idnits 2.17.1 draft-templin-intarea-6706bis-61.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 (September 18, 2020) is 1310 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) -- Looks like a reference, but probably isn't: '1' on line 913 -- Looks like a reference, but probably isn't: '2' on line 913 -- Looks like a reference, but probably isn't: '3' on line 886 == Unused Reference: 'BGP' is defined on line 3191, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-intarea-frag-fragile' is defined on line 3204, but no explicit reference was found in the text == Unused Reference: 'I-D.templin-6man-dhcpv6-ndopt' is defined on line 3221, but no explicit reference was found in the text == Outdated reference: A later version (-99) exists of draft-templin-6man-omni-interface-34 == Outdated reference: A later version (-31) exists of draft-bonica-6man-comp-rtg-hdr-22 == Outdated reference: A later version (-04) exists of draft-bonica-6man-crh-helper-opt-01 == Outdated reference: A later version (-13) exists of draft-ietf-intarea-tunnels-10 == Outdated reference: A later version (-26) exists of draft-ietf-rtgwg-atn-bgp-06 == Outdated reference: A later version (-11) exists of draft-templin-6man-dhcpv6-ndopt-10 == Outdated reference: A later version (-27) exists of draft-templin-v6ops-pdhost-26 -- Obsolete informational reference (is this intentional?): RFC 3330 (Obsoleted by RFC 5735) -- Obsolete informational reference (is this intentional?): RFC 6106 (Obsoleted by RFC 8106) -- Obsolete informational reference (is this intentional?): RFC 6347 (Obsoleted by RFC 9147) Summary: 0 errors (**), 0 flaws (~~), 12 warnings (==), 7 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group F. Templin, Ed. 3 Internet-Draft Boeing Research & Technology 4 Obsoletes: rfc5320, rfc5558, rfc5720, September 18, 2020 5 rfc6179, rfc6706 (if 6 approved) 7 Intended status: Standards Track 8 Expires: March 22, 2021 10 Asymmetric Extended Route Optimization (AERO) 11 draft-templin-intarea-6706bis-61 13 Abstract 15 This document specifies the operation of IP over Overlay Multilink 16 Network (OMNI) interfaces using the Asymmetric Extended Route 17 Optimization (AERO) internetworking and mobility management service. 18 AERO/OMNI use an IPv6 link-local address format that supports 19 operation of the IPv6 Neighbor Discovery (ND) protocol and links ND 20 to IP forwarding. Prefix delegation/registration services are 21 employed for network admission and to manage the routing system. 22 Multilink operation, mobility management, quality of service (QoS) 23 signaling and route optimization are naturally supported through 24 dynamic neighbor cache updates. Standard IP multicasting services 25 are also supported. AERO is a widely-applicable mobile 26 internetworking service especially well-suited to aviation services, 27 intelligent transportation systems, mobile Virtual Private Networks 28 (VPNs) and many other applications. 30 Status of This Memo 32 This Internet-Draft is submitted in full conformance with the 33 provisions of BCP 78 and BCP 79. 35 Internet-Drafts are working documents of the Internet Engineering 36 Task Force (IETF). Note that other groups may also distribute 37 working documents as Internet-Drafts. The list of current Internet- 38 Drafts is at https://datatracker.ietf.org/drafts/current/. 40 Internet-Drafts are draft documents valid for a maximum of six months 41 and may be updated, replaced, or obsoleted by other documents at any 42 time. It is inappropriate to use Internet-Drafts as reference 43 material or to cite them other than as "work in progress." 45 This Internet-Draft will expire on March 22, 2021. 47 Copyright Notice 49 Copyright (c) 2020 IETF Trust and the persons identified as the 50 document authors. All rights reserved. 52 This document is subject to BCP 78 and the IETF Trust's Legal 53 Provisions Relating to IETF Documents 54 (https://trustee.ietf.org/license-info) in effect on the date of 55 publication of this document. Please review these documents 56 carefully, as they describe your rights and restrictions with respect 57 to this document. Code Components extracted from this document must 58 include Simplified BSD License text as described in Section 4.e of 59 the Trust Legal Provisions and are provided without warranty as 60 described in the Simplified BSD License. 62 Table of Contents 64 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4 65 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 5 66 3. Asymmetric Extended Route Optimization (AERO) . . . . . . . . 10 67 3.1. AERO Node Types . . . . . . . . . . . . . . . . . . . . . 