idnits 2.17.1 draft-templin-intarea-6706bis-96.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 (March 19, 2021) is 1133 days in the past. Is this intentional? -- Found something which looks like a code comment -- if you have code sections in the document, please surround them with '' and '' lines. 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 1063, but not defined == Missing Reference: 'ULA0' is mentioned on line 1063, but not defined == Missing Reference: 'ULA1' is mentioned on line 1063, but not defined == Missing Reference: 'ULA2' is mentioned on line 1063, but not defined == Unused Reference: 'RFC3971' is defined on line 3502, but no explicit reference was found in the text == Unused Reference: 'RFC3972' is defined on line 3507, but no explicit reference was found in the text == Unused Reference: 'RFC4193' is defined on line 3515, but no explicit reference was found in the text == Unused Reference: 'BGP' is defined on line 3564, but no explicit reference was found in the text == Unused Reference: 'I-D.bonica-6man-comp-rtg-hdr' is defined on line 3567, but no explicit reference was found in the text == Unused Reference: 'I-D.bonica-6man-crh-helper-opt' is defined on line 3573, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-intarea-frag-fragile' is defined on line 3578, but no explicit reference was found in the text == Unused Reference: 'I-D.templin-6man-dhcpv6-ndopt' is defined on line 3601, but no explicit reference was found in the text == Unused Reference: 'RFC2003' is defined on line 3640, but no explicit reference was found in the text == Unused Reference: 'RFC4122' is defined on line 3679, but no explicit reference was found in the text == Unused Reference: 'RFC4982' is defined on line 3728, but no explicit reference was found in the text == Unused Reference: 'RFC6273' is defined on line 3784, but no explicit reference was found in the text == Unused Reference: 'RFC6355' is defined on line 3793, but no explicit reference was found in the text == Outdated reference: A later version (-99) exists of draft-templin-6man-omni-interface-69 ** Downref: Normative reference to an Informational RFC: RFC 7739 == 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: 1 error (**), 0 flaws (~~), 25 warnings (==), 5 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, March 19, 2021 5 rfc6179, rfc6706 (if 6 approved) 7 Intended status: Standards Track 8 Expires: September 20, 2021 10 Automatic Extended Route Optimization (AERO) 11 draft-templin-intarea-6706bis-96 13 Abstract 15 This document specifies an Automatic 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 September 20, 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. Automatic Extended Route Optimization (AERO) . . . . . . . . 12 65 3.1. AERO Node Types . . . . . . . . . . . . . . . . . . . . . 12 66 3.2. The AERO Service over OMNI Links . . . . . . . . . . . . 13 67 3.2.1. AERO/OMNI Reference Model . . . . . . . . . . . . . . 13 68 3.2.2. Addressing and Node Identification . . . . . . . . . 16 69 3.2.3. AERO Routing System . . . . . . . . . . . . . . . . . 17 70 3.2.4. OMNI Link Segment Routing . . . . . . . . . . . . . . 19 71 3.2.5. Segment Routing Topologies (SRTs) . . . . . . . . . . 23 72 3.2.6. Segment Routing For OMNI Link Selection . . . . . . . 24 73 3.2.7. Segment Routing Within the OMNI Link . . . . . . . . 24 74 3.3. OMNI Interface Characteristics . . . . . . . . . . . . . 25 75 3.4. OMNI Interface Initialization . . . . . . . . . . . . . . 27 76 3.4.1. AERO Server/Relay Behavior . . . . . . . . . . . . . 27 77 3.4.2. AERO Proxy Behavior . . . . . . . . . . . . . . . . . 27 78 3.4.3. AERO Client Behavior . . . . . . . . . . . . . . . . 28 79 3.4.4. AERO Bridge Behavior . . . . . . . . . . . . . . . . 28 80 3.5. OMNI Interface Neighbor Cache Maintenance . . . . . . . . 28 81 3.5.1. OMNI Neighbor Interface Attributes . . . . . . . . . 30 82 3.5.2. OMNI Neighbor Advertisement Message Flags . . . . . . 30 83 3.6. OMNI Interface Encapsulation and Re-encapsulation . . . . 31 84 3.7. OMNI Interface Decapsulation . . . . . . . . . . . . . . 31 85 3.8. OMNI Interface Data Origin Authentication . . . . . . . . 32 86 3.9. 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/Server Forwarding Algorithm . . . . . . . . . 35 90 3.10.3. Bridge Forwarding Algorithm . . . . . . . . . . . . 37 91 3.11. OMNI Interface Error Handling . . . . . . . . . . . . . . 38 92 3.12. AERO Router Discovery, Prefix Delegation and 93 Autoconfiguration . . . . . . . . . . . . . . . . . . . . 41 94 3.12.1. AERO Service Model . . . . . . . . . . . . . . . . . 41 95 3.12.2. AERO Client Behavior . . . . . . . . . . . . . . . . 42 96 3.12.3. AERO Server Behavior . . . . . . . . . . . . . . . . 43 97 3.13. The AERO Proxy . . . . . . . . . . . . . . . . . . . . . 46 98 3.13.1. Combined Proxy/Servers . . . . . . . . . . . . . . . 49 99 3.13.2. Detecting and Responding to Server Failures . . . . 49 100 3.13.3. Point-to-Multipoint Server Coordination . . . . . . 50 101 3.14. AERO Address Resolution . . . . . . . . . . . . . . . . . 51 102 3.14.1. Route Optimization Initiation . . . . . . . . . . . 51 103 3.14.2. Relaying the NS(AR) *NET Packet(s) . . . . . . . . . 52 104 3.14.3. Processing the NS(AR) and Sending the NA(AR) . . . . 52 105 3.14.4. Relaying the NA(AR) . . . . . . . . . . . . . . . . 54 106 3.14.5. Processing the NA(AR) . . . . . . . . . . . . . . . 54 107 3.14.6. Route Optimization Maintenance . . . . . . . . . . . 54 108 3.15. Neighbor Unreachability Detection (NUD) . . . . . . . . . 56 109 3.16. Mobility Management and Quality of Service (QoS) . . . . 57 110 3.16.1. Mobility Update Messaging . . . . . . . . . . . . . 58 111 3.16.2. Announcing Link-Layer Address and/or QoS Preference 112 Changes . . . . . . . . . . . . . . . . . . . . . . 59 113 3.16.3. Bringing New Links Into Service . . . . . . . . . . 59 114 3.16.4. Deactivating Existing Links . . . . . . . . . . . . 60 115 3.16.5. Moving Between Servers . . . . . . . . . . . . . . . 60 116 3.17. Multicast . . . . . . . . . . . . . . . . . . . . . . . . 61 117 3.17.1. Source-Specific Multicast (SSM) . . . . . . . . . . 62 118 3.17.2. Any-Source Multicast (ASM) . . . . . . . . . . . . . 63 119 3.17.3. Bi-Directional PIM (BIDIR-PIM) . . . . . . . . . . . 64 120 3.18. Operation over Multiple OMNI Links . . . . . . . . . . . 64 121 3.19. DNS Considerations . . . . . . . . . . . . . . . . . . . 65 122 3.20. Transition Considerations . . . . . . . . . . . . . . . . 65 123 3.21. Detecting and Reacting to Server and Bridge Failures . . 66 124 3.22. AERO Clients on the Open Internet . . . . . . . . . . . . 66 125 3.23. Time-Varying MNPs . . . . . . . . . . . . . . . . . . . . 70 126 4. Implementation Status . . . . . . . . . . . . . . . . . . . . 70 127 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 71 128 6. Security Considerations . . . . . . . . . . . . . . . . . . . 71 129 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 73 130 8. References . . . . . . . . . . . . . . . . . . . . . . . . . 74 131 8.1. Normative References . . . . . . . . . . . . . . . . . . 74 132 8.2. Informative References . . . . . . . . . . . . . . . . . 76 133 Appendix A. Non-Normative Considerations . . . . . . . . . . . . 82 134 A.1. Implementation Strategies for Route Optimization . . . . 82 135 A.2. Implicit Mobility Management . . . . . . . . . . . . . . 83 136 A.3. Direct Underlying Interfaces . . . . . . . . . . . . . . 83 137 A.4. AERO Critical Infrastructure Considerations . . . . . . . 83 138 A.5. AERO Server Failure Implications . . . . . . . . . . . . 84 139 A.6. AERO Client / Server Architecture . . . . . . . . . . . . 85 140 Appendix B. Change Log . . . . . . . . . . . . . . . . . . . . . 87 141 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 88 143 1. Introduction 145 Automatic Extended Route Optimization (AERO) fulfills the 146 requirements of Distributed Mobility Management (DMM) [RFC7333] and 147 route optimization [RFC5522] for aeronautical networking and other 148 network mobility use cases such as intelligent transportation 149 systems. AERO is an internetworking and mobility management service 150 that employs the Overlay Multilink Network Interface (OMNI) 151 [I-D.templin-6man-omni-interface] Non-Broadcast, Multiple Access 152 (NBMA) virtual link model. The OMNI link is a virtual overlay 153 configured over one or more underlying Internetworks, and nodes on 154 the link can exchange original IP packets as single-hop neighbors via 155 encapsulation and fragmentation. The OMNI Adaptation Layer (OAL) 156 supports multilink operation for increased reliability, bandwidth 157 optimization and traffic path selection while performing 158 fragmentation and reassembly to accommodate Maximum Transmission Unit 159 (MTU) diversity. 161 The AERO service comprises Clients, Proxys, Servers and Relays that 162 are seen as OMNI link neighbors as well as Bridges that interconnect 163 OMNI link segments through OAL forwarding at a layer below IP. Each 164 node's OMNI interface uses an IPv6 link-local address format that 165 supports operation of the IPv6 Neighbor Discovery (ND) protocol 166 [RFC4861] and links ND to IP forwarding. A node's OMNI interface can 167 be configured over multiple underlying interfaces, and therefore 168 appears as a single interface with multiple link-layer addresses. 169 Each link-layer address is subject to change due to mobility and/or 170 QoS fluctuations, and link-layer address changes are signaled by ND 171 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 original IP packets destined to other AERO nodes via 177 the nearest Relay, which forwards them through the cloud. A mobile 178 node's initial packets are forwarded through the Server, and direct 179 routing is supported through automatic extended route optimization 180 while packets are flowing. Both unicast and multicast communications 181 are supported, and mobile nodes may efficiently move between 182 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 to provide an OAL routing/bridging service that 187 joins the underlying Internetworks of multiple disjoint 188 administrative domains into a single unified OMNI link at a layer 189 below IP. Each OMNI link instance is characterized by the set of 190 Mobility Service Prefixes (MSPs) common to all mobile nodes. The 191 link extends to the point where a Relay/Server is on the optimal 192 route from any correspondent node on the link, and provides a conduit 193 between the underlying Internetwork and the OMNI link. To the 194 underlying Internetwork, the Relay/Server is the source of a route to 195 the MSP, and hence uplink traffic to the mobile node is naturally 196 routed to 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 (e.g., "protected 254 spectrum") to prevent unauthorized access internally and with 255 border network-layer security services such as firewalls and 256 proxys that prevent 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 interface 271 a node's attachment to a link in an INET. 273 *NET 274 a "wildcard" term referring to either ANET or INET when it is not 275 necessary to draw a distinction between the two. 277 *NET interface 278 a node's attachment to a link in a *NET. 280 *NET Partition 281 frequently, *NETs such as large corporate enterprise networks are 282 sub-divided internally into separate isolated partitions. Each 283 partition is fully connected internally but disconnected from 284 other partitions, and there is no requirement that separate 285 partitions maintain consistent Internet Protocol and/or addressing 286 plans. (Each *NET partition is seen as a separate OMNI link 287 segment as discussed below.) 289 *NET address 290 an IP address assigned to a node's interface connection to a *NET. 292 *NET encapsulation 293 the encapsulation of a packet in an outer header or headers that 294 can be routed within the scope of the local *NET partition. 296 OMNI link 297 the same as defined in [I-D.templin-6man-omni-interface], and 298 manifested by IPv6 encapsulation [RFC2473]. The OMNI link spans 299 underlying INET segments joined by virtual bridges in a spanning 300 tree the same as a bridged campus LAN. AERO nodes on the OMNI 301 link appear as single-hop neighbors even though they may be 302 separated by multiple underlying INET hops, and can use Segment 303 Routing [RFC8402] to cause packets to visit selected waypoints on 304 the link. 306 OMNI Interface 307 a node's attachment to an OMNI link. Since the addresses assigned 308 to an OMNI interface are managed for uniqueness, OMNI interfaces 309 do not require Duplicate Address Detection (DAD) and therefore set 310 the administrative variable 'DupAddrDetectTransmits' to zero 311 [RFC4862]. 313 OMNI Adaptation Layer (OAL) 314 an OMNI interface process whereby original IP packets admitted 315 into the interface are wrapped in a mid-layer IPv6 header and 316 subject to fragmentation and reassembly. The OAL is also 317 responsible for generating MTU-related control messages as 318 necessary, and for providing addressing context for spanning 319 multiple segments of a bridged OMNI link. 321 original IP packet 322 a whole IP packet or fragment admitted into the OMNI interface by 323 the network layer prior to OAL encapsulation and fragmentation, or 324 an IP packet delivered to the network layer by the OMNI interface 325 following OAL decapsulation and reassembly. 327 OAL packet 328 an original IP packet encapsulated in OAL headers and trailers 329 before OAL fragmentation, or following OAL reassembly. 331 OAL fragment 332 a portion of an OAL packet following fragmentation but prior to 333 *NET encapsulation, or following *NET encapsulation but prior to 334 OAL reassembly. 336 (OAL) carrier packet 337 an encapsulated OAL fragment following *NET encapsulation or prior 338 to *NET decapsulation. OAL sources and destinations exchange 339 carrier packets over underlying interfaces, and may be separated 340 by one or more OAL intermediate nodes. OAL intermediate nodes may 341 perform re-encapsulation on carrier packets by removing the *NET 342 headers of the first hop network and replacing them with new *NET 343 headers for the next hop network. 345 OAL source 346 an OMNI interface acts as an OAL source when it encapsulates 347 original IP packets to form OAL packets, then performs OAL 348 fragmentation and *NET encapsulation to create carrier packets. 350 OAL destination 351 an OMNI interface acts as an OAL destination when it decapsulates 352 carrier packets, then performs OAL reassembly and decapsulation to 353 derive the original IP packet. 355 OAL intermediate node 356 an OMNI interface acts as an OAL intermediate node when it removes 357 the *NET headers of carrier packes received on a first segment, 358 then re-encapsulates the carrier packets in new *NET headers and 359 forwards them into the next segment. 361 underlying interface 362 a *NET interface over which an OMNI interface is configured. 364 Mobility Service Prefix (MSP) 365 an aggregated IP Global Unicast Address (GUA) prefix (e.g., 366 2001:db8::/32, 192.0.2.0/24, etc.) assigned to the OMNI link and 367 from which more-specific Mobile Network Prefixes (MNPs) are 368 delegated. OMNI link administrators typically obtain MSPs from an 369 Internet address registry, however private-use prefixes can 370 alternatively be used subject to certain limitations (see: 371 [I-D.templin-6man-omni-interface]). OMNI links that connect to 372 the global Internet advertise their MSPs to their interdomain 373 routing peers. 375 Mobile Network Prefix (MNP) 376 a longer IP prefix delegated from an MSP (e.g., 377 2001:db8:1000:2000::/56, 192.0.2.8/30, etc.) and delegated to an 378 AERO Client or Relay. 380 Mobile Network Prefix Link Local Address (MNP-LLA) 381 an IPv6 Link Local Address that embeds the most significant 64 382 bits of an MNP in the lower 64 bits of fe80::/64, as specified in 383 [I-D.templin-6man-omni-interface]. 385 Mobile Network Prefix Unique Local Address (MNP-ULA) 386 an IPv6 Unique-Local Address derived from an MNP-LLA. 388 Administrative Link Local Address (ADM-LLA) 389 an IPv6 Link Local Address that embeds a 32-bit administratively- 390 assigned identification value in the lower 32 bits of fe80::/96, 391 as specified in [I-D.templin-6man-omni-interface]. 393 Administrative Unique Local Address (ADM-ULA) 394 an IPv6 Unique-Local Address derived from an ADM-LLA. 396 AERO node 397 a node that is connected to an OMNI link and participates in the 398 AERO internetworking and mobility service. 400 AERO Client ("Client") 401 an AERO node that connects over one or more underlying interfaces 402 and requests MNP delegation/registration service from AERO 403 Servers. The Client assigns an MNP-LLA to the OMNI interface for 404 use in ND exchanges with other AERO nodes and forwards original IP 405 packets to correspondents according to OMNI interface neighbor 406 cache state. 408 AERO Server ("Server") 409 an INET node that configures an OMNI interface to provide default 410 forwarding and Mobility Anchor Point (MAP) services for AERO 411 Clients. The Server assigns an ADM-LLA to its OMNI interface to 412 support the operation of the ND services, and advertises all of 413 its associated MNPs via BGP peerings with Bridges. 415 AERO Relay ("Relay") 416 an AERO Server that also provides forwarding services between 417 nodes reached via the OMNI link and correspondents on other links. 418 AERO Relays are provisioned with MNPs (i.e., the same as for an 419 AERO Client) and run a dynamic routing protocol to discover any 420 non-MNP IP GUA routes in service on its connected INET links. In 421 both cases, the Relay advertises the MSP(s) to its downstream 422 networks, and distributes all of its associated MNPs and non-MNP 423 IP GUA routes via BGP peerings with Bridges (i.e., the same as for 424 an AERO Server). 426 AERO Bridge ("Bridge") 427 a node that provides hybrid routing/bridging services (as well as 428 a security trust anchor) for nodes on an OMNI link. The Bridge 429 forwards carrier packets between OMNI link segments as OAL 430 intermediate nodes without decrementing the network layer IP TTL/ 431 Hop Limit. AERO Bridges peer with Servers and other Bridges to 432 discover the full set of MNPs for the link as well as any non-MNP 433 IP GUA routes that are reachable via Relays. 435 AERO Proxy ("Proxy") 436 a node that provides proxying services between Clients in an ANET 437 and Servers in external INETs. The AERO Proxy is a conduit 438 between the ANET and external INETs in the same manner as for 439 common web proxys, and behaves in a similar fashion as for ND 440 proxys [RFC4389]. A node may be configured to act as either a 441 Proxy and/or a Server, depending on Client Server selection 442 criteria. 444 ingress tunnel endpoint (ITE) 445 an OMNI interface endpoint that injects encapsulated packets into 446 an OMNI link. 448 egress tunnel endpoint (ETE) 449 an OMNI interface endpoint that receives encapsulated packets from 450 an OMNI link. 452 link-layer address 453 an IP address used as an encapsulation header source or 454 destination address from the perspective of the OMNI interface. 455 When an upper layer protocol (e.g., UDP) is used as part of the 456 encapsulation, the port number is also considered as part of the 457 link-layer address. 459 network layer address 460 the source or destination address of an original IP packet 461 presented to the OMNI interface. 463 end user network (EUN) 464 an internal virtual or external edge IP network that an AERO 465 Client or Relay connects to the rest of the network via the OMNI 466 interface. The Client/Relay sees each EUN as a "downstream" 467 network, and sees the OMNI interface as the point of attachment to 468 the "upstream" network. 470 Mobile Node (MN) 471 an AERO Client and all of its downstream-attached networks that 472 move together as a single unit, i.e., an end system that connects 473 an Internet of Things. 475 Mobile Router (MR) 476 a MN's on-board router that forwards origianl IP packets between 477 any downstream-attached networks and the OMNI link. The MR is the 478 MN entity that hosts the AERO Client. 480 Route Optimization Source (ROS) 481 the AERO node nearest the source that initiates route 482 optimization. The ROS may be a Server or Proxy acting on behalf 483 of the source Client, or may be the Client itself if the Client is 484 connected to the INET either directly or through a NAT. 486 Route Optimization responder (ROR) 487 the AERO node nearest the target destination that responds to 488 route optimization requests. The ROR may be a Proxy/Server acting 489 on behalf of a target MNP Client, a Relay for a non-MNP 490 destination or may be the target Client itself. 492 MAP List 493 a geographically and/or topologically referenced list of addresses 494 of all Servers within the same OMNI link. There is a single MAP 495 list for the entire OMNI link. 497 Distributed Mobility Management (DMM) 498 a BGP-based overlay routing service coordinated by Servers and 499 Bridges that tracks all Server-to-Client associations. 501 Mobility Service (MS) 502 the collective set of all Servers, Proxys, Bridges and Relays that 503 provide the AERO Service to Clients. 505 Mobility Service Endpoint MSE) 506 an individual Server, Proxy, Bridge or Relay in the Mobility 507 Service. 509 Throughout the document, the simple terms "Client", "Server", 510 "Bridge", "Proxy" and "Relay" refer to "AERO Client", "AERO Server", 511 "AERO Bridge", "AERO Proxy" and "AERO Relay", respectively. 512 Capitalization is used to distinguish these terms from other common 513 Internetworking uses in which they appear without capitalization. 515 The terminology of DHCPv6 [RFC8415] and IPv6 ND [RFC4861] (including 516 the names of node variables, messages and protocol constants) is used 517 throughout this document. The terms "All-Routers multicast", "All- 518 Nodes multicast", "Solicited-Node multicast" and "Subnet-Router 519 anycast" are defined in [RFC4291]. Also, the term "IP" is used to 520 generically refer to either Internet Protocol version, i.e., IPv4 521 [RFC0791] or IPv6 [RFC8200]. 523 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 524 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 525 "OPTIONAL" in this document are to be interpreted as described in BCP 526 14 [RFC2119][RFC8174] when, and only when, they appear in all 527 capitals, as shown here. 529 3. Automatic Extended Route Optimization (AERO) 531 The following sections specify the operation of IP over OMNI links 532 using the AERO service: 534 3.1. AERO Node Types 536 AERO Clients are Mobile Nodes (MNs) that configure OMNI interfaces 537 over underlying interfaces with addresses that may change when the 538 Client moves to a new network connection point. AERO Clients 539 register their Mobile Network Prefixes (MNPs) with the AERO service, 540 and distribute the MNPs to nodes on EUNs. AERO Bridges, Servers, 541 Proxys and Relays are critical infrastructure elements in fixed 542 (i.e., non-mobile) INET deployments and hence have permanent and 543 unchanging INET addresses. Together, they constitute the AERO 544 service which provides an OMNI link virtual overlay for connecting 545 AERO Clients. 547 AERO Bridges provide hybrid routing/bridging services (as well as a 548 security trust anchor) for nodes on an OMNI link. Bridges use 549 standard IPv6 routing to forward carrier packets both within the same 550 *NET partition and between disjoint *NET partitions based on an IPv6 551 encapsulation mid-layer known as the OMNI Adaptation Layer (OAL) 552 [I-D.templin-6man-omni-interface]. During forwarding, the inner IP 553 layer experiences a virtual bridging service since the inner IP TTL/ 554 Hop Limit is not decremented. Each Bridge also peers with Servers 555 and other Bridges in a dynamic routing protocol instance to provide a 556 Distributed Mobility Management (DMM) service for the list of active 557 MNPs (see Section 3.2.3). Bridges present the OMNI link as a set of 558 one or more Mobility Service Prefixes (MSPs) and configure secured 559 tunnels with Servers, Relays, Proxys and other Bridges; they further 560 maintain IP forwarding table entries for each MNP and any other 561 reachable non-MNP prefixes. 563 AERO Servers in distributed INET locations provide default forwarding 564 and mobility/multilink services for AERO Client Mobile Nodes (MNs). 565 Each Server also peers with Bridges in a dynamic routing protocol 566 instance to advertise its list of associated MNPs (see 567 Section 3.2.3). Servers facilitate prefix delegation/registration 568 exchanges with Clients, where each delegated prefix becomes an MNP 569 taken from an MSP. Servers forward carrier packets between OMNI 570 interface neighbors and track each Client's mobility profiles. 572 Servers may further act as Servers for some sets of Clients and as 573 Proxys for others. 575 AERO Proxys provide a conduit for ANET Clients to associate with 576 Servers in external INETs. Client and Servers exchange control plane 577 messages via the Proxy acting as a bridge between the ANET/INET 578 boundary. The Proxy forwards original IP packets between Clients and 579 the OMNI link according to forwarding information in the neighbor 580 cache. The Proxy function is specified in Section 3.13. Proxys may 581 further act as Proxys for some sets of Clients and as Servers for 582 others. 584 AERO Relays are Servers that provide forwarding services to exchange 585 original IP packets between the OMNI interface and INET/EUN 586 interfaces. Relays are provisioned with MNPs the same as for an AERO 587 Client, and also run a dynamic routing protocol to discover any non- 588 MNP IP routes. The Relay advertises the MSP(s) to its connected 589 networks, and distributes all of its associated MNPs and non-MNP 590 routes via BGP peerings with Bridges 592 3.2. The AERO Service over OMNI Links 594 3.2.1. AERO/OMNI Reference Model 596 Figure 1 presents the basic OMNI link reference model: 598 +----------------+ 599 | AERO Bridge B1 | 600 | Nbr: S1, S2, P1| 601 |(X1->S1; X2->S2)| 602 | MSP M1 | 603 +-+---------+--+-+ 604 +--------------+ | Secured | | +--------------+ 605 |AERO Server S1| | tunnels | | |AERO Server S2| 606 | Nbr: C1, B1 +-----+ | +-----+ Nbr: C2, B1 | 607 | default->B1 | | | default->B1 | 608 | X1->C1 | | | X2->C2 | 609 +-------+------+ | +------+-------+ 610 | OMNI link | | 611 X===+===+===================+==)===============+===+===X 612 | | | | 613 +-----+--------+ +--------+--+-----+ +--------+-----+ 614 |AERO Client C1| | AERO Proxy P1 | |AERO Client C2| 615 | Nbr: S1 | |(Proxy Nbr Cache)| | Nbr: S2 | 616 | default->S1 | +--------+--------+ | default->S2 | 617 | MNP X1 | | | MNP X2 | 618 +------+-------+ .