (When the ROS is a Client, it instead securely sends the NS(AR) to 2422 one of its current Servers as specified in Section 3.22. The Server 2423 then forwards an adapted version of the NS(AR) into the spanning tree 2424 on behalf of the Client.) 2426 Note: The ROS can apply packing as discussed in Section 5.2 of 2427 [I-D.templin-6man-omni-interface] to concatenate the NS(AR) onto the 2428 same OAL packet used to carry one or more data packets that triggered 2429 route optimization. 2431 3.14.2. Relaying the NS(AR) 2433 When the Bridge receives the NS(AR) message from the ROS, it discards 2434 the INET header and determines that the ROR is the next hop by 2435 consulting its standard IPv6 forwarding table for the OAL header 2436 destination address. The Bridge then forwards the message toward the 2437 ROR via the spanning tree the same as for any IPv6 router. The 2438 final-hop Bridge in the spanning tree will deliver the message via a 2439 secured tunnel to the ROR. 2441 3.14.3. Processing the NS(AR) and Sending the NA(AR) 2443 When the ROR receives the NS(AR) message, it examines the Target 2444 Address to determine whether it has a neighbor cache entry and/or 2445 route that matches the target. If there is no match, the ROR drops 2446 the message. Otherwise, the ROR continues processing as follows: 2448 o if the target belongs to an MNP Client neighbor in the DEPARTED 2449 state the ROR changes the NS(AR) message OAL destination address 2450 to the ADM-ULA of the Client's new Server, forwards the message 2451 into the spanning tree and returns from processing. 2453 o If the target belongs to an MNP Client neighbor in the REACHABLE 2454 state, the ROR instead adds the AERO source address to the target 2455 Client's Report List with time set to ReportTime. 2457 o If the target belongs to a non-MNP route, the ROR continues 2458 processing without adding an entry to the Report List. 2460 The ROR then prepares a (solicited) NA(AR) message to send back to 2461 the ROS but does not create a neighbor cache entry. The ROR sets the 2462 NA(AR) source address to the MNP-LLA corresponding to the target, 2463 sets the Target Address to the target of the solicitation, and sets 2464 the destination address to the source of the solicitation. The ROR 2465 then includes an OMNI option with Prefix Length set to the length 2466 associated with the MNP-LLA. 2468 If the target is an MNP Client, the ROR next includes Interface 2469 Attributes in the OMNI option for each of the target Client's 2470 underlying interfaces with current information for each interface and 2471 with the S/T-ifIndex field in the OMNI header set to 0 to indicate 2472 that the message originated from the ROR and not the Client. 2474 For each Interface Attributes sub-option, the ROR sets the L2ADDR 2475 according to its own INET address for VPNed or Direct interfaces, to 2476 the INET address of the Proxy or to the Client's INET address for 2477 INET interfaces. The ROR then includes the lower 32 bits of its own 2478 ADM-ULA (or the ADM-ULA of the Proxy) as the LHS, encodes the ADM-ULA 2479 prefix length code in the SRT field and sets the FMT code accordingly 2480 as specified in Section 3.3. 2482 The ROR then sets the NA(AR) message R flag to 1 (as a router), S 2483 flag to 1 (as a response to a solicitation), and O flag to 0 (as a 2484 proxy). The ROR finally encapsulates the NA(AR) message in an OAL 2485 header with source set to its own ULA and destination set to the 2486 source ULA of the NS(AR) message, then forwards the message into the 2487 spanning tree without decrementing the network-layer TTL/Hop Limit 2488 field. 2490 3.14.4. Relaying the NA(AR) 2492 When the Bridge receives the NA(AR) message from the ROR, it discards 2493 the INET header and determines that the ROS is the next hop by 2494 consulting its standard IPv6 forwarding table for the OAL header 2495 destination address. The Bridge then forwards the OAL-encapsulated 2496 NA(AR) message toward the ROS the same as for any IPv6 router. The 2497 final-hop Bridge in the spanning tree will deliver the message via a 2498 secured tunnel to the ROS. 2500 3.14.5. Processing the NA(AR) 2502 When the ROS receives the NA(AR) message, it processes the message 2503 the same as for standard IPv6 Address Resolution [RFC4861]. In the 2504 process, it caches the source MNP-ULA then creates an asymmetric 2505 neighbor cache entry for the target and caches all information found 2506 in the OMNI option. The ROS finally sets the asymmetric neighbor 2507 cache entry lifetime to ReachableTime seconds. (When the ROS is a 2508 Client, the solicited NA(AR) message will first be delivered via the 2509 spanning tree to one of its current Servers, which then securely 2510 forwards the message to the Client as discussed in Section 3.22.) 2512 3.14.6. Route Optimization Maintenance 2514 Following route optimization, the ROS forwards future data packets 2515 destined to the target via the addresses found in the cached link- 2516 layer information. The route optimization is shared by all sources 2517 that send packets to the target via the ROS, i.e., and not just the 2518 source on behalf of which the route optimization was initiated. 2520 While new data packets destined to the target are flowing through the 2521 ROS, it sends additional NS(AR) messages to the ROR before 2522 ReachableTime expires to receive a fresh NA(AR) message the same as 2523 described in the previous sections (route optimization refreshment 2524 strategies are an implementation matter, with a non-normative example 2525 given in Appendix A.1). The ROS uses the cached ADM-ULA of the ROR 2526 as the NS(AR) OAL destination address (i.e., instead of using the ULA 2527 corresponding to the target as was the case for the initial NS(AR)), 2528 and sends up to MAX_MULTICAST_SOLICIT NS(AR) messages separated by 1 2529 second until an NA(AR) is received. If no NA(AR) is received, the 2530 ROS assumes that the current ROR has become unreachable and deletes 2531 the target neighbor cache entry. Subsequent data packets will 2532 trigger a new route optimization with an NS with OAL destination 2533 address set to the MNP-ULA corresponding to the target per 2534 Section 3.14.1 to discover a new ROR while initial data packets 2535 travel over a suboptimal route. 2537 If an NA(AR) is received, the ROS then updates the asymmetric 2538 neighbor cache entry to refresh ReachableTime, while (for MNP 2539 destinations) the ROR adds or updates the ROS address to the target's 2540 Report List and with time set to ReportTime. While no data packets 2541 are flowing, the ROS instead allows ReachableTime for the asymmetric 2542 neighbor cache entry to expire. When ReachableTime expires, the ROS 2543 deletes the asymmetric neighbor cache entry. Any future data packets 2544 flowing through the ROS will again trigger a new route optimization. 2546 The ROS may also receive unsolicited NA messages from the ROR at any 2547 time (see: Section 3.16). If there is an asymmetric neighbor cache 2548 entry for the target, the ROS updates the link-layer information but 2549 does not update ReachableTime since the receipt of an unsolicited NA 2550 does not confirm that any forward paths are working. If there is no 2551 asymmetric neighbor cache entry, the ROS simply discards the 2552 unsolicited NA. 2554 In this arrangement, the ROS holds an asymmetric neighbor cache entry 2555 for the target via the ROR, but the ROR does not hold an asymmetric 2556 neighbor cache entry for the ROS. The route optimization neighbor 2557 relationship is therefore asymmetric and unidirectional. If the 2558 target node also has packets to send back to the source node, then a 2559 separate route optimization procedure is performed in the reverse 2560 direction. But, there is no requirement that the forward and reverse 2561 paths be symmetric. 2563 3.15. Neighbor Unreachability Detection (NUD) 2565 AERO nodes perform Neighbor Unreachability Detection (NUD) per 2566 [RFC4861] either reactively in response to persistent link-layer 2567 errors (see Section 3.11) or proactively to confirm reachability and/ 2568 or establish NAT state. The NUD algorithm is based on periodic 2569 control message exchanges. The algorithm may further be seeded by ND 2570 hints of forward progress, but care must be taken to avoid inferring 2571 reachability based on spoofed information. For example, authentic 2572 IPv6 ND message exchanges may be considered as acceptable hints of 2573 forward progress, while spurious data packets should not be. 2575 AERO nodes can use (OAL/ORH-encapsulated) standard NS/NA exchanges 2576 sent over the OMNI link spanning tree to securely test reachability 2577 without risk of DoS attacks from nodes pretending to be a neighbor 2578 (these NS/NA(NUD) messages use the unicast LLAs and ULAs of the two 2579 parties involved in the NUD test the same as for standard IPv6 ND, 2580 and both messages flow over the spanning tree). Proxys can further 2581 perform NUD to securely verify Server reachability on behalf of their 2582 proxyed Clients. However, a means for an ROS to test the unsecured 2583 target route optimized paths is also necessary. 2585 When an ROR directs an ROS to a target neighbor with one or more 2586 link-layer addresses, the ROS can proactively test each such 2587 unsecured route optimized path through secured NS(NUD) messages over 2588 the spanning tree that invoke an unsecured NA(NUD) reply that travels 2589 over the route optimized path.. (The NS(NUD) messages must therefore 2590 include Nonce and Timestamp options that will be echoed in the 2591 unsecured NA(NUD) replies.) While testing the paths, the ROS can 2592 optionally continue to send packets via the spanning tree, maintain a 2593 small queue of packets until target reachability is confirmed, or 2594 (optimistically) allow packets to flow via the route optimized paths. 2596 When the ROS sends an NS(NUD) message, it sets the IPv6 source to its 2597 own address, sets the destination to the MNP-LLA of the target, and 2598 sets the target's MNP Subnet-Router anycast address as the Target 2599 Address. The ROS also includes an OMNI option with a single 2600 Interface Attributes sub-option with the SRT, FMT, LHS and L2ADDR 2601 information for its own underlying interface it wishes to test, but 2602 sets the S/T-ifIndex field to the index for target's underlying 2603 interface to be tested. The ROS includes a Nonce and Timestamp 2604 option, then encapsulates the message in OAL/INET headers with its 2605 own ULA as the source and the ULA of the target as the destination. 2606 The ROS then forwards the NS(NUD) message toward the target via a 2607 Server or Bridge. 