10 68 3.2. The AERO Service over OMNI Links . . . . . . . . . . . . 11 69 3.2.1. AERO/OMNI Reference Model . . . . . . . . . . . . . . 11 70 3.2.2. Link-Local Addresses (LLAs) and Unique Local 71 Addresses (ULAs) . . . . . . . . . . . . . . . . . . 14 72 3.2.3. AERO Routing System . . . . . . . . . . . . . . . . . 15 73 3.2.4. OMNI Link Encapsulation . . . . . . . . . . . . . . . 16 74 3.2.5. Segment Routing Topologies (SRTs) . . . . . . . . . . 18 75 3.2.6. Segment Routing To the OMNI Link . . . . . . . . . . 18 76 3.2.7. Segment Routing Within the OMNI Link . . . . . . . . 19 77 3.2.8. Segment Routing Header Compression . . . . . . . . . 21 78 3.3. OMNI Interface Characteristics . . . . . . . . . . . . . 21 79 3.4. OMNI Interface Initialization . . . . . . . . . . . . . . 23 80 3.4.1. AERO Server/Relay Behavior . . . . . . . . . . . . . 23 81 3.4.2. AERO Proxy Behavior . . . . . . . . . . . . . . . . . 24 82 3.4.3. AERO Client Behavior . . . . . . . . . . . . . . . . 24 83 3.4.4. AERO Bridge Behavior . . . . . . . . . . . . . . . . 24 84 3.5. OMNI Interface Neighbor Cache Maintenance . . . . . . . . 24 85 3.6. OMNI Interface Encapsulation and Re-encapsulation . . . . 26 86 3.7. OMNI Interface Decapsulation . . . . . . . . . . . . . . 28 87 3.8. OMNI Interface Data Origin Authentication . . . . . . . . 28 88 3.9. OMNI Adaptation Layer (OAL) and OMNI Interface MTU . . . 28 89 3.10. OMNI Interface Forwarding Algorithm . . . . . . . . . . . 28 90 3.10.1. Client Forwarding Algorithm . . . . . . . . . . . . 29 91 3.10.2. Proxy Forwarding Algorithm . . . . . . . . . . . . . 30 92 3.10.3. Server/Relay Forwarding Algorithm . . . . . . . . . 31 93 3.10.4. Bridge Forwarding Algorithm . . . . . . . . . . . . 32 94 3.11. OMNI Interface Error Handling . . . . . . . . . . . . . . 33 95 3.12. AERO Router Discovery, Prefix Delegation and 96 Autoconfiguration . . . . . . . . . . . . . . . . . . . . 35 97 3.12.1. AERO Service Model . . . . . . . . . . . . . . . . . 35 98 3.12.2. AERO Client Behavior . . . . . . . . . . . . . . . . 36 99 3.12.3. AERO Server Behavior . . . . . . . . . . . . . . . . 38 100 3.13. The AERO Proxy . . . . . . . . . . . . . . . . . . . . . 41 101 3.13.1. Combined Proxy/Servers . . . . . . . . . . . . . . . 43 102 3.13.2. Detecting and Responding to Server Failures . . . . 43 103 3.13.3. Point-to-Multipoint Server Coordination . . . . . . 44 104 3.14. AERO Route Optimization / Address Resolution . . . . . . 45 105 3.14.1. Route Optimization Initiation . . . . . . . . . . . 45 106 3.14.2. Relaying the NS . . . . . . . . . . . . . . . . . . 46 107 3.14.3. Processing the NS and Sending the NA . . . . . . . . 46 108 3.14.4. Relaying the NA . . . . . . . . . . . . . . . . . . 47 109 3.14.5. Processing the NA . . . . . . . . . . . . . . . . . 47 110 3.14.6. Route Optimization Maintenance . . . . . . . . . . . 47 111 3.15. Neighbor Unreachability Detection (NUD) . . . . . . . . . 48 112 3.16. Mobility Management and Quality of Service (QoS) . . . . 50 113 3.16.1. Mobility Update Messaging . . . . . . . . . . . . . 50 114 3.16.2. Announcing Link-Layer Address and/or QoS Preference 115 Changes . . . . . . . . . . . . . . . . . . . . . . 51 116 3.16.3. Bringing New Links Into Service . . . . . . . . . . 51 117 3.16.4. Removing Existing Links from Service . . . . . . . . 51 118 3.16.5. Moving to a New Server . . . . . . . . . . . . . . . 52 119 3.17. Multicast . . . . . . . . . . . . . . . . . . . . . . . . 53 120 3.17.1. Source-Specific Multicast (SSM) . . . . . . . . . . 53 121 3.17.2. Any-Source Multicast (ASM) . . . . . . . . . . . . . 55 122 3.17.3. Bi-Directional PIM (BIDIR-PIM) . . . . . . . . . . . 55 123 3.18. Operation over Multiple OMNI Links . . . . . . . . . . . 55 124 3.19. DNS Considerations . . . . . . . . . . . . . . . . . . . 56 125 3.20. Transition Considerations . . . . . . . . . . . . . . . . 57 126 3.21. Detecting and Reacting to Server and Bridge Failures . . 57 127 3.22. AERO Clients on the Open Internet . . . . . . . . . . . . 58 128 3.22.1. Use of SEND and CGA . . . . . . . . . . . . . . . . 60 129 3.23. Time-Varying MNPs . . . . . . . . . . . . . . . . . . . . 62 130 4. Implementation Status . . . . . . . . . . . . . . . . . . . . 62 131 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 62 132 6. Security Considerations . . . . . . . . . . . . . . . . . . . 63 133 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 65 134 8. References . . . . . . . . . . . . . . . . . . . . . . . . . 66 135 8.1. Normative References . . . . . . . . . . . . . . . . . . 