--------+------. +-----+--------+ 619 | (- Proxyed Clients -) | 620 .-. `---------------' .-. 621 ,-( _)-. ,-( _)-. 622 .-(_ IP )-. +-------+ +-------+ .-(_ IP )-. 623 (__ EUN )--|Host H1| |Host H2|--(__ EUN ) 624 `-(______)-' +-------+ +-------+ `-(______)-' 626 Figure 1: AERO/OMNI Reference Model 628 In this model: 630 o the OMNI link is an overlay network service configured over one or 631 more underlying *NET partitions which may be managed by different 632 administrative authorities and have incompatible protocols and/or 633 addressing plans. 635 o AERO Bridge B1 aggregates Mobility Service Prefix (MSP) M1, 636 discovers Mobile Network Prefixes (MNPs) X* and advertises the MSP 637 via BGP peerings over secured tunnels to Servers (S1, S2). 638 Bridges connect the disjoint segments of a partitioned OMNI link. 640 o AERO Servers/Relays S1 and S2 configure secured tunnels with 641 Bridge B1 and also provide mobility, multilink and default router 642 services for their associated Clients C1 and C2. 644 o AERO Clients C1 and C2 associate with Servers S1 and S2, 645 respectively. They receive Mobile Network Prefix (MNP) 646 delegations X1 and X2, and also act as default routers for their 647 associated physical or internal virtual EUNs. Simple hosts H1 and 648 H2 attach to the EUNs served by Clients C1 and C2, respectively. 650 o AERO Proxy P1 configures a secured tunnel with Bridge B1 and 651 provides proxy services for AERO Clients in secured enclaves that 652 cannot associate directly with other OMNI link neighbors. 654 An OMNI link configured over a single *NET appears as a single 655 unified link with a consistent underlying network addressing plan. 656 In that case, all nodes on the link can exchange carrier packets via 657 simple *NET encapsulation, since the underlying *NET is connected. 658 In common practice, however, an OMNI link may be partitioned into 659 multiple "segments", where each segment is a distinct *NET 660 potentially managed under a different administrative authority (e.g., 661 as for worldwide aviation service providers such as ARINC, SITA, 662 Inmarsat, etc.). Individual *NETs may also themselves be partitioned 663 internally, in which case each internal partition is seen as a 664 separate segment. 666 The addressing plan of each segment is consistent internally but will 667 often bear no relation to the addressing plans of other segments. 668 Each segment is also likely to be separated from others by network 669 security devices (e.g., firewalls, proxys, packet filtering gateways, 670 etc.), and in many cases disjoint segments may not even have any 671 common physical link connections. Therefore, nodes can only be 672 assured of exchanging carrier packets directly with correspondents in 673 the same segment, and not with those in other segments. The only 674 means for joining the segments therefore is through inter-domain 675 peerings between AERO Bridges. 677 The same as for traditional campus LANs, multiple OMNI link segments 678 can be joined into a single unified link via a virtual bridging 679 service using the OMNI Adaptation Layer (OAL) 680 [I-D.templin-6man-omni-interface] which inserts a mid-layer IPv6 681 encapsulation header that supports inter-segment forwarding (i.e., 682 bridging) without decrementing the network-layer TTL/Hop Limit. This 683 bridging of OMNI link segments is shown in Figure 2: 685 . . . . . . . . . . . . . . . . . . . . . . . 686 . . 687 . .-(::::::::) . 688 . .-(::::::::::::)-. +-+ . 689 . (:::: Segment A :::)--|B|---+ . 690 . `-(::::::::::::)-' +-+ | . 691 . `-(::::::)-' | . 692 . | . 693 . .-(::::::::) | . 694 . .-(::::::::::::)-. +-+ | . 695 . (:::: Segment B :::)--|B|---+ . 696 . `-(::::::::::::)-' +-+ | . 697 . `-(::::::)-' | . 698 . | . 699 . .-(::::::::) | . 700 . .-(::::::::::::)-. +-+ | . 701 . (:::: Segment C :::)--|B|---+ . 702 . `-(::::::::::::)-' +-+ | . 703 . `-(::::::)-' | . 704 . | . 705 . ..(etc).. x . 706 . . 707 . . 708 . <- OMNI link Bridged by encapsulation -> . 709 . . . . . . . . . . . . . .. . . . . . . . . 711 Figure 2: Bridging OMNI Link Segments 713 Bridges, Servers, Relays and Proxys connect via secured INET tunnels 714 over their respective segments in a spanning tree topology rooted at 715 the Bridges. The secured spanning tree supports strong 716 authentication for IPv6 ND control messages and may also be used to 717 convey the initial carrier packets in a flow. Route optimization can 718 then be employed to cause carrier packets to take more direct paths 719 between OMNI link neighbors without having to strictly follow the 720 spanning tree. 722 3.2.2. Addressing and Node Identification 724 AERO nodes on OMNI links use the Link-Local Address (LLA) prefix 725 fe80::/64 [RFC4291] to assign LLAs used for network-layer addresses 726 in link-scoped IPv6 ND and data messages. AERO Clients use LLAs 727 constructed from MNPs (i.e., "MNP-LLAs") while other AERO nodes use 728 LLAs constructed from administrative identification values ("ADM- 729 LLAs") as specified in [I-D.templin-6man-omni-interface]. 731 AERO nodes also use the Unique Local Address (ULA) prefix fd00::/8 732 followed by a pseudo-random 40-bit OMNI domain identifier to form the 733 prefix [ULA]::/48, then include a 16-bit OMNI link identifier '*' to 734 form the prefix [ULA*]::/64 [RFC4291]. The AERO node then uses the 735 prefix [ULA*]::/64 to form "MNP-ULAs" or "ADM-ULA"s as specified in 736 [I-D.templin-6man-omni-interface] to support OAL addressing. AERO 737 Clients also use Temporary ULAs constructed per 738 [I-D.templin-6man-omni-interface], where the addresses are typically 739 used only in initial control message exchanges until a stable MNP- 740 LLA/ULA is assigned. 742 AERO MSPs, MNPs and non-MNP routes are typically based on Global 743 Unicast Addresses (GUAs), but in some cases may be based on private- 744 use addresses. See [I-D.templin-6man-omni-interface] for a full 745 specification of LLAs, ULAs and GUAs used by AERO nodes on OMNI 746 links. 748 Finally, AERO Clients and Servers configure node identification 749 values as specified in [I-D.templin-6man-omni-interface]. 751 3.2.3. AERO Routing System 753 The AERO routing system comprises a private instance of the Border 754 Gateway Protocol (BGP) [RFC4271] that is coordinated between Bridges 755 and Servers and does not interact with either the public Internet BGP 756 routing system or any underlying INET routing systems. 758 In a reference deployment, each Server is configured as an Autonomous 759 System Border Router (ASBR) for a stub Autonomous System (AS) using a 760 32-bit AS Number (ASN) [RFC4271] that is unique within the BGP 761 instance, and each Server further uses eBGP to peer with one or more 762 Bridges but does not peer with other Servers. Each *NET of a multi- 763 segment OMNI link must include one or more Bridges, which peer with 764 the Servers and Proxys within that *NET. All Bridges within the same 765 *NET are members of the same hub AS, and use iBGP to maintain a 766 consistent view of all active routes currently in service. The 767 Bridges of different *NETs peer with one another using eBGP. 769 Bridges maintain forwarding table entries only for the ULAs 770 correseponding to MNP and non-MNP routes that are currently active, 771 and carrier packets destined to all other ULAs will correctly incur 772 Destination Unreachable messages due to the black-hole route. In 773 this way, Servers and Relays have only partial topology knowledge 774 (i.e., they know only about the routes their directly associated 775 Clients and non-AERO links) and they forward all other carrier 776 packets to Bridges which have full topology knowledge. 778 Each OMNI link segment assigns a unique ADM-ULA sub-prefix of 779 [ULA*]::/96. For example, a first segment could assign 780 [ULA*]::1000/116, a second could assign [ULA*]::2000/116, a third 781 could assign [ULA*]::3000/116, etc. Within each segment, each Proxy/ 782 Server configures an ADM-ULA within the segment's prefix, e.g., the 783 Proxy/Servers within [ULA*]::2000/116 could assign the ADM-ULAs 784 [ULA*]::2011/116, [ULA*]::2026/116, [ULA*]::2003/116, etc. 786 The administrative authorities for each segment must therefore 787 coordinate to assure mutually-exclusive ADM-ULA prefix assignments, 788 but internal provisioning of ADM-ULAs an independent local 789 consideration for each administrative authority. For each ADM-ULA 790 prefix, the Bridge(s) that connect that segment assign the all-zero's 791 address of the prefix as a Subnet Router Anycast address. For 792 example, the Subnet Router Anycast address for [ULA*]::1023/116 is 793 simply [ULA*]::1000. 795 ADM-ULA prefixes are statically represented in Bridge forwarding 796 tables. Bridges join multiple segments into a unified OMNI link over 797 multiple diverse administrative domains. They support a bridging 798 function by first establishing forwarding table entries for their 799 ADM-ULA prefixes either via standard BGP routing or static routes. 800 For example, if three Bridges ('A', 'B' and 'C') from different 801 segments serviced [ULA*]::1000/116, [ULA*]::2000/116 and 802 [ULA*]::3000/116 respectively, then the forwarding tables in each 803 Bridge are as follows: 805 A: [ULA*]::1000/116->local, [ULA*]::2000/116->B, [ULA*]::3000/116->C 807 B: [ULA*]::1000/116->A, [ULA*]::2000/116->local, [ULA*]::3000/116->C 809 C: [ULA*]::1000/116->A, [ULA*]::2000/116->B, [ULA*]::3000/116->local 811 These forwarding table entries are permanent and never change, since 812 they correspond to fixed infrastructure elements in their respective 813 segments. 815 MNP ULAs are instead dynamically advertised in the AERO routing 816 system by Servers and Relays that provide service for their 817 corresponding MNPs. For example, if three Servers ('D', 'E' and 'F') 818 service the MNPs 2001:db8:1000:2000::/56, 2001:db8:3000:4000::/56 and 819 2001:db8:5000:6000::/56 then the routing system would include: 821 D: [ULA*]:2001:db8:1000:2000/120 823 E: [ULA*]:2001:db8:3000:4000/120 825 F: [ULA*]:2001:db8:5000:6000/120 827 A full discussion of the BGP-based routing system used by AERO is 828 found in [I-D.ietf-rtgwg-atn-bgp]. 830 3.2.4. OMNI Link Segment Routing 832 With the Client and partition prefixes in place in Bridge forwarding 833 tables, the OMNI interface sends control and data messages 834 encapsulated in carrier packets toward AERO destination nodes located 835 in different OMNI link segments over the spanning tree. The OMNI 836 interface uses the OMNI Adaptation Layer (OAL) encapsulation service 837 [I-D.templin-6man-omni-interface], and includes an OMNI Routing 838 Header (ORH) as an extension to the OAL header if final segment 839 forwarding information is available, e.g., in the neighbor cache. 840 (For nodes located in the same OMNI link segment, or when no final 841 segment forwarding information is available, the ORH may be omitted.) 842 The ORH is formatted as shown in Figure 3: 844 0 1 2 3 845 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 1 846 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 847 | Next Header | Hdr Ext Len | Routing Type | SRT | FMT | 848 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 849 | Last Hop Segment-id (LHS) | 850 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 851 ~ Link Layer Address (L2ADDR) ~ 852 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 853 ~ Destination Suffix (if necessary) ~ 854 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 855 ~ Null Padding (if necessary) ~ 856 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 858 Figure 3: OMNI Routing Header (ORH) Format 860 In this format: 862 o Next Header identifies the type of header immediately following 863 the ORH. 865 o Hdr Ext Len is the length of the Routing header in 8-octet units 866 (not including the first 8 octets), with trailing padding added if 867 necessary to produce an integral number of 8-octet units. 869 o Routing Type is set to TBD1 (see IANA Considerations). 871 o Segments Left is omitted, and replaced by a 5-bit SRT and 3-bit 872 FMT field. 874 o SRT - a 5-bit Segment Routing Topology prefix length value that 875 (when added to 96) determines the prefix length to apply to the 876 ADM-ULA formed from concatenating [ULA*]::/96 with the 32 bit LHS 877 value that follows (for example, the value 16 corresponds to the 878 prefix length 112). 880 o FMT - a 3-bit "Framework/Mode/Type" code corresponding to the 881 included Link Layer Address as follows: 883 * When the most significant bit (i.e., "Framework") is set to 1, 884 L2ADDR is the *NET encapsulation address for the target Client 885 itself; otherwise L2ADDR is the address of the Server/Proxy 886 named in the LHS. 888 * When the next most significant bit (i.e., "Mode") is set to 1, 889 the Framework node is (likely) located behind a *NET Network 890 Address Translator (NAT); otherwise, it is on the open *NET. 892 * When the least significant bit (i.e., "Type") is set to 0, 893 L2ADDR includes a UDP Port Number followed by an IPv4 address; 894 otherwise, it includes a UDP Port Number followed by an IPv6 895 address. 897 o LHS - the 32 bit ID of a node in the Last Hop Segment that 898 services the target. When SRT and LHS are both set to 0, the LHS 899 is considered unspecified. When SRT is set to 0 and LHS is non- 900 zero, the prefix length is set to 128. SRT and LHS provide 901 guidance to the OMNI interface forwarding algorithm. 902 Specifically, if SRT/LHS is located in the local OMNI link 903 segment, the OAL source can omit the ORH and (following any 904 necessary NAT traversal messaging) send directly to the OAL 905 destination according to FMT/L2ADDR. Otherwise, it includes the 906 ORH and forwards according to the OMNI link spanning tree. 908 o Link Layer Address (L2ADDR) - Formatted according to FMT, and 909 identifies the link-layer address (i.e., the encapsulation 910 address) of the target. The UDP Port Number appears in the first 911 two octets and the IP address appears in the next 4 octets for 912 IPv4 or 16 octets for IPv6. The Port Number and IP address are 913 recorded in network byte order, and in ones-compliment 914 "obfuscated" form per [RFC4380]. The OMNI interface forwarding 915 algorithm uses FMT/L2ADDR to determine the *NET encapsulation 916 address for local forwarding when SRT/LHS is located in the same 917 OMNI link segment. Note that if the target is behind a NAT, 918 L2ADDR will contain the mapped *NET address stored in the NAT; 919 otherwise, L2ADDR will contain the native *NET information of the 920 target itself. 922 o Destination Suffix is a 64-bit field included only for OAL non- 923 first-fragments. Present only when Hdr Ext Len indicates that at 924 least 8 bytes follow L2ADDR. When present, encodes the 64-bit 925 MNP-ULA suffix for the target Client. 927 o Null Padding contains zero-valued octets as necessary to pad the 928 ORH to an integral number of 8-octet units. 930 AERO neighbors use OAL encapsulation and fragmentation to exchange 931 OAL packets as specified in [I-D.templin-6man-omni-interface]. When 932 an AERO node's OMNI interface (acting as an OAL source) uses OAL 933 encapsulation for an original IP packet with source address 934 2001:db8:1:2::1 and destination address 2001:db8:1234:5678::1, it 935 sets the OAL header source address to its own ULA (e.g., 936 [ULA*]::2001:db8:1:2), sets the destination address to the MNP-ULA 937 corresponding to the IP destination address (e.g., 938 [ULA*]::2001:db8:1234:5678), selects an Identification and appends an 939 OAL checksum. 941 If the neighbor cache information indicates that the target is in a 942 different segment, the OAL source next inserts an ORH immediately 943 following the OAL header while including the correct SRT, FMT, LHS, 944 L2ADDR and Destination Suffix if fragmentation if needed (in this 945 case, the Destination Suffix is 2001:db8:1234:5678). Next, the OAL 946 source overwrites the OAL header destination address with the LHS 947 Subnet Router Anycast address (for example, for LHS 3000:4567 with 948 SRT 16, the Subnet Router Anycast address is [ULA*]::3000:0000). 949 (Note: if the ADM-ULA of the last-hop Proxy/Server is known but the 950 SRT, FMT, LHS and L2ADDR are not (yet) known, the OAL source instead 951 overwrites the OAL header destination address with the ADM-ULA.) 953 The OAL source then fragments the OAL packet, with each resulting OAL 954 fragment including the OAL/ORH headers while only the first fragment 955 includes the original IPv6 header. (Note that the packet is prepared 956 as an "atomic" OAL fragment even if no actual fragmentation was 957 required.) The OAL source finally encapsulates each resulting OAL 958 fragment in an *NET header to form an OAL carrier packet, with source 959 address set to its own *NET address (e.g., 192.0.2.100) and 960 destination set to the *NET address of a Bridge (e.g., 192.0.2.1). 962 The carrier packet encapsulation format in the above example is shown 963 in Figure 4: 965 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 966 | *NET Header | 967 | src = 192.0.2.100 | 968 | dst = 192.0.2.1 | 969 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 970 | OAL IPv6 Header | 971 | src = [ULA*]::2001:db8:1:2 | 972 | dst= [ULA*]::3000:0000 | 973 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 974 | ORH (if necessary) 975 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 976 | OAL Fragment Header | 977 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 978 | Original IP Header | 979 | (first-fragment only) | 980 | src = 2001:db8:1:2::1 | 981 | dst = 2001:db8:1234:5678::1 | 982 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 983 | | 984 ~ ~ 985 ~ Original Packet Body/Fragment ~ 986 ~ ~ 987 | | 988 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 990 Figure 4: Carrier Packet Format 992 In this format, the origianl IP header and packet body/fragment are 993 from the original IP packet, the OAL header is an IPv6 header 994 prepared according to [RFC2473], the ORH is a Routing Header 995 extension of the OAL header the Fragment Header identifies each 996 fragment, and the INET header is prepared as discussed in 997 Section 3.6. When the OAL source transmits the resulting carrier 998 packets, they are forwarded over possibly multiple OMNI link spanning 999 tree hops until they arrive at the OAL destination. 1001 This gives rise to a routing system that contains both Client MNP-ULA 1002 routes that may change dynamically due to regional node mobility and 1003 per-partition ADM-ULA routes that rarely if ever change. The Bridges 1004 can therefore provide link-layer bridging by sending carrier packets 1005 over the spanning tree instead of network-layer routing according to 1006 MNP routes. As a result, opportunities for loss due to node mobility 1007 between different segments are mitigated. 1009 In normal operations, IPv6 ND messages are conveyed over secured 1010 paths between OMNI link neighbors so that specific Proxys, Servers or 1011 Relays can be addressed without being subject to mobility events. 1012 Conversely, only the first few carrier packets destined to Clients 1013 need to traverse secured paths until route optimization can determine 1014 a more direct path. 1016 Note: When the OAL source and destination are on the same INET 1017 segment, the ORH is not needed and OAL header compression can be used 1018 to significantly reduce encapsulation overhead 1019 [I-D.templin-6man-omni-interface]. 1021 Note: When the OAL source has multiple original IP packets to send to 1022 the same OAL destination, it can perform "packing" to generate a 1023 "super-packet" [I-D.templin-6man-omni-interface]. In that case, the 1024 OAL/ORH super-packet may include up to N original IP packets as long 1025 as the total length of the super-packet does not exceed the OMNI 1026 interface MTU. 1028 Note: Use of an IPv6 "minimal encapsulation" format (i.e., an IPv6 1029 variant of [RFC2004]) based on extensions to the ORH was considered 1030 and abandoned. In the approach, the ORH would be inserted as an 1031 extension header to the original IPv6 packet header. The IPv6 1032 destination address would then be written into the ORH, and the ULA 1033 corresponding to the destination would be overwritten in the IPv6 1034 destination address. This would seemingly convey enough forwarding 1035 information so that OAL encapsulation could be avoided. However, 1036 this "minimal encapsulation" IPv6 packet would then have a non-ULA 1037 source address and ULA destination address, an incorrect value in 1038 upper layer protocol checksums, and a Hop Limit that is decremented 1039 within the spanning tree when it should not be. The insertion and 1040 removal of the ORH would also entail rewriting the Payload Length and 1041 Next Header fields - again, invalidating upper layer checksums. 1042 These irregularities would result in implementation challenges and 1043 the potential for operational issues, e.g., since actionable ICMPv6 1044 error reports could not be delivered to the original source. In 1045 order to address the issues, still more information such as the 1046 original IPv6 source address could be written into the ORH. However, 1047 with the additional information the benefit of the "minimal 1048 encapsulation" savings quickly diminishes, and becomes overshadowed 1049 by the implementation and operational irregularities. 1051 3.2.5. Segment Routing Topologies (SRTs) 1053 The 64-bit sub-prefixes of [ULA]::/48 identify up to 2^16 distinct 1054 Segment Routing Topologies (SRTs). Each SRT is a mutually-exclusive 1055 OMNI link overlay instance using a distinct set of ULAs, and emulates 1056 a Virtual LAN (VLAN) service for the OMNI link. In some cases (e.g., 1057 when redundant topologies are needed for fault tolerance and 1058 reliability) it may be beneficial to deploy multiple SRTs that act as 1059 independent overlay instances. A communication failure in one 1060 instance therefore will not affect communications in other instances. 1062 Each SRT is identified by a distinct value in bits 48-63 of 1063 [ULA]::/48, i.e., as [ULA0]::/64, [ULA1]::/64, [ULA2]::/64, etc. 1064 Each OMNI interface is identified by a unique interface name (e.g., 1065 omni0, omni1, omni2, etc.) and assigns an anycast ADM-ULA 1066 corresponding to its SRT prefix length. The anycast ADM-ULA is used 1067 for OMNI interface determination in Safety-Based Multilink (SBM) as 1068 discussed in [I-D.templin-6man-omni-interface]. Each OMNI interface 1069 further applies Performance-Based Multilink (PBM) internally. 1071 3.2.6. Segment Routing For OMNI Link Selection 1073 An original IPv6 source can direct an IPv6 packet to an AERO node by 1074 including a standard IPv6 Segment Routing Header (SRH) [RFC8754] with 1075 the anycast ADM-ULA for the selected SRT as either the IPv6 1076 destination or as an intermediate hop within the SRH. This allows 1077 the original source to determine the specific OMNI link topology an 1078 original IPv6 packet will traverse when there may be multiple 1079 alternatives. 1081 When the AERO node processes the SRH and forwards the original IPv6 1082 packet to the correct OMNI interface, the OMNI interface writes the 1083 next IPv6 address from the SRH into the IPv6 destination address and 1084 decrements Segments Left. If decrementing would cause Segments Left 1085 to become 0, the OMNI interface deletes the SRH before forwarding. 1086 This form of Segment Routing supports Safety-Based Multilink (SBM). 1088 3.2.7. Segment Routing Within the OMNI Link 1090 OAL sources can insert an ORH for Segment Routing within the OMNI 1091 link to influence the paths of OAL packets sent to OAL destinations 1092 in remote segments without requiring all carrier packets to traverse 1093 strict spanning tree paths. 1095 When an AERO node's OMNI interface has an original IP packet to send 1096 to a target discovered through route optimization located in the same 1097 OMNI link segment, it acts as an OAL source to perform OAL 1098 encapsulation and fragmentation. The node then uses the target's 1099 Link Layer Address (L2ADDR) information for *NET encapsulation. 1101 When an AERO node's OMNI interface has an original IP packet to send 1102 to a route optimization target located in a remote OMNI link segment, 1103 it acts as an OAL source the same as above but also includes an ORH 1104 while setting the OAL destination to the Subnet Router Anycast 1105 address for the final OMNI link segment, then forwards the resulting 1106 carrier packets to a Bridge. 1108 When a Bridge receives a carrier packet destined to its Subnet Router 1109 Anycast address with an ORH with SRT/LHS values corresponding to the 1110 local segment, it examines the L2ADDR according to FMT and removes 1111 the ORH from the carrier packet. The Bridge then writes the MNP-ULA 1112 corresponding to the ORH Destination Suffix into the OAL destination 1113 address, re-encapsulates the carrier packet according to L2ADDR and 1114 forwards the carrier packet either to the LHS Server/Proxy or 1115 directly to the target Client itself. In this way, the Bridge 1116 participates in route optimization to reduce traffic load and 1117 suboptimal routing through strict spanning tree paths. 1119 3.3. OMNI Interface Characteristics 1121 OMNI interfaces are virtual interfaces configured over one or more 1122 underlying interfaces classified as follows: 1124 o INET interfaces connect to an INET either natively or through one 1125 or several NATs. Native INET interfaces have global IP addresses 1126 that are reachable from any INET correspondent. All Server, Relay 1127 and Bridge interfaces are native interfaces, as are INET-facing 1128 interfaces of Proxys. NATed INET interfaces connect to a private 1129 network behind one or more NATs that provide INET access. Clients 1130 that are behind a NAT are required to send periodic keepalive 1131 messages to keep NAT state alive when there are no carrier packets 1132 flowing. 1134 o ANET interfaces connect to an ANET that is separated from the open 1135 INET by a Proxy. Proxys can actively issue control messages over 1136 the INET on behalf of the Client to reduce ANET congestion. 1138 o VPNed interfaces use security encapsulation over the INET to a 1139 Virtual Private Network (VPN) server that also acts as a Server or 1140 Proxy. Other than the link-layer encapsulation format, VPNed 1141 interfaces behave the same as Direct interfaces. 