2609 When the target receives the NS(NUD) message, it creates an NA(NUD) 2610 by reversing the OAL and IPv6 addresses and including an Interface 2611 Attributes sub-option with attributes for its own interface 2612 identified by the NS(NUD) S/T-ifIndex. The target sets the NA(NUD) 2613 S/T-ifIndex to the index of the ROS, sets the Target Address to the 2614 same value that was in the NS(NUD), and returns the message using its 2615 own underlying interface identified by S/T-ifIndex and destined to 2616 the ROS's interface identified by the original Interface Attributes 2617 sub-option. 2619 When the ROS receives the NA(NUD) message, it can determine from the 2620 Nonce, Timestamp and Target Address that the message matched its 2621 NS(NUD) and that it transited the direct path from the ROR using the 2622 selected underlying interface pair. The ROS marks route optimization 2623 target paths that pass these NUD tests as "reachable", and those that 2624 do not as "unreachable". These markings inform the OMNI interface 2625 forwarding algorithm specified in Section 3.10. 2627 Note: If the target determines that the OMNI option Interface 2628 Attributes in the NS(NUD) is located in a different OMNI link segment 2629 than its own interface named in the S/T-ifIndex, it instead returns 2630 the NA(NUD) via the spanning tree while including an ORH and setting 2631 the OAL destination address to the Subnet Router Anycast address used 2632 by Bridges on the ROS segment. When a Bridge on the ROS segment 2633 receives the NA(NUD), it replaces the Interface Attributes with 2634 information for its own interface while using the ifIndex value 2635 specific to the target. 2637 3.16. Mobility Management and Quality of Service (QoS) 2639 AERO is a Distributed Mobility Management (DMM) service. Each Server 2640 is responsible for only a subset of the Clients on the OMNI link, as 2641 opposed to a Centralized Mobility Management (CMM) service where 2642 there is a single network mobility collective entity for all Clients. 2643 Clients coordinate with their associated Servers via RS/RA exchanges 2644 to maintain the DMM profile, and the AERO routing system tracks all 2645 current Client/Server peering relationships. 2647 Servers provide default routing and mobility/multilink services for 2648 their dependent Clients. Clients are responsible for maintaining 2649 neighbor relationships with their Servers through periodic RS/RA 2650 exchanges, which also serves to confirm neighbor reachability. When 2651 a Client's underlying interface address and/or QoS information 2652 changes, the Client is responsible for updating the Server with this 2653 new information. Note that when there is a Proxy in the path, the 2654 Proxy can also perform some RS/RA exchanges on the Client's behalf. 2656 Mobility management messaging is based on the transmission and 2657 reception of unsolicited Neighbor Advertisement (uNA) messages. Each 2658 uNA message sets the IPv6 destination address to (link-local) All- 2659 Nodes multicast to convey a general update of Interface Attributes to 2660 (possibly) multiple recipients, or to a specific unicast LLA to 2661 announce a departure event to a specific recipient. Implementations 2662 must therefore examine the destination address to determine the 2663 nature of the mobility event (i.e., update vs departure). 2665 Mobility management considerations are specified in the following 2666 sections. 2668 3.16.1. Mobility Update Messaging 2670 Servers accommodate Client mobility, multilink and/or QoS change 2671 events by sending unsolicited NA (uNA) messages to each ROS in the 2672 target Client's Report List. When a Server sends a uNA message, it 2673 sets the IPv6 source address to the Client's MNP-LLA, sets the 2674 destination address to (link-local) All-Nodes multicast and sets the 2675 Target Address to the Client's Subnet-Router anycast address. The 2676 Server also includes an OMNI option with Prefix Length set to the 2677 length associated with the Client's MNP-LLA, with Interface 2678 Attributes for the target Client's underlying interfaces and with the 2679 OMNI header S/T-ifIndex set to 0. The Server then sets the NA R flag 2680 to 1, the S flag to 0 and the O flag to 1, then encapsulates the 2681 message in an OAL header with source set to its own ADM-ULA and 2682 destination set to the ULA of the ROS and sends the message into the 2683 spanning tree. 2685 As discussed in Section 7.2.6 of [RFC4861], the transmission and 2686 reception of uNA messages is unreliable but provides a useful 2687 optimization. In well-connected Internetworks with robust data links 2688 uNA messages will be delivered with high probability, but in any case 2689 the Server can optionally send up to MAX_NEIGHBOR_ADVERTISEMENT uNAs 2690 to each ROS to increase the likelihood that at least one will be 2691 received. 2693 When the ROS receives a uNA message prepared as above, it ignores the 2694 message if there is no existing neighbor cache entry for the Client. 2695 Otherwise, it uses the included OMNI option information to update the 2696 neighbor cache entry, but does not reset ReachableTime since the 2697 receipt of an unsolicited NA message from the target Server does not 2698 provide confirmation that any forward paths to the target Client are 2699 working. 2701 If uNA messages are lost, the ROS may be left with stale address and/ 2702 or QoS information for the Client for up to ReachableTime seconds. 2703 During this time, the ROS can continue sending packets according to 2704 its stale neighbor cache information. When ReachableTime is close to 2705 expiring, the ROS will re-initiate route optimization and receive 2706 fresh link-layer address information. 2708 In addition to sending uNA messages to the current set of ROSs for 2709 the Client, the Server also sends uNAs to the MNP-ULA associated with 2710 the link-layer address for any underlying interface for which the 2711 link-layer address has changed. These uNA messages update an old 2712 Proxy/Server that cannot easily detect (e.g., without active probing) 2713 when a formerly-active Client has departed. When the Server sends 2714 the uNA, it sets the IPv6 source address to the Client's MNP-LLA, 2715 sets the destination address to the old Proxy/Server's ADM-LLA, and 2716 sets the Target Address to the Client's Subnet-Router anycast 2717 address. The Server also includes an OMNI option with Prefix Length 2718 set to the length associated with the Client's MNP-LLA, with 2719 Interface Attributes for the changed underlying interface, and with 2720 the OMNI header S/T-ifIndex set to 0. The Server then sets the NA R 2721 flag to 1, the S flag to 0 and the O flag to 1, then encapsulates the 2722 message in an OAL header with source set to its own ADM-ULA and 2723 destination set to the ADM-ULA of the old Proxy/Server and sends the 2724 message into the spanning tree. 2726 3.16.2. Announcing Link-Layer Address and/or QoS Preference Changes 2728 When a Client needs to change its underlying interface addresses and/ 2729 or QoS preferences (e.g., due to a mobility event), either the Client 2730 or its Proxys send RS messages to the Server via the spanning tree 2731 with an OMNI option that includes Interface attributes with the new 2732 link quality and address information. 2734 Up to MAX_RTR_SOLICITATIONS RS messages MAY be sent in parallel with 2735 sending actual data packets in case one or more RAs are lost. If all 2736 RAs are lost, the Client SHOULD re-associate with a new Server. 2738 When the Server receives the Client's changes, it sends uNA messages 2739 to all nodes in the Report List the same as described in the previous 2740 section. 2742 3.16.3. Bringing New Links Into Service 2744 When a Client needs to bring new underlying interfaces into service 2745 (e.g., when it activates a new data link), it sends an RS message to 2746 the Server via the underlying interface with an OMNI option that 2747 includes Interface Attributes with appropriate link quality values 2748 and with link-layer address information for the new link. 2750 3.16.4. Deactivating Existing Links 2752 When a Client needs to deactivate an existing underlying interface, 2753 it sends an RS or uNA message to its Server with an OMNI option with 2754 appropriate Interface Attribute values - in particular, the link 2755 quality value 0 assures that neighbors will cease to use the link. 2757 If the Client needs to send RS/uNA messages over an underlying 2758 interface other than the one being deactivated, it MUST include 2759 Interface Attributes with appropriate link quality values for any 2760 underlying interfaces being deactivated. 2762 Note that when a Client deactivates an underlying interface, 2763 neighbors that have received the RS/uNA messages need not purge all 2764 references for the underlying interface from their neighbor cache 2765 entries. The Client may reactivate or reuse the underlying interface 2766 and/or its ifIndex at a later point in time, when it will send RS/uNA 2767 messages with fresh Interface Attributes to update any neighbors. 2769 3.16.5. Moving Between Servers 2771 The Client performs the procedures specified in Section 3.12.2 when 2772 it first associates with a new Server or renews its association with 2773 an existing Server. The Client also includes MS-Release identifiers 2774 in the RS message OMNI option per [I-D.templin-6man-omni-interface] 2775 if it wants the new Server to notify any old Servers from which the 2776 Client is departing. 2778 When the new Server receives the Client's RS message, it returns an 2779 RA as specified in Section 3.12.3 and sends up to 2780 MAX_NEIGHBOR_ADVERTIISEMENT uNA messages to any old Servers listed in 2781 OMNI option MS-Release identifiers. When the new Server sends a uNA 2782 message, it sets the IPv6 source address to the Client's MNP-LLA, 2783 sets the destination address to the old Server's ADM-LLA, and sets 2784 the Target Address to the Client's Subnet-Router anycast address. 2785 The new Server also includes an OMNI option with Prefix Length set to 2786 the length associated with the Client's MNP-LLA, with Interface 2787 Attributes for its own underlying interface, and with the OMNI header 2788 S/T-ifIndex set to 0. The new Server then sets the NA R flag to 1, 2789 the S flag to 0 and the O flag to 1, then encapsulates the message in 2790 an OAL header with source set to its own ADM-ULA and destination set 2791 to the ADM-ULA of the old Server and sends the message into the 2792 spanning tree. 2794 When an old Server receives the uNA, it changes the Client's neighbor 2795 cache entry state to DEPARTED, sets the link-layer address of the 2796 Client to the new Server's ADM-ULA, and resets DepartTime. After a 2797 short delay (e.g., 2 seconds) the old Server withdraws the Client's 2798 MNP from the routing system. After DepartTime expires, the old 2799 Server deletes the Client's neighbor cache entry. 