66 136 8.2. Informative References . . . . . . . . . . . . . . . . . 68 137 Appendix A. Non-Normative Considerations . . . . . . . . . . . . 73 138 A.1. Implementation Strategies for Route Optimization . . . . 73 139 A.2. Implicit Mobility Management . . . . . . . . . . . . . . 74 140 A.3. Direct Underlying Interfaces . . . . . . . . . . . . . . 74 141 A.4. AERO Critical Infrastructure Considerations . . . . . . . 74 142 A.5. AERO Server Failure Implications . . . . . . . . . . . . 75 143 A.6. AERO Client / Server Architecture . . . . . . . . . . . . 76 144 Appendix B. Change Log . . . . . . . . . . . . . . . . . . . . . 78 145 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 79 147 1. Introduction 149 Asymmetric Extended Route Optimization (AERO) fulfills the 150 requirements of Distributed Mobility Management (DMM) [RFC7333] and 151 route optimization [RFC5522] for aeronautical networking and other 152 network mobility use cases such as intelligent transportation 153 systems. AERO is an internetworking and mobility management service 154 based on the Overlay Multilink Network Interface (OMNI) 155 [I-D.templin-6man-omni-interface] Non-Broadcast, Multiple Access 156 (NBMA) virtual link model. The OMNI link is a virtual overlay 157 configured over one or more underlying Internetworks, and nodes on 158 the link can exchange IP packets via tunneling. The OMNI Adaptation 159 Layer (OAL) supports multilink operation for increased reliability, 160 bandwidth optimization and traffic path diversity while accommodating 161 Maximum Transmission Unit (MTU) diversity. 163 The AERO service comprises Clients, Proxys, Servers and Relays that 164 are seen as OMNI link neighbors as well as Bridges that interconnect 165 OMNI link segments. Each node's OMNI interface uses an IPv6 link- 166 local address format that supports operation of the IPv6 Neighbor 167 Discovery (ND) protocol [RFC4861] and links ND to IP forwarding. A 168 node's OMNI interface can be configured over multiple underlying 169 interfaces, and may therefore appear as a single interface with 170 multiple link-layer addresses. Each link-layer address is subject to 171 change due to mobility and/or QoS fluctuations, and link-layer 172 address changes are signaled by ND messaging the same as for any IPv6 173 link. 175 AERO provides a cloud-based service where mobile nodes may use any 176 Server acting as a Mobility Anchor Point (MAP) and fixed nodes may 177 use any Relay on the link for efficient communications. Fixed nodes 178 forward packets destined to other AERO nodes to the nearest Relay, 179 which forwards them through the cloud. A mobile node's initial 180 packets are forwarded through the Server, while direct routing is 181 supported through asymmetric extended route optimization while data 182 packets are flowing. Both unicast and multicast communications are 183 supported, and mobile nodes may efficiently move between locations 184 while maintaining continuous communications with correspondents and 185 without changing their IP Address. 187 AERO Bridges are interconnected in a secured private BGP overlay 188 routing instance using encapsulation to provide a hybrid routing/ 189 bridging service that joins the underlying Internetworks of multiple 190 disjoint administrative domains into a single unified OMNI link. 192 Each OMNI link instance is characterized by the set of Mobility 193 Service Prefixes (MSPs) common to all mobile nodes. The link extends 194 to the point where a Relay/Server is on the optimal route from any 195 correspondent node on the link, and provides a conduit between the 196 underlying Internetwork and the OMNI link. To the underlying 197 Internetwork, the Relay/Server is the source of a route to the MSP, 198 and hence uplink traffic to the mobile node is naturally routed to 199 the nearest Relay/Server. 201 AERO assumes the use of PIM Sparse Mode in support of multicast 202 communication. In support of Source Specific Multicast (SSM) when a 203 Mobile Node is the source, AERO route optimization ensures that a 204 shortest-path multicast tree is established with provisions for 205 mobility and multilink operation. In all other multicast scenarios 206 there are no AERO dependencies. 208 AERO was designed for aeronautical networking for both manned and 209 unmanned aircraft, where the aircraft is treated as a mobile node 210 that can connect an Internet of Things (IoT). AERO is also 211 applicable to a wide variety of other use cases. For example, it can 212 be used to coordinate the Virtual Private Network (VPN) links of 213 mobile nodes (e.g., cellphones, tablets, laptop computers, etc.) that 214 connect into a home enterprise network via public access networks 215 using services such as OpenVPN [OVPN]. It can also be used to 216 facilitate vehicular and pedestrian communications services for 217 intelligent transportation systems. Other applicable use cases are 218 also in scope. 