1143 o Direct (i.e., single-hop point-to-point) interfaces connect a 1144 Client directly to a Server or Proxy without crossing any ANET/ 1145 INET paths. An example is a line-of-sight link between a remote 1146 pilot and an unmanned aircraft. The same Client considerations 1147 apply as for VPNed interfaces. 1149 OMNI interfaces use OAL encapsulation and fragmentation as discussed 1150 in Section 3.2.4. OMNI interfaces use *NET encapsulation (see: 1151 Section 3.6) to exchange carrier packets with OMNI link neighbors 1152 over INET or VPNed interfaces as well as over ANET interfaces for 1153 which the Client and Proxy may be multiple IP hops away. OMNI 1154 interfaces do not use link-layer encapsulation over Direct underlying 1155 interfaces or ANET interfaces when the Client and Proxy are known to 1156 be on the same underlying link. 1158 OMNI interfaces maintain a neighbor cache for tracking per-neighbor 1159 state the same as for any interface. OMNI interfaces use ND messages 1160 including Router Solicitation (RS), Router Advertisement (RA), 1161 Neighbor Solicitation (NS) and Neighbor Advertisement (NA) for 1162 neighbor cache management. 1164 OMNI interfaces send ND messages with an OMNI option formatted as 1165 specified in [I-D.templin-6man-omni-interface]. The OMNI option 1166 includes prefix registration information and Interface Attributes 1167 containing link information parameters for the OMNI interface's 1168 underlying interfaces. Each OMNI option may include multiple 1169 Interface Attributes sub-options, each identified by an ifIndex 1170 value. 1172 A Client's OMNI interface may be configured over multiple underlying 1173 interface connections. For example, common mobile handheld devices 1174 have both wireless local area network ("WLAN") and cellular wireless 1175 links. These links are often used "one at a time" with low-cost WLAN 1176 preferred and highly-available cellular wireless as a standby, but a 1177 simultaneous-use capability could provide benefits. In a more 1178 complex example, aircraft frequently have many wireless data link 1179 types (e.g. satellite-based, cellular, terrestrial, air-to-air 1180 directional, etc.) with diverse performance and cost properties. 1182 If a Client's multiple underlying interfaces are used "one at a time" 1183 (i.e., all other interfaces are in standby mode while one interface 1184 is active), then ND message OMNI options include only a single 1185 Interface Attributes sub-option set to constant values. In that 1186 case, the Client would appear to have a single interface but with a 1187 dynamically changing link-layer address. 1189 If the Client has multiple active underlying interfaces, then from 1190 the perspective of ND it would appear to have multiple link-layer 1191 addresses. In that case, ND message OMNI options MAY include 1192 multiple Interface Attributes sub-options - each with values that 1193 correspond to a specific interface. Every ND message need not 1194 include Interface Attributes for all underlying interfaces; for any 1195 attributes not included, the neighbor considers the status as 1196 unchanged. 1198 Bridge, Server and Proxy OMNI interfaces may be configured over one 1199 or more secured tunnel interfaces. The OMNI interface configures 1200 both an ADM-LLA and its corresponding ADM-ULA, while the underlying 1201 secured tunnel interfaces are either unnumbered or configure the same 1202 ULA. The OMNI interface acting as an OAL source encapsulates and 1203 fragments each original IP packet, then and presents the resulting 1204 carrier packets to the underlying secured tunnel interface. Routing 1205 protocols such as BGP that run over the OMNI interface do not employ 1206 OAL encapsulation, but rather present the routing protocol messages 1207 directly to the underlying secured tunnels while using the ULA as the 1208 source address. This distinction must be honored consistently 1209 according to each node's configuration so that the IP forwarding 1210 table will associate discovered IP routes with the correct interface. 1212 3.4. OMNI Interface Initialization 1214 AERO Servers, Proxys and Clients configure OMNI interfaces as their 1215 point of attachment to the OMNI link. AERO nodes assign the MSPs for 1216 the link to their OMNI interfaces (i.e., as a "route-to-interface") 1217 to ensure that original IP packets with destination addresses covered 1218 by an MNP not explicitly assigned to a non-OMNI interface are 1219 directed to the OMNI interface. 1221 OMNI interface initialization procedures for Servers, Proxys, Clients 1222 and Bridges are discussed in the following sections. 1224 3.4.1. AERO Server/Relay Behavior 1226 When a Server enables an OMNI interface, it assigns an ADM-{LLA,ULA} 1227 appropriate for the given OMNI link segment. The Server also 1228 configures secured tunnels with one or more neighboring Bridges and 1229 engages in a BGP routing protocol session with each Bridge. 1231 The OMNI interface provides a single interface abstraction to the IP 1232 layer, but internally comprises multiple secured tunnels as well as 1233 an NBMA nexus for sending carrier packets to OMNI interface 1234 neighbors. The Server further configures a service to facilitate ND 1235 exchanges with AERO Clients and manages per-Client neighbor cache 1236 entries and IP forwarding table entries based on control message 1237 exchanges. 1239 Relays are simply Servers that run a dynamic routing protocol to 1240 redistribute routes between the OMNI interface and INET/EUN 1241 interfaces (see: Section 3.2.3). The Relay provisions MNPs to 1242 networks on the INET/EUN interfaces (i.e., the same as a Client would 1243 do) and advertises the MSP(s) for the OMNI link over the INET/EUN 1244 interfaces. The Relay further provides an attachment point of the 1245 OMNI link to a non-MNP-based global topology. 1247 3.4.2. AERO Proxy Behavior 1249 When a Proxy enables an OMNI interface, it assigns an ADM-{LLA, ULA} 1250 and configures secured tunnels with neighboring Bridges the same as 1251 for Servers. The Proxy also configures secured tunnels with one or 1252 more neighboring Bridges and maintains per-Client neighbor cache 1253 entries based on control message exchanges. Importantly Proxys are 1254 often configured to act as Servers, and vice-versa. 1256 3.4.3. AERO Client Behavior 1258 When a Client enables an OMNI interface, it assigns either an 1259 MNP-{LLA, ULA} or a Temporary ULA and sends RS messages with ND 1260 parameters over its underlying interfaces to a Server, which returns 1261 an RA message with corresponding parameters. The RS/RA messages may 1262 pass through a Proxy in the case of a Client's ANET interface, or 1263 through one or more NATs in the case of a Client's INET interface. 1264 (Note: if the Client used a Temporary ULA in its initial RS message, 1265 it will discover an MNP-{LLA, ULA} in the corresponding RA that it 1266 receives from the Server and begin using these new addresses. If the 1267 Client is operating outside the context of AERO infrastructure such 1268 as in a Mobile Ad-hoc Network (MANET), however, it may continue using 1269 Temporary ULAs for Client-to-Client communications until it 1270 encounters an infrastructure element that can provide an MNP.) 1272 3.4.4. AERO Bridge Behavior 1274 AERO Bridges configure an OMNI interface and assign the ADM-ULA 1275 Subnet Router Anycast address for each OMNI link segment they connect 1276 to. Bridges configure secured tunnels with Servers, Proxys and other 1277 Bridges, and engage in a BGP routing protocol session with neighbors 1278 on the spanning tree (see: Section 3.2.3). 1280 3.5. OMNI Interface Neighbor Cache Maintenance 1282 Each OMNI interface maintains a conceptual neighbor cache that 1283 includes an entry for each neighbor it communicates with on the OMNI 1284 link per [RFC4861]. Proxy neighbor cache entries are created and 1285 maintained by AERO Proxys when they process Client/Server ND 1286 exchanges, and remain in place for durations bounded by ND and prefix 1287 lifetimes. AERO Proxys maintain proxy neighbor cache entries for 1288 each of their associated Clients. Proxy neighbor cache entries track 1289 the Client state and the address of the Client's associated 1290 Server(s). 1292 To the list of neighbor cache entry states in Section 7.3.2 of 1293 [RFC4861], Proxy and Server OMNI interfaces add an additional state 1294 DEPARTED that applies to Clients that have recently departed. The 1295 interface sets a "DepartTime" variable for the neighbor cache entry 1296 to "DEPART_TIME" seconds. DepartTime is decremented unless a new ND 1297 message causes the state to return to REACHABLE. While a neighbor 1298 cache entry is in the DEPARTED state, carrier packets destined to the 1299 target Client are forwarded to the Client's new location instead of 1300 being dropped. When DepartTime decrements to 0, the neighbor cache 1301 entry is deleted. It is RECOMMENDED that DEPART_TIME be set to the 1302 default constant value REACHABLE_TIME plus 10 seconds (40 seconds by 1303 default) to allow a window for carrier packets in flight to be 1304 delivered while stale route optimization state may be present. 1306 Servers can act as RORs on behalf of multilink and/or disadvantaged 1307 Clients according to the Proxy Neighbor Advertisement specification 1308 in Section 7.2.8 of [RFC4861]. (Well-connected Clients can act as an 1309 ROR on their own behalf, and perform normal IPv6 ND operations the 1310 same as for any node.) When a Server acting as an ROR receives an 1311 authentic NS message used for route optimization, it searches for a 1312 neighbor cache entry for the target Client. The ROR then returns a 1313 solicited NA message while creating a neighbor cache entry for the 1314 ROS, and also creates or updates a target Client "Report List" entry 1315 for the ROS and sets a "ReportTime" variable for the entry to 1316 REPORT_TIME seconds. The ROR resets ReportTime when it receives a 1317 new authentic NS message, and otherwise decrements ReportTime while 1318 no authentic NS messages have been received. It is RECOMMENDED that 1319 REPORT_TIME be set to the default constant value REACHABLE_TIME plus 1320 10 seconds (40 seconds by default) to allow a window for route 1321 optimization to converge before ReportTime decrements below 1322 REACHABLE_TIME. 1324 When the ROS receives a solicited NA message response to its NS 1325 message used for route optimization, it creates or updates a neighbor 1326 cache entry for the target network-layer and link-layer addresses. 1327 The ROS then (re)sets ReachableTime for the neighbor cache entry to 1328 REACHABLE_TIME seconds and uses this value to determine whether 1329 carrier packets can be forwarded directly to the target, i.e., 1330 instead of via a default route. The ROS otherwise decrements 1331 ReachableTime while no further solicited NA messages arrive. It is 1332 RECOMMENDED that REACHABLE_TIME be set to the default constant value 1333 30 seconds as specified in [RFC4861]. 1335 AERO nodes also use the value MAX_UNICAST_SOLICIT to limit the number 1336 of NS keepalives sent when a correspondent may have gone unreachable, 1337 the value MAX_RTR_SOLICITATIONS to limit the number of RS messages 1338 sent without receiving an RA and the value MAX_NEIGHBOR_ADVERTISEMENT 1339 to limit the number of unsolicited NAs that can be sent based on a 1340 single event. It is RECOMMENDED that MAX_UNICAST_SOLICIT, 1341 MAX_RTR_SOLICITATIONS and MAX_NEIGHBOR_ADVERTISEMENT be set to 3 the 1342 same as specified in [RFC4861]. 1344 Different values for DEPART_TIME, REPORT_TIME, REACHABLE_TIME, 1345 MAX_UNICAST_SOLICIT, MAX_RTR_SOLCITATIONS and 1346 MAX_NEIGHBOR_ADVERTISEMENT MAY be administratively set; however, if 1347 different values are chosen, all nodes on the link MUST consistently 1348 configure the same values. Most importantly, DEPART_TIME and 1349 REPORT_TIME SHOULD be set to a value that is sufficiently longer than 1350 REACHABLE_TIME to avoid packet loss due to stale route optimization 1351 state. 1353 3.5.1. OMNI Neighbor Interface Attributes 1355 OMNI interface IPv6 ND messages include OMNI options 1356 [I-D.templin-6man-omni-interface] with Interface Attributes that 1357 provide Link-Layer Address and QoS Preference information for the 1358 neighbor's underlying interfaces. This information is stored in the 1359 neighbor cache and provides the basis for the forwarding algorithm 1360 specified in Section 3.10. The information is cumulative and 1361 reflects the union of the OMNI information from the most recent ND 1362 messages received from the neighbor; it is therefore not required 1363 that each ND message contain all neighbor information. 1365 The OMNI option Interface Attributes for each underlying interface 1366 includes a two-part "Link-Layer Address" consisting of a simple IP 1367 encapsulation address determined by the FMT and L2ADDR fields and an 1368 ADM-ULA determined by the SRT and LHS fields. Underlying interfaces 1369 are further selected based on their associated preference values 1370 "high", "medium", "low" or "disabled". 1372 Note: the OMNI option is distinct from any Source/Target Link-Layer 1373 Address Options (S/TLLAOs) that may appear in an ND message according 1374 to the appropriate IPv6 over specific link layer specification (e.g., 1375 [RFC2464]). If both an OMNI option and S/TLLAO appear, the former 1376 pertains to encapsulation addresses while the latter pertains to the 1377 native L2 address format of the underlying media. 1379 3.5.2. OMNI Neighbor Advertisement Message Flags 1381 As discussed in Section 4.4 of [RFC4861] NA messages include three 1382 flag bits R, S and O. OMNI interface NA messages treat the flags as 1383 follows: 1385 o R: The R ("Router") flag is set to 1 in the NA messages sent by 1386 all AERO/OMNI node types. Simple hosts that would set R to 0 do 1387 not occur on the OMNI link itself, but may occur on the downstream 1388 links of Clients and Relays. 1390 o S: The S ("Solicited") flag is set exactly as specified in 1391 Section 4.4. of [RFC4861], i.e., it is set to 1 for Solicited NAs 1392 and set to 0 for Unsolicited NAs (both unicast and multicast). 1394 o O: The O ("Override") flag is set to 0 for solicited proxy NAs and 1395 set to 1 for all other solicited and unsolicited NAs. For further 1396 study is whether solicited NAs for anycast targets apply for OMNI 1397 links. Since MNP-LLAs must be uniquely assigned to Clients to 1398 support correct ND protocol operation, however, no role is 1399 currently seen for assigning the same MNP-LLA to multiple Clients. 1401 3.6. OMNI Interface Encapsulation and Re-encapsulation 1403 The OMNI interface admits original IP packets then (acting as an OAL 1404 source) performs OAL encapsulation and fragmentation as specified in 1405 Section 3.2.4[I-D.templin-6man-omni-interface]. For original IP 1406 packets entering the OMNI interface from the network layer, the OMNI 1407 interface copies the "TTL/Hop Limit", "Type of Service/Traffic Class" 1408 [RFC2983], "Flow Label"[RFC6438] (for IPv6) and "Congestion 1409 Experienced" [RFC3168] values in the original IP packet's header into 1410 the corresponding fields in the OAL and *NET encapsulation header(s). 1411 For carrier packets undergoing re-encapsulation, the OMNI interface 1412 instead copies these values from the original *NET encapsulation 1413 header into the new *NET encapsulation header, i.e., the values are 1414 transferred between *NET encapsulation headers and *not* copied from 1415 the carrier packet's network-layer header. (Note especially that by 1416 copying the TTL/Hop Limit between encapsulation headers the value 1417 will eventually decrement to 0 if there is a (temporary) routing 1418 loop.) 1420 The OAL source sets the *NET header UDP source port to a constant 1421 value that it will use in each successive carrier packet it sends. 1422 For carrier packets sent to a Server, Relay or Bridge, the OAL source 1423 sets the UDP destination port to 8060, i.e., the IANA-registered port 1424 number for AERO. For carrier packets sent to a Client, the OAL 1425 source sets the UDP destination port to the port value stored in the 1426 neighbor cache entry for this Client. The OAL source finally 1427 includes/omits the UDP checksum according to [RFC6935][RFC6936]. 1429 When a Proxy, Relay or Server re-encapsulates a carrier packet 1430 received from a Client that includes an OAL but no ORH, it inserts an 1431 ORH immediately following the OAL header and adjusts the OAL payload 1432 length and destination address field. The inserted ORH will be 1433 removed by the final-hop Bridge, but its insertion and removal will 1434 not interfere with reassembly at the final destination. For this 1435 reason, Clients must reserve 40 bytes for a maximum-length ORH when 1436 they perform OAL encapsulation (see: Section 3.9). 1438 3.7. OMNI Interface Decapsulation 1440 OMNI interfaces (acting as OAL destinations) decapsulate and 1441 reassemble OAL packets into original IP packets destined either to 1442 the AERO node itself or to a destination reached via an interface 1443 other than the OMNI interface the original IP packet was received on. 1445 When carrier packets containing OAL fragments arrive, the OMNI 1446 interface reassembles as discussed in Section 3.9. 1448 3.8. OMNI Interface Data Origin Authentication 1450 AERO nodes employ simple data origin authentication procedures. In 1451 particular: 1453 o AERO Bridges, Servers and Proxys accept carrier packets (including 1454 either data or control messages) received from the (secured) 1455 spanning tree. 1457 o AERO Proxys and Clients accept carrier packets and original IP 1458 packets that originate from within the same secured ANET. 1460 o AERO Clients and Relays accept original IP packets from downstream 1461 network correspondents based on ingress filtering. 1463 o AERO Clients, Relays and Servers verify carrier packet UDP/IP 1464 encapsulation addresses according to [RFC4380]. 1466 o AERO Clients, Relays and Servers as OAL destinations accept OAL 1467 packets/fragments with Identification values within the current 1468 window for the OAL source. 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 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) that 1480 accommodates multiple underlying links with diverse MTUs while 1481 observing both a minimum and per-path Maximum Payload Size (MPS). 1482 The functions of the OAL and the OMNI interface MTU/MRU/MPS are 1483 specified in Section 5 of [I-D.templin-6man-omni-interface] with MTU/ 1484 MRU both set to the constant value 9180 bytes, with minimum MPS set 1485 to 400 bytes, and with path MPS set to a potentially larger value 1486 depending on the underlying path. 1488 When the network layer presents an original IP packet to the OMNI 1489 interface, the OAL source encapsulates and fragments the original IP 1490 packet if necessary. When the network layer presents the OMNI 1491 interface with multiple original IP packets bound to the same OAL 1492 destination, the OAL source can concatenate them together into a 1493 single OAL super-packet as discussed in 1494 [I-D.templin-6man-omni-interface]. The OAL source then fragments the 1495 OAL packet if necessary according to the minimum/path MPS such that 1496 the OAL headers appear in each fragment while the original IP packet 1497 header appears only in the first fragment. The OAL source then 1498 encapsulates each OAL fragment in *NET headers for transmission as 1499 carrier packets over an underlying interface connected to either a 1500 physical link such as Ethernet, WiFi and the like or a virtual link 1501 such as an Internet or higher-layer tunnel (see the definition of 1502 link in [RFC8200]). 1504 Note: A Client that does not (yet) have neighbor cache state for a 1505 target may omit the ORH in carrier packets with the understanding 1506 that a Proxy/Server may insert an ORH on its behalf. For this 1507 reason, Clients reserve 40 bytes for the largest possible ORH in 1508 their OAL fragment size calculations. 1510 Note: Although the ORH may be removed by a Bridge on the path (see: 1511 Section 3.10.3), this does not interfere with the destination's 1512 ability to reassemble. This is due to the fact that the ORH is not 1513 included in the fragmentable part; therefore, its removal does not 1514 invalidate the offset values in any fragment headers. 1516 3.10. OMNI Interface Forwarding Algorithm 1518 Original IP packets enter a node's OMNI interface either from the 1519 network layer (i.e., from a local application or the IP forwarding 1520 system) while carrier packets enter from the link layer (i.e., from 1521 an OMNI interface neighbor). All original IP packets and carrier 1522 packets entering a node's OMNI interface first undergo data origin 1523 authentication as discussed in Section 3.8. Those that satisfy data 1524 origin authentication are processed further, while all others are 1525 dropped silently. 1527 Original IP packets that enter the OMNI interface from the network 1528 layer are forwarded to an OMNI interface neighbor using OAL 1529 encapsulation and fragmentation to produce carrier packets for 1530 transmission over underlying interfaces. (If routing indicates that 1531 the original IP packet should instead be forwarded back to the 1532 network layer, the packet is dropped to avoid looping). Carrier 1533 packets that enter the OMNI interface from the link layer are either 1534 re-encapsulated and re-admitted into the OMNI link, or reassembled 1535 and forwarded to the network layer where they are subject to either 1536 local delivery or IP forwarding. In all cases, the OAL MUST NOT 1537 decrement the network layer TTL/Hop-count since its forwarding 1538 actions occur below the network layer. 1540 OMNI interfaces may have multiple underlying interfaces and/or 1541 neighbor cache entries for neighbors with multiple underlying 1542 interfaces (see Section 3.3). The OAL uses interface attributes and/ 1543 or traffic classifiers (e.g., DSCP value, port number, flow 1544 specification, etc.) to select an outgoing underlying interface for 1545 each OAL packet based on the node's own QoS preferences, and also to 1546 select a destination link-layer address based on the neighbor's 1547 underlying interface with the highest preference. AERO 1548 implementations SHOULD allow for QoS preference values to be modified 1549 at runtime through network management. 1551 If multiple outgoing interfaces and/or neighbor interfaces have a 1552 preference of "high", the AERO node replicates the OAL packet and 1553 sends one copy via each of the (outgoing / neighbor) interface pairs; 1554 otherwise, the node sends a single copy of the OAL packet via an 1555 interface with the highest preference. (While not strictly required, 1556 successful delivery may be more likely when all OAL fragments of the 1557 same OAL packet are sent over the same underlying interface.) AERO 1558 nodes keep track of which underlying interfaces are currently 1559 "reachable" or "unreachable", and only use "reachable" interfaces for 1560 forwarding purposes. 1562 The following sections discuss the OMNI interface forwarding 1563 algorithms for Clients, Proxys, Servers and Bridges. In the 1564 following discussion, an original IP packet's destination address is 1565 said to "match" if it is the same as a cached address, or if it is 1566 covered by a cached prefix (which may be encoded in an MNP-LLA). 1568 3.10.1. Client Forwarding Algorithm 1570 When an original IP packet enters a Client's OMNI interface from the 1571 network layer the Client searches for a neighbor cache entry that 1572 matches the destination. If there is a match, the Client selects one 1573 or more "reachable" neighbor interfaces in the entry for forwarding 1574 purposes. If there is no neighbor cache entry, the Client instead 1575 forwards the original IP packet toward a Server (possibly via a Proxy 1576 on the path). The Client (acting as an OAL source) performs OAL 1577 encapsulation and sets the OAL destination address to the MNP-ULA if 1578 there is a matching neighbor cache entry; otherwise, it sets the OAL 1579 destination to the ADM-ULA of the Proxy/Server. If the Client has 1580 multiple original IP packets to send to the same neighbor, it can 1581 concatenate them in a single super-packet 1582 [I-D.templin-6man-omni-interface]. The OAL source then performs 1583 fragmentation to create OAL fragments (see: Section 3.9), appends any 1584 *NET encapsulation, and sends the resulting carrier packets over 1585 underlying interfaces to the neighbor acting as an OAL destination. 1587 If the neighbor interface selected for forwarding is located on the 1588 same OMNI link segment and not behind a NAT, the Client forwards the 1589 carrier packets directly according to the L2ADDR information for the 1590 neighbor. If the neighbor interface is behind a NAT on the same OMNI 1591 link segment, the Client instead forwards the initial carrier packets 1592 to its Server and initiates NAT traversal procedures. If the 1593 Client's intended source underlying interface is also behind a NAT 1594 and located on the same OMNI link segment, it sends a "direct bubble" 1595 over the interface per [RFC6081][RFC4380] to the L2ADDR found in the 1596 neighbor cache in order to establish state in its own NAT by 1597 generating traffic toward the neighbor (note that no response to the 1598 bubble is expected). 1600 The Client next sends an NS(NUD) message toward the MNP-ULA of the 1601 neighbor via its Server as discussed in Section 3.15. If the Client 1602 receives an NA(NUD) from the neighbor over the underlying interface, 1603 it marks the neighbor interface as "trusted" and sends future carrier 1604 packets directly to the L2ADDR information for the neighbor instead 1605 of indirectly via the Server. The Client must honor the neighbor 1606 cache maintenance procedure by sending additional direct bubbles and/ 1607 or NS/NA(NUD) messages as discussed in [RFC6081][RFC4380] in order to 1608 keep NAT state alive as long as carrier packets are still flowing. 1610 When an carrier packet enters a Client's OMNI interface from the 1611 link-layer, if the OAL destination matches one of the Client's MNPs 1612 or LLAs the Client (acting as an OAL destination) reassembles and 1613 decapsulates as necessary and delivers the original IP packet to the 1614 network layer. Otherwise, the Client drops the original IP packet 1615 and MAY return a network-layer ICMP Destination Unreachable message 1616 subject to rate limiting (see: Section 3.11). 