2801 The old Server also iteratively forwards a copy of the uNA message to 2802 each ROS in the Client's Report List by changing the OAL destination 2803 address to the ULA of the ROS while leaving all other fields of the 2804 message unmodified. When the ROS receives the uNA, it examines the 2805 Target address to determine the correct asymmetric neighbor cache 2806 entry and verifies that the IPv6 destination address matches the old 2807 Server. The ROS then caches the IPv6 source address as the new 2808 Server for the existing asymmetric neighbor cache entry and marks the 2809 entry as STALE. While in the STALE state, the ROS allows new data 2810 packets to flow according to any existing cached link-layer 2811 information and sends new NS(AR) messages using its own ULA as the 2812 OAL source and the ADM-ULA of the new Server as the OAL destination 2813 address to elicit NA messages that reset the asymmetric neighbor 2814 cache entry state to REACHABLE. If no new NA message is received for 2815 10 seconds while in the STALE state, the ROS deletes the neighbor 2816 cache entry. 2818 Clients SHOULD NOT move rapidly between Servers in order to avoid 2819 causing excessive oscillations in the AERO routing system. Examples 2820 of when a Client might wish to change to a different Server include a 2821 Server that has gone unreachable, topological movements of 2822 significant distance, movement to a new geographic region, movement 2823 to a new OMNI link segment, etc. 2825 When a Client moves to a new Server, some of the fragments of a 2826 multiple fragment packet may have already arrived at the old Server 2827 while others are en route to the new Server, however no special 2828 attention in the reassembly algorithm is necessary when re-routed 2829 fragments are simply treated as loss. 2831 3.17. Multicast 2833 The AERO Client provides an IGMP (IPv4) [RFC2236] or MLD (IPv6) 2834 [RFC3810] proxy service for its EUNs and/or hosted applications 2835 [RFC4605]. The Client forwards IGMP/MLD messages over any of its 2836 underlying interfaces for which group membership is required. The 2837 IGMP/MLD messages may be further forwarded by a first-hop ANET access 2838 router acting as an IGMP/MLD-snooping switch [RFC4541], then 2839 ultimately delivered to an AERO Proxy/Server acting as a Protocol 2840 Independent Multicast - Sparse-Mode (PIM-SM, or simply "PIM") 2841 Designated Router (DR) [RFC7761]. AERO Relays also act as PIM 2842 routers (i.e., the same as AERO Proxys/Servers) on behalf of nodes on 2843 INET/EUN networks. The behaviors identified in the following 2844 sections correspond to Source-Specific Multicast (SSM) and Any-Source 2845 Multicast (ASM) operational modes. 2847 3.17.1. Source-Specific Multicast (SSM) 2849 When an ROS (i.e., an AERO Proxy/Server/Relay) "X" acting as PIM 2850 router receives a Join/Prune message from a node on its downstream 2851 interfaces containing one or more ((S)ource, (G)roup) pairs, it 2852 updates its Multicast Routing Information Base (MRIB) accordingly. 2853 For each S belonging to a prefix reachable via X's non-OMNI 2854 interfaces, X then forwards the (S, G) Join/Prune to any PIM routers 2855 on those interfaces per [RFC7761]. 2857 For each S belonging to a prefix reachable via X's OMNI interface, X 2858 originates a separate copy of the Join/Prune for each (S,G) in the 2859 message using its own LLA as the source address and ALL-PIM-ROUTERS 2860 as the destination address. X then encapsulates each message in an 2861 OAL header with source address set to the ULA of X and destination 2862 address set to S then forwards the message into the spanning tree, 2863 which delivers it to AERO Server/Relay "Y" that services S. At the 2864 same time, if the message was a Join, X sends a route-optimization NS 2865 message toward each S the same as discussed in Section 3.14. The 2866 resulting NAs will return the LLA for the prefix that matches S as 2867 the network-layer source address and with an OMNI option with the ULA 2868 corresponding to any underlying interfaces that are currently 2869 servicing S. 2871 When Y processes the Join/Prune message, if S located behind any 2872 INET, Direct, or VPNed interfaces Y acts as a PIM router and updates 2873 its MRIB to list X as the next hop in the reverse path. If S is 2874 located behind any Proxys "Z"*, Y also forwards the message to each 2875 Z* over the spanning tree while continuing to use the LLA of X as the 2876 source address. Each Z* then updates its MRIB accordingly and 2877 maintains the LLA of X as the next hop in the reverse path. Since 2878 the Bridges do not examine network layer control messages, this means 2879 that the (reverse) multicast tree path is simply from each Z* (and/or 2880 Y) to X with no other multicast-aware routers in the path. If any Z* 2881 (and/or Y) is located on the same OMNI link segment as X, the 2882 multicast data traffic sent to X directly using OAL/INET 2883 encapsulation instead of via a Bridge. 2885 Following the initial Join/Prune and NS/NA messaging, X maintains an 2886 asymmetric neighbor cache entry for each S the same as if X was 2887 sending unicast data traffic to S. In particular, X performs 2888 additional NS/NA exchanges to keep the neighbor cache entry alive for 2889 up to t_periodic seconds [RFC7761]. If no new Joins are received 2890 within t_periodic seconds, X allows the neighbor cache entry to 2891 expire. Finally, if X receives any additional Join/Prune messages 2892 for (S,G) it forwards the messages to each Y and Z* in the neighbor 2893 cache entry over the spanning tree. 2895 At some later time, Client C that holds an MNP for source S may 2896 depart from a first Proxy Z1 and/or connect via a new Proxy Z2. In 2897 that case, Y sends an unsolicited NA message to X the same as 2898 specified for unicast mobility in Section 3.16. When X receives the 2899 unsolicited NA message, it updates its asymmetric neighbor cache 2900 entry for the LLA for source S and sends new Join messages to any new 2901 Proxys Z2. There is no requirement to send any Prune messages to old 2902 Proxys Z1 since source S will no longer source any multicast data 2903 traffic via Z1. Instead, the multicast state for (S,G) in Proxy Z1 2904 will soon time out since no new Joins will arrive. 2906 After some later time, C may move to a new Server Y2 and depart from 2907 old Sever Y1. In that case, Y1 sends Join messages for any of C's 2908 active (S,G) groups to Y2 while including its own LLA as the source 2909 address. This causes Y2 to include Y1 in the multicast forwarding 2910 tree during the interim time that Y1's symmetric neighbor cache entry 2911 for C is in the DEPARTED state. At the same time, Y1 sends an 2912 unsolicited NA message to X with an OMNI option with S/T-ifIndex in 2913 the header set to 0 and a release indication to cause X to release 2914 its asymmetric neighbor cache entry. X then sends a new Join message 2915 to S via the spanning tree and re-initiates route optimization the 2916 same as if it were receiving a fresh Join message from a node on a 2917 downstream link. 2919 3.17.2. Any-Source Multicast (ASM) 2921 When an ROS X acting as a PIM router receives a Join/Prune from a 2922 node on its downstream interfaces containing one or more (*,G) pairs, 2923 it updates its Multicast Routing Information Base (MRIB) accordingly. 2924 X then forwards a copy of the message to the Rendezvous Point (RP) R 2925 for each G over the spanning tree. X uses its own LLA as the source 2926 address and ALL-PIM-ROUTERS as the destination address, then 2927 encapsulates each message in an OAL header with source address set to 2928 the ULA of X and destination address set to R, then sends the message 2929 into the spanning tree. At the same time, if the message was a Join 2930 X initiates NS/NA route optimization the same as for the SSM case 2931 discussed in Section 3.17.1. 2933 For each source S that sends multicast traffic to group G via R, the 2934 Proxy/Server Z* for the Client that aggregates S encapsulates the 2935 packets in PIM Register messages and forwards them to R via the 2936 spanning tree, which may then elect to send a PIM Join to Z*. This 2937 will result in an (S,G) tree rooted at Z* with R as the next hop so 2938 that R will begin to receive two copies of the packet; one native 2939 copy from the (S, G) tree and a second copy from the pre-existing (*, 2940 G) tree that still uses PIM Register encapsulation. R can then issue 2941 a PIM Register-stop message to suppress the Register-encapsulated 2942 stream. At some later time, if C moves to a new Proxy/Server Z*, it 2943 resumes sending packets via PIM Register encapsulation via the new 2944 Z*. 2946 At the same time, as multicast listeners discover individual S's for 2947 a given G, they can initiate an (S,G) Join for each S under the same 2948 procedures discussed in Section 3.17.1. Once the (S,G) tree is 2949 established, the listeners can send (S, G) Prune messages to R so 2950 that multicast packets for group G sourced by S will only be 2951 delivered via the (S, G) tree and not from the (*, G) tree rooted at 2952 R. All mobility considerations discussed for SSM apply. 2954 3.17.3. Bi-Directional PIM (BIDIR-PIM) 2956 Bi-Directional PIM (BIDIR-PIM) [RFC5015] provides an alternate 2957 approach to ASM that treats the Rendezvous Point (RP) as a Designated 2958 Forwarder (DF). Further considerations for BIDIR-PIM are out of 2959 scope. 2961 3.18. Operation over Multiple OMNI Links 2963 An AERO Client can connect to multiple OMNI links the same as for any 2964 data link service. In that case, the Client maintains a distinct 2965 OMNI interface for each link, e.g., 'omni0' for the first link, 2966 'omni1' for the second, 'omni2' for the third, etc. Each OMNI link 2967 would include its own distinct set of Bridges, Servers and Proxys, 2968 thereby providing redundancy in case of failures. 2970 Each OMNI link could utilize the same or different ANET connections. 2971 The links can be distinguished at the link-layer via the SRT prefix 2972 in a similar fashion as for Virtual Local Area Network (VLAN) tagging 2973 (e.g., IEEE 802.1Q) and/or through assignment of distinct sets of 2974 MSPs on each link. This gives rise to the opportunity for supporting 2975 multiple redundant networked paths, with each VLAN distinguished by a 2976 different SRT "color" (see: Section 3.2.5). 