220 The following numbered sections present the AERO specification. The 221 appendices at the end of the document are non-normative. 223 2. Terminology 225 The terminology in the normative references applies; especially, the 226 terminology in the OMNI specification 227 [I-D.templin-6man-omni-interface] is used extensively throughout. 228 The following terms are defined within the scope of this document: 230 IPv6 Neighbor Discovery (ND) 231 an IPv6 control message service for coordinating neighbor 232 relationships between nodes connected to a common link. AERO uses 233 the ND service specified in [RFC4861]. 235 IPv6 Prefix Delegation 236 a networking service for delegating IPv6 prefixes to nodes on the 237 link. The nominal service is DHCPv6 [RFC8415], however alternate 238 services (e.g., based on ND messaging) are also in scope. Most 239 notably, a minimal form of prefix delegation known as "prefix 240 registration" can be used if the Client knows its prefix in 241 advance and can represent it in the IPv6 source address of an ND 242 message. 244 Access Network (ANET) 245 a node's first-hop data link service network (e.g., a radio access 246 network, cellular service provider network, corporate enterprise 247 network, etc.) that often provides link-layer security services 248 such as IEEE 802.1X and physical-layer security prevent 249 unauthorized access internally and with border network-layer 250 security services such as firewalls and proxies that prevent 251 unauthorized outside access. 253 ANET interface 254 a node's attachment to a link in an ANET. 256 Internetwork (INET) 257 a connected IP network topology with a coherent routing and 258 addressing plan and that provides a transit backbone service for 259 ANET end systems. INETs also provide an underlay service over 260 which the AERO virtual link is configured. Example INETs include 261 corporate enterprise networks, aviation networks, and the public 262 Internet itself. When there is no administrative boundary between 263 an ANET and the INET, the ANET and INET are one and the same. 265 INET Partition 266 frequently, INETs such as large corporate enterprise networks are 267 sub-divided internally into separate isolated partitions. Each 268 partition is fully connected internally but disconnected from 269 other partitions, and there is no requirement that separate 270 partitions maintain consistent Internet Protocol and/or addressing 271 plans. (Each INET partition is seen as a separate OMNI link 272 segment as discussed below.) 274 INET interface 275 a node's attachment to a link in an INET. 277 INET address 278 an IP address assigned to a node's interface connection to an 279 INET. 281 INET encapsulation 282 the encapsulation of a packet in an outer header or headers that 283 can be routed within the scope of the local INET partition. 285 OMNI link 286 the same as defined in [I-D.templin-6man-omni-interface], and 287 manifested by IPv6 encapsulation [RFC2473]. The OMNI link spans 288 underlying INET segments joined by virtual bridges in a spanning 289 tree the same as a bridged campus LAN. AERO nodes on the OMNI 290 link appear as single-hop neighbors even though they may be 291 separated by multiple underlying INET hops, and can use Segment 292 Routing [RFC8402] to cause packets to visit selected waypoints on 293 the link. 295 OMNI Interface 296 a node's attachment to an OMNI link. Since the addresses assigned 297 to an OMNI interface are managed for uniqueness, OMNI interfaces 298 do not require Duplicate Address Detection (DAD) and therefore set 299 the administrative variable 'DupAddrDetectTransmits' to zero 300 [RFC4862]. 302 OMNI Adaptation Layer (OAL) 303 an OMNI interface process whereby packets admitted into the 304 interface are wrapped in a mid-layer IPv6 header and fragmented/ 305 reassembled if necessary to support the OMNI link Maximum 306 Transmission Unit (MTU). The OAL is also responsible for 307 generating MTU-related control messages as necessary, and for 308 providing addressing context for spanning multiple segments of a 309 bridged OMNI link. 311 OMNI Link-Local Address (LLA) 312 a link local IPv6 address per [RFC4291] constructed as specified 313 in Section 3.2.2. 