1618 3.10.2. Proxy/Server Forwarding Algorithm 1620 For IPv6 ND control messages originating from or destined to a 1621 Client, the Proxy/Server intercepts the message and updates its 1622 neighbor cache entry for the Client. If acting in Proxy-only mode, 1623 it then forwards a (proxyed) copy of the control message to one or 1624 more neighbors. (For example, the Proxy forwards a proxyed version 1625 of a Client's NS/RS message to the neighbor, and forwards a proxyed 1626 version of the neighbor's NA/RA reply to the Client.) 1628 When the Proxy/Server receives an original IP packet from the network 1629 layer, it drops the packet if routing indicates that it should be 1630 forwarded back to the network layer to avoid looping. Otherwise, the 1631 Proxy/Server regards the original IP packet the same as if it had 1632 arrived as carrier packets with OAL destination set to its own ADM- 1633 ULA. When the Proxy/Server receives carrier packets with OAL 1634 destination set to its own ADM-ULA, it performs OAL reassembly if 1635 necessary to obtain the original IP packet. 1637 The Proxy/Server next searches for a neighbor cache entry that 1638 matches the original IP destination and proceeds as follows: 1640 o if the original IP packet destination matches a neighbor cache 1641 entry, the Proxy/Sever uses one or more "reachable" neighbor 1642 interfaces in the entry for packet forwarding using OAL 1643 encapsulation and fragmentation according to the cached link-layer 1644 address information. If the neighbor interface is in a different 1645 OMNI link segment, the Proxy/Server forwards the resulting carrier 1646 packets to a Bridge; otherwise, it forwards the carrier packets 1647 directly to the neighbor. If the neighbor is behind a NAT, the 1648 Proxy/Server instead forwards initial carrier packets via a Bridge 1649 while sending an NS(NUD) to the neighbor. When the Proxy/Server 1650 receives the NA(NUD), it can begin forwarding carrier packets 1651 directly to the neighbor the same as discussed in Section 3.10.1 1652 while sending additional NS(NUD) messages as necessary to maintain 1653 NAT state. Note that no direct bubbles are necessary since the 1654 Proxy/Server is by definition not located behind a NAT. 1656 o else, if the original IP destination matches a non-MNP route in 1657 the IP forwarding table or an ADM-LLA assigned to the Proxy/ 1658 Server's OMNI interface, the Proxy/Server presents the original IP 1659 packet to the network layer for local delivery or IP forwarding. 1661 o else, the Proxy/Server initiates address resolution as discussed 1662 in Section 3.14, while retaining initial original IP packets in a 1663 small queue awaiting address resolution completion. 1665 When the Proxy/Server receives a carrier packet with OAL destination 1666 set to a non-MNP ULA, it accepts the carrier packet only if data 1667 origin authentication succeeds and if there is a network layer 1668 routing table for a GUA route that matches the non-MNP ULA. If there 1669 is no route, the Proxy/Server drops the carrier packet; otherwise, it 1670 reassembles and decapsulates to obtain the original IP packet and 1671 presents it to the network layer where it will be delivered according 1672 to standard IP forwarding. 1674 When the Proxy/Server receives a carrier packet with OAL destination 1675 set to an MNP-ULA, it accepts the carrier packet only if data origin 1676 authentication succeeds and if there is a neighbor cache entry that 1677 matches the OAL destination. If the neighbor cache entry state is 1678 DEPARTED, the Proxy/Server inserts an ORH that encodes the MNP-ULA 1679 destination suffix and changes the OAL destination address to the 1680 ADM-ULA of the new Server, then re-encapsulates the carrier packet 1681 and forwards it to a Bridge which will eventually deliver it to the 1682 new Server. 1684 If the neighbor cache state is REACHABLE, the Proxy/Server can 1685 instead either reassemble first and then re-encapsulate/re-fragment 1686 before forwarding to the Client or forward the raw fragments on to 1687 the Client which then must reassemble. In the former case, the 1688 Proxy/Server can re-fragment to a size that better matches the link 1689 MTU for the Client, which may be important for low-end links with 1690 large MTUs. In the latter case, the Client may receive fragments 1691 that are smaller than its link MTU but can still be reassembled; this 1692 case may provide an important performance benefit to Proxy/Servers by 1693 permitting them to avoid excessive reassembly and re-fragmentation 1694 overhead. In either case, the Proxy/Server can return a PTB if 1695 necessary (see: [I-D.templin-6man-omni-interface]) when it receives a 1696 carrier packet containing an OAL first fragment. 1698 Note: If the Proxy/Server has multiple original IP packets to send to 1699 the same neighbor, it can concatenate them in a single OAL super- 1700 packet [I-D.templin-6man-omni-interface]. 1702 3.10.3. Bridge Forwarding Algorithm 1704 Bridges forward OAL fragments over secured tunnels the same as any 1705 IPv6 router. When the Bridge receives a carrier packet or an 1706 original IP packet via a secured tunnel, it removes the outer *NET 1707 header and searches for a forwarding table entry that matches the OAL 1708 destination address. The Bridge then processes the packet as 1709 follows: 1711 o if the packet is a carrier packet with a destination that matches 1712 its ADM-ULA Subnet Router Anycast address the Bridge processes the 1713 carrier packet locally before forwarding. The Bridge drops the 1714 carrier packet if it does not include an ORH; otherwise, for 1715 NA(NUD) messages the Bridge replaces the OMNI option Interface 1716 Attributes sub-option with information for its own interface while 1717 retaining the ifIndex value supplied by the NA(NUD) message 1718 source. The Bridge next examines the ORH FMT code. If the code 1719 indicates the destination is a Client on the open *NET (or, a 1720 Client behind a NAT for which NAT traversal procedures have 1721 already converged) the Bridge removes the ORH then writes the MNP- 1722 ULA formed from the ORH Destination Suffix into the OAL 1723 destination. The Bridge then re-encapsulates the carrier packet 1724 and forwards it to the ORH L2ADDR. For all other destination 1725 cases, the Bridge instead writes the ADM-ULA formed from the ORH 1726 SRT/LHS into the OAL destination address and forwards the carrier 1727 packet to the ADM-ULA Server while invoking NAT traversal 1728 procedures the same as for Proxys and Servers if necessary, noting 1729 that no direct bubbles are necessary since only the target Client 1730 and not the Bridge is behind a NAT. 1732 o else, if the packet is a carrier packet with a destination that 1733 matches a forwarding table entry the Bridge forwards the carrier 1734 packet via a secured tunnel to the next hop. If the destination 1735 matches an MSP without matching an MNP, however, the Bridge 1736 instead drops the packet and returns an ICMP Destination 1737 Unreachable message subject to rate limiting (see: Section 3.11). 1739 o else, if the packet is an original IP packet with a destination 1740 that matches one of the Bridge's own addresses, the Bridge submits 1741 the original IP packet for local delivery to support local 1742 applications such as routing protocols. 1744 o else, the Bridge drops the packet and returns an ICMP Destination 1745 Unreachable as above. 1747 As for any IP router, the Bridge decrements the TTL/Hop Limit when it 1748 forwards the carrier packet. Therefore, when an OAL header is 1749 present only the Hop Limit in the OAL header is decremented and not 1750 the TTL/Hop Limit in the original IP packet header. Bridges do not 1751 insert OAL/ORH headers themselves; instead, they act as IPv6 routers 1752 and forward carrier packets based on their destination addresses. 1754 3.11. OMNI Interface Error Handling 1756 When an AERO node admits a original IP packet into the OMNI 1757 interface, it may receive link-layer or network-layer error 1758 indications. 1760 A link-layer error indication is an ICMP error message generated by a 1761 router in the INET on the path to the neighbor or by the neighbor 1762 itself. The message includes an IP header with the address of the 1763 node that generated the error as the source address and with the 1764 link-layer address of the AERO node as the destination address. 1766 The IP header is followed by an ICMP header that includes an error 1767 Type, Code and Checksum. Valid type values include "Destination 1768 Unreachable", "Time Exceeded" and "Parameter Problem" 1769 [RFC0792][RFC4443]. (OMNI interfaces ignore all link-layer IPv4 1770 "Fragmentation Needed" and IPv6 "Packet Too Big" messages since they 1771 only emit carrier packets that are guaranteed to be no larger than 1772 the IP minimum link MTU as discussed in Section 3.9.) 1774 The ICMP header is followed by the leading portion of the original IP 1775 packet that generated the error, also known as the "packet-in-error". 1776 For ICMPv6, [RFC4443] specifies that the packet-in-error includes: 1778 "As much of invoking packet as possible without the ICMPv6 packet 1779 exceeding the minimum IPv6 MTU" (i.e., no more than 1280 bytes). For 1780 ICMPv4, [RFC0792] specifies that the packet-in-error includes: 1781 "Internet Header + 64 bits of Original Data Datagram", however 1782 [RFC1812] Section 4.3.2.3 updates this specification by stating: "the 1783 ICMP datagram SHOULD contain as much of the original datagram as 1784 possible without the length of the ICMP datagram exceeding 576 1785 bytes". 1787 The link-layer error message format is shown in Figure 5 (where, "L2" 1788 and "L3" refer to link-layer and network-layer, respectively): 1790 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1791 ~ ~ 1792 | L2 IP Header of | 1793 | error message | 1794 ~ ~ 1795 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1796 | L2 ICMP Header | 1797 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1798 ~ ~ P 1799 | IP and other encapsulation | a 1800 | headers of original IP packet | c 1801 ~ ~ k 1802 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ e 1803 ~ ~ t 1804 | IP header of | 1805 | original IP packet | i 1806 ~ ~ n 1807 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1808 ~ ~ e 1809 | Upper layer headers and | r 1810 | leading portion of body | r 1811 | of the original IP packet | o 1812 ~ ~ r 1813 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1815 Figure 5: OMNI Interface Link-Layer Error Message Format 1817 The AERO node rules for processing these link-layer error messages 1818 are as follows: 1820 o When an AERO node receives a link-layer Parameter Problem message, 1821 it processes the message the same as described as for ordinary 1822 ICMP errors in the normative references [RFC0792][RFC4443]. 1824 o When an AERO node receives persistent link-layer Time Exceeded 1825 messages, the IP ID field may be wrapping before earlier fragments 1826 awaiting reassembly have been processed. In that case, the node 1827 should begin including integrity checks and/or institute rate 1828 limits for subsequent packets. 1830 o When an AERO node receives persistent link-layer Destination 1831 Unreachable messages in response to carrier packets that it sends 1832 to one of its neighbor correspondents, the node should process the 1833 message as an indication that a path may be failing, and 1834 optionally initiate NUD over that path. If it receives 1835 Destination Unreachable messages over multiple paths, the node 1836 should allow future carrier packets destined to the correspondent 1837 to flow through a default route and re-initiate route 1838 optimization. 1840 o When an AERO Client receives persistent link-layer Destination 1841 Unreachable messages in response to carrier packets that it sends 1842 to one of its neighbor Servers, the Client should mark the path as 1843 unusable and use another path. If it receives Destination 1844 Unreachable messages on many or all paths, the Client should 1845 associate with a new Server and release its association with the 1846 old Server as specified in Section 3.16.5. 1848 o When an AERO Server receives persistent link-layer Destination 1849 Unreachable messages in response to carrier packets that it sends 1850 to one of its neighbor Clients, the Server should mark the 1851 underlying path as unusable and use another underlying path. 1853 o When an AERO Server or Proxy receives link-layer Destination 1854 Unreachable messages in response to a carrier packet that it sends 1855 to one of its permanent neighbors, it treats the messages as an 1856 indication that the path to the neighbor may be failing. However, 1857 the dynamic routing protocol should soon reconverge and correct 1858 the temporary outage. 1860 When an AERO Bridge receives a carrier packet for which the network- 1861 layer destination address is covered by an MSP, the Bridge drops the 1862 packet if there is no more-specific routing information for the 1863 destination and returns a network-layer Destination Unreachable 1864 message subject to rate limiting. The Bridge writes the network- 1865 layer source address of the original IP packet as the destination 1866 address and uses one of its non link-local addresses as the source 1867 address of the message. 1869 When an AERO node receives a carrier packet for which the reassembly 1870 buffer it too small, it drops the packet and returns a network-layer 1871 Packet Too Big (PTB) message. The node first writes the MRU value 1872 into the PTB message MTU field, writes the network-layer source 1873 address of the original IP packet as the destination address and 1874 writes one of its non link-local addresses as the source address. 1876 3.12. AERO Router Discovery, Prefix Delegation and Autoconfiguration 1878 AERO Router Discovery, Prefix Delegation and Autoconfiguration are 1879 coordinated as discussed in the following Sections. 1881 3.12.1. AERO Service Model 1883 Each AERO Server on the OMNI link is configured to facilitate Client 1884 prefix delegation/registration requests. Each Server is provisioned 1885 with a database of MNP-to-Client ID mappings for all Clients enrolled 1886 in the AERO service, as well as any information necessary to 1887 authenticate each Client. The Client database is maintained by a 1888 central administrative authority for the OMNI link and securely 1889 distributed to all Servers, e.g., via the Lightweight Directory 1890 Access Protocol (LDAP) [RFC4511], via static configuration, etc. 1891 Clients receive the same service regardless of the Servers they 1892 select. 1894 AERO Clients and Servers use ND messages to maintain neighbor cache 1895 entries. AERO Servers configure their OMNI interfaces as advertising 1896 NBMA interfaces, and therefore send unicast RA messages with a short 1897 Router Lifetime value (e.g., ReachableTime seconds) in response to a 1898 Client's RS message. Thereafter, Clients send additional RS messages 1899 to keep Server state alive. 1901 AERO Clients and Servers include prefix delegation and/or 1902 registration parameters in RS/RA messages (see 1903 [I-D.templin-6man-omni-interface]). The ND messages are exchanged 1904 between Client and Server according to the prefix management schedule 1905 required by the service. If the Client knows its MNP in advance, it 1906 can employ prefix registration by including its MNP-LLA as the source 1907 address of an RS message and with an OMNI option with valid prefix 1908 registration information for the MNP. If the Server (and Proxy) 1909 accept the Client's MNP assertion, they inject the prefix into the 1910 routing system and establish the necessary neighbor cache state. If 1911 the Client does not have a pre-assigned MNP, it can instead employ 1912 prefix delegation by including the unspecified address (::) as the 1913 source address of an RS message and with an OMNI option with prefix 1914 delegation parameters to request an MNP. 1916 The following sections specify the Client and Server behavior. 1918 3.12.2. AERO Client Behavior 1920 AERO Clients discover the addresses of Servers in a similar manner as 1921 described in [RFC5214]. Discovery methods include static 1922 configuration (e.g., from a flat-file map of Server addresses and 1923 locations), or through an automated means such as Domain Name System 1924 (DNS) name resolution [RFC1035]. Alternatively, the Client can 1925 discover Server addresses through a layer 2 data link login exchange, 1926 or through a unicast RA response to a multicast/anycast RS as 1927 described below. In the absence of other information, the Client can 1928 resolve the DNS Fully-Qualified Domain Name (FQDN) 1929 "linkupnetworks.[domainname]" where "linkupnetworks" is a constant 1930 text string and "[domainname]" is a DNS suffix for the OMNI link 1931 (e.g., "example.com"). 1933 To associate with a Server, the Client acts as a requesting router to 1934 request MNPs. The Client prepares an RS message with prefix 1935 management parameters and includes a Nonce and Timestamp option if 1936 the Client needs to correlate RA replies. If the Client already 1937 knows the Server's ADM-LLA, it includes the LLA as the network-layer 1938 destination address; otherwise, the Client includes the (link-local) 1939 All-Routers multicast as the network-layer destination. If the 1940 Client already knows its own MNP-LLA, it can use the MNP-LLA as the 1941 network-layer source address and include an OMNI option with prefix 1942 registration information. Otherwise, the Client uses the unspecified 1943 address (::) as the network-layer source address and includes prefix 1944 delegation parameters in the OMNI option (see: 1945 [I-D.templin-6man-omni-interface]). The Client includes Interface 1946 Attributes corresponding to the underlying interface over which it 1947 will send the RS message, and MAY include additional Interface 1948 Attributes specific to other underlying interfaces. 1950 The Client then sends the RS message (either directly via Direct 1951 interfaces, via a VPN for VPNed interfaces, via a Proxy for ANET 1952 interfaces or via INET encapsulation for INET interfaces) while using 1953 OAL encapsulation/fragmentation, then waits for an RA message reply 1954 (see Section 3.12.3). The Client retries up to MAX_RTR_SOLICITATIONS 1955 times until an RA is received. If the Client receives no RAs, or if 1956 it receives an RA with Router Lifetime set to 0, the Client SHOULD 1957 abandon this Server and try another Server. Otherwise, the Client 1958 processes the prefix information found in the RA message. 1960 When the Client processes an RA, it first performs OAL reassembly and 1961 decapsulation then creates a neighbor cache entry with the Server's 1962 ADM-LLA as the network-layer address and the Server's encapsulation 1963 and/or link-layer addresses as the link-layer address. The Client 1964 next records the RA Router Lifetime field value in the neighbor cache 1965 entry as the time for which the Server has committed to maintaining 1966 the MNP in the routing system via this underlying interface, and 1967 caches the other RA configuration information including Cur Hop 1968 Limit, M and O flags, Reachable Time and Retrans Timer. The Client 1969 then autoconfigures MNP-LLAs for any delegated MNPs and assigns them 1970 to the OMNI interface. The Client also caches any MSPs included in 1971 Route Information Options (RIOs) [RFC4191] as MSPs to associate with 1972 the OMNI link, and assigns the MTU value in the MTU option to the 1973 underlying interface. 1975 The Client then registers additional underlying interfaces with the 1976 Server by sending RS messages via each additional interface as 1977 described above. The RS messages include the same parameters as for 1978 the initial RS/RA exchange, but with destination address set to the 1979 Server's ADM-LLA. The Client finally sub-delegates the MNPs to its 1980 attached EUNs and/or the Client's own internal virtual interfaces as 1981 described in [I-D.templin-v6ops-pdhost] to support the Client's 1982 downstream attached "Internet of Things (IoT)". The Client then 1983 sends additional RS messages over each underlying interface before 1984 the Router Lifetime received for that interface expires. 1986 After the Client registers its underlying interfaces, it may wish to 1987 change one or more registrations, e.g., if an interface changes 1988 address or becomes unavailable, if QoS preferences change, etc. To 1989 do so, the Client prepares an RS message to send over any available 1990 underlying interface as above. The RS includes an OMNI option with 1991 prefix registration/delegation information, with Interface Attributes 1992 specific to the selected underlying interface, and with any 1993 additional Interface Attributes specific to other underlying 1994 interfaces. When the Client receives the Server's RA response, it 1995 has assurance that the Server has been updated with the new 1996 information. 1998 If the Client wishes to discontinue use of a Server it issues an RS 1999 message over any underlying interface with an OMNI option with a 2000 prefix release indication. When the Server processes the message, it 2001 releases the MNP, sets the neighbor cache entry state for the Client 2002 to DEPARTED and returns an RA reply with Router Lifetime set to 0. 2003 After a short delay (e.g., 2 seconds), the Server withdraws the MNP 2004 from the routing system. 2006 3.12.3. AERO Server Behavior 2008 AERO Servers act as IP routers and support a prefix delegation/ 2009 registration service for Clients. Servers arrange to add their ADM- 2010 LLAs to a static map of Server addresses for the link and/or the DNS 2011 resource records for the FQDN "linkupnetworks.[domainname]" before 2012 entering service. Server addresses should be geographically and/or 2013 topologically referenced, and made available for discovery by Clients 2014 on the OMNI link. 2016 When a Server receives a prospective Client's RS message on its OMNI 2017 interface, it SHOULD return an immediate RA reply with Router 2018 Lifetime set to 0 if it is currently too busy or otherwise unable to 2019 service the Client. Otherwise, the Server performs OAL reassembly 2020 and decapsulation, then authenticates the RS message and processes 2021 the prefix delegation/registration parameters. The Server first 2022 determines the correct MNPs to provide to the Client by processing 2023 the MNP-LLA prefix parameters and/or the DHCPv6 OMNI sub-option. 2024 When the Server returns the MNPs, it also creates a forwarding table 2025 entry for the MNP-ULA corresponding to each MNP so that the MNPs are 2026 propagated into the routing system (see: Section 3.2.3). For IPv6, 2027 the Server creates an IPv6 forwarding table entry for each MNP. For 2028 IPv4, the Server creates an IPv6 forwarding table entry with the 2029 IPv4-compatibility MNP-ULA prefix corresponding to the IPv4 address. 2031 The Server next creates a neighbor cache entry for the Client using 2032 the base MNP-LLA as the network-layer address and with lifetime set 2033 to no more than the smallest prefix lifetime. Next, the Server 2034 updates the neighbor cache entry by recording the information in each 2035 Interface Attributes sub-option in the RS OMNI option. The Server 2036 also records the actual OAL/INET addresses in the neighbor cache 2037 entry. 2039 Next, the Server prepares an RA message using its ADM-LLA as the 2040 network-layer source address and the network-layer source address of 2041 the RS message as the network-layer destination address. The Server 2042 sets the Router Lifetime to the time for which it will maintain both 2043 this underlying interface individually and the neighbor cache entry 2044 as a whole. The Server also sets Cur Hop Limit, M and O flags, 2045 Reachable Time and Retrans Timer to values appropriate for the OMNI 2046 link. The Server includes the MNPs, any other prefix management 2047 parameters and an OMNI option with no Interface Attributes. The 2048 Server then includes one or more RIOs that encode the MSPs for the 2049 OMNI link, plus an MTU option (see Section 3.9). The Server finally 2050 forwards the message to the Client using OAL encapsulation/ 2051 fragmentation as necessary. 2053 After the initial RS/RA exchange, the Server maintains a 2054 ReachableTime timer for each of the Client's underlying interfaces 2055 individually (and for the Client's neighbor cache entry collectively) 2056 set to expire after ReachableTime seconds. If the Client (or Proxy) 2057 issues additional RS messages, the Server sends an RA response and 2058 resets ReachableTime. If the Server receives an ND message with a 2059 prefix release indication it sets the Client's neighbor cache entry 2060 to the DEPARTED state and withdraws the MNP from the routing system 2061 after a short delay (e.g., 2 seconds). If ReachableTime expires 2062 before a new RS is received on an individual underlying interface, 2063 the Server marks the interface as DOWN. If ReachableTime expires 2064 before any new RS is received on any individual underlying interface, 2065 the Server sets the neighbor cache entry state to STALE and sets a 10 2066 second timer. If the Server has not received a new RS or ND message 2067 with a prefix release indication before the 10 second timer expires, 2068 it deletes the neighbor cache entry and withdraws the MNP from the 2069 routing system. 2071 The Server processes any ND messages pertaining to the Client and 2072 returns an NA/RA reply in response to solicitations. The Server may 2073 also issue unsolicited RA messages, e.g., with reconfigure parameters 2074 to cause the Client to renegotiate its prefix delegation/ 2075 registrations, with Router Lifetime set to 0 if it can no longer 2076 service this Client, etc. Finally, If the neighbor cache entry is in 2077 the DEPARTED state, the Server deletes the entry after DepartTime 2078 expires. 2080 Note: Clients SHOULD notify former Servers of their departures, but 2081 Servers are responsible for expiring neighbor cache entries and 2082 withdrawing routes even if no departure notification is received 2083 (e.g., if the Client leaves the network unexpectedly). Servers 2084 SHOULD therefore set Router Lifetime to ReachableTime seconds in 2085 solicited RA messages to minimize persistent stale cache information 2086 in the absence of Client departure notifications. A short Router 2087 Lifetime also ensures that proactive Client/Server RS/RA messaging 2088 will keep any NAT state alive (see above). 2090 Note: All Servers on an OMNI link MUST advertise consistent values in 2091 the RA Cur Hop Limit, M and O flags, Reachable Time and Retrans Timer 2092 fields the same as for any link, since unpredictable behavior could 2093 result if different Servers on the same link advertised different 2094 values. 