2978 The Client's IP layer can select the outgoing OMNI interface 2979 appropriate for a given traffic profile while (in the reverse 2980 direction) correspondent nodes must have some way of steering their 2981 packets destined to a target via the correct OMNI link. 2983 In a first alternative, if each OMNI link services different MSPs, 2984 then the Client can receive a distinct MNP from each of the links. 2985 IP routing will therefore assure that the correct OMNI link is used 2986 for both outbound and inbound traffic. This can be accomplished 2987 using existing technologies and approaches, and without requiring any 2988 special supporting code in correspondent nodes or Bridges. 2990 In a second alternative, if each OMNI link services the same MSP(s) 2991 then each link could assign a distinct "OMNI link Anycast" address 2992 that is configured by all Bridges on the link. Correspondent nodes 2993 can then perform Segment Routing to select the correct SRT, which 2994 will then direct the packet over multiple hops to the target. 2996 3.19. DNS Considerations 2998 AERO Client MNs and INET correspondent nodes consult the Domain Name 2999 System (DNS) the same as for any Internetworking node. When 3000 correspondent nodes and Client MNs use different IP protocol versions 3001 (e.g., IPv4 correspondents and IPv6 MNs), the INET DNS must maintain 3002 A records for IPv4 address mappings to MNs which must then be 3003 populated in Relay NAT64 mapping caches. In that way, an IPv4 3004 correspondent node can send packets to the IPv4 address mapping of 3005 the target MN, and the Relay will translate the IPv4 header and 3006 destination address into an IPv6 header and IPv6 destination address 3007 of the MN. 3009 When an AERO Client registers with an AERO Server, the Server can 3010 return the address(es) of DNS servers in RDNSS options [RFC6106]. 3011 The DNS server provides the IP addresses of other MNs and 3012 correspondent nodes in AAAA records for IPv6 or A records for IPv4. 3014 3.20. Transition Considerations 3016 OAL encapsulation ensures that dissimilar INET partitions can be 3017 joined into a single unified OMNI link, even though the partitions 3018 themselves may have differing protocol versions and/or incompatible 3019 addressing plans. However, a commonality can be achieved by 3020 incrementally distributing globally routable (i.e., native) IP 3021 prefixes to eventually reach all nodes (both mobile and fixed) in all 3022 OMNI link segments. This can be accomplished by incrementally 3023 deploying AERO Relays on each INET partition, with each Relay 3024 distributing its MNPs and/or discovering non-MNP IP GUA prefixes on 3025 its INET links. 3027 This gives rise to the opportunity to eventually distribute native IP 3028 addresses to all nodes, and to present a unified OMNI link view even 3029 if the INET partitions remain in their current protocol and 3030 addressing plans. In that way, the OMNI link can serve the dual 3031 purpose of providing a mobility/multilink service and a transition 3032 service. Or, if an INET partition is transitioned to a native IP 3033 protocol version and addressing scheme that is compatible with the 3034 OMNI link MNP-based addressing scheme, the partition and OMNI link 3035 can be joined by Relays. 3037 Relays that connect INETs/EUNs with dissimilar IP protocol versions 3038 may need to employ a network address and protocol translation 3039 function such as NAT64 [RFC6146]. 3041 3.21. Detecting and Reacting to Server and Bridge Failures 3043 In environments where rapid failure recovery is required, Servers and 3044 Bridges SHOULD use Bidirectional Forwarding Detection (BFD) 3045 [RFC5880]. Nodes that use BFD can quickly detect and react to 3046 failures so that cached information is re-established through 3047 alternate nodes. BFD control messaging is carried only over well- 3048 connected ground domain networks (i.e., and not low-end radio links) 3049 and can therefore be tuned for rapid response. 3051 Servers and Bridges maintain BFD sessions in parallel with their BGP 3052 peerings. If a Server or Bridge fails, BGP peers will quickly re- 3053 establish routes through alternate paths the same as for common BGP 3054 deployments. Similarly, Proxys maintain BFD sessions with their 3055 associated Bridges even though they do not establish BGP peerings 3056 with them. 3058 Proxys SHOULD use proactive NUD for Servers for which there are 3059 currently active ANET Clients in a manner that parallels BFD, i.e., 3060 by sending unicast NS messages in rapid succession to receive 3061 solicited NA messages. When the Proxy is also sending RS messages on 3062 behalf of ANET Clients, the RS/RA messaging can be considered as 3063 equivalent hints of forward progress. This means that the Proxy need 3064 not also send a periodic NS if it has already sent an RS within the 3065 same period. If a Server fails, the Proxy will cease to receive 3066 advertisements and can quickly inform Clients of the outage by 3067 sending multicast RA messages on the ANET interface. 3069 The Proxy sends multicast RA messages with source address set to the 3070 Server's address, destination address set to (link-local) All-Nodes 3071 multicast, and Router Lifetime set to 0. The Proxy SHOULD send 3072 MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small delays 3073 [RFC4861]. Any Clients on the ANET interface that have been using 3074 the (now defunct) Server will receive the RA messages and associate 3075 with a new Server. 3077 3.22. AERO Clients on the Open Internet 3079 AERO Clients that connect to the open Internet via INET interfaces 3080 can establish a VPN or direct link to securely connect to a Server in 3081 a "tethered" arrangement with all of the Client's traffic transiting 3082 the Server. Alternatively, the Client can associate with an INET 3083 Server using UDP/IP encapsulation and control message securing 3084 services as discussed in the following sections. 3086 When a Client's OMNI interface enables an INET underlying interface, 3087 it first determines whether the interface is likely to be behind a 3088 NAT. For IPv4, the Client assumes it is on the open Internet if the 3089 INET address is not a special-use IPv4 address per [RFC3330]. 3090 Similarly for IPv6, the Client assumes it is on the open Internet if 3091 the INET address is a Global Unicast Address (GUA) [RFC4291]. 3092 Otherwise, the Client assumes it may be behind one or several NATs. 3094 The Client then prepares an RS message with IPv6 source address set 3095 to its MNP-LLA, with IPv6 destination set to (link-local) All-Routers 3096 multicast and with an OMNI option with underlying interface 3097 attributes. If the Client believes that it is on the open Internet, 3098 it SHOULD include an L2ADDR in the Interface Attributes sub-option 3099 corresponding to the underlying interface; otherwise, it MAY omit 3100 L2ADDR. If the underlying address is IPv4, the Client includes the 3101 Port Number and IPv4 address written in obfuscated form [RFC4380] as 3102 discussed in Section 3.3. If the underlying interface address is 3103 IPv6, the Client instead includes the Port Number and IPv6 address in 3104 obfuscated form. The Client finally includes a HIP "Initiator" 3105 message sub-option in the OMNI option 3106 [I-D.templin-6man-omni-interface] to provide message authentication 3107 and performs OAL encapsulation and fragmentation if necessary. The 3108 Client then encapsulates each fragment in UDP/IP headers, sets the 3109 UDP/IP source to its INET address and UDP port, sets the UDP/IP 3110 destination to the Server's INET address and the AERO service port 3111 number (8060), then sends the (whole or fragmented) message to the 3112 Server. 3114 When the Server receives the RS, it performs OAL reassembly if 3115 necessary then authenticates the message and registers the Client's 3116 MNP and INET interface information according to the OMNI option 3117 parameters. If the RS message OMNI option includes Interface 3118 Attributes with an L2ADDR, the Server compares the encapsulation IP 3119 address and UDP port number with the (unobfuscated) values. If the 3120 values are the same, the Server caches the Client's information as 3121 "INET" addresses meaning that the Client is likely to accept direct 3122 messages without requiring NAT traversal exchanges. If the values 3123 are different (or, if the OMNI option did not include an L2ADDR) the 3124 Server instead caches the Client's information as "mapped" addresses 3125 meaning that NAT traversal exchanges may be necessary. 3127 The Server then prepares an RA message with IPv6 source and 3128 destination set corresponding to the addresses in the RS, and with an 3129 OMNI option with an Origin Indication sub-option per 3130 [I-D.templin-6man-omni-interface] with the mapped and obfuscated Port 3131 Number and IP address observed in the encapsulation headers. The 3132 Server also includes a HIP "Responder" message sub-option per 3133 [I-D.templin-6man-omni-interface] that contains an acknowledgement of 3134 the update sent by the Client. The Server then performs OAL 3135 encapsulation and fragmentation if necessary, and encapsulates each 3136 fragment in UDP/IP headers with adresses set per the L2ADDR 3137 information in the neighbor cache entry for the Client. 3139 When the Client receives the RA message, it performs OAL reassembly 3140 if necessary, authenticates the HIP "Responder" message, then 3141 compares the mapped Port Number and IP address from the Origin 3142 Indication sub-option with its own address. If the addresses are the 3143 same, the Client assumes the open Internet / Cone NAT principle; if 3144 the addresses are different, the Client instead assumes that further 3145 qualification procedures are necessary to detect the type of NAT and 3146 proceeds according to standard procedures [RFC6081][RFC4380]. 3148 After the Client has registered its INET interfaces in such RS/RA 3149 exchanges it sends periodic RS messages to receive fresh RA messages 3150 before the Router Lifetime received on each INET interface expires. 3151 The Client also maintains default routes via its Servers, i.e., the 3152 same as described in earlier sections. 3154 When the Client sends messages to target IP addresses, it also 3155 invokes route optimization per Section 3.14 using IPv6 ND address 3156 resolution messaging. The Client sends the NS(AR) message to the 3157 Server with an OMNI option with a HIP "Update/Sequence" message sub- 3158 option. The Client sets the NS source address to the Client's MNP- 3159 LLA and destination address to the target solicited node multicast 3160 address. The Client wraps the NS message in an OAL header with 3161 source address set to its own MNP-ULA and destination address set to 3162 the Server's ADM-ULA. The Client then performs OAL fragmentation if 3163 necessary, wraps each fragment in a UDP/IP header and sends it to the 3164 Server. 