315 OMNI Unique-Local Address (ULA) 316 a unique local IPv6 address per [RFC4193] constructed as specified 317 in Section 3.2.2. OMNI ULAs are statelessly derived from OMNI 318 LLAs, and vice-versa. 320 underlying interface 321 an ANET or INET interface over which an OMNI interface is 322 configured. 324 Mobility Service Prefix (MSP) 325 an IP prefix assigned to the OMNI link and from which more- 326 specific Mobile Network Prefixes (MNPs) are derived. 328 Mobile Network Prefix (MNP) 329 an IP prefix allocated from an MSP and delegated to an AERO Client 330 or Relay. 332 AERO node 333 a node that is connected to an OMNI link and participates in the 334 AERO internetworking and mobility service. 336 AERO Client ("Client") 337 an AERO node that connects over one or more underlying interfaces 338 and requests MNP delegation/registration service from AERO 339 Servers. The Client assigns a Client LLA to the OMNI interface 340 for use in ND exchanges with other AERO nodes and forwards packets 341 to correspondents according to OMNI interface neighbor cache 342 state. 344 AERO Server ("Server") 345 an INET node that configures an OMNI interface to provide default 346 forwarding and mobility/multilink services for AERO Clients. The 347 Server assigns an administratively-provisioned LLA to its OMNI 348 interface to support the operation of the ND services, and 349 advertises all of its associated MNPs via BGP peerings with 350 Bridges. 352 AERO Relay ("Relay") 353 an AERO Server that also provides forwarding services between 354 nodes reached via the OMNI link and correspondents on other links. 355 AERO Relays are provisioned with MNPs (i.e., the same as for an 356 AERO Client) and run a dynamic routing protocol to discover any 357 non-MNP IP routes. In both cases, the Relay advertises the MSP(s) 358 to its downstream networks, and distributes all of its associated 359 MNPs and non-MNP IP routes via BGP peerings with Bridges (i.e., 360 the same as for an AERO Server). 362 AERO Bridge ("Bridge") 363 a node that provides hybrid routing/bridging services (as well as 364 a security trust anchor) for nodes on an OMNI link. As a router, 365 the Bridge forwards packets using standard IP forwarding. As a 366 bridge, the Bridge forwards packets over the OMNI link without 367 decrementing the IPv6 Hop Limit. AERO Bridges peer with Servers 368 and other Bridges to discover the full set of MNPs for the link as 369 well as any non-MNPs that are reachable via Relays. 371 AERO Proxy ("Proxy") 372 a node that provides proxying services between Clients in an ANET 373 and Servers in external INETs. The AERO Proxy is a conduit 374 between the ANET and external INETs in the same manner as for 375 common web proxies, and behaves in a similar fashion as for ND 376 proxies [RFC4389]. A node may be configured to act as either a 377 Proxy and/or a Server, depending on Client Server selection 378 criteria. 380 ingress tunnel endpoint (ITE) 381 an OMNI interface endpoint that injects encapsulated packets into 382 an OMNI link. 384 egress tunnel endpoint (ETE) 385 an OMNI interface endpoint that receives encapsulated packets from 386 an OMNI link. 388 link-layer address 389 an IP address used as an encapsulation header source or 390 destination address from the perspective of the OMNI interface. 391 When an upper layer protocol (e.g., UDP) is used as part of the 392 encapsulation, the port number is also considered as part of the 393 link-layer address. 395 network layer address 396 the source or destination address of an encapsulated IP packet 397 presented to the OMNI interface. 399 end user network (EUN) 400 an internal virtual or external edge IP network that an AERO 401 Client or Relay connects to the rest of the network via the OMNI 402 interface. The Client/Relay sees each EUN as a "downstream" 403 network, and sees the OMNI interface as the point of attachment to 404 the "upstream" network. 406 Mobile Node (MN) 407 an AERO Client and all of its downstream-attached networks that 408 move together as a single unit, i.e., an end system that connects 409 an Internet of Things. 411 Mobile Router (MR) 412 a MN's on-board router that forwards packets between any 413 downstream-attached networks and the OMNI link. 415 Route Optimization Source (ROS) 416 the AERO node nearest the source that initiates route 417 optimization. The ROS may be a Server or Proxy acting on behalf 418 of the source Client. 420 Route Optimization responder (ROR) 421 the AERO node nearest the target destination that responds to 422 route optimization requests. The ROR may be a Server acting on 423 behalf of a target MNP Client, or a Relay for a non-MNP 424 destination. 