2096 3.12.3.1. DHCPv6-Based Prefix Registration 2098 When a Client is not pre-provisioned with an MNP-LLA, it will need 2099 for the Server to select one or more MNPs on its behalf and set up 2100 the correct state in the AERO routing service. (A Client with a pre- 2101 provisioned MNP may also request the Server to select additional 2102 MNPs.) The DHCPv6 service [RFC8415] is used to support this 2103 requirement. 2105 When a Client needs to have the Server select MNPs, it sends an RS 2106 message with source address set to the unspecified address (::) and 2107 with an OMNI option that includes a DHCPv6 message sub-option with 2108 DHCPv6 Prefix Delegation (DHCPv6-PD) parameters. When the Server 2109 receives the RS message, it extracts the DHCPv6-PD message from the 2110 OMNI option. 2112 The Server then acts as a "Proxy DHCPv6 Client" in a message exchange 2113 with the locally-resident DHCPv6 server, which delegates MNPs and 2114 returns a DHCPv6-PD Reply message. (If the Server wishes to defer 2115 creation of MN state until the DHCPv6-PD Reply is received, it can 2116 instead act as a Lightweight DHCPv6 Relay Agent per [RFC6221] by 2117 encapsulating the DHCPv6-PD message in a Relay-forward/reply exchange 2118 with Relay Message and Interface ID options.) 2120 When the Server receives the DHCPv6-PD Reply, it adds a route to the 2121 routing system and creates an MNP-LLA based on the delegated MNP. 2122 The Server then sends an RA back to the Client with the (newly- 2123 created) MNP-LLA as the destination address and with the DHCPv6-PD 2124 Reply message coded in the OMNI option. When the Client receives the 2125 RA, it creates a default route, assigns the Subnet Router Anycast 2126 address and sets its MNP-LLA based on the delegated MNP. 2128 Note: See [I-D.templin-6man-omni-interface] for an MNP delegation 2129 alternative in which the Client can optionally avoid including a 2130 DHCPv6 message sub-option. Namely, when the Client requests a single 2131 MNP it can set the RS source to the unspecified address (::) and 2132 include a Node Identification sub-option and Preflen in the OMNI 2133 option (but with no DHCPv6 message sub-option). When the Server 2134 receives the RS message, it forwards a self-generated DHCPv6 Solicit 2135 message to the DHCPv6 server on behalf of the Client. When the 2136 Server receives the DHCPv6 Reply, it prepares an RA message with an 2137 OMNI option with Preflen information (but with no DHCPv6 message sub- 2138 option), then places the (newly-created) MNP-LLA in the RA 2139 destination address and returns the message to the Client. 2141 3.13. The AERO Proxy 2143 Clients may connect to protected-spectrum ANETs that employ physical 2144 and/or link-layer security services to facilitate communications to 2145 Servers in outside INETs. In that case, the ANET can employ an AERO 2146 Proxy. The Proxy is located at the ANET/INET border and listens for 2147 RS messages originating from or RA messages destined to ANET Clients. 2148 The Proxy acts on these control messages as follows: 2150 o when the Proxy receives an RS message from a new ANET Client, it 2151 first authenticates the message then examines the network-layer 2152 destination address. If the destination address is a Server's 2153 ADM-LLA, the Proxy proceeds to the next step. Otherwise, if the 2154 destination is (link-local) All-Routers multicast, the Proxy 2155 selects a "nearby" Server that is likely to be a good candidate to 2156 serve the Client and replaces the destination address with the 2157 Server's ADM-LLA. Next, the Proxy creates a proxy neighbor cache 2158 entry and caches the Client and Server link-layer addresses along 2159 with the OMNI option information and any other identifying 2160 information including Transaction IDs, Client Identifiers, Nonce 2161 values, etc. The Proxy finally encapsulates the (proxyed) RS 2162 message in an OAL header with source set to the Proxy's ADM-ULA 2163 and destination set to the Server's ADM-ULA. The Proxy also 2164 includes an OMNI header with an Interface Attributes option that 2165 includes its own INET address plus a unique Port Number for this 2166 Client, then forwards the message into the OMNI link spanning 2167 tree. (Note: including a unique Port Number allows the Server to 2168 distinguish different Clients located behind the same Proxy at the 2169 link-layer, whereas the link-layer addresses would otherwise be 2170 indistinguishable) 2172 o when the Server receives the RS, it authenticates the message then 2173 creates or updates a neighbor cache entry for the Client with the 2174 Proxy's ADM-ULA, INET address and Port Number as the link-layer 2175 address information. The Server then sends an RA message back to 2176 the Proxy via the spanning tree. 2178 o when the Proxy receives the RA, it authenticates the message and 2179 matches it with the proxy neighbor cache entry created by the RS. 2180 The Proxy then caches the prefix information as a mapping from the 2181 Client's MNPs to the Client's link-layer address, caches the 2182 Server's advertised Router Lifetime and sets the neighbor cache 2183 entry state to REACHABLE. The Proxy then optionally rewrites the 2184 Router Lifetime and forwards the (proxyed) message to the Client. 2185 The Proxy finally includes an MTU option (if necessary) with an 2186 MTU to use for the underlying ANET interface. 2188 After the initial RS/RA exchange, the Proxy forwards any of the 2189 Client's carrier packets for which there is no matching neighbor 2190 cache entry to a Bridge using OAL encapsulation with its own ADM-ULA 2191 as the source and the MNP-ULA corresponding to the Client as the 2192 destination. The Proxy instead forwards any carrier packets destined 2193 to a neighbor cache target directly to the target according to the 2194 OAL/link-layer information - the process of establishing neighbor 2195 cache entries is specified in Section 3.14. 2197 While the Client is still attached to the ANET, the Proxy sends NS, 2198 RS and/or unsolicited NA messages to update the Server's neighbor 2199 cache entries on behalf of the Client and/or to convey QoS updates. 2200 This allows for higher-frequency Proxy-initiated RS/RA messaging over 2201 well-connected INET infrastructure supplemented by lower-frequency 2202 Client-initiated RS/RA messaging over constrained ANET data links. 2204 If the Server ceases to send solicited advertisements, the Proxy 2205 sends unsolicited RAs on the ANET interface with destination set to 2206 (link-local) All-Nodes multicast and with Router Lifetime set to zero 2207 to inform Clients that the Server has failed. Although the Proxy 2208 engages in ND exchanges on behalf of the Client, the Client can also 2209 send ND messages on its own behalf, e.g., if it is in a better 2210 position than the Proxy to convey QoS changes, etc. For this reason, 2211 the Proxy marks any Client-originated solicitation messages (e.g. by 2212 inserting a Nonce option) so that it can return the solicited 2213 advertisement to the Client instead of processing it locally. 2215 If the Client becomes unreachable, the Proxy sets the neighbor cache 2216 entry state to DEPARTED and retains the entry for DepartTime seconds. 2217 While the state is DEPARTED, the Proxy forwards any carrier packets 2218 destined to the Client to a Bridge via OAL/ORH encapsulation. When 2219 DepartTime expires, the Proxy deletes the neighbor cache entry and 2220 discards any further carrier packets destined to this (now forgotten) 2221 Client. 2223 In some ANETs that employ a Proxy, the Client's MNP can be injected 2224 into the ANET routing system. In that case, the Client can send 2225 original IP packets without the OAL so that the ANET routing system 2226 transports the oiginal IP packets to the Proxy. This can be very 2227 beneficial, e.g., if the Client connects to the ANET via low-end data 2228 links such as some aviation wireless links. 2230 If the first-hop ANET access router is on the same underlying link 2231 and recognizes the AERO/OMNI protocol, the Client can avoid OAL 2232 encapsulation for both its control and data messages. When the 2233 Client connects to the link, it can send an unencapsulated RS message 2234 with source address set to its own MNP-LLA (or to a Temporary LLA), 2235 and with destination address set to the ADM-LLA of the Client's 2236 selected Server or to (link-local) All-Routers multicast. The Client 2237 includes an OMNI option formatted as specified in 2238 [I-D.templin-6man-omni-interface]. 2240 The Client then sends the unencapsulated RS message, which will be 2241 intercepted by the AERO-Aware access router. The access router then 2242 performs OAL encapsulation on the RS message with its own OAL/*NET 2243 addresses as the source and the OAL/*NET addresses of a Proxy as the 2244 destination. The access router further remembers the address of the 2245 Proxy so that it can encapsulate future original IP packets from the 2246 Client via the same Proxy. If the access router needs to change to a 2247 new Proxy, it simply sends another RS message toward the Server via 2248 the new Proxy on behalf of the Client. 2250 In some cases, the access router and Proxy may be one and the same 2251 node. In that case, the node would be located on the same physical 2252 link as the Client, but its message exchanges with the Server would 2253 need to pass through a security gateway at the ANET/INET border. The 2254 method for deploying access routers and Proxys (i.e. as a single node 2255 or multiple nodes) is an ANET-local administrative consideration. 2257 Note: The Proxy can apply packing as discussed in 2258 [I-D.templin-6man-omni-interface] if an opportunity arises to 2259 concatenate multiple original IP packets that will be destined to the 2260 same neighbor. 2262 3.13.1. Combined Proxy/Servers 2264 Clients may need to connect directly to Servers via INET, Direct and 2265 VPNed interfaces (i.e., non-ANET interfaces). If the Client's 2266 underlying interfaces all connect via the same INET partition, then 2267 it can connect to a single controlling Server via all interfaces. 2269 If some Client interfaces connect via different INET partitions, 2270 however, the Client still selects a set of controlling Servers and 2271 sends RS messages via their directly-connected Servers while using 2272 the ADM-LLA of the controlling Server as the destination. 2274 When a Server receives an RS with destination set to the ADM-LLA of a 2275 controlling Server, it acts as a Proxy to forward the message to the 2276 controlling Server while forwarding the corresponding RA reply to the 2277 Client. 2279 3.13.2. Detecting and Responding to Server Failures 2281 In environments where fast recovery from Server failure is required, 2282 Proxys SHOULD use proactive Neighbor Unreachability Detection (NUD) 2283 to track Server reachability in a similar fashion as for 2284 Bidirectional Forwarding Detection (BFD) [RFC5880]. Proxys can then 2285 quickly detect and react to failures so that cached information is 2286 re-established through alternate paths. The NUD control messaging is 2287 carried only over well-connected ground domain networks (i.e., and 2288 not low-end aeronautical radio links) and can therefore be tuned for 2289 rapid response. 2291 Proxys perform proactive NUD with Servers for which there are 2292 currently active ANET Clients by sending continuous NS messages in 2293 rapid succession, e.g., one message per second. The Proxy sends the 2294 NS message via the spanning tree with the Proxy's ADM-LLA as the 2295 source and the ADM-LLA of the Server as the destination. When the 2296 Proxy is also sending RS messages to the Server on behalf of ANET 2297 Clients, the resulting RA responses can be considered as equivalent 2298 hints of forward progress. This means that the Proxy need not also 2299 send a periodic NS if it has already sent an RS within the same 2300 period. If the Server fails (i.e., if the Proxy ceases to receive 2301 advertisements), the Proxy can quickly inform Clients by sending 2302 multicast RA messages on the ANET interface. 2304 The Proxy sends RA messages on the ANET interface with source address 2305 set to the Server's address, destination address set to (link-local) 2306 All-Nodes multicast, and Router Lifetime set to 0. The Proxy SHOULD 2307 send MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small 2308 delays [RFC4861]. Any Clients on the ANET that had been using the 2309 failed Server will receive the RA messages and associate with a new 2310 Server. 2312 3.13.3. Point-to-Multipoint Server Coordination 2314 In environments where Client messaging over ANETs is bandwidth- 2315 limited and/or expensive, Clients can enlist the services of the 2316 Proxy to coordinate with multiple Servers in a single RS/RA message 2317 exchange. The Client can send a single RS message to (link-local) 2318 All-Routers multicast that includes the ID's of multiple Servers in 2319 MS-Register sub-options of the OMNI option. 2321 When the Proxy receives the RS and processes the OMNI option, it 2322 sends a separate RS to each MS-Register Server ID. When the Proxy 2323 receives an RA, it can optionally return an immediate "singleton" RA 2324 to the Client or record the Server's ID for inclusion in a pending 2325 "aggregate" RA message. The Proxy can then return aggregate RA 2326 messages to the Client including multiple Server IDs in order to 2327 conserve bandwidth. Each RA includes a proper subset of the Server 2328 IDs from the original RS message, and the Proxy must ensure that the 2329 message contents of each RA are consistent with the information 2330 received from the (aggregated) Servers. 2332 Clients can thereafter employ efficient point-to-multipoint Server 2333 coordination under the assistance of the Proxy to reduce the number 2334 of messages sent over the ANET while enlisting the support of 2335 multiple Servers for fault tolerance. Clients can further include 2336 MS-Release sub-options in IPv6 ND messages to request the Proxy to 2337 release from former Servers via the procedures discussed in 2338 Section 3.16.5. 2340 The OMNI interface specification [I-D.templin-6man-omni-interface] 2341 provides further discussion of the Client/Proxy RS/RA messaging 2342 involved in point-to-multipoint coordination. 2344 3.14. AERO Address Resolution 2346 While carrier packets are flowing between a source and target node, 2347 route optimization SHOULD be used. Route optimization is initiated 2348 by the first eligible Route Optimization Source (ROS) closest to the 2349 source as follows: 2351 o For Clients on VPNed and Direct interfaces, the Server is the ROS. 2353 o For Clients on ANET interfaces, the Proxy is the ROS. 2355 o For Clients on INET interfaces, the Client itself is the ROS. 2357 o For correspondent nodes on INET/EUN interfaces serviced by a 2358 Relay, the Relay is the ROS. 2360 The route optimization procedure is conducted between the ROS and 2361 with the target Proxy/Server/Relay or the target Client itself acting 2362 as a Route Optimization Responder (ROR) in the same manner as for 2363 IPv6 ND Address Resolution and using the same NS/NA messaging. The 2364 target may either be a MNP Client serviced by a Proxy/Server, or a 2365 non-MNP correspondent reachable via a Relay. 2367 The procedures are specified in the following sections. 2369 3.14.1. Route Optimization Initiation 2371 When an original IP packet from a source node destined to a target 2372 node arrives, the ROS checks for a neighbor cache entry for the LLA 2373 that matches the target destination. If there is a neighbor cache 2374 entry in the REACHABLE state, the ROS forwards the packets according 2375 to the cached state and returns from processing. Otherwise, if there 2376 is already a neighbor cache entry in the STALE state the ROS 2377 continues, and if there or no neighbor cache entry the ROS creates 2378 one in the INCOMPLETE state. 2380 The ROS next places the original IP packet on a short queue then 2381 sends an NS message for Address Resolution (NS(AR)) to receive a 2382 solicited NA(AR) message from a ROR. The NS(AR) message must be no 2383 larger than the OAL minimum MPS so that its entire contents will fit 2384 in an OAL first fragment (i.e., as an "atomic fragment"). The ROS 2385 prepares an NS(AR) that includes: 2387 o the LLA of the ROS as the source address. 2389 o the LLA corresponding to the original IP packet's destination as 2390 the Target Address, e.g., for 2001:db8:1:2::10:2000 the LLA is 2391 fe80::2001:db8:1:2. 2393 o the Solicited-Node multicast address [RFC4291] formed from the 2394 lower 24 bits of the original IP packet's destination as the 2395 destination address, e.g., for 2001:db8:1:2::10:2000 the NS 2396 destination address is ff02:0:0:0:0:1:ff10:2000. 2398 The NS(AR) message also includes an OMNI option with an Interface 2399 Attributes entry for the sending interface, and with Prefix Length 2400 set to the length associated with the ROS's LLA. The ROS then 2401 submits the NS(AR) message for OAL encapsulation and fragmentation, 2402 with OAL source set to its own ULA and OAL destination set to the ULA 2403 corresponding to the target, and with an unpredictable initial 2404 Identification value selected according to [RFC7739]. The source 2405 caches the initial Identification value in the (newly-created) 2406 neighbor cache entry as the starting sequence number for the "send" 2407 window for future carrier packets sent to this target. 2409 The ROS then sends the resulting carrier packet into the spanning 2410 tree without decrementing the network-layer TTL/Hop Limit field. 2411 (When the ROS is a Client, it instead must first sign the NS(AR) 2412 message and send the resulting carrier packet to one of its current 2413 Servers which then verifies the NS(AR) signature and forwards the 2414 carrier packet into the spanning tree on behalf of the Client.) 2416 3.14.2. Relaying the NS(AR) *NET Packet(s) 2418 When the Bridge receives the carrier packet from the ROS, it discards 2419 the *NET headers and determines the next hop by consulting its 2420 standard IPv6 forwarding table for the OAL header destination 2421 address. The Bridge then re-encapsulates and forwards the carrier 2422 packet via the spanning tree the same as for any IPv6 router, where 2423 it may traverse multiple OMNI link segments. The final-hop Bridge in 2424 the spanning tree will deliver the carrier packet via a secured 2425 tunnel to a Server or Relay that services the target. 2427 3.14.3. Processing the NS(AR) and Sending the NA(AR) 2429 When the Proxy/Server/Relay receives the carrier packet, it examines 2430 the enclosed atomic OAL fragment to determine that it contains an 2431 NS(AR) then examines the NS(AR) target to determine whether it has a 2432 matching neighbor cache entry and/or non-MNP route. If there is no 2433 match, the Proxy/Server/Relay drops the message. Otherwise, the ROR 2434 continues processing as follows: 2436 o if the NS(AR) target matches a Client neighbor cache entry in the 2437 DEPARTED state, the Proxy/Server inserts an ORH with destination 2438 prefix set to the lower 64 bits of the Client's MNP-ULA and sets 2439 the destination address to the ADM-ULA of the Client's new Server. 2441 The (old) Proxy/Server then re-encapsulates the carrier packet, 2442 forwards it into the spanning tree and returns from processing. 2444 o If the NS(AR) target matches a Client neighbor cache entry in the 2445 REACHABLE state, the Proxy/Server proceeds according to whether 2446 the Client requires Proxy services. For "disadvantaged" Clients 2447 (e.g., those on low-end links) and Clients with complex multilink 2448 arrangements the Proxy/Server nominates itself as the ROR; 2449 otherwise, the Proxy/Server re-encapsulates the carrier packet and 2450 forwards it to the target Client which will act as an ROR on its 2451 own behalf. 2453 o If the NS(AR) target matches one of its non-MNP routes, the Relay 2454 acts the ROR. 2456 The ROR next checks for a neighbor cache entry that matches the 2457 NS(AR) source LLA. If there is a neighbor cache entry in the 2458 REACHABLE state, the ROR accepts the NS(AR) only if the OAL 2459 Identification value is within the "accept" window for this NS(AR) 2460 source LLA or if the NS(AR) was forwarded by the (secured) spanning 2461 tree. Otherwise, the ROR creates a neighbor cache entry for the 2462 NS(AR) source LLA if necessary with state set to STALE. If the ROR 2463 is a Proxy/Server, it next adds a Report List entry to the target 2464 Client neighbor cache entry for the NS(AR) source LLA while caching 2465 the OAL Identification value in the (newly-created) neighbor cache 2466 entry as the starting sequence number for the "accept" window for 2467 future carrier packets received from this LLA source. 2469 The ROR then prepares a (solicited) NA(AR) message to send back to 2470 the ROS using the same Identification value received in the NS(AR) 2471 (unlike the NS(AR), the NA(AR) need not fit in a single OAL 2472 fragment). The ROR sets the NA(AR) source address to its own LLA, 2473 sets the destination address to the NS(AR) LLA source address and 2474 sets the Target Address to the same value that was in the NS(AR). 2475 The ROR then includes an OMNI option with Prefix Length set to the 2476 length associated with the MNP-LLA. 2478 If the NS(AR) target was an MNP Client, the ROR next includes 2479 Interface Attributes in the OMNI option for each of the target 2480 Client's underlying interfaces with current information for each 2481 interface. If the ROR is a Proxy/Server/Relay, it then sets the the 2482 S/T-ifIndex field in the OMNI header set to 0. If the ROR is the 2483 Client itself, it instead sets S/T-ifIndex to the index of the 2484 underlying interface that will send the NS(AR). 2486 For each Interface Attributes sub-option, the ROR sets the L2ADDR 2487 according to the Proxy/Server's INET address for VPNed or Direct 2488 interfaces, to the INET address of the Proxy for Proxyed interfaces 2489 or to the Client's INET address for INET interfaces. The ROR then 2490 includes the lower 32 bits of the Proxy/Server's ADM-ULA as the LHS, 2491 encodes the ADM-ULA prefix length code in the SRT field and sets the 2492 FMT code accordingly as specified in Section 3.3. 2494 The ROR then sets the NA(AR) message R flag to 1 (as a router) and S 2495 flag to 1 (as a response to a solicitation). If the ROR is the 2496 Client itself, it sets the O flag to 1; if the ROR is the Server, it 2497 instead sets the O flag to 0 (as a proxy). The ROR finally submits 2498 the NA(AR) for OAL encapsulation with source set to its own ULA and 2499 destination set to the source ULA of the NS(AR) message, then 2500 performs OAL fragmentation using the same Identification value that 2501 appeared in the NS(AR) and forwards the resulting (*NET-encapsulated) 2502 carrier packets into the spanning tree without decrementing the 2503 network-layer TTL/Hop Limit field. 2505 3.14.4. Relaying the NA(AR) 2507 When the Bridge receives the carrier packets from the ROR, it 2508 discards the *NET header and determines the next hop by consulting 2509 its standard IPv6 forwarding table for the OAL header destination 2510 address. The Bridge then re-encapsulates and forwards the carrier 2511 packets via the spanning tree the same as for any IPv6 router, where 2512 it may traverse multiple OMNI link segments. The final-hop Bridge in 2513 the spanning tree will deliver the carrier packet via a secured 2514 tunnel to a Server for the ROS. 2516 3.14.5. Processing the NA(AR) 2518 When the ROS receives the NA(AR) message, it first searches for a 2519 neighbor cache entry that matches the NA(AR) target address. If 2520 there is an entry in the INCOMPLETE or STALE state, the ROS matches 2521 the OAL Identification value with the value it had included in the 2522 corresponding NS(AR). If the values match, the ROS processes the 2523 message the same as for standard IPv6 Address Resolution [RFC4861]. 2524 In the process, it caches the target LLA and all information found in 2525 the OMNI option in the neighbor cache entry for the target. The ROS 2526 finally sets the neighbor cache entry state to REACHABLE and sets its 2527 lifetime to ReachableTime seconds. (When the ROS is a Client, the 2528 solicited NA(AR) message will first be delivered via the spanning 2529 tree to one of its current Servers, which then securely forwards the 2530 message to the Client.) 2532 3.14.6. Route Optimization Maintenance 2534 Following route optimization, the ROS forwards future carrier packets 2535 with user data destined to the target via the addresses found in the 2536 cached link-layer information. The route optimization is shared by 2537 all sources that send original IP packets to the target via the ROS, 2538 i.e., and not just the source on behalf of which the route 2539 optimization was initiated. Note that route optimization is 2540 performed only for original IP packets that contain user data, and 2541 not for those that contain other IPv6 ND control messages. 2543 While the ROS continues to forward additional original IP packets 2544 destined to the target, it sends additional NS(AR) messages to the 2545 ROR before ReachableTime expires to receive a fresh NA(AR) message 2546 the same as described in the previous sections (route optimization 2547 refreshment strategies are an implementation matter, with a non- 2548 normative example given in Appendix A.1). The ROS may supply a new 2549 unpredictable OAL Identification value if it wishes to reset the 2550 neighbor's "accept" Identification window. If the ROS is a Client, 2551 it mst sign the NS(AR) message so that the Server can authenticate. 2553 The ROS uses the cached ULA of the ROR (i.e., either the ADM-ULA of 2554 the Server or the MNP-ULA of the Client itself) as the NS(AR) OAL 2555 destination address, and sends up to MAX_MULTICAST_SOLICIT NS(AR) 2556 messages separated by 1 second until an NA(AR) is received. If no 2557 NA(AR) is received, the ROS assumes that the current ROR has become 2558 unreachable and deletes the target neighbor cache entry. Subsequent 2559 original IP packets will trigger a new route optimization event (see: 2560 Section 3.14.1). 