3166 When the Server receives the OAL-encapsulated NS, it performs OAL 3167 reassembly if necessary. The Server then authenticates the message 3168 by processing the HIP message sub-option and sends a corresponding 3169 NS(AR) message over the spanning tree the same as if it were the ROS, 3170 but with the OAL source address set to the Server's ADM-ULA, with 3171 destination set to the MNP-ULA of the target, and with an OMNI option 3172 that includes no sub-options. When the ROR receives the NS(AR), it 3173 adds the Server's ADM-ULA and Client's MNP-LLA to the target's Report 3174 List, and returns an NA(AR) with OMNI option information for the 3175 target including all of the target's Interface Attributes. The ROR 3176 sets the NA(AR) source address to the target's MNP-LLA and sets the 3177 destination address to the Client's MNP-LLA, then sets the OAL source 3178 address to the ADM-ULA of the ROR and the destination to the ADM-ULA 3179 of the Server. When the Server receives the NA(AR) message, it 3180 rewrites the OAL source address to its own ADM-ULA and the 3181 destination address to the MNP-ULA of the Client, then includes a HIP 3182 "Update/Acknowledge" message sub-option in the OMNI option, performs 3183 OAL fragmentation if necessary, wraps each fragment in UDP/IP 3184 headers, and sends it to the Client. 3186 Following route optimization for targets in the same OMNI link 3187 segment, if the target's L2ADDR is on the open INET, the Client 3188 forwards data packets directly to the target INET address. If the 3189 target is behind a NAT, the Client first establishes NAT state for 3190 the L2ADDR using the "direct bubble" and NUD mechanisms discussed in 3191 Section 3.10.1. The Client continues to send data packets via its 3192 Server until NAT state is populated, then begins forwarding packets 3193 via the direct path through the NAT to the target. For targets in 3194 different OMNI link segments, the Client uses OAL/ORH encapsulation 3195 and forwards data packets to the Bridge that returned the NA message. 3197 The ROR may return uNAs via the Server if the target moves, and the 3198 Server will send corresponding uNAs to the Client with a HIP "Notify" 3199 authentication message. The Client can also send NUD messages to 3200 test forward path reachability even though there is no security 3201 association between the Client and the target. 3203 The Client sends UDP/IP encapsulated IPv6 packets to route-optimized 3204 neighbors in the same OMNI link segment no larger than 576 bytes in 3205 one piece and without OAL encapsulation. Otherwise, the Client 3206 inserts an OAL header with source set to its own MNP-ULA and 3207 destination set to the MNP-ULA of the target and uses OAL 3208 fragmentation if necessary according to Section 3.9. The Client then 3209 encapsulates each fragment in a UDP/IP header and sends the fragments 3210 to the next hop. 3212 Note: The NAT traversal procedures specified in this document are 3213 applicable for Cone, Address-Restricted and Port-Restricted NATs 3214 only. While future updates to this document may specify procedures 3215 for other NAT variations (e.g., hairpinning and various forms of 3216 Symmetric NATs), it should be noted that continuous communications 3217 are always possible through forwarding via a Server even if NAT 3218 traversal is not employed. 3220 Note: Following the initial HIP Initiator/Responder exchange, AERO 3221 Clients with OMNI interfaces configured over the open Internet 3222 maintain HIP associations through the transmission of IPv6 ND 3223 messages that include OMNI options with HIP "Update" and "Notify" 3224 messages. OMNI interfaces use the HIP "Update" message when an 3225 acknowledgement is required, and use the "Notify" message in 3226 unacknowledged isolated IPv6 ND messages (e.g., unsolicited NAs). 3228 3.23. Time-Varying MNPs 3230 In some use cases, it is desirable, beneficial and efficient for the 3231 Client to receive a constant MNP that travels with the Client 3232 wherever it moves. For example, this would allow air traffic 3233 controllers to easily track aircraft, etc. In other cases, however 3234 (e.g., intelligent transportation systems), the MN may be willing to 3235 sacrifice a modicum of efficiency in order to have time-varying MNPs 3236 that can be changed every so often to defeat adversarial tracking. 3238 The DHCPv6 service offers a way for Clients that desire time-varying 3239 MNPs to obtain short-lived prefixes (e.g., on the order of a small 3240 number of minutes). In that case, the identity of the Client would 3241 not be bound to the MNP but rather the Client's identity would be 3242 bound to the DHCPv6 Device Unique Identifier (DUID) and used as the 3243 seed for Prefix Delegation. The Client would then be obligated to 3244 renumber its internal networks whenever its MNP (and therefore also 3245 its MNP-LLA) changes. This should not present a challenge for 3246 Clients with automated network renumbering services, however presents 3247 limits for the durations of ongoing sessions that would prefer to use 3248 a constant address. 3250 4. Implementation Status 3252 An early AERO implementation based on OpenVPN (https://openvpn.net/) 3253 was announced on the v6ops mailing list on January 10, 2018 and an 3254 initial public release of the AERO proof-of-concept source code was 3255 announced on the intarea mailing list on August 21, 2015. 3257 AERO Release-3.0.2 was tagged on October 15, 2020, and is undergoing 3258 internal testing. Additional internal releases expected within the 3259 coming months, with first public release expected end of 1H2021. 3261 5. IANA Considerations 3263 The IANA is instructed to assign a new type value TBD in the IPv6 3264 Routing Types registry. 3266 The IANA has assigned the UDP port number "8060" for an earlier 3267 experimental first version of AERO [RFC6706]. This document 3268 obsoletes [RFC6706], and together with 3269 [I-D.templin-6man-omni-interface] reclaims the UDP port number "8060" 3270 for 'aero' as the service port for UDP/IP encapsualtion. (Note that, 3271 although [RFC6706] was not widely implemented or deployed, any 3272 messages coded to that specification can be easily distinguished and 3273 ignored since they use the invalid ICMPv6 message type number '0'.) 3274 This document makes no request of IANA, since 3275 [I-D.templin-6man-omni-interface] already provides instructions. 3277 No further IANA actions are required. 3279 6. Security Considerations 3281 AERO Bridges configure secured tunnels with AERO Servers, Relays and 3282 Proxys within their local OMNI link segments. Applicable secured 3283 tunnel alternatives include IPsec [RFC4301], TLS/SSL [RFC8446], DTLS 3284 [RFC6347], WireGuard [WG], etc. The AERO Bridges of all OMNI link 3285 segments in turn configure secured tunnels for their neighboring AERO 3286 Bridges in a spanning tree topology. Therefore, control messages 3287 exchanged between any pair of OMNI link neighbors on the spanning 3288 tree are already secured. 3290 AERO Servers, Relays and Proxys targeted by a route optimization may 3291 also receive data packets directly from arbitrary nodes in INET 3292 partitions instead of via the spanning tree. For INET partitions 3293 that apply effective ingress filtering to defeat source address 3294 spoofing, the simple data origin authentication procedures in 3295 Section 3.8 can be applied. 3297 For INET partitions that require strong security in the data plane, 3298 two options for securing communications include 1) disable route 3299 optimization so that all traffic is conveyed over secured tunnels, or 3300 2) enable on-demand secure tunnel creation between INET partition 3301 neighbors. Option 1) would result in longer routes than necessary 3302 and traffic concentration on critical infrastructure elements. 3303 Option 2) could be coordinated by establishing a secured tunnel on- 3304 demand instead of performing an NS/NA exchange in the route 3305 optimization procedures. Procedures for establishing on-demand 3306 secured tunnels are out of scope. 3308 AERO Clients that connect to secured ANETs need not apply security to 3309 their ND messages, since the messages will be intercepted by a 3310 perimeter Proxy that applies security on its INET-facing interface as 3311 part of the spanning tree (see above). AERO Clients connected to the 3312 open INET can use symmetric network and/or transport layer security 3313 services such as VPNs or can by some other means establish a direct 3314 link. When a VPN or direct link may be impractical, however, the 3315 HIP-based authentication services specified in [RFC7401] should be 3316 applied. 3318 Application endpoints SHOULD use application-layer security services 3319 such as TLS/SSL, DTLS or SSH [RFC4251] to assure the same level of 3320 protection as for critical secured Internet services. AERO Clients 3321 that require host-based VPN services SHOULD use symmetric network 3322 and/or transport layer security services such as IPsec, TLS/SSL, 3323 DTLS, etc. AERO Proxys and Servers can also provide a network-based 3324 VPN service on behalf of the Client, e.g., if the Client is located 3325 within a secured enclave and cannot establish a VPN on its own 3326 behalf. 3328 AERO Servers and Bridges present targets for traffic amplification 3329 Denial of Service (DoS) attacks. This concern is no different than 3330 for widely-deployed VPN security gateways in the Internet, where 3331 attackers could send spoofed packets to the gateways at high data 3332 rates. This can be mitigated by connecting Servers and Bridges over 3333 dedicated links with no connections to the Internet and/or when 3334 connections to the Internet are only permitted through well-managed 3335 firewalls. Traffic amplification DoS attacks can also target an AERO 3336 Client's low data rate links. This is a concern not only for Clients 3337 located on the open Internet but also for Clients in secured 3338 enclaves. AERO Servers and Proxys can institute rate limits that 3339 protect Clients from receiving packet floods that could DoS low data 3340 rate links. 