426 MAP List 427 a geographically and/or topologically referenced list of addresses 428 of all Servers within the same OMNI link. There is a single MAP 429 list for the entire OMNI link. 431 Distributed Mobility Management (DMM) 432 a BGP-based overlay routing service coordinated by Servers and 433 Bridges that tracks all Server-to-Client associations. 435 Mobility Service (MS) 436 the collective set of all Servers, Proxys, Bridges and Relays that 437 provide the AERO Service to Clients. 439 Mobility Service Endpoint MSE) 440 an individual Server, Proxy, Bridge or Relay in the Mobility 441 Service. 443 Throughout the document, the simple terms "Client", "Server", 444 "Bridge", "Proxy" and "Relay" refer to "AERO Client", "AERO Server", 445 "AERO Bridge", "AERO Proxy" and "AERO Relay", respectively. 446 Capitalization is used to distinguish these terms from other common 447 Internetworking uses in which they appear without capitalization. 449 The terminology of DHCPv6 [RFC8415] and IPv6 ND [RFC4861] (including 450 the names of node variables, messages and protocol constants) is used 451 throughout this document. The terms "All-Routers multicast", "All- 452 Nodes multicast", "Solicited-Node multicast" and "Subnet-Router 453 anycast" are defined in [RFC4291]. Also, the term "IP" is used to 454 generically refer to either Internet Protocol version, i.e., IPv4 455 [RFC0791] or IPv6 [RFC8200]. 457 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 458 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 459 "OPTIONAL" in this document are to be interpreted as described in BCP 460 14 [RFC2119][RFC8174] when, and only when, they appear in all 461 capitals, as shown here. 463 3. Asymmetric Extended Route Optimization (AERO) 465 The following sections specify the operation of IP over OMNI links 466 using the AERO service: 468 3.1. AERO Node Types 470 AERO Clients are Mobile Nodes (MNs) that connect via underlying 471 interfaces with addresses that may change when the Client moves to a 472 new network connection point. AERO Clients register their Mobile 473 Network Prefixes (MNPs) with the AERO service, and distribute the 474 MNPs to nodes on EUNs. AERO Bridges, Servers, Proxys and Relays are 475 critical infrastructure elements in fixed (i.e., non-mobile) INET 476 deployments and hence have permanent and unchanging INET addresses. 477 Together, they constitute the AERO service which provides an OMNI 478 link virtual overlay for connecting AERO Clients. 480 AERO Bridges provide hybrid routing/bridging services (as well as a 481 security trust anchor) for nodes on an OMNI link. Bridges use 482 standard IPv6 routing to forward packets both within the same INET 483 partitions and between disjoint INET partitions based on a mid-layer 484 IPv6 encapsulation per [RFC2473]. The inner IP layer experiences a 485 virtual bridging service since the inner IP TTL/Hop Limit is not 486 decremented during forwarding. Each Bridge also peers with Servers 487 and other Bridges in a dynamic routing protocol instance to provide a 488 Distributed Mobility Management (DMM) service for the list of active 489 MNPs (see Section 3.2.3). Bridges present the OMNI link as a set of 490 one or more Mobility Service Prefixes (MSPs) and configure secured 491 tunnels with Servers, Relays, Proxys and other Bridges; they further 492 maintain IP forwarding table entries for each MNP and any other 493 reachable non-MNP prefixes. 495 AERO Servers provide default forwarding and mobility/multilink 496 services for AERO Client Mobile Nodes (MNs). Each Server also peers 497 with Bridges in a dynamic routing protocol instance to advertise its 498 list of associated MNPs (see Section 3.2.3). Servers facilitate 499 prefix delegation/registration exchanges with Clients, where each 500 delegated prefix becomes an MNP taken from an MSP. Servers forward 501 packets between OMNI interface neighbors and track each Client's 502 mobility profiles. Servers may further act as Servers for some sets 503 of Clients and as Proxies for others. 505 AERO Proxys provide a conduit for ANET Clients to associate with 506 Servers in external INETs. Client and Servers exchange control plane 507 messages via the Proxy acting as a bridge between the ANET/INET 508 boundary. The Proxy forwards data packets between Clients and the 509 OMNI link according to forwarding information in the neighbor cache. 510 The Proxy function is specified in Section 3.13. Proxys may further 511 act as Proxys for some sets of Clients and as Servers for others. 