2562 If an NA(AR) is received, the ROS then updates the neighbor cache 2563 entry to refresh ReachableTime, while (for MNP destinations) the ROR 2564 adds or updates the ROS address to the target's Report List and with 2565 time set to ReportTime. While no carrier packets are flowing, the 2566 ROS instead allows ReachableTime for the neighbor cache entry to 2567 expire. When ReachableTime expires, the ROS deletes the neighbor 2568 cache entry. Any future carrier packets flowing through the ROS will 2569 again trigger a new route optimization. 2571 The ROS may also receive unsolicited NA messages from the ROR at any 2572 time (see: Section 3.16). If there is a neighbor cache entry for the 2573 target, the ROS updates the link-layer information but does not 2574 update ReachableTime since the receipt of an unsolicited NA does not 2575 confirm that any forward paths are working. If there is no neighbor 2576 cache entry, the ROS simply discards the unsolicited NA. 2578 In this arrangement, the ROS holds a neighbor cache entry with only a 2579 "send" Identification window value for the target, while the ROR 2580 holds a neighbor cache entry with only an "accept" Identification 2581 window value for the ROS. The route optimization neighbor 2582 relationship is therefore asymmetric and unidirectional. If the 2583 target node also has carrier packets to send back to the source node, 2584 then a separate route optimization procedure is performed in the 2585 reverse direction. But, there is no requirement that the forward and 2586 reverse paths be symmetric. 2588 3.15. Neighbor Unreachability Detection (NUD) 2590 AERO nodes perform Neighbor Unreachability Detection (NUD) per 2591 [RFC4861] either reactively in response to persistent link-layer 2592 errors (see Section 3.11) or proactively to confirm reachability and/ 2593 or establish NAT state. The NUD algorithm is based on periodic 2594 control message exchanges. The algorithm may further be seeded by ND 2595 hints of forward progress, but care must be taken to avoid inferring 2596 reachability based on spoofed information. For example, authentic 2597 IPv6 ND message exchanges may be considered as acceptable hints of 2598 forward progress, while spurious random packets should not be. 2600 AERO nodes can use standard NS/NA exchanges sent over the OMNI link 2601 spanning tree to securely test reachability without risk of DoS 2602 attacks from nodes pretending to be a neighbor (these NS/NA(NUD) 2603 messages use the unicast LLAs and ULAs of the two parties involved in 2604 the NUD test the same as for standard IPv6 ND, and both messages flow 2605 over the spanning tree). Proxys can further perform NUD to securely 2606 verify Server reachability on behalf of their proxyed Clients. 2607 However, a means for an ROS to test the unsecured target route 2608 optimized paths is also necessary. 2610 When an ROR directs an ROS to a target neighbor with one or more 2611 link-layer addresses, the ROS can proactively test each such 2612 unsecured route optimized path through secured NS(NUD) messages over 2613 the spanning tree that invoke an unsecured NA(NUD) reply that travels 2614 over the route optimized path.. (The NS(NUD) messages must therefore 2615 include Nonce and Timestamp options that will be echoed in the 2616 unsecured NA(NUD) replies.) While testing the paths, the ROS can 2617 optionally continue to send carrier packets via the spanning tree, 2618 maintain a small queue of carrier packets until target reachability 2619 is confirmed, or (optimistically) allow carrier packets to flow via 2620 the route optimized paths. 2622 When the ROS sends an NS(NUD) message, it sets the IPv6 source to its 2623 own address, sets the destination to the MNP-LLA of the target, and 2624 sets the target's MNP Subnet-Router anycast address as the Target 2625 Address. The ROS also includes an OMNI option with a single 2626 Interface Attributes sub-option with the SRT, FMT, LHS and L2ADDR 2627 information for its own underlying interface it wishes to test, but 2628 sets the S/T-ifIndex field to the index for target's underlying 2629 interface to be tested. The ROS includes a Nonce and Timestamp 2630 option, then encapsulates the message in OAL/INET headers with its 2631 own ULA as the source and the ULA of the target as the destination. 2633 The ROS then forwards the NS(NUD) message toward the target via a 2634 Server or Bridge. 2636 When the target receives the NS(NUD) message, it creates an NA(NUD) 2637 by reversing the OAL and IPv6 addresses and including an Interface 2638 Attributes sub-option with attributes for its own interface 2639 identified by the NS(NUD) S/T-ifIndex. The target sets the NA(NUD) 2640 S/T-ifIndex to the index of the ROS, sets the Target Address to the 2641 same value that was in the NS(NUD), and returns the message using its 2642 own underlying interface identified by S/T-ifIndex and destined to 2643 the ROS's interface identified by the original Interface Attributes 2644 sub-option. 2646 When the ROS receives the NA(NUD) message, it can determine from the 2647 Nonce, Timestamp and Target Address that the message matched its 2648 NS(NUD) and that it transited the direct path from the ROR using the 2649 selected underlying interface pair. The ROS marks route optimization 2650 target paths that pass these NUD tests as "reachable", and those that 2651 do not as "unreachable". These markings inform the OMNI interface 2652 forwarding algorithm specified in Section 3.10. 2654 Note: If the target determines that the OMNI option Interface 2655 Attributes in the NS(NUD) is located in a different OMNI link segment 2656 than its own interface named in the S/T-ifIndex, it instead returns 2657 the NA(NUD) via the spanning tree while including an ORH and setting 2658 the OAL destination address to the Subnet Router Anycast address used 2659 by Bridges on the ROS segment. When a Bridge on the ROS segment 2660 receives the NA(NUD), it replaces the Interface Attributes with 2661 information for its own interface while using the ifIndex value 2662 specific to the target. 2664 3.16. Mobility Management and Quality of Service (QoS) 2666 AERO is a Distributed Mobility Management (DMM) service. Each Server 2667 is responsible for only a subset of the Clients on the OMNI link, as 2668 opposed to a Centralized Mobility Management (CMM) service where 2669 there is a single network mobility collective entity for all Clients. 2670 Clients coordinate with their associated Servers via RS/RA exchanges 2671 to maintain the DMM profile, and the AERO routing system tracks all 2672 current Client/Server peering relationships. 2674 Servers provide default routing and mobility/multilink services for 2675 their dependent Clients. Clients are responsible for maintaining 2676 neighbor relationships with their Servers through periodic RS/RA 2677 exchanges, which also serves to confirm neighbor reachability. When 2678 a Client's underlying interface address and/or QoS information 2679 changes, the Client is responsible for updating the Server with this 2680 new information. Note that when there is a Proxy in the path, the 2681 Proxy can also perform some RS/RA exchanges on the Client's behalf. 2683 Mobility management messaging is based on the transmission and 2684 reception of unsolicited Neighbor Advertisement (uNA) messages. Each 2685 uNA message sets the IPv6 destination address to (link-local) All- 2686 Nodes multicast to convey a general update of Interface Attributes to 2687 (possibly) multiple recipients, or to a specific unicast LLA to 2688 announce a departure event to a specific recipient. Implementations 2689 must therefore examine the destination address to determine the 2690 nature of the mobility event (i.e., update vs departure). 2692 Mobility management considerations are specified in the following 2693 sections. 2695 3.16.1. Mobility Update Messaging 2697 Servers accommodate Client mobility, multilink and/or QoS change 2698 events by sending unsolicited NA (uNA) messages to each ROS in the 2699 target Client's Report List. When a Server sends a uNA message, it 2700 sets the IPv6 source address to the Client's MNP-LLA, sets the 2701 destination address to (link-local) All-Nodes multicast and sets the 2702 Target Address to the Client's Subnet-Router anycast address. The 2703 Server also includes an OMNI option with Prefix Length set to the 2704 length associated with the Client's MNP-LLA, with Interface 2705 Attributes for the target Client's underlying interfaces and with the 2706 OMNI header S/T-ifIndex set to 0. The Server then sets the NA R flag 2707 to 1, the S flag to 0 and the O flag to 1, then encapsulates the 2708 message in an OAL header with source set to its own ADM-ULA and 2709 destination set to the ULA of the ROS and sends the message into the 2710 spanning tree. 2712 As discussed in Section 7.2.6 of [RFC4861], the transmission and 2713 reception of uNA messages is unreliable but provides a useful 2714 optimization. In well-connected Internetworks with robust data links 2715 uNA messages will be delivered with high probability, but in any case 2716 the Server can optionally send up to MAX_NEIGHBOR_ADVERTISEMENT uNAs 2717 to each ROS to increase the likelihood that at least one will be 2718 received. 2720 When the ROS receives a uNA message prepared as above, it ignores the 2721 message if there is no existing neighbor cache entry for the Client. 2722 Otherwise, it uses the included OMNI option information to update the 2723 neighbor cache entry, but does not reset ReachableTime since the 2724 receipt of an unsolicited NA message from the target Server does not 2725 provide confirmation that any forward paths to the target Client are 2726 working. 2728 If uNA messages are lost, the ROS may be left with stale address and/ 2729 or QoS information for the Client for up to ReachableTime seconds. 2730 During this time, the ROS can continue sending carrier packets 2731 according to its stale neighbor cache information. When 2732 ReachableTime is close to expiring, the ROS will re-initiate route 2733 optimization and receive fresh link-layer address information. 2735 In addition to sending uNA messages to the current set of ROSs for 2736 the Client, the Server also sends uNAs to the MNP-ULA associated with 2737 the link-layer address for any underlying interface for which the 2738 link-layer address has changed. These uNA messages update an old 2739 Proxy/Server that cannot easily detect (e.g., without active probing) 2740 when a formerly-active Client has departed. When the Server sends 2741 the uNA, it sets the IPv6 source address to the Client's MNP-LLA, 2742 sets the destination address to the old Proxy/Server's ADM-LLA, and 2743 sets the Target Address to the Client's Subnet-Router anycast 2744 address. The Server also includes an OMNI option with Prefix Length 2745 set to the length associated with the Client's MNP-LLA, with 2746 Interface Attributes for the changed underlying interface, and with 2747 the OMNI header S/T-ifIndex set to 0. The Server then sets the NA R 2748 flag to 1, the S flag to 0 and the O flag to 1, then encapsulates the 2749 message in an OAL header with source set to its own ADM-ULA and 2750 destination set to the ADM-ULA of the old Proxy/Server and sends the 2751 message into the spanning tree. 2753 3.16.2. Announcing Link-Layer Address and/or QoS Preference Changes 2755 When a Client needs to change its underlying interface addresses and/ 2756 or QoS preferences (e.g., due to a mobility event), either the Client 2757 or its Proxys send RS messages to the Server via the spanning tree 2758 with an OMNI option that includes Interface attributes with the new 2759 link quality and address information. 2761 Up to MAX_RTR_SOLICITATIONS RS messages MAY be sent in parallel with 2762 sending carrier packets containing user data in case one or more RAs 2763 are lost. If all RAs are lost, the Client SHOULD re-associate with a 2764 new Server. 2766 When the Server receives the Client's changes, it sends uNA messages 2767 to all nodes in the Report List the same as described in the previous 2768 section. 2770 3.16.3. Bringing New Links Into Service 2772 When a Client needs to bring new underlying interfaces into service 2773 (e.g., when it activates a new data link), it sends an RS message to 2774 the Server via the underlying interface with an OMNI option that 2775 includes Interface Attributes with appropriate link quality values 2776 and with link-layer address information for the new link. 2778 3.16.4. Deactivating Existing Links 2780 When a Client needs to deactivate an existing underlying interface, 2781 it sends an RS or uNA message to its Server with an OMNI option with 2782 appropriate Interface Attribute values - in particular, the link 2783 quality value 0 assures that neighbors will cease to use the link. 2785 If the Client needs to send RS/uNA messages over an underlying 2786 interface other than the one being deactivated, it MUST include 2787 Interface Attributes with appropriate link quality values for any 2788 underlying interfaces being deactivated. 2790 Note that when a Client deactivates an underlying interface, 2791 neighbors that have received the RS/uNA messages need not purge all 2792 references for the underlying interface from their neighbor cache 2793 entries. The Client may reactivate or reuse the underlying interface 2794 and/or its ifIndex at a later point in time, when it will send RS/uNA 2795 messages with fresh Interface Attributes to update any neighbors. 2797 3.16.5. Moving Between Servers 2799 The Client performs the procedures specified in Section 3.12.2 when 2800 it first associates with a new Server or renews its association with 2801 an existing Server. The Client also includes MS-Release identifiers 2802 in the RS message OMNI option per [I-D.templin-6man-omni-interface] 2803 if it wants the new Server to notify any old Servers from which the 2804 Client is departing. 2806 When the new Server receives the Client's RS message, it returns an 2807 RA as specified in Section 3.12.3 and sends up to 2808 MAX_NEIGHBOR_ADVERTIISEMENT uNA messages to any old Servers listed in 2809 OMNI option MS-Release identifiers. When the new Server sends a uNA 2810 message, it sets the IPv6 source address to the Client's MNP-LLA, 2811 sets the destination address to the old Server's ADM-LLA, and sets 2812 the Target Address to the Client's Subnet-Router anycast address. 2813 The new Server also includes an OMNI option with Prefix Length set to 2814 the length associated with the Client's MNP-LLA, with Interface 2815 Attributes for its own underlying interface, and with the OMNI header 2816 S/T-ifIndex set to 0. The new Server then sets the NA R flag to 1, 2817 the S flag to 0 and the O flag to 1, then encapsulates the message in 2818 an OAL header with source set to its own ADM-ULA and destination set 2819 to the ADM-ULA of the old Server and sends the message into the 2820 spanning tree. 2822 When an old Server receives the uNA, it changes the Client's neighbor 2823 cache entry state to DEPARTED, sets the link-layer address of the 2824 Client to the new Server's ADM-ULA, and resets DepartTime. After a 2825 short delay (e.g., 2 seconds) the old Server withdraws the Client's 2826 MNP from the routing system. After DepartTime expires, the old 2827 Server deletes the Client's neighbor cache entry. 2829 The old Server also iteratively forwards a copy of the uNA message to 2830 each ROS in the Client's Report List by changing the OAL destination 2831 address to the ULA of the ROS while leaving all other fields of the 2832 message unmodified. When the ROS receives the uNA, it examines the 2833 Target address to determine the correct neighbor cache entry and 2834 verifies that the IPv6 destination address matches the old Server. 2835 The ROS then caches the IPv6 source address as the new Server for the 2836 existing neighbor cache entry and marks the entry as STALE. While in 2837 the STALE state, the ROS allows new carrier packets to flow according 2838 to any existing cached link-layer information and sends new NS(AR) 2839 messages using its own ULA as the OAL source and the ADM-ULA of the 2840 new Server as the OAL destination address to elicit NA messages that 2841 reset the neighbor cache entry state to REACHABLE. If no new NA 2842 message is received for 10 seconds while in the STALE state, the ROS 2843 deletes the neighbor cache entry. 2845 Clients SHOULD NOT move rapidly between Servers in order to avoid 2846 causing excessive oscillations in the AERO routing system. Examples 2847 of when a Client might wish to change to a different Server include a 2848 Server that has gone unreachable, topological movements of 2849 significant distance, movement to a new geographic region, movement 2850 to a new OMNI link segment, etc. 2852 When a Client moves to a new Server, some of the fragments of a 2853 multiple fragment OAL packet may have already arrived at the old 2854 Server while others are en route to the new Server, however no 2855 special attention in the reassembly algorithm is necessary since all 2856 fragments will eventually be delivered to the Client which can then 2857 reassemble. 2859 3.17. Multicast 2861 The AERO Client provides an IGMP (IPv4) [RFC2236] or MLD (IPv6) 2862 [RFC3810] proxy service for its EUNs and/or hosted applications 2863 [RFC4605]. The Client forwards IGMP/MLD messages over any of its 2864 underlying interfaces for which group membership is required. The 2865 IGMP/MLD messages may be further forwarded by a first-hop ANET access 2866 router acting as an IGMP/MLD-snooping switch [RFC4541], then 2867 ultimately delivered to an AERO Proxy/Server acting as a Protocol 2868 Independent Multicast - Sparse-Mode (PIM-SM, or simply "PIM") 2869 Designated Router (DR) [RFC7761]. AERO Relays also act as PIM 2870 routers (i.e., the same as AERO Proxys/Servers) on behalf of nodes on 2871 INET/EUN networks. The behaviors identified in the following 2872 sections correspond to Source-Specific Multicast (SSM) and Any-Source 2873 Multicast (ASM) operational modes. 2875 3.17.1. Source-Specific Multicast (SSM) 2877 When an ROS (i.e., an AERO Proxy/Server/Relay) "X" acting as PIM 2878 router receives a Join/Prune message from a node on its downstream 2879 interfaces containing one or more ((S)ource, (G)roup) pairs, it 2880 updates its Multicast Routing Information Base (MRIB) accordingly. 2881 For each S belonging to a prefix reachable via X's non-OMNI 2882 interfaces, X then forwards the (S, G) Join/Prune to any PIM routers 2883 on those interfaces per [RFC7761]. 2885 For each S belonging to a prefix reachable via X's OMNI interface, X 2886 originates a separate copy of the Join/Prune for each (S,G) in the 2887 message using its own LLA as the source address and ALL-PIM-ROUTERS 2888 as the destination address. X then encapsulates each message in an 2889 OAL header with source address set to the ULA of X and destination 2890 address set to S then forwards the message into the spanning tree, 2891 which delivers it to AERO Server/Relay "Y" that services S. At the 2892 same time, if the message was a Join, X sends a route-optimization NS 2893 message toward each S the same as discussed in Section 3.14. The 2894 resulting NAs will return the LLA for the prefix that matches S as 2895 the network-layer source address and with an OMNI option with the ULA 2896 corresponding to any underlying interfaces that are currently 2897 servicing S. 2899 When Y processes the Join/Prune message, if S located behind any 2900 INET, Direct, or VPNed interfaces Y acts as a PIM router and updates 2901 its MRIB to list X as the next hop in the reverse path. If S is 2902 located behind any Proxys "Z"*, Y also forwards the message to each 2903 Z* over the spanning tree while continuing to use the LLA of X as the 2904 source address. Each Z* then updates its MRIB accordingly and 2905 maintains the LLA of X as the next hop in the reverse path. Since 2906 the Bridges do not examine network layer control messages, this means 2907 that the (reverse) multicast tree path is simply from each Z* (and/or 2908 Y) to X with no other multicast-aware routers in the path. If any Z* 2909 (and/or Y) is located on the same OMNI link segment as X, the 2910 multicast data traffic sent to X directly using OAL/INET 2911 encapsulation instead of via a Bridge. 2913 Following the initial Join/Prune and NS/NA messaging, X maintains a 2914 neighbor cache entry for each S the same as if X was sending unicast 2915 data traffic to S. In particular, X performs additional NS/NA 2916 exchanges to keep the neighbor cache entry alive for up to t_periodic 2917 seconds [RFC7761]. If no new Joins are received within t_periodic 2918 seconds, X allows the neighbor cache entry to expire. Finally, if X 2919 receives any additional Join/Prune messages for (S,G) it forwards the 2920 messages to each Y and Z* in the neighbor cache entry over the 2921 spanning tree. 2923 At some later time, Client C that holds an MNP for source S may 2924 depart from a first Proxy Z1 and/or connect via a new Proxy Z2. In 2925 that case, Y sends an unsolicited NA message to X the same as 2926 specified for unicast mobility in Section 3.16. When X receives the 2927 unsolicited NA message, it updates its neighbor cache entry for the 2928 LLA for source S and sends new Join messages to any new Proxys Z2. 2929 There is no requirement to send any Prune messages to old Proxys Z1 2930 since source S will no longer source any multicast data traffic via 2931 Z1. Instead, the multicast state for (S,G) in Proxy Z1 will soon 2932 time out since no new Joins will arrive. 2934 After some later time, C may move to a new Server Y2 and depart from 2935 old Sever Y1. In that case, Y1 sends Join messages for any of C's 2936 active (S,G) groups to Y2 while including its own LLA as the source 2937 address. This causes Y2 to include Y1 in the multicast forwarding 2938 tree during the interim time that Y1's neighbor cache entry for C is 2939 in the DEPARTED state. At the same time, Y1 sends an unsolicited NA 2940 message to X with an OMNI option with S/T-ifIndex in the header set 2941 to 0 and a release indication to cause X to release its neighbor 2942 cache entry. X then sends a new Join message to S via the spanning 2943 tree and re-initiates route optimization the same as if it were 2944 receiving a fresh Join message from a node on a downstream link. 2946 3.17.2. Any-Source Multicast (ASM) 2948 When an ROS X acting as a PIM router receives a Join/Prune from a 2949 node on its downstream interfaces containing one or more (*,G) pairs, 2950 it updates its Multicast Routing Information Base (MRIB) accordingly. 2951 X then forwards a copy of the message to the Rendezvous Point (RP) R 2952 for each G over the spanning tree. X uses its own LLA as the source 2953 address and ALL-PIM-ROUTERS as the destination address, then 2954 encapsulates each message in an OAL header with source address set to 2955 the ULA of X and destination address set to R, then sends the message 2956 into the spanning tree. At the same time, if the message was a Join 2957 X initiates NS/NA route optimization the same as for the SSM case 2958 discussed in Section 3.17.1. 2960 For each source S that sends multicast traffic to group G via R, the 2961 Proxy/Server Z* for the Client that aggregates S encapsulates the 2962 original IP packets in PIM Register messages and forwards them to R 2963 via the spanning tree, which may then elect to send a PIM Join to Z*. 2964 This will result in an (S,G) tree rooted at Z* with R as the next hop 2965 so that R will begin to receive two copies of the original IP packet; 2966 one native copy from the (S, G) tree and a second copy from the pre- 2967 existing (*, G) tree that still uses PIM Register encapsulation. R 2968 can then issue a PIM Register-stop message to suppress the Register- 2969 encapsulated stream. At some later time, if C moves to a new Proxy/ 2970 Server Z*, it resumes sending original IP packets via PIM Register 2971 encapsulation via the new Z*. 2973 At the same time, as multicast listeners discover individual S's for 2974 a given G, they can initiate an (S,G) Join for each S under the same 2975 procedures discussed in Section 3.17.1. Once the (S,G) tree is 2976 established, the listeners can send (S, G) Prune messages to R so 2977 that multicast original IP packets for group G sourced by S will only 2978 be delivered via the (S, G) tree and not from the (*, G) tree rooted 2979 at R. All mobility considerations discussed for SSM apply. 2981 3.17.3. Bi-Directional PIM (BIDIR-PIM) 2983 Bi-Directional PIM (BIDIR-PIM) [RFC5015] provides an alternate 2984 approach to ASM that treats the Rendezvous Point (RP) as a Designated 2985 Forwarder (DF). Further considerations for BIDIR-PIM are out of 2986 scope. 2988 3.18. Operation over Multiple OMNI Links 2990 An AERO Client can connect to multiple OMNI links the same as for any 2991 data link service. In that case, the Client maintains a distinct 2992 OMNI interface for each link, e.g., 'omni0' for the first link, 2993 'omni1' for the second, 'omni2' for the third, etc. Each OMNI link 2994 would include its own distinct set of Bridges, Servers and Proxys, 2995 thereby providing redundancy in case of failures. 2997 Each OMNI link could utilize the same or different ANET connections. 2998 The links can be distinguished at the link-layer via the SRT prefix 2999 in a similar fashion as for Virtual Local Area Network (VLAN) tagging 3000 (e.g., IEEE 802.1Q) and/or through assignment of distinct sets of 3001 MSPs on each link. This gives rise to the opportunity for supporting 3002 multiple redundant networked paths, with each VLAN distinguished by a 3003 different SRT "color" (see: Section 3.2.5). 3005 The Client's IP layer can select the outgoing OMNI interface 3006 appropriate for a given traffic profile while (in the reverse 3007 direction) correspondent nodes must have some way of steering their 3008 original IP packets destined to a target via the correct OMNI link. 3010 In a first alternative, if each OMNI link services different MSPs, 3011 then the Client can receive a distinct MNP from each of the links. 3012 IP routing will therefore assure that the correct OMNI link is used 3013 for both outbound and inbound traffic. This can be accomplished 3014 using existing technologies and approaches, and without requiring any 3015 special supporting code in correspondent nodes or Bridges. 3017 In a second alternative, if each OMNI link services the same MSP(s) 3018 then each link could assign a distinct "OMNI link Anycast" address 3019 that is configured by all Bridges on the link. Correspondent nodes 3020 can then perform Segment Routing to select the correct SRT, which 3021 will then direct the original IP packet over multiple hops to the 3022 target. 