3342 AERO Relays must implement ingress filtering to avoid a spoofing 3343 attack in which spurious messages with ULA addresses are injected 3344 into an OMNI link from an outside attacker. AERO Clients MUST ensure 3345 that their connectivity is not used by unauthorized nodes on their 3346 EUNs to gain access to a protected network, i.e., AERO Clients that 3347 act as routers MUST NOT provide routing services for unauthorized 3348 nodes. (This concern is no different than for ordinary hosts that 3349 receive an IP address delegation but then "share" the address with 3350 other nodes via some form of Internet connection sharing such as 3351 tethering.) 3353 The MAP list MUST be well-managed and secured from unauthorized 3354 tampering, even though the list contains only public information. 3355 The MAP list can be conveyed to the Client in a similar fashion as in 3356 [RFC5214] (e.g., through layer 2 data link login messaging, secure 3357 upload of a static file, DNS lookups, etc.). 3359 SRH authentication facilities are specified in [RFC8754]. 3361 Security considerations for accepting link-layer ICMP messages and 3362 reflected packets are discussed throughout the document. 3364 Security considerations for IPv6 fragmentation and reassembly are 3365 discussed in [I-D.templin-6man-omni-interface]. 3367 7. Acknowledgements 3369 Discussions in the IETF, aviation standards communities and private 3370 exchanges helped shape some of the concepts in this work. 3371 Individuals who contributed insights include Mikael Abrahamsson, Mark 3372 Andrews, Fred Baker, Bob Braden, Stewart Bryant, Brian Carpenter, 3373 Wojciech Dec, Pavel Drasil, Ralph Droms, Adrian Farrel, Nick Green, 3374 Sri Gundavelli, Brian Haberman, Bernhard Haindl, Joel Halpern, Tom 3375 Herbert, Sascha Hlusiak, Lee Howard, Zdenek Jaron, Andre Kostur, 3376 Hubert Kuenig, Ted Lemon, Andy Malis, Satoru Matsushima, Tomek 3377 Mrugalski, Madhu Niraula, Alexandru Petrescu, Behcet Saikaya, Michal 3378 Skorepa, Joe Touch, Bernie Volz, Ryuji Wakikawa, Tony Whyman, Lloyd 3379 Wood and James Woodyatt. Members of the IESG also provided valuable 3380 input during their review process that greatly improved the document. 3381 Special thanks go to Stewart Bryant, Joel Halpern and Brian Haberman 3382 for their shepherding guidance during the publication of the AERO 3383 first edition. 3385 This work has further been encouraged and supported by Boeing 3386 colleagues including Kyle Bae, M. Wayne Benson, Dave Bernhardt, Cam 3387 Brodie, John Bush, Balaguruna Chidambaram, Irene Chin, Bruce Cornish, 3388 Claudiu Danilov, Don Dillenburg, Joe Dudkowski, Wen Fang, Samad 3389 Farooqui, Anthony Gregory, Jeff Holland, Seth Jahne, Brian Jaury, 3390 Greg Kimberly, Ed King, Madhuri Madhava Badgandi, Laurel Matthew, 3391 Gene MacLean III, Rob Muszkiewicz, Sean O'Sullivan, Vijay 3392 Rajagopalan, Greg Saccone, Rod Santiago, Kent Shuey, Brian Skeen, 3393 Mike Slane, Carrie Spiker, Katie Tran, Brendan Williams, Amelia 3394 Wilson, Julie Wulff, Yueli Yang, Eric Yeh and other members of the 3395 Boeing mobility, networking and autonomy teams. Kyle Bae, Wayne 3396 Benson, Katie Tran and Eric Yeh are especially acknowledged for 3397 implementing the AERO functions as extensions to the public domain 3398 OpenVPN distribution. 3400 Earlier works on NBMA tunneling approaches are found in 3401 [RFC2529][RFC5214][RFC5569]. 3403 Many of the constructs presented in this second edition of AERO are 3404 based on the author's earlier works, including: 3406 o The Internet Routing Overlay Network (IRON) 3407 [RFC6179][I-D.templin-ironbis] 3409 o Virtual Enterprise Traversal (VET) 3410 [RFC5558][I-D.templin-intarea-vet] 3412 o The Subnetwork Encapsulation and Adaptation Layer (SEAL) 3413 [RFC5320][I-D.templin-intarea-seal] 3415 o AERO, First Edition [RFC6706] 3417 Note that these works cite numerous earlier efforts that are not also 3418 cited here due to space limitations. The authors of those earlier 3419 works are acknowledged for their insights. 3421 This work is aligned with the NASA Safe Autonomous Systems Operation 3422 (SASO) program under NASA contract number NNA16BD84C. 3424 This work is aligned with the FAA as per the SE2025 contract number 3425 DTFAWA-15-D-00030. 3427 This work is aligned with the Boeing Commercial Airplanes (BCA) 3428 Internet of Things (IoT) and autonomy programs. 3430 This work is aligned with the Boeing Information Technology (BIT) 3431 MobileNet program. 3433 8. References 3435 8.1. Normative References 3437 [I-D.templin-6man-omni-interface] 3438 Templin, F. and T. Whyman, "Transmission of IP Packets 3439 over Overlay Multilink Network (OMNI) Interfaces", draft- 3440 templin-6man-omni-interface-69 (work in progress), January 3441 2021. 3443 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 3444 DOI 10.17487/RFC0791, September 1981, 3445 . 3447 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 3448 RFC 792, DOI 10.17487/RFC0792, September 1981, 3449 . 3451 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 3452 Requirement Levels", BCP 14, RFC 2119, 3453 DOI 10.17487/RFC2119, March 1997, 3454 . 3456 [RFC2473] Conta, A. and S. Deering, "Generic Packet Tunneling in 3457 IPv6 Specification", RFC 2473, DOI 10.17487/RFC2473, 3458 December 1998, . 3460 [RFC3971] Arkko, J., Ed., Kempf, J., Zill, B., and P. Nikander, 3461 "SEcure Neighbor Discovery (SEND)", RFC 3971, 3462 DOI 10.17487/RFC3971, March 2005, 3463 . 3465 [RFC3972] Aura, T., "Cryptographically Generated Addresses (CGA)", 3466 RFC 3972, DOI 10.17487/RFC3972, March 2005, 3467 . 3469 [RFC4191] Draves, R. and D. Thaler, "Default Router Preferences and 3470 More-Specific Routes", RFC 4191, DOI 10.17487/RFC4191, 3471 November 2005, . 3473 [RFC4193] Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast 3474 Addresses", RFC 4193, DOI 10.17487/RFC4193, October 2005, 3475 . 3477 [RFC4380] Huitema, C., "Teredo: Tunneling IPv6 over UDP through 3478 Network Address Translations (NATs)", RFC 4380, 3479 DOI 10.17487/RFC4380, February 2006, 3480 . 3482 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 3483 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 3484 DOI 10.17487/RFC4861, September 2007, 3485 . 3487 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 3488 Address Autoconfiguration", RFC 4862, 3489 DOI 10.17487/RFC4862, September 2007, 3490 . 3492 [RFC6081] Thaler, D., "Teredo Extensions", RFC 6081, 3493 DOI 10.17487/RFC6081, January 2011, 3494 . 3496 [RFC7401] Moskowitz, R., Ed., Heer, T., Jokela, P., and T. 3497 Henderson, "Host Identity Protocol Version 2 (HIPv2)", 3498 RFC 7401, DOI 10.17487/RFC7401, April 2015, 3499 . 3501 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 3502 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 3503 May 2017, . 3505 [RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6 3506 (IPv6) Specification", STD 86, RFC 8200, 3507 DOI 10.17487/RFC8200, July 2017, 3508 . 3510 [RFC8415] Mrugalski, T., Siodelski, M., Volz, B., Yourtchenko, A., 3511 Richardson, M., Jiang, S., Lemon, T., and T. Winters, 3512 "Dynamic Host Configuration Protocol for IPv6 (DHCPv6)", 3513 RFC 8415, DOI 10.17487/RFC8415, November 2018, 3514 . 3516 8.2. Informative References 3518 [BGP] Huston, G., "BGP in 2015, http://potaroo.net", January 3519 2016. 3521 [I-D.bonica-6man-comp-rtg-hdr] 3522 Bonica, R., Kamite, Y., Alston, A., Henriques, D., and L. 3523 Jalil, "The IPv6 Compact Routing Header (CRH)", draft- 3524 bonica-6man-comp-rtg-hdr-24 (work in progress), January 3525 2021. 3527 [I-D.bonica-6man-crh-helper-opt] 3528 Li, X., Bao, C., Ruan, E., and R. Bonica, "Compressed 3529 Routing Header (CRH) Helper Option", draft-bonica-6man- 3530 crh-helper-opt-02 (work in progress), October 2020. 3532 [I-D.ietf-intarea-frag-fragile] 3533 Bonica, R., Baker, F., Huston, G., Hinden, R., Troan, O., 3534 and F. Gont, "IP Fragmentation Considered Fragile", draft- 3535 ietf-intarea-frag-fragile-17 (work in progress), September 3536 2019. 3538 [I-D.ietf-intarea-tunnels] 3539 Touch, J. and M. Townsley, "IP Tunnels in the Internet 3540 Architecture", draft-ietf-intarea-tunnels-10 (work in 3541 progress), September 2019. 3543 [I-D.ietf-ipwave-vehicular-networking] 3544 Jeong, J., "IPv6 Wireless Access in Vehicular Environments 3545 (IPWAVE): Problem Statement and Use Cases", draft-ietf- 3546 ipwave-vehicular-networking-19 (work in progress), July 3547 2020. 3549 [I-D.ietf-rtgwg-atn-bgp] 3550 Templin, F., Saccone, G., Dawra, G., Lindem, A., and V. 3551 Moreno, "A Simple BGP-based Mobile Routing System for the 3552 Aeronautical Telecommunications Network", draft-ietf- 3553 rtgwg-atn-bgp-10 (work in progress), January 2021. 3555 [I-D.templin-6man-dhcpv6-ndopt] 3556 Templin, F., "A Unified Stateful/Stateless Configuration 3557 Service for IPv6", draft-templin-6man-dhcpv6-ndopt-11 3558 (work in progress), January 2021. 3560 [I-D.templin-intarea-seal] 3561 Templin, F., "The Subnetwork Encapsulation and Adaptation 3562 Layer (SEAL)", draft-templin-intarea-seal-68 (work in 3563 progress), January 2014. 3565 [I-D.templin-intarea-vet] 3566 Templin, F., "Virtual Enterprise Traversal (VET)", draft- 3567 templin-intarea-vet-40 (work in progress), May 2013. 3569 [I-D.templin-ipwave-uam-its] 3570 Templin, F., "Urban Air Mobility Implications for 3571 Intelligent Transportation Systems", draft-templin-ipwave- 3572 uam-its-04 (work in progress), January 2021. 3574 [I-D.templin-ironbis] 3575 Templin, F., "The Interior Routing Overlay Network 3576 (IRON)", draft-templin-ironbis-16 (work in progress), 3577 March 2014. 3579 [I-D.templin-v6ops-pdhost] 3580 Templin, F., "IPv6 Prefix Delegation and Multi-Addressing 3581 Models", draft-templin-v6ops-pdhost-27 (work in progress), 3582 January 2021. 3584 [OVPN] OpenVPN, O., "http://openvpn.net", October 2016. 3586 [RFC1035] Mockapetris, P., "Domain names - implementation and 3587 specification", STD 13, RFC 1035, DOI 10.17487/RFC1035, 3588 November 1987, . 3590 [RFC1812] Baker, F., Ed., "Requirements for IP Version 4 Routers", 3591 RFC 1812, DOI 10.17487/RFC1812, June 1995, 3592 . 3594 [RFC2003] Perkins, C., "IP Encapsulation within IP", RFC 2003, 3595 DOI 10.17487/RFC2003, October 1996, 3596 . 3598 [RFC2004] Perkins, C., "Minimal Encapsulation within IP", RFC 2004, 3599 DOI 10.17487/RFC2004, October 1996, 3600 . 3602 [RFC2236] Fenner, W., "Internet Group Management Protocol, Version 3603 2", RFC 2236, DOI 10.17487/RFC2236, November 1997, 3604 . 3606 [RFC2464] Crawford, M., "Transmission of IPv6 Packets over Ethernet 3607 Networks", RFC 2464, DOI 10.17487/RFC2464, December 1998, 3608 . 3610 [RFC2529] Carpenter, B. and C. Jung, "Transmission of IPv6 over IPv4 3611 Domains without Explicit Tunnels", RFC 2529, 3612 DOI 10.17487/RFC2529, March 1999, 3613 . 3615 [RFC2983] Black, D., "Differentiated Services and Tunnels", 3616 RFC 2983, DOI 10.17487/RFC2983, October 2000, 3617 . 3619 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 3620 of Explicit Congestion Notification (ECN) to IP", 3621 RFC 3168, DOI 10.17487/RFC3168, September 2001, 3622 . 