513 AERO Relays are Servers that provide forwarding services between the 514 OMNI interface and INET/EUN interfaces. Relays are provisioned with 515 MNPs the same as for an AERO Client, and also run a dynamic routing 516 protocol to discover any non-MNP IP routes. The Relay advertises the 517 MSP(s) to its connected networks, and distributes all of its 518 associated MNPs and non-MNP IP routes via BGP peerings with Bridges 520 3.2. The AERO Service over OMNI Links 522 3.2.1. AERO/OMNI Reference Model 524 Figure 1 presents the basic OMNI link reference model: 526 +----------------+ 527 | AERO Bridge B1 | 528 | Nbr: S1, S2, P1| 529 |(X1->S1; X2->S2)| 530 | MSP M1 | 531 +-+---------+--+-+ 532 +--------------+ | Secured | | +--------------+ 533 |AERO Server S1| | tunnels | | |AERO Server S2| 534 | Nbr: C1, B1 +-----+ | +-----+ Nbr: C2, B1 | 535 | default->B1 | | | default->B1 | 536 | X1->C1 | | | X2->C2 | 537 +-------+------+ | +------+-------+ 538 | OMNI link | | 539 X===+===+===================+==)===============+===+===X 540 | | | | 541 +-----+--------+ +--------+--+-----+ +--------+-----+ 542 |AERO Client C1| | AERO Proxy P1 | |AERO Client C2| 543 | Nbr: S1 | |(Proxy Nbr Cache)| | Nbr: S2 | 544 | default->S1 | +--------+--------+ | default->S2 | 545 | MNP X1 | | | MNP X2 | 546 +------+-------+ .--------+------. +-----+--------+ 547 | (- Proxyed Clients -) | 548 .-. `---------------' .-. 549 ,-( _)-. ,-( _)-. 550 .-(_ IP )-. +-------+ +-------+ .-(_ IP )-. 551 (__ EUN )--|Host H1| |Host H2|--(__ EUN ) 552 `-(______)-' +-------+ +-------+ `-(______)-' 554 Figure 1: AERO/OMNI Reference Model 556 In this model: 558 o the OMNI link is an overlay network service configured over one or 559 more underlying INET partitions which may be managed by different 560 administrative authorities and have incompatible protocols and/or 561 addressing plans. 563 o AERO Bridge B1 aggregates Mobility Service Prefix (MSP) M1, 564 discovers Mobile Network Prefixes (MNPs) X* and advertises the MSP 565 via BGP peerings over secured tunnels to Servers (S1, S2). 566 Bridges connect the disjoint segments of a partitioned OMNI link. 568 o AERO Servers/Relays S1 and S2 configure secured tunnels with 569 Bridge B1 and also provide mobility, multilink and default router 570 services for their associated Clients C1 and C2. 572 o AERO Clients C1 and C2 associate with Servers S1 and S2, 573 respectively. They receive Mobile Network Prefix (MNP) 574 delegations X1 and X2, and also act as default routers for their 575 associated physical or internal virtual EUNs. Simple hosts H1 and 576 H2 attach to the EUNs served by Clients C1 and C2, respectively. 578 o AERO Proxy P1 configures a secured tunnel with Bridge B1 and 579 provides proxy services for AERO Clients in secured enclaves that 580 cannot associate directly with other OMNI link neighbors. 582 An OMNI link configured over a single INET appears as a single 583 unified link with a consistent underlying network addressing plan. 584 In that case, all nodes on the link can exchange packets via simple 585 INET encapsulation, since the underlying INET is connected. In 586 common practice, however, an OMNI link may be partitioned into 587 multiple "segments", where each segment is a distinct INET 588 potentially managed under a different administrative authority (e.g., 589 as for worldwide aviation service providers such as ARINC, SITA, 590 Inmarsat, etc.). Individual INETs may also themselves be partitioned 591 internally, in which case each internal partition is seen as a 592 separate segment. 594 The addressing plan of each segment is consistent internally but will 595 often bear no relation to the addressing plans of other segments. 596 Each segment is also likely to be separated from others by network 597 security devices (e.g., firewalls, proxies, packet filtering 598 gateways, etc.), and in many cases disjoint segments may not even 599 have any common physical link connections. Therefore, nodes can only 600 be assured of exchanging packets directly with correspondents in the 601 same segment, and not with those in other segments. The only means 602 for joining the segments therefore is through inter-domain peerings 603 between AERO Bridges. 605 The same as for traditional campus LANs, multiple OMNI link segments 606 can be joined into a single unified link via a virtual bridging 607 service using the OMNI Adaptation Layer (OAL) which inserts a mid- 608 layer IPv6 encpasulation per [RFC2473] that supports inter-segment 609 forwarding (i.e., bridging) without decrementing the network-layer 610 TTL/Hop Limit. This bridging of OMNI link segments is shown in 611 Figure 2: 613 . . . . . . . . . . . . . . . . . . . . . . . 