3024 3.19. DNS Considerations 3026 AERO Client MNs and INET correspondent nodes consult the Domain Name 3027 System (DNS) the same as for any Internetworking node. When 3028 correspondent nodes and Client MNs use different IP protocol versions 3029 (e.g., IPv4 correspondents and IPv6 MNs), the INET DNS must maintain 3030 A records for IPv4 address mappings to MNs which must then be 3031 populated in Relay NAT64 mapping caches. In that way, an IPv4 3032 correspondent node can send original IPv4 packets to the IPv4 address 3033 mapping of the target MN, and the Relay will translate the IPv4 3034 header and destination address into an IPv6 header and IPv6 3035 destination address of the MN. 3037 When an AERO Client registers with an AERO Server, the Server can 3038 return the address(es) of DNS servers in RDNSS options [RFC6106]. 3039 The DNS server provides the IP addresses of other MNs and 3040 correspondent nodes in AAAA records for IPv6 or A records for IPv4. 3042 3.20. Transition Considerations 3044 OAL encapsulation ensures that dissimilar INET partitions can be 3045 joined into a single unified OMNI link, even though the partitions 3046 themselves may have differing protocol versions and/or incompatible 3047 addressing plans. However, a commonality can be achieved by 3048 incrementally distributing globally routable (i.e., native) IP 3049 prefixes to eventually reach all nodes (both mobile and fixed) in all 3050 OMNI link segments. This can be accomplished by incrementally 3051 deploying AERO Relays on each INET partition, with each Relay 3052 distributing its MNPs and/or discovering non-MNP IP GUA prefixes on 3053 its INET links. 3055 This gives rise to the opportunity to eventually distribute native IP 3056 addresses to all nodes, and to present a unified OMNI link view even 3057 if the INET partitions remain in their current protocol and 3058 addressing plans. In that way, the OMNI link can serve the dual 3059 purpose of providing a mobility/multilink service and a transition 3060 service. Or, if an INET partition is transitioned to a native IP 3061 protocol version and addressing scheme that is compatible with the 3062 OMNI link MNP-based addressing scheme, the partition and OMNI link 3063 can be joined by Relays. 3065 Relays that connect INETs/EUNs with dissimilar IP protocol versions 3066 may need to employ a network address and protocol translation 3067 function such as NAT64 [RFC6146]. 3069 3.21. Detecting and Reacting to Server and Bridge Failures 3071 In environments where rapid failure recovery is required, Servers and 3072 Bridges SHOULD use Bidirectional Forwarding Detection (BFD) 3073 [RFC5880]. Nodes that use BFD can quickly detect and react to 3074 failures so that cached information is re-established through 3075 alternate nodes. BFD control messaging is carried only over well- 3076 connected ground domain networks (i.e., and not low-end radio links) 3077 and can therefore be tuned for rapid response. 3079 Servers and Bridges maintain BFD sessions in parallel with their BGP 3080 peerings. If a Server or Bridge fails, BGP peers will quickly re- 3081 establish routes through alternate paths the same as for common BGP 3082 deployments. Similarly, Proxys maintain BFD sessions with their 3083 associated Bridges even though they do not establish BGP peerings 3084 with them. 3086 Proxys SHOULD use proactive NUD for Servers for which there are 3087 currently active ANET Clients in a manner that parallels BFD, i.e., 3088 by sending unicast NS messages in rapid succession to receive 3089 solicited NA messages. When the Proxy is also sending RS messages on 3090 behalf of ANET Clients, the RS/RA messaging can be considered as 3091 equivalent hints of forward progress. This means that the Proxy need 3092 not also send a periodic NS if it has already sent an RS within the 3093 same period. If a Server fails, the Proxy will cease to receive 3094 advertisements and can quickly inform Clients of the outage by 3095 sending multicast RA messages on the ANET interface. 3097 The Proxy sends multicast RA messages with source address set to the 3098 Server's address, destination address set to (link-local) All-Nodes 3099 multicast, and Router Lifetime set to 0. The Proxy SHOULD send 3100 MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small delays 3101 [RFC4861]. Any Clients on the ANET interface that have been using 3102 the (now defunct) Server will receive the RA messages and associate 3103 with a new Server. 3105 3.22. AERO Clients on the Open Internet 3107 AERO Clients that connect to the open Internet via INET interfaces 3108 can establish a VPN or direct link to securely connect to a Server in 3109 a "tethered" arrangement with all of the Client's traffic transiting 3110 the Server. Alternatively, the Client can associate with an INET 3111 Server using UDP/IP encapsulation and control message securing 3112 services as discussed in the following sections. 3114 When a Client's OMNI interface enables an INET underlying interface, 3115 it first determines whether the interface is likely to be behind a 3116 NAT. For IPv4, the Client assumes it is on the open Internet if the 3117 INET address is not a special-use IPv4 address per [RFC3330]. 3118 Similarly for IPv6, the Client assumes it is on the open Internet if 3119 the INET address is a Global Unicast Address (GUA) [RFC4291]. 3120 Otherwise, the Client assumes it may be behind one or several NATs. 3122 The Client then prepares an RS message with IPv6 source address set 3123 to its MNP-LLA, with IPv6 destination set to (link-local) All-Routers 3124 multicast and with an OMNI option with underlying interface 3125 attributes. If the Client believes that it is on the open Internet, 3126 it SHOULD include an L2ADDR in the Interface Attributes sub-option 3127 corresponding to the underlying interface; otherwise, it MAY omit 3128 L2ADDR. If the underlying address is IPv4, the Client includes the 3129 Port Number and IPv4 address written in obfuscated form [RFC4380] as 3130 discussed in Section 3.3. If the underlying interface address is 3131 IPv6, the Client instead includes the Port Number and IPv6 address in 3132 obfuscated form. The Client finally includes a HIP "Initiator" 3133 message sub-option in the OMNI option 3134 [I-D.templin-6man-omni-interface] to provide message authentication 3135 and performs OAL encapsulation and fragmentation if necessary. The 3136 Client then encapsulates each fragment in UDP/IP headers, sets the 3137 UDP/IP source to its INET address and UDP port, sets the UDP/IP 3138 destination to the Server's INET address and the AERO service port 3139 number (8060), then sends the (whole or fragmented) message to the 3140 Server. 3142 When the Server receives the RS, it performs OAL reassembly if 3143 necessary then authenticates the message and registers the Client's 3144 MNP and INET interface information according to the OMNI option 3145 parameters. If the RS message OMNI option includes Interface 3146 Attributes with an L2ADDR, the Server compares the encapsulation IP 3147 address and UDP port number with the (unobfuscated) values. If the 3148 values are the same, the Server caches the Client's information as 3149 "INET" addresses meaning that the Client is likely to accept direct 3150 messages without requiring NAT traversal exchanges. If the values 3151 are different (or, if the OMNI option did not include an L2ADDR) the 3152 Server instead caches the Client's information as "mapped" addresses 3153 meaning that NAT traversal exchanges may be necessary. 3155 The Server then prepares an RA message with IPv6 source and 3156 destination set corresponding to the addresses in the RS, and with an 3157 OMNI option with an Origin Indication sub-option per 3159 [I-D.templin-6man-omni-interface] with the mapped and obfuscated Port 3160 Number and IP address observed in the encapsulation headers. The 3161 Server also includes a HIP "Responder" message sub-option per 3162 [I-D.templin-6man-omni-interface] that contains an acknowledgement of 3163 the update sent by the Client. The Server then performs OAL 3164 encapsulation and fragmentation if necessary, and encapsulates each 3165 fragment in UDP/IP headers with addresses set per the L2ADDR 3166 information in the neighbor cache entry for the Client. 3168 When the Client receives the RA message, it performs OAL reassembly 3169 if necessary, authenticates the HIP "Responder" message, then 3170 compares the mapped Port Number and IP address from the Origin 3171 Indication sub-option with its own address. If the addresses are the 3172 same, the Client assumes the open Internet / Cone NAT principle; if 3173 the addresses are different, the Client instead assumes that further 3174 qualification procedures are necessary to detect the type of NAT and 3175 proceeds according to standard procedures [RFC6081][RFC4380]. 3177 After the Client has registered its INET interfaces in such RS/RA 3178 exchanges it sends periodic RS messages to receive fresh RA messages 3179 before the Router Lifetime received on each INET interface expires. 3180 The Client also maintains default routes via its Servers, i.e., the 3181 same as described in earlier sections. 3183 When the Client sends messages to target IP addresses, it also 3184 invokes route optimization per Section 3.14 using IPv6 ND address 3185 resolution messaging. The Client sends the NS(AR) message to the 3186 Server with an OMNI option with a HIP "Update/Sequence" message sub- 3187 option. The Client sets the NS source address to the Client's MNP- 3188 LLA and destination address to the target solicited node multicast 3189 address. The Client wraps the NS message in an OAL header with 3190 source address set to its own MNP-ULA and destination address set to 3191 the Server's ADM-ULA. The Client then performs OAL fragmentation if 3192 necessary, wraps each fragment in a UDP/IP header and sends it to the 3193 Server. 3195 When the Server receives the OAL-encapsulated NS, it performs OAL 3196 reassembly if necessary. The Server then authenticates the message 3197 by processing the HIP message sub-option and sends a corresponding 3198 NS(AR) message over the spanning tree the same as if it were the ROS, 3199 but with the OAL source address set to the Server's ADM-ULA, with 3200 destination set to the MNP-ULA of the target, and with an OMNI option 3201 that includes no sub-options. When the ROR receives the NS(AR), it 3202 adds the Server's ADM-ULA and Client's MNP-LLA to the target's Report 3203 List, and returns an NA(AR) with OMNI option information for the 3204 target including all of the target's Interface Attributes. The ROR 3205 sets the NA(AR) source address to the target's MNP-LLA and sets the 3206 destination address to the Client's MNP-LLA, then sets the OAL source 3207 address to the ADM-ULA of the ROR and the destination to the ADM-ULA 3208 of the Server. When the Server receives the NA(AR) message, it 3209 rewrites the OAL source address to its own ADM-ULA and the 3210 destination address to the MNP-ULA of the Client, then includes a HIP 3211 "Update/Acknowledge" message sub-option in the OMNI option, performs 3212 OAL fragmentation if necessary, wraps each fragment in UDP/IP 3213 headers, and sends it to the Client. 3215 Following route optimization for targets in the same OMNI link 3216 segment, if the target's L2ADDR is on the open INET, the Client 3217 forwards carrier packets directly to the target INET address. If the 3218 target is behind a NAT, the Client first establishes NAT state for 3219 the L2ADDR using the "direct bubble" and NUD mechanisms discussed in 3220 Section 3.10.1. The Client continues to send carrier packets via its 3221 Server until NAT state is populated, then begins forwarding carrier 3222 packets via the direct path through the NAT to the target. For 3223 targets in different OMNI link segments, the Client uses OAL/ORH 3224 encapsulation and forwards carrier packets to the Bridge that 3225 returned the NA message. 3227 The ROR may return uNAs via the Server if the target moves, and the 3228 Server will send corresponding uNAs to the Client with a HIP "Notify" 3229 authentication message. The Client can also send NUD messages to 3230 test forward path reachability even though there is no security 3231 association between the Client and the target. 3233 The Client can send original IP packets to route-optimized neighbors 3234 in the same OMNI link segment no larger than the path MPS in one 3235 piece and without OAL encapsulation and fragmentation, but with *NET 3236 encapsulation. For larger original IP packets, the Client applies 3237 OAL encapsulation and fragmentation if necessary according to 3238 Section 3.9, with OAL header with source set to its own MNP-ULA and 3239 destination set to the MNP-ULA of the target. The Client then 3240 encapsulates each original IP packet or OAL fragment in UDP/IP *NET 3241 headers and sends them to the next hop. 3243 Note: The NAT traversal procedures specified in this document are 3244 applicable for Cone, Address-Restricted and Port-Restricted NATs 3245 only. While future updates to this document may specify procedures 3246 for other NAT variations (e.g., hairpinning and various forms of 3247 Symmetric NATs), it should be noted that continuous communications 3248 are always possible through forwarding via a Server even if NAT 3249 traversal is not employed. 3251 Note: Following the initial HIP Initiator/Responder exchange, AERO 3252 Clients with OMNI interfaces configured over the open Internet 3253 maintain HIP associations through the transmission of IPv6 ND 3254 messages that include OMNI options with HIP "Update" and "Notify" 3255 messages. OMNI interfaces use the HIP "Update" message when an 3256 acknowledgement is required, and use the "Notify" message in 3257 unacknowledged isolated IPv6 ND messages (e.g., unsolicited NAs). 3259 Note: Servers on the open Internet that act as Proxys authenticate 3260 and remove OMNI option HIP message sub-options from RSes they forward 3261 from the MN to another Server, and insert and sign HIP message and 3262 Origin Indication sub-options in RAs they forward from another Server 3263 to the MN. Conversely, Servers that act as Proxys forward without 3264 processing any MNP registration/delegation information in RS/RA 3265 message exchanges between MNs and other Servers. The Server acting 3266 as a Proxy is therefore responsible for MN authentication, while the 3267 other Servers are responsible for registering/delegating MNPs (noting 3268 that the same node can act as both Proxy and Server). 3270 3.23. Time-Varying MNPs 3272 In some use cases, it is desirable, beneficial and efficient for the 3273 Client to receive a constant MNP that travels with the Client 3274 wherever it moves. For example, this would allow air traffic 3275 controllers to easily track aircraft, etc. In other cases, however 3276 (e.g., intelligent transportation systems), the MN may be willing to 3277 sacrifice a modicum of efficiency in order to have time-varying MNPs 3278 that can be changed every so often to defeat adversarial tracking. 3280 The DHCPv6 service offers a way for Clients that desire time-varying 3281 MNPs to obtain short-lived prefixes (e.g., on the order of a small 3282 number of minutes). In that case, the identity of the Client would 3283 not be bound to the MNP but rather to a Node Identification value 3284 (see: [I-D.templin-6man-omni-interface]) to be used as the Client ID 3285 seed for MNP prefix delegation. The Client would then be obligated 3286 to renumber its internal networks whenever its MNP (and therefore 3287 also its MNP-LLA) changes. This should not present a challenge for 3288 Clients with automated network renumbering services, however presents 3289 limits for the durations of ongoing sessions that would prefer to use 3290 a constant address. 3292 4. Implementation Status 3294 An early AERO implementation based on OpenVPN (https://openvpn.net/) 3295 was announced on the v6ops mailing list on January 10, 2018 and an 3296 initial public release of the AERO proof-of-concept source code was 3297 announced on the intarea mailing list on August 21, 2015. 3299 AERO Release-3.0.2 was tagged on October 15, 2020, and is undergoing 3300 internal testing. Additional internal releases expected within the 3301 coming months, with first public release expected end of 1H2021. 3303 5. IANA Considerations 3305 The IANA is instructed to assign a new type value TBD1 in the IPv6 3306 Routing Types registry. 3308 The IANA has assigned the UDP port number "8060" for an earlier 3309 experimental first version of AERO [RFC6706]. This document 3310 obsoletes [RFC6706], and together with 3311 [I-D.templin-6man-omni-interface] reclaims the UDP port number "8060" 3312 for 'aero' as the service port for UDP/IP encapsulation. (Note that, 3313 although [RFC6706] was not widely implemented or deployed, any 3314 messages coded to that specification can be easily distinguished and 3315 ignored since they use the invalid ICMPv6 message type number '0'.) 3316 This document makes no request of IANA, since 3317 [I-D.templin-6man-omni-interface] already provides instructions. 3319 No further IANA actions are required. 3321 6. Security Considerations 3323 AERO Bridges configure secured tunnels with AERO Servers, Relays and 3324 Proxys within their local OMNI link segments. Applicable secured 3325 tunnel alternatives include IPsec [RFC4301], TLS/SSL [RFC8446], DTLS 3326 [RFC6347], WireGuard [WG], etc. The AERO Bridges of all OMNI link 3327 segments in turn configure secured tunnels for their neighboring AERO 3328 Bridges in a spanning tree topology. Therefore, control messages 3329 exchanged between any pair of OMNI link neighbors on the spanning 3330 tree are already secured. 3332 AERO nodes acting as Route Optimization Responders (RORs) may also 3333 receive packets directly from arbitrary nodes in INET partitions 3334 instead of via the spanning tree. For INET partitions that apply 3335 effective ingress filtering to defeat source address spoofing, the 3336 simple data origin authentication procedures in Section 3.8 can be 3337 applied. 3339 For INET partitions that require strong security in the data plane, 3340 two options for securing communications include 1) disable route 3341 optimization so that all traffic is conveyed over secured tunnels, or 3342 2) enable on-demand secure tunnel creation between INET partition 3343 neighbors. Option 1) would result in longer routes than necessary 3344 and traffic concentration on critical infrastructure elements. 3345 Option 2) could be coordinated by establishing a secured tunnel on- 3346 demand instead of performing an NS/NA exchange in the route 3347 optimization procedures. Procedures for establishing on-demand 3348 secured tunnels are out of scope. 3350 AERO Clients that connect to secured ANETs need not apply security to 3351 their ND messages, since the messages will be intercepted by a 3352 perimeter Proxy that applies security on its INET-facing interface as 3353 part of the spanning tree (see above). AERO Clients connected to the 3354 open INET can use network and/or transport layer security services 3355 such as VPNs or can by some other means establish a direct link. 3356 When a VPN or direct link may be impractical, however, the 3357 authentication services specified in [RFC7401] and/or [RFC4380] 3358 should be applied. 3360 Application endpoints SHOULD use application-layer security services 3361 such as TLS/SSL, DTLS or SSH [RFC4251] to assure the same level of 3362 protection as for critical secured Internet services. AERO Clients 3363 that require host-based VPN services SHOULD use network and/or 3364 transport layer security services such as IPsec, TLS/SSL, DTLS, etc. 3365 AERO Proxys and Servers can also provide a network-based VPN service 3366 on behalf of the Client, e.g., if the Client is located within a 3367 secured enclave and cannot establish a VPN on its own behalf. 3369 AERO Servers and Bridges present targets for traffic amplification 3370 Denial of Service (DoS) attacks. This concern is no different than 3371 for widely-deployed VPN security gateways in the Internet, where 3372 attackers could send spoofed packets to the gateways at high data 3373 rates. This can be mitigated by connecting Servers and Bridges over 3374 dedicated links with no connections to the Internet and/or when 3375 connections to the Internet are only permitted through well-managed 3376 firewalls. Traffic amplification DoS attacks can also target an AERO 3377 Client's low data rate links. This is a concern not only for Clients 3378 located on the open Internet but also for Clients in secured 3379 enclaves. AERO Servers and Proxys can institute rate limits that 3380 protect Clients from receiving packet floods that could DoS low data 3381 rate links. 3383 AERO Relays must implement ingress filtering to avoid a spoofing 3384 attack in which spurious messages with ULA addresses are injected 3385 into an OMNI link from an outside attacker. AERO Clients MUST ensure 3386 that their connectivity is not used by unauthorized nodes on their 3387 EUNs to gain access to a protected network, i.e., AERO Clients that 3388 act as routers MUST NOT provide routing services for unauthorized 3389 nodes. (This concern is no different than for ordinary hosts that 3390 receive an IP address delegation but then "share" the address with 3391 other nodes via some form of Internet connection sharing such as 3392 tethering.) 3394 The MAP list MUST be well-managed and secured from unauthorized 3395 tampering, even though the list contains only public information. 3396 The MAP list can be conveyed to the Client in a similar fashion as in 3398 [RFC5214] (e.g., through layer 2 data link login messaging, secure 3399 upload of a static file, DNS lookups, etc.). 3401 SRH authentication facilities are specified in [RFC8754]. 3403 Security considerations for accepting link-layer ICMP messages and 3404 reflected packets are discussed throughout the document. 3406 Security considerations for IPv6 fragmentation and reassembly are 3407 discussed in [I-D.templin-6man-omni-interface]. 3409 7. Acknowledgements 3411 Discussions in the IETF, aviation standards communities and private 3412 exchanges helped shape some of the concepts in this work. 3413 Individuals who contributed insights include Mikael Abrahamsson, Mark 3414 Andrews, Fred Baker, Bob Braden, Stewart Bryant, Brian Carpenter, 3415 Wojciech Dec, Pavel Drasil, Ralph Droms, Adrian Farrel, Nick Green, 3416 Sri Gundavelli, Brian Haberman, Bernhard Haindl, Joel Halpern, Tom 3417 Herbert, Sascha Hlusiak, Lee Howard, Zdenek Jaron, Andre Kostur, 3418 Hubert Kuenig, Ted Lemon, Andy Malis, Satoru Matsushima, Tomek 3419 Mrugalski, Madhu Niraula, Alexandru Petrescu, Behcet Saikaya, Michal 3420 Skorepa, Joe Touch, Bernie Volz, Ryuji Wakikawa, Tony Whyman, Lloyd 3421 Wood and James Woodyatt. Members of the IESG also provided valuable 3422 input during their review process that greatly improved the document. 3423 Special thanks go to Stewart Bryant, Joel Halpern and Brian Haberman 3424 for their shepherding guidance during the publication of the AERO 3425 first edition. 3427 This work has further been encouraged and supported by Boeing 3428 colleagues including Kyle Bae, M. Wayne Benson, Dave Bernhardt, Cam 3429 Brodie, John Bush, Balaguruna Chidambaram, Irene Chin, Bruce Cornish, 3430 Claudiu Danilov, Don Dillenburg, Joe Dudkowski, Wen Fang, Samad 3431 Farooqui, Anthony Gregory, Jeff Holland, Seth Jahne, Brian Jaury, 3432 Greg Kimberly, Ed King, Madhuri Madhava Badgandi, Laurel Matthew, 3433 Gene MacLean III, Rob Muszkiewicz, Sean O'Sullivan, Vijay 3434 Rajagopalan, Greg Saccone, Rod Santiago, Kent Shuey, Brian Skeen, 3435 Mike Slane, Carrie Spiker, Katie Tran, Brendan Williams, Amelia 3436 Wilson, Julie Wulff, Yueli Yang, Eric Yeh and other members of the 3437 Boeing mobility, networking and autonomy teams. Kyle Bae, Wayne 3438 Benson, Katie Tran and Eric Yeh are especially acknowledged for 3439 implementing the AERO functions as extensions to the public domain 3440 OpenVPN distribution. 3442 Earlier works on NBMA tunneling approaches are found in 3443 [RFC2529][RFC5214][RFC5569]. 3445 Many of the constructs presented in this second edition of AERO are 3446 based on the author's earlier works, including: 3448 o The Internet Routing Overlay Network (IRON) 3449 [RFC6179][I-D.templin-ironbis] 3451 o Virtual Enterprise Traversal (VET) 3452 [RFC5558][I-D.templin-intarea-vet] 3454 o The Subnetwork Encapsulation and Adaptation Layer (SEAL) 3455 [RFC5320][I-D.templin-intarea-seal] 3457 o AERO, First Edition [RFC6706] 3459 Note that these works cite numerous earlier efforts that are not also 3460 cited here due to space limitations. The authors of those earlier 3461 works are acknowledged for their insights. 3463 This work is aligned with the NASA Safe Autonomous Systems Operation 3464 (SASO) program under NASA contract number NNA16BD84C. 3466 This work is aligned with the FAA as per the SE2025 contract number 3467 DTFAWA-15-D-00030. 3469 This work is aligned with the Boeing Commercial Airplanes (BCA) 3470 Internet of Things (IoT) and autonomy programs. 3472 This work is aligned with the Boeing Information Technology (BIT) 3473 MobileNet program. 3475 8. References 3477 8.1. Normative References 3479 [I-D.templin-6man-omni-interface] 3480 Templin, F. and T. Whyman, "Transmission of IP Packets 3481 over Overlay Multilink Network (OMNI) Interfaces", draft- 3482 templin-6man-omni-interface-69 (work in progress), January 3483 2021. 3485 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 3486 DOI 10.17487/RFC0791, September 1981, 3487 . 3489 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 3490 RFC 792, DOI 10.17487/RFC0792, September 1981, 3491 . 3493 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 3494 Requirement Levels", BCP 14, RFC 2119, 3495 DOI 10.17487/RFC2119, March 1997, 3496 . 3498 [RFC2473] Conta, A. and S. Deering, "Generic Packet Tunneling in 3499 IPv6 Specification", RFC 2473, DOI 10.17487/RFC2473, 3500 December 1998, . 3502 [RFC3971] Arkko, J., Ed., Kempf, J., Zill, B., and P. Nikander, 3503 "SEcure Neighbor Discovery (SEND)", RFC 3971, 3504 DOI 10.17487/RFC3971, March 2005, 3505 . 3507 [RFC3972] Aura, T., "Cryptographically Generated Addresses (CGA)", 3508 RFC 3972, DOI 10.17487/RFC3972, March 2005, 3509 . 3511 [RFC4191] Draves, R. and D. Thaler, "Default Router Preferences and 3512 More-Specific Routes", RFC 4191, DOI 10.17487/RFC4191, 3513 November 2005, . 3515 [RFC4193] Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast 3516 Addresses", RFC 4193, DOI 10.17487/RFC4193, October 2005, 3517 . 