3624 [RFC3330] IANA, "Special-Use IPv4 Addresses", RFC 3330, 3625 DOI 10.17487/RFC3330, September 2002, 3626 . 3628 [RFC3810] Vida, R., Ed. and L. Costa, Ed., "Multicast Listener 3629 Discovery Version 2 (MLDv2) for IPv6", RFC 3810, 3630 DOI 10.17487/RFC3810, June 2004, 3631 . 3633 [RFC4122] Leach, P., Mealling, M., and R. Salz, "A Universally 3634 Unique IDentifier (UUID) URN Namespace", RFC 4122, 3635 DOI 10.17487/RFC4122, July 2005, 3636 . 3638 [RFC4251] Ylonen, T. and C. Lonvick, Ed., "The Secure Shell (SSH) 3639 Protocol Architecture", RFC 4251, DOI 10.17487/RFC4251, 3640 January 2006, . 3642 [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A 3643 Border Gateway Protocol 4 (BGP-4)", RFC 4271, 3644 DOI 10.17487/RFC4271, January 2006, 3645 . 3647 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 3648 Architecture", RFC 4291, DOI 10.17487/RFC4291, February 3649 2006, . 3651 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 3652 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 3653 December 2005, . 3655 [RFC4389] Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery 3656 Proxies (ND Proxy)", RFC 4389, DOI 10.17487/RFC4389, April 3657 2006, . 3659 [RFC4443] Conta, A., Deering, S., and M. Gupta, Ed., "Internet 3660 Control Message Protocol (ICMPv6) for the Internet 3661 Protocol Version 6 (IPv6) Specification", STD 89, 3662 RFC 4443, DOI 10.17487/RFC4443, March 2006, 3663 . 3665 [RFC4511] Sermersheim, J., Ed., "Lightweight Directory Access 3666 Protocol (LDAP): The Protocol", RFC 4511, 3667 DOI 10.17487/RFC4511, June 2006, 3668 . 3670 [RFC4541] Christensen, M., Kimball, K., and F. Solensky, 3671 "Considerations for Internet Group Management Protocol 3672 (IGMP) and Multicast Listener Discovery (MLD) Snooping 3673 Switches", RFC 4541, DOI 10.17487/RFC4541, May 2006, 3674 . 3676 [RFC4605] Fenner, B., He, H., Haberman, B., and H. Sandick, 3677 "Internet Group Management Protocol (IGMP) / Multicast 3678 Listener Discovery (MLD)-Based Multicast Forwarding 3679 ("IGMP/MLD Proxying")", RFC 4605, DOI 10.17487/RFC4605, 3680 August 2006, . 3682 [RFC4982] Bagnulo, M. and J. Arkko, "Support for Multiple Hash 3683 Algorithms in Cryptographically Generated Addresses 3684 (CGAs)", RFC 4982, DOI 10.17487/RFC4982, July 2007, 3685 . 3687 [RFC5015] Handley, M., Kouvelas, I., Speakman, T., and L. Vicisano, 3688 "Bidirectional Protocol Independent Multicast (BIDIR- 3689 PIM)", RFC 5015, DOI 10.17487/RFC5015, October 2007, 3690 . 3692 [RFC5214] Templin, F., Gleeson, T., and D. Thaler, "Intra-Site 3693 Automatic Tunnel Addressing Protocol (ISATAP)", RFC 5214, 3694 DOI 10.17487/RFC5214, March 2008, 3695 . 3697 [RFC5320] Templin, F., Ed., "The Subnetwork Encapsulation and 3698 Adaptation Layer (SEAL)", RFC 5320, DOI 10.17487/RFC5320, 3699 February 2010, . 3701 [RFC5522] Eddy, W., Ivancic, W., and T. Davis, "Network Mobility 3702 Route Optimization Requirements for Operational Use in 3703 Aeronautics and Space Exploration Mobile Networks", 3704 RFC 5522, DOI 10.17487/RFC5522, October 2009, 3705 . 3707 [RFC5558] Templin, F., Ed., "Virtual Enterprise Traversal (VET)", 3708 RFC 5558, DOI 10.17487/RFC5558, February 2010, 3709 . 3711 [RFC5569] Despres, R., "IPv6 Rapid Deployment on IPv4 3712 Infrastructures (6rd)", RFC 5569, DOI 10.17487/RFC5569, 3713 January 2010, . 3715 [RFC5880] Katz, D. and D. Ward, "Bidirectional Forwarding Detection 3716 (BFD)", RFC 5880, DOI 10.17487/RFC5880, June 2010, 3717 . 3719 [RFC6106] Jeong, J., Park, S., Beloeil, L., and S. Madanapalli, 3720 "IPv6 Router Advertisement Options for DNS Configuration", 3721 RFC 6106, DOI 10.17487/RFC6106, November 2010, 3722 . 3724 [RFC6146] Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful 3725 NAT64: Network Address and Protocol Translation from IPv6 3726 Clients to IPv4 Servers", RFC 6146, DOI 10.17487/RFC6146, 3727 April 2011, . 3729 [RFC6179] Templin, F., Ed., "The Internet Routing Overlay Network 3730 (IRON)", RFC 6179, DOI 10.17487/RFC6179, March 2011, 3731 . 3733 [RFC6221] Miles, D., Ed., Ooghe, S., Dec, W., Krishnan, S., and A. 3734 Kavanagh, "Lightweight DHCPv6 Relay Agent", RFC 6221, 3735 DOI 10.17487/RFC6221, May 2011, 3736 . 3738 [RFC6273] Kukec, A., Krishnan, S., and S. Jiang, "The Secure 3739 Neighbor Discovery (SEND) Hash Threat Analysis", RFC 6273, 3740 DOI 10.17487/RFC6273, June 2011, 3741 . 3743 [RFC6347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer 3744 Security Version 1.2", RFC 6347, DOI 10.17487/RFC6347, 3745 January 2012, . 3747 [RFC6355] Narten, T. and J. Johnson, "Definition of the UUID-Based 3748 DHCPv6 Unique Identifier (DUID-UUID)", RFC 6355, 3749 DOI 10.17487/RFC6355, August 2011, 3750 . 3752 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 3753 for Equal Cost Multipath Routing and Link Aggregation in 3754 Tunnels", RFC 6438, DOI 10.17487/RFC6438, November 2011, 3755 . 3757 [RFC6706] Templin, F., Ed., "Asymmetric Extended Route Optimization 3758 (AERO)", RFC 6706, DOI 10.17487/RFC6706, August 2012, 3759 . 3761 [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and 3762 UDP Checksums for Tunneled Packets", RFC 6935, 3763 DOI 10.17487/RFC6935, April 2013, 3764 . 3766 [RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement 3767 for the Use of IPv6 UDP Datagrams with Zero Checksums", 3768 RFC 6936, DOI 10.17487/RFC6936, April 2013, 3769 . 3771 [RFC7333] Chan, H., Ed., Liu, D., Seite, P., Yokota, H., and J. 3772 Korhonen, "Requirements for Distributed Mobility 3773 Management", RFC 7333, DOI 10.17487/RFC7333, August 2014, 3774 . 3776 [RFC7761] Fenner, B., Handley, M., Holbrook, H., Kouvelas, I., 3777 Parekh, R., Zhang, Z., and L. Zheng, "Protocol Independent 3778 Multicast - Sparse Mode (PIM-SM): Protocol Specification 3779 (Revised)", STD 83, RFC 7761, DOI 10.17487/RFC7761, March 3780 2016, . 3782 [RFC8402] Filsfils, C., Ed., Previdi, S., Ed., Ginsberg, L., 3783 Decraene, B., Litkowski, S., and R. Shakir, "Segment 3784 Routing Architecture", RFC 8402, DOI 10.17487/RFC8402, 3785 July 2018, . 3787 [RFC8446] Rescorla, E., "The Transport Layer Security (TLS) Protocol 3788 Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018, 3789 . 3791 [RFC8754] Filsfils, C., Ed., Dukes, D., Ed., Previdi, S., Leddy, J., 3792 Matsushima, S., and D. Voyer, "IPv6 Segment Routing Header 3793 (SRH)", RFC 8754, DOI 10.17487/RFC8754, March 2020, 3794 . 3796 [WG] Wireguard, "Wireguard, https://www.wireguard.com", August 3797 2020. 3799 Appendix A. Non-Normative Considerations 3801 AERO can be applied to a multitude of Internetworking scenarios, with 3802 each having its own adaptations. The following considerations are 3803 provided as non-normative guidance: 3805 A.1. Implementation Strategies for Route Optimization 3807 Route optimization as discussed in Section 3.14 results in the route 3808 optimization source (ROS) creating an asymmetric neighbor cache entry 3809 for the target neighbor. The neighbor cache entry is maintained for 3810 at most ReachableTime seconds and then deleted unless updated. In 3811 order to refresh the neighbor cache entry lifetime before the 3812 ReachableTime timer expires, the specification requires 3813 implementations to issue a new NS/NA exchange to reset ReachableTime 3814 while data packets are still flowing. However, the decision of when 3815 to initiate a new NS/NA exchange and to perpetuate the process is 3816 left as an implementation detail. 3818 One possible strategy may be to monitor the neighbor cache entry 3819 watching for data packets for (ReachableTime - 5) seconds. If any 3820 data packets have been sent to the neighbor within this timeframe, 3821 then send an NS to receive a new NA. If no data packets have been 3822 sent, wait for 5 additional seconds and send an immediate NS if any 3823 data packets are sent within this "expiration pending" 5 second 3824 window. If no additional data packets are sent within the 5 second 3825 window, delete the neighbor cache entry. 3827 The monitoring of the neighbor data packet traffic therefore becomes 3828 an asymmetric ongoing process during the neighbor cache entry 3829 lifetime. If the neighbor cache entry expires, future data packets 3830 will trigger a new NS/NA exchange while the packets themselves are 3831 delivered over a longer path until route optimization state is re- 3832 established. 3834 A.2. Implicit Mobility Management 3836 OMNI interface neighbors MAY provide a configuration option that 3837 allows them to perform implicit mobility management in which no ND 3838 messaging is used. In that case, the Client only transmits packets 3839 over a single interface at a time, and the neighbor always observes 3840 packets arriving from the Client from the same link-layer source 3841 address. 3843 If the Client's underlying interface address changes (either due to a 3844 readdressing of the original interface or switching to a new 3845 interface) the neighbor immediately updates the neighbor cache entry 3846 for the Client and begins accepting and sending packets according to 3847 the Client's new address. This implicit mobility method applies to 3848 use cases such as cellphones with both WiFi and Cellular interfaces 3849 where only one of the interfaces is active at a given time, and the 3850 Client automatically switches over to the backup interface if the 3851 primary interface fails. 3853 A.3. Direct Underlying Interfaces 3855 When a Client's OMNI interface is configured over a Direct interface, 3856 the neighbor at the other end of the Direct link can receive packets 3857 without any encapsulation. In that case, the Client sends packets 3858 over the Direct link according to QoS preferences. If the Direct 3859 interface has the highest QoS preference, then the Client's IP 3860 packets are transmitted directly to the peer without going through an 3861 ANET/INET. If other interfaces have higher QoS preferences, then the 3862 Client's IP packets are transmitted via a different interface, which 3863 may result in the inclusion of Proxys, Servers and Bridges in the 3864 communications path. Direct interfaces must be tested periodically 3865 for reachability, e.g., via NUD. 3867 A.4. AERO Critical Infrastructure Considerations 3869 AERO Bridges can be either Commercial off-the Shelf (COTS) standard 3870 IP routers or virtual machines in the cloud. Bridges must be 3871 provisioned, supported and managed by the INET administrative 3872 authority, and connected to the Bridges of other INETs via inter- 3873 domain peerings. Cost for purchasing, configuring and managing 3874 Bridges is nominal even for very large OMNI links. 3876 AERO Servers can be standard dedicated server platforms, but most 3877 often will be deployed as virtual machines in the cloud. The only 3878 requirements for Servers are that they can run the AERO user-level 3879 code and have at least one network interface connection to the INET. 3880 As with Bridges, Servers must be provisioned, supported and managed 3881 by the INET administrative authority. Cost for purchasing, 3882 configuring and managing Servers is nominal especially for virtual 3883 Servers hosted in the cloud. 3885 AERO Proxys are most often standard dedicated server platforms with 3886 one network interface connected to the ANET and a second interface 3887 connected to an INET. As with Servers, the only requirements are 3888 that they can run the AERO user-level code and have at least one 3889 interface connection to the INET. Proxys must be provisioned, 3890 supported and managed by the ANET administrative authority. Cost for 3891 purchasing, configuring and managing Proxys is nominal, and borne by 3892 the ANET administrative authority. 3894 AERO Relays can be any dedicated server or COTS router platform 3895 connected to INETs and/or EUNs. The Relay connects to the OMNI link 3896 and engages in eBGP peering with one or more Bridges as a stub AS. 3897 The Relay then injects its MNPs and/or non-MNP prefixes into the BGP 3898 routing system, and provisions the prefixes to its downstream- 3899 attached networks. The Relay can perform ROS/ROR services the same 3900 as for any Server, and can route between the MNP and non-MNP address 3901 spaces. 3903 A.5. AERO Server Failure Implications 3905 AERO Servers may appear as a single point of failure in the 3906 architecture, but such is not the case since all Servers on the link 3907 provide identical services and loss of a Server does not imply 3908 immediate and/or comprehensive communication failures. Although 3909 Clients typically associate with a single Server at a time, Server 3910 failure is quickly detected and conveyed by Bidirectional Forward 3911 Detection (BFD) and/or proactive NUD allowing Clients to migrate to 3912 new Servers. 3914 If a Server fails, ongoing packet forwarding to Clients will continue 3915 by virtue of the asymmetric neighbor cache entries that have already 3916 been established in route optimization sources (ROSs). If a Client 3917 also experiences mobility events at roughly the same time the Server 3918 fails, unsolicited NA messages may be lost but proxy neighbor cache 3919 entries in the DEPARTED state will ensure that packet forwarding to 3920 the Client's new locations will continue for up to DepartTime 3921 seconds. 3923 If a Client is left without a Server for an extended timeframe (e.g., 3924 greater than ReachableTime seconds) then existing asymmetric neighbor 3925 cache entries will eventually expire and both ongoing and new 3926 communications will fail. The original source will continue to 3927 retransmit until the Client has established a new Server 3928 relationship, after which time continuous communications will resume. 3930 Therefore, providing many Servers on the link with high availability 3931 profiles provides resilience against loss of individual Servers and 3932 assurance that Clients can establish new Server relationships quickly 3933 in event of a Server failure. 3935 A.6. AERO Client / Server Architecture 3937 The AERO architectural model is client / server in the control plane, 3938 with route optimization in the data plane. The same as for common 3939 Internet services, the AERO Client discovers the addresses of AERO 3940 Servers and selects one Server to connect to. The AERO service is 3941 analogous to common Internet services such as google.com, yahoo.com, 3942 cnn.com, etc. However, there is only one AERO service for the link 3943 and all Servers provide identical services. 3945 Common Internet services provide differing strategies for advertising 3946 server addresses to clients. The strategy is conveyed through the 3947 DNS resource records returned in response to name resolution queries. 3949 As of January 2020 Internet-based 'nslookup' services were used to 3950 determine the following: 3952 o When a client resolves the domainname "google.com", the DNS always 3953 returns one A record (i.e., an IPv4 address) and one AAAA record 3954 (i.e., an IPv6 address). The client receives the same addresses 3955 each time it resolves the domainname via the same DNS resolver, 3956 but may receive different addresses when it resolves the 3957 domainname via different DNS resolvers. But, in each case, 3958 exactly one A and one AAAA record are returned. 3960 o When a client resolves the domainname "ietf.org", the DNS always 3961 returns one A record and one AAAA record with the same addresses 3962 regardless of which DNS resolver is used. 3964 o When a client resolves the domainname "yahoo.com", the DNS always 3965 returns a list of 4 A records and 4 AAAA records. Each time the 3966 client resolves the domainname via the same DNS resolver, the same 3967 list of addresses are returned but in randomized order (i.e., 3968 consistent with a DNS round-robin strategy). But, interestingly, 3969 the same addresses are returned (albeit in randomized order) when 3970 the domainname is resolved via different DNS resolvers. 3972 o When a client resolves the domainname "amazon.com", the DNS always 3973 returns a list of 3 A records and no AAAA records. As with 3974 "yahoo.com", the same three A records are returned from any 3975 worldwide Internet connection point in randomized order. 3977 The above example strategies show differing approaches to Internet 3978 resilience and service distribution offered by major Internet 3979 services. The Google approach exposes only a single IPv4 and a 3980 single IPv6 address to clients. Clients can then select whichever IP 3981 protocol version offers the best response, but will always use the 3982 same IP address according to the current Internet connection point. 3983 This means that the IP address offered by the network must lead to a 3984 highly-available server and/or service distribution point. In other 3985 words, resilience is predicated on high availability within the 3986 network and with no client-initiated failovers expected (i.e., it is 3987 all-or-nothing from the client's perspective). However, Google does 3988 provide for worldwide distributed service distribution by virtue of 3989 the fact that each Internet connection point responds with a 3990 different IPv6 and IPv4 address. The IETF approach is like google 3991 (all-or-nothing from the client's perspective), but provides only a 3992 single IPv4 or IPv6 address on a worldwide basis. This means that 3993 the addresses must be made highly-available at the network level with 3994 no client failover possibility, and if there is any worldwide service 3995 distribution it would need to be conducted by a network element that 3996 is reached via the IP address acting as a service distribution point. 3998 In contrast to the Google and IETF philosophies, Yahoo and Amazon 3999 both provide clients with a (short) list of IP addresses with Yahoo 4000 providing both IP protocol versions and Amazon as IPv4-only. The 4001 order of the list is randomized with each name service query 4002 response, with the effect of round-robin load balancing for service 4003 distribution. With a short list of addresses, there is still 4004 expectation that the network will implement high availability for 4005 each address but in case any single address fails the client can 4006 switch over to using a different address. The balance then becomes 4007 one of function in the network vs function in the end system. 4009 The same implications observed for common highly-available services 4010 in the Internet apply also to the AERO client/server architecture. 4011 When an AERO Client connects to one or more ANETs, it discovers one 4012 or more AERO Server addresses through the mechanisms discussed in 4013 earlier sections. Each Server address presumably leads to a fault- 4014 tolerant clustering arrangement such as supported by Linux-HA, 4015 Extended Virtual Synchrony or Paxos. Such an arrangement has 4016 precedence in common Internet service deployments in lightweight 4017 virtual machines without requiring expensive hardware deployment. 4018 Similarly, common Internet service deployments set service IP 4019 addresses on service distribution points that may relay requests to 4020 many different servers. 4022 For AERO, the expectation is that a combination of the Google/IETF 4023 and Yahoo/Amazon philosophies would be employed. The AERO Client 4024 connects to different ANET access points and can receive 1-2 Server 4025 ADM-LLAs at each point. It then selects one AERO Server address, and 4026 engages in RS/RA exchanges with the same Server from all ANET 4027 connections. The Client remains with this Server unless or until the 4028 Server fails, in which case it can switch over to an alternate 4029 Server. The Client can likewise switch over to a different Server at 4030 any time if there is some reason for it to do so. So, the AERO 4031 expectation is for a balance of function in the network and end 4032 system, with fault tolerance and resilience at both levels. 4034 Appendix B. Change Log 4036 << RFC Editor - remove prior to publication >> 4038 Changes from draft-templin-intarea-6706bis-61 to draft-templin- 4039 intrea-6706bis-62: 4041 o New sub-section on OMNI Neighbor Interface Attributes 4043 Changes from draft-templin-intarea-6706bis-59 to draft-templin- 4044 intrea-6706bis-60: 4046 o Removed all references to S/TLLAO - all Interface Attributes are 4047 now maintained completely in the OMNI option. 4049 Changes from draft-templin-intarea-6706bis-58 to draft-templin- 4050 intrea-6706bis-59: 4052 o The term "Relay" used in older draft versions is now "Bridge". 4053 "Relay" now refers to what was formally called: "Gateway". 4055 o Fine-grained cleanup of Forwarding Algorithm; IPv6 ND message 4056 addressing; OMNI Prefix Lengths, etc. 4058 Changes from draft-templin-intarea-6706bis-54 to draft-templin- 4059 intrea-6706bis-55: 4061 o Updates on Segment Routing and S/TLLAO contents. 4063 o Various editorials and addressing cleanups. 4065 Changes from draft-templin-intarea-6706bis-52 to draft-templin- 4066 intrea-6706bis-53: 4068 o Normative reference to the OMNI spec, and remove portions that are 4069 already specified in OMNI. 4071 o Renamed "AERO interface/link" to "OMIN interface/link" throughout 4072 the document. 4074 o Truncated obsolete back section matter. 4076 Author's Address 4078 Fred L. Templin (editor) 4079 Boeing Research & Technology 4080 P.O. Box 3707 4081 Seattle, WA 98124 4082 USA 4084 Email: fltemplin@acm.org