3519 [RFC4380] Huitema, C., "Teredo: Tunneling IPv6 over UDP through 3520 Network Address Translations (NATs)", RFC 4380, 3521 DOI 10.17487/RFC4380, February 2006, 3522 . 3524 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 3525 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 3526 DOI 10.17487/RFC4861, September 2007, 3527 . 3529 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 3530 Address Autoconfiguration", RFC 4862, 3531 DOI 10.17487/RFC4862, September 2007, 3532 . 3534 [RFC6081] Thaler, D., "Teredo Extensions", RFC 6081, 3535 DOI 10.17487/RFC6081, January 2011, 3536 . 3538 [RFC7401] Moskowitz, R., Ed., Heer, T., Jokela, P., and T. 3539 Henderson, "Host Identity Protocol Version 2 (HIPv2)", 3540 RFC 7401, DOI 10.17487/RFC7401, April 2015, 3541 . 3543 [RFC7739] Gont, F., "Security Implications of Predictable Fragment 3544 Identification Values", RFC 7739, DOI 10.17487/RFC7739, 3545 February 2016, . 3547 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 3548 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 3549 May 2017, . 3551 [RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6 3552 (IPv6) Specification", STD 86, RFC 8200, 3553 DOI 10.17487/RFC8200, July 2017, 3554 . 3556 [RFC8415] Mrugalski, T., Siodelski, M., Volz, B., Yourtchenko, A., 3557 Richardson, M., Jiang, S., Lemon, T., and T. Winters, 3558 "Dynamic Host Configuration Protocol for IPv6 (DHCPv6)", 3559 RFC 8415, DOI 10.17487/RFC8415, November 2018, 3560 . 3562 8.2. Informative References 3564 [BGP] Huston, G., "BGP in 2015, http://potaroo.net", January 3565 2016. 3567 [I-D.bonica-6man-comp-rtg-hdr] 3568 Bonica, R., Kamite, Y., Alston, A., Henriques, D., and L. 3569 Jalil, "The IPv6 Compact Routing Header (CRH)", draft- 3570 bonica-6man-comp-rtg-hdr-24 (work in progress), January 3571 2021. 3573 [I-D.bonica-6man-crh-helper-opt] 3574 Li, X., Bao, C., Ruan, E., and R. Bonica, "Compressed 3575 Routing Header (CRH) Helper Option", draft-bonica-6man- 3576 crh-helper-opt-02 (work in progress), October 2020. 3578 [I-D.ietf-intarea-frag-fragile] 3579 Bonica, R., Baker, F., Huston, G., Hinden, R., Troan, O., 3580 and F. Gont, "IP Fragmentation Considered Fragile", draft- 3581 ietf-intarea-frag-fragile-17 (work in progress), September 3582 2019. 3584 [I-D.ietf-intarea-tunnels] 3585 Touch, J. and M. Townsley, "IP Tunnels in the Internet 3586 Architecture", draft-ietf-intarea-tunnels-10 (work in 3587 progress), September 2019. 3589 [I-D.ietf-ipwave-vehicular-networking] 3590 Jeong, J., "IPv6 Wireless Access in Vehicular Environments 3591 (IPWAVE): Problem Statement and Use Cases", draft-ietf- 3592 ipwave-vehicular-networking-19 (work in progress), July 3593 2020. 3595 [I-D.ietf-rtgwg-atn-bgp] 3596 Templin, F., Saccone, G., Dawra, G., Lindem, A., and V. 3597 Moreno, "A Simple BGP-based Mobile Routing System for the 3598 Aeronautical Telecommunications Network", draft-ietf- 3599 rtgwg-atn-bgp-10 (work in progress), January 2021. 3601 [I-D.templin-6man-dhcpv6-ndopt] 3602 Templin, F., "A Unified Stateful/Stateless Configuration 3603 Service for IPv6", draft-templin-6man-dhcpv6-ndopt-11 3604 (work in progress), January 2021. 3606 [I-D.templin-intarea-seal] 3607 Templin, F., "The Subnetwork Encapsulation and Adaptation 3608 Layer (SEAL)", draft-templin-intarea-seal-68 (work in 3609 progress), January 2014. 3611 [I-D.templin-intarea-vet] 3612 Templin, F., "Virtual Enterprise Traversal (VET)", draft- 3613 templin-intarea-vet-40 (work in progress), May 2013. 3615 [I-D.templin-ipwave-uam-its] 3616 Templin, F., "Urban Air Mobility Implications for 3617 Intelligent Transportation Systems", draft-templin-ipwave- 3618 uam-its-04 (work in progress), January 2021. 3620 [I-D.templin-ironbis] 3621 Templin, F., "The Interior Routing Overlay Network 3622 (IRON)", draft-templin-ironbis-16 (work in progress), 3623 March 2014. 3625 [I-D.templin-v6ops-pdhost] 3626 Templin, F., "IPv6 Prefix Delegation and Multi-Addressing 3627 Models", draft-templin-v6ops-pdhost-27 (work in progress), 3628 January 2021. 3630 [OVPN] OpenVPN, O., "http://openvpn.net", October 2016. 3632 [RFC1035] Mockapetris, P., "Domain names - implementation and 3633 specification", STD 13, RFC 1035, DOI 10.17487/RFC1035, 3634 November 1987, . 3636 [RFC1812] Baker, F., Ed., "Requirements for IP Version 4 Routers", 3637 RFC 1812, DOI 10.17487/RFC1812, June 1995, 3638 . 3640 [RFC2003] Perkins, C., "IP Encapsulation within IP", RFC 2003, 3641 DOI 10.17487/RFC2003, October 1996, 3642 . 3644 [RFC2004] Perkins, C., "Minimal Encapsulation within IP", RFC 2004, 3645 DOI 10.17487/RFC2004, October 1996, 3646 . 3648 [RFC2236] Fenner, W., "Internet Group Management Protocol, Version 3649 2", RFC 2236, DOI 10.17487/RFC2236, November 1997, 3650 . 3652 [RFC2464] Crawford, M., "Transmission of IPv6 Packets over Ethernet 3653 Networks", RFC 2464, DOI 10.17487/RFC2464, December 1998, 3654 . 3656 [RFC2529] Carpenter, B. and C. Jung, "Transmission of IPv6 over IPv4 3657 Domains without Explicit Tunnels", RFC 2529, 3658 DOI 10.17487/RFC2529, March 1999, 3659 . 3661 [RFC2983] Black, D., "Differentiated Services and Tunnels", 3662 RFC 2983, DOI 10.17487/RFC2983, October 2000, 3663 . 3665 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 3666 of Explicit Congestion Notification (ECN) to IP", 3667 RFC 3168, DOI 10.17487/RFC3168, September 2001, 3668 . 3670 [RFC3330] IANA, "Special-Use IPv4 Addresses", RFC 3330, 3671 DOI 10.17487/RFC3330, September 2002, 3672 . 3674 [RFC3810] Vida, R., Ed. and L. Costa, Ed., "Multicast Listener 3675 Discovery Version 2 (MLDv2) for IPv6", RFC 3810, 3676 DOI 10.17487/RFC3810, June 2004, 3677 . 3679 [RFC4122] Leach, P., Mealling, M., and R. Salz, "A Universally 3680 Unique IDentifier (UUID) URN Namespace", RFC 4122, 3681 DOI 10.17487/RFC4122, July 2005, 3682 . 3684 [RFC4251] Ylonen, T. and C. Lonvick, Ed., "The Secure Shell (SSH) 3685 Protocol Architecture", RFC 4251, DOI 10.17487/RFC4251, 3686 January 2006, . 3688 [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A 3689 Border Gateway Protocol 4 (BGP-4)", RFC 4271, 3690 DOI 10.17487/RFC4271, January 2006, 3691 . 3693 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 3694 Architecture", RFC 4291, DOI 10.17487/RFC4291, February 3695 2006, . 3697 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 3698 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 3699 December 2005, . 3701 [RFC4389] Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery 3702 Proxies (ND Proxy)", RFC 4389, DOI 10.17487/RFC4389, April 3703 2006, . 3705 [RFC4443] Conta, A., Deering, S., and M. Gupta, Ed., "Internet 3706 Control Message Protocol (ICMPv6) for the Internet 3707 Protocol Version 6 (IPv6) Specification", STD 89, 3708 RFC 4443, DOI 10.17487/RFC4443, March 2006, 3709 . 3711 [RFC4511] Sermersheim, J., Ed., "Lightweight Directory Access 3712 Protocol (LDAP): The Protocol", RFC 4511, 3713 DOI 10.17487/RFC4511, June 2006, 3714 . 3716 [RFC4541] Christensen, M., Kimball, K., and F. Solensky, 3717 "Considerations for Internet Group Management Protocol 3718 (IGMP) and Multicast Listener Discovery (MLD) Snooping 3719 Switches", RFC 4541, DOI 10.17487/RFC4541, May 2006, 3720 . 3722 [RFC4605] Fenner, B., He, H., Haberman, B., and H. Sandick, 3723 "Internet Group Management Protocol (IGMP) / Multicast 3724 Listener Discovery (MLD)-Based Multicast Forwarding 3725 ("IGMP/MLD Proxying")", RFC 4605, DOI 10.17487/RFC4605, 3726 August 2006, . 3728 [RFC4982] Bagnulo, M. and J. Arkko, "Support for Multiple Hash 3729 Algorithms in Cryptographically Generated Addresses 3730 (CGAs)", RFC 4982, DOI 10.17487/RFC4982, July 2007, 3731 . 3733 [RFC5015] Handley, M., Kouvelas, I., Speakman, T., and L. Vicisano, 3734 "Bidirectional Protocol Independent Multicast (BIDIR- 3735 PIM)", RFC 5015, DOI 10.17487/RFC5015, October 2007, 3736 . 3738 [RFC5214] Templin, F., Gleeson, T., and D. Thaler, "Intra-Site 3739 Automatic Tunnel Addressing Protocol (ISATAP)", RFC 5214, 3740 DOI 10.17487/RFC5214, March 2008, 3741 . 3743 [RFC5320] Templin, F., Ed., "The Subnetwork Encapsulation and 3744 Adaptation Layer (SEAL)", RFC 5320, DOI 10.17487/RFC5320, 3745 February 2010, . 3747 [RFC5522] Eddy, W., Ivancic, W., and T. Davis, "Network Mobility 3748 Route Optimization Requirements for Operational Use in 3749 Aeronautics and Space Exploration Mobile Networks", 3750 RFC 5522, DOI 10.17487/RFC5522, October 2009, 3751 . 3753 [RFC5558] Templin, F., Ed., "Virtual Enterprise Traversal (VET)", 3754 RFC 5558, DOI 10.17487/RFC5558, February 2010, 3755 . 3757 [RFC5569] Despres, R., "IPv6 Rapid Deployment on IPv4 3758 Infrastructures (6rd)", RFC 5569, DOI 10.17487/RFC5569, 3759 January 2010, . 3761 [RFC5880] Katz, D. and D. Ward, "Bidirectional Forwarding Detection 3762 (BFD)", RFC 5880, DOI 10.17487/RFC5880, June 2010, 3763 . 3765 [RFC6106] Jeong, J., Park, S., Beloeil, L., and S. Madanapalli, 3766 "IPv6 Router Advertisement Options for DNS Configuration", 3767 RFC 6106, DOI 10.17487/RFC6106, November 2010, 3768 . 3770 [RFC6146] Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful 3771 NAT64: Network Address and Protocol Translation from IPv6 3772 Clients to IPv4 Servers", RFC 6146, DOI 10.17487/RFC6146, 3773 April 2011, . 3775 [RFC6179] Templin, F., Ed., "The Internet Routing Overlay Network 3776 (IRON)", RFC 6179, DOI 10.17487/RFC6179, March 2011, 3777 . 3779 [RFC6221] Miles, D., Ed., Ooghe, S., Dec, W., Krishnan, S., and A. 3780 Kavanagh, "Lightweight DHCPv6 Relay Agent", RFC 6221, 3781 DOI 10.17487/RFC6221, May 2011, 3782 . 3784 [RFC6273] Kukec, A., Krishnan, S., and S. Jiang, "The Secure 3785 Neighbor Discovery (SEND) Hash Threat Analysis", RFC 6273, 3786 DOI 10.17487/RFC6273, June 2011, 3787 . 3789 [RFC6347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer 3790 Security Version 1.2", RFC 6347, DOI 10.17487/RFC6347, 3791 January 2012, . 3793 [RFC6355] Narten, T. and J. Johnson, "Definition of the UUID-Based 3794 DHCPv6 Unique Identifier (DUID-UUID)", RFC 6355, 3795 DOI 10.17487/RFC6355, August 2011, 3796 . 3798 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 3799 for Equal Cost Multipath Routing and Link Aggregation in 3800 Tunnels", RFC 6438, DOI 10.17487/RFC6438, November 2011, 3801 . 3803 [RFC6706] Templin, F., Ed., "Asymmetric Extended Route Optimization 3804 (AERO)", RFC 6706, DOI 10.17487/RFC6706, August 2012, 3805 . 3807 [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and 3808 UDP Checksums for Tunneled Packets", RFC 6935, 3809 DOI 10.17487/RFC6935, April 2013, 3810 . 3812 [RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement 3813 for the Use of IPv6 UDP Datagrams with Zero Checksums", 3814 RFC 6936, DOI 10.17487/RFC6936, April 2013, 3815 . 3817 [RFC7333] Chan, H., Ed., Liu, D., Seite, P., Yokota, H., and J. 3818 Korhonen, "Requirements for Distributed Mobility 3819 Management", RFC 7333, DOI 10.17487/RFC7333, August 2014, 3820 . 3822 [RFC7761] Fenner, B., Handley, M., Holbrook, H., Kouvelas, I., 3823 Parekh, R., Zhang, Z., and L. Zheng, "Protocol Independent 3824 Multicast - Sparse Mode (PIM-SM): Protocol Specification 3825 (Revised)", STD 83, RFC 7761, DOI 10.17487/RFC7761, March 3826 2016, . 3828 [RFC8402] Filsfils, C., Ed., Previdi, S., Ed., Ginsberg, L., 3829 Decraene, B., Litkowski, S., and R. Shakir, "Segment 3830 Routing Architecture", RFC 8402, DOI 10.17487/RFC8402, 3831 July 2018, . 3833 [RFC8446] Rescorla, E., "The Transport Layer Security (TLS) Protocol 3834 Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018, 3835 . 3837 [RFC8754] Filsfils, C., Ed., Dukes, D., Ed., Previdi, S., Leddy, J., 3838 Matsushima, S., and D. Voyer, "IPv6 Segment Routing Header 3839 (SRH)", RFC 8754, DOI 10.17487/RFC8754, March 2020, 3840 . 3842 [WG] Wireguard, "Wireguard, https://www.wireguard.com", August 3843 2020. 3845 Appendix A. Non-Normative Considerations 3847 AERO can be applied to a multitude of Internetworking scenarios, with 3848 each having its own adaptations. The following considerations are 3849 provided as non-normative guidance: 3851 A.1. Implementation Strategies for Route Optimization 3853 Route optimization as discussed in Section 3.14 results in the route 3854 optimization source (ROS) creating a neighbor cache entry for the 3855 target neighbor. The neighbor cache entry is maintained for at most 3856 ReachableTime seconds and then deleted unless updated. In order to 3857 refresh the neighbor cache entry lifetime before the ReachableTime 3858 timer expires, the specification requires implementations to issue a 3859 new NS/NA exchange to reset ReachableTime while data packets are 3860 still flowing. However, the decision of when to initiate a new NS/NA 3861 exchange and to perpetuate the process is left as an implementation 3862 detail. 3864 One possible strategy may be to monitor the neighbor cache entry 3865 watching for data packets for (ReachableTime - 5) seconds. If any 3866 data packets have been sent to the neighbor within this timeframe, 3867 then send an NS to receive a new NA. If no data packets have been 3868 sent, wait for 5 additional seconds and send an immediate NS if any 3869 data packets are sent within this "expiration pending" 5 second 3870 window. If no additional data packets are sent within the 5 second 3871 window, delete the neighbor cache entry. 3873 The monitoring of the neighbor data packet traffic therefore becomes 3874 an asymmetric ongoing process during the neighbor cache entry 3875 lifetime. If the neighbor cache entry expires, future data packets 3876 will trigger a new NS/NA exchange while the packets themselves are 3877 delivered over a longer path until route optimization state is re- 3878 established. 3880 A.2. Implicit Mobility Management 3882 OMNI interface neighbors MAY provide a configuration option that 3883 allows them to perform implicit mobility management in which no ND 3884 messaging is used. In that case, the Client only transmits packets 3885 over a single interface at a time, and the neighbor always observes 3886 packets arriving from the Client from the same link-layer source 3887 address. 3889 If the Client's underlying interface address changes (either due to a 3890 readdressing of the original interface or switching to a new 3891 interface) the neighbor immediately updates the neighbor cache entry 3892 for the Client and begins accepting and sending packets according to 3893 the Client's new address. This implicit mobility method applies to 3894 use cases such as cellphones with both WiFi and Cellular interfaces 3895 where only one of the interfaces is active at a given time, and the 3896 Client automatically switches over to the backup interface if the 3897 primary interface fails. 3899 A.3. Direct Underlying Interfaces 3901 When a Client's OMNI interface is configured over a Direct interface, 3902 the neighbor at the other end of the Direct link can receive packets 3903 without any encapsulation. In that case, the Client sends packets 3904 over the Direct link according to QoS preferences. If the Direct 3905 interface has the highest QoS preference, then the Client's IP 3906 packets are transmitted directly to the peer without going through an 3907 ANET/INET. If other interfaces have higher QoS preferences, then the 3908 Client's IP packets are transmitted via a different interface, which 3909 may result in the inclusion of Proxys, Servers and Bridges in the 3910 communications path. Direct interfaces must be tested periodically 3911 for reachability, e.g., via NUD. 3913 A.4. AERO Critical Infrastructure Considerations 3915 AERO Bridges can be either Commercial off-the Shelf (COTS) standard 3916 IP routers or virtual machines in the cloud. Bridges must be 3917 provisioned, supported and managed by the INET administrative 3918 authority, and connected to the Bridges of other INETs via inter- 3919 domain peerings. Cost for purchasing, configuring and managing 3920 Bridges is nominal even for very large OMNI links. 3922 AERO Servers can be standard dedicated server platforms, but most 3923 often will be deployed as virtual machines in the cloud. The only 3924 requirements for Servers are that they can run the AERO user-level 3925 code and have at least one network interface connection to the INET. 3926 As with Bridges, Servers must be provisioned, supported and managed 3927 by the INET administrative authority. Cost for purchasing, 3928 configuring and managing Servers is nominal especially for virtual 3929 Servers hosted in the cloud. 3931 AERO Proxys are most often standard dedicated server platforms with 3932 one network interface connected to the ANET and a second interface 3933 connected to an INET. As with Servers, the only requirements are 3934 that they can run the AERO user-level code and have at least one 3935 interface connection to the INET. Proxys must be provisioned, 3936 supported and managed by the ANET administrative authority. Cost for 3937 purchasing, configuring and managing Proxys is nominal, and borne by 3938 the ANET administrative authority. 3940 AERO Relays can be any dedicated server or COTS router platform 3941 connected to INETs and/or EUNs. The Relay connects to the OMNI link 3942 and engages in eBGP peering with one or more Bridges as a stub AS. 3943 The Relay then injects its MNPs and/or non-MNP prefixes into the BGP 3944 routing system, and provisions the prefixes to its downstream- 3945 attached networks. The Relay can perform ROS/ROR services the same 3946 as for any Server, and can route between the MNP and non-MNP address 3947 spaces. 3949 A.5. AERO Server Failure Implications 3951 AERO Servers may appear as a single point of failure in the 3952 architecture, but such is not the case since all Servers on the link 3953 provide identical services and loss of a Server does not imply 3954 immediate and/or comprehensive communication failures. Although 3955 Clients typically associate with a single Server at a time, Server 3956 failure is quickly detected and conveyed by Bidirectional Forward 3957 Detection (BFD) and/or proactive NUD allowing Clients to migrate to 3958 new Servers. 3960 If a Server fails, ongoing packet forwarding to Clients will continue 3961 by virtue of the neighbor cache entries that have already been 3962 established in route optimization sources (ROSs). If a Client also 3963 experiences mobility events at roughly the same time the Server 3964 fails, unsolicited NA messages may be lost but proxy neighbor cache 3965 entries in the DEPARTED state will ensure that packet forwarding to 3966 the Client's new locations will continue for up to DepartTime 3967 seconds. 3969 If a Client is left without a Server for an extended timeframe (e.g., 3970 greater than ReachableTime seconds) then existing neighbor cache 3971 entries will eventually expire and both ongoing and new 3972 communications will fail. The original source will continue to 3973 retransmit until the Client has established a new Server 3974 relationship, after which time continuous communications will resume. 3976 Therefore, providing many Servers on the link with high availability 3977 profiles provides resilience against loss of individual Servers and 3978 assurance that Clients can establish new Server relationships quickly 3979 in event of a Server failure. 3981 A.6. AERO Client / Server Architecture 3983 The AERO architectural model is client / server in the control plane, 3984 with route optimization in the data plane. The same as for common 3985 Internet services, the AERO Client discovers the addresses of AERO 3986 Servers and selects one Server to connect to. The AERO service is 3987 analogous to common Internet services such as google.com, yahoo.com, 3988 cnn.com, etc. However, there is only one AERO service for the link 3989 and all Servers provide identical services. 3991 Common Internet services provide differing strategies for advertising 3992 server addresses to clients. The strategy is conveyed through the 3993 DNS resource records returned in response to name resolution queries. 3994 As of January 2020 Internet-based 'nslookup' services were used to 3995 determine the following: 3997 o When a client resolves the domainname "google.com", the DNS always 3998 returns one A record (i.e., an IPv4 address) and one AAAA record 3999 (i.e., an IPv6 address). The client receives the same addresses 4000 each time it resolves the domainname via the same DNS resolver, 4001 but may receive different addresses when it resolves the 4002 domainname via different DNS resolvers. But, in each case, 4003 exactly one A and one AAAA record are returned. 4005 o When a client resolves the domainname "ietf.org", the DNS always 4006 returns one A record and one AAAA record with the same addresses 4007 regardless of which DNS resolver is used. 4009 o When a client resolves the domainname "yahoo.com", the DNS always 4010 returns a list of 4 A records and 4 AAAA records. Each time the 4011 client resolves the domainname via the same DNS resolver, the same 4012 list of addresses are returned but in randomized order (i.e., 4013 consistent with a DNS round-robin strategy). But, interestingly, 4014 the same addresses are returned (albeit in randomized order) when 4015 the domainname is resolved via different DNS resolvers. 4017 o When a client resolves the domainname "amazon.com", the DNS always 4018 returns a list of 3 A records and no AAAA records. As with 4019 "yahoo.com", the same three A records are returned from any 4020 worldwide Internet connection point in randomized order. 4022 The above example strategies show differing approaches to Internet 4023 resilience and service distribution offered by major Internet 4024 services. The Google approach exposes only a single IPv4 and a 4025 single IPv6 address to clients. Clients can then select whichever IP 4026 protocol version offers the best response, but will always use the 4027 same IP address according to the current Internet connection point. 4028 This means that the IP address offered by the network must lead to a 4029 highly-available server and/or service distribution point. In other 4030 words, resilience is predicated on high availability within the 4031 network and with no client-initiated failovers expected (i.e., it is 4032 all-or-nothing from the client's perspective). However, Google does 4033 provide for worldwide distributed service distribution by virtue of 4034 the fact that each Internet connection point responds with a 4035 different IPv6 and IPv4 address. The IETF approach is like google 4036 (all-or-nothing from the client's perspective), but provides only a 4037 single IPv4 or IPv6 address on a worldwide basis. This means that 4038 the addresses must be made highly-available at the network level with 4039 no client failover possibility, and if there is any worldwide service 4040 distribution it would need to be conducted by a network element that 4041 is reached via the IP address acting as a service distribution point. 4043 In contrast to the Google and IETF philosophies, Yahoo and Amazon 4044 both provide clients with a (short) list of IP addresses with Yahoo 4045 providing both IP protocol versions and Amazon as IPv4-only. The 4046 order of the list is randomized with each name service query 4047 response, with the effect of round-robin load balancing for service 4048 distribution. With a short list of addresses, there is still 4049 expectation that the network will implement high availability for 4050 each address but in case any single address fails the client can 4051 switch over to using a different address. The balance then becomes 4052 one of function in the network vs function in the end system. 4054 The same implications observed for common highly-available services 4055 in the Internet apply also to the AERO client/server architecture. 4056 When an AERO Client connects to one or more ANETs, it discovers one 4057 or more AERO Server addresses through the mechanisms discussed in 4058 earlier sections. Each Server address presumably leads to a fault- 4059 tolerant clustering arrangement such as supported by Linux-HA, 4060 Extended Virtual Synchrony or Paxos. Such an arrangement has 4061 precedence in common Internet service deployments in lightweight 4062 virtual machines without requiring expensive hardware deployment. 4063 Similarly, common Internet service deployments set service IP 4064 addresses on service distribution points that may relay requests to 4065 many different servers. 4067 For AERO, the expectation is that a combination of the Google/IETF 4068 and Yahoo/Amazon philosophies would be employed. The AERO Client 4069 connects to different ANET access points and can receive 1-2 Server 4070 ADM-LLAs at each point. It then selects one AERO Server address, and 4071 engages in RS/RA exchanges with the same Server from all ANET 4072 connections. The Client remains with this Server unless or until the 4073 Server fails, in which case it can switch over to an alternate 4074 Server. The Client can likewise switch over to a different Server at 4075 any time if there is some reason for it to do so. So, the AERO 4076 expectation is for a balance of function in the network and end 4077 system, with fault tolerance and resilience at both levels. 4079 Appendix B. Change Log 4081 << RFC Editor - remove prior to publication >> 4083 Changes from draft-templin-intarea-6706bis-61 to draft-templin- 4084 intrea-6706bis-62: 4086 o New sub-section on OMNI Neighbor Interface Attributes 4088 Changes from draft-templin-intarea-6706bis-59 to draft-templin- 4089 intrea-6706bis-60: 4091 o Removed all references to S/TLLAO - all Interface Attributes are 4092 now maintained completely in the OMNI option. 4094 Changes from draft-templin-intarea-6706bis-58 to draft-templin- 4095 intrea-6706bis-59: 4097 o The term "Relay" used in older draft versions is now "Bridge". 4098 "Relay" now refers to what was formally called: "Gateway". 4100 o Fine-grained cleanup of Forwarding Algorithm; IPv6 ND message 4101 addressing; OMNI Prefix Lengths, etc. 4103 Changes from draft-templin-intarea-6706bis-54 to draft-templin- 4104 intrea-6706bis-55: 4106 o Updates on Segment Routing and S/TLLAO contents. 4108 o Various editorials and addressing cleanups. 4110 Changes from draft-templin-intarea-6706bis-52 to draft-templin- 4111 intrea-6706bis-53: 4113 o Normative reference to the OMNI spec, and remove portions that are 4114 already specified in OMNI. 4116 o Renamed "AERO interface/link" to "OMIN interface/link" throughout 4117 the document. 4119 o Truncated obsolete back section matter. 4121 Author's Address 4123 Fred L. Templin (editor) 4124 Boeing Research & Technology 4125 P.O. Box 3707 4126 Seattle, WA 98124 4127 USA 4129 Email: fltemplin@acm.org