614 . . 615 . .-(::::::::) . 616 . .-(::::::::::::)-. +-+ . 617 . (:::: Segment A :::)--|B|---+ . 618 . `-(::::::::::::)-' +-+ | . 619 . `-(::::::)-' | . 620 . | . 621 . .-(::::::::) | . 622 . .-(::::::::::::)-. +-+ | . 623 . (:::: Segment B :::)--|B|---+ . 624 . `-(::::::::::::)-' +-+ | . 625 . `-(::::::)-' | . 626 . | . 627 . .-(::::::::) | . 628 . .-(::::::::::::)-. +-+ | . 629 . (:::: Segment C :::)--|B|---+ . 630 . `-(::::::::::::)-' +-+ | . 631 . `-(::::::)-' | . 632 . | . 633 . ..(etc).. x . 634 . . 635 . . 636 . <- OMNI link Bridged by encapsulation -> . 637 . . . . . . . . . . . . . .. . . . . . . . . 639 Figure 2: Bridging OMNI Link Segments 641 Bridges, Servers, Relays and Proxys connect via secured INET tunnels 642 over their respecitve segments in a spanning tree topology rooted at 643 the Bridges. The secured spanning tree supports strong 644 authentication for IPv6 ND control messages and may also be used to 645 convey the initial data packets in a flow. Route optimization can 646 then be employed to cause data packets to take more direct paths 647 between OMNI link neighbors without having to strictly follow the 648 spanning tree. 650 3.2.2. Link-Local Addresses (LLAs) and Unique Local Addresses (ULAs) 652 AERO nodes on OMNI links use the Link-Local Address (LLA) prefix 653 fe80::/10 [RFC4193] to assign LLAs used for network-layer addresses 654 in IPv6 ND and data messages. They also use the Unique Local Address 655 (ULA) prefix fc80::/10 [RFC4193] to form ULAs used for OAL header 656 source and desitnation addresses. See 657 [I-D.templin-6man-omni-interface] for a full specification of the 658 LLAs and ULAs used by AERO nodes on OMNI links. 660 For routing system organization (see Section 3.2.3), ULAs are 661 organized in partition prefixes, e.g., fc80::1000/116. For each such 662 partition prefix, the Bridge(s) that connect that segment assign the 663 all-zero's address of the prefix as a Subnet Router Anycast address. 664 For example, the Subnet Router Anycast address for fc80::1000/116 is 665 simply fc80::1000. 667 3.2.3. AERO Routing System 669 The AERO routing system comprises a private instance of the Border 670 Gateway Protocol (BGP) [RFC4271] that is coordinated between Bridges 671 and Servers and does not interact with either the public Internet BGP 672 routing system or any underlying INET routing systems. 674 In a reference deployment, each Server is configured as an Autonomous 675 System Border Router (ASBR) for a stub Autonomous System (AS) using 676 an AS Number (ASN) that is unique within the BGP instance, and each 677 Server further uses eBGP to peer with one or more Bridges but does 678 not peer with other Servers. Each INET of a multi-segment OMNI link 679 must include one or more Bridges, which peer with the Servers and 680 Proxys within that INET. All Bridges within the same INET are 681 members of the same hub AS using a common ASN, and use iBGP to 682 maintain a consistent view of all active MNPs currently in service. 683 The Bridges of different INETs peer with one another using eBGP. 685 Bridges advertise the OMNI link's MSPs and any non-MNP routes to each 686 of their Servers. This means that any aggregated non-MNPs (including 687 "default") are advertised to all Servers. Each Bridge configures a 688 black-hole route for each of its MSPs. By black-holing the MSPs, the 689 Bridge will maintain forwarding table entries only for the MNPs that 690 are currently active, and packets destined to all other MNPs will 691 correctly incur Destination Unreachable messages due to the black- 692 hole route. In this way, Servers have only partial topology 693 knowledge (i.e., they know only about the MNPs of their directly 694 associated Clients) and they forward all other packets to Bridges 695 which have full topology knowledge. 697 Each OMNI link segment assigns a unique sub-prefix of fc80::/96 known 698 as the ULA partition prefix. For example, a first segment could 699 assign fc80::1000/116, a second could assign fc80::2000/116, a third 700 could assign fc80::3000/116, etc. The administrative authorities for 701 each segment must therefore coordinate to assure mutually-exclusive 702 partiton prefix assignments, but internal provisioning of each prefix 703 is an independent local consideration for each administrative 704 authority. 706 ULA partition prefixes are statitcally represented in Bridge 707 forwarding tables. Bridges join multiple segments into a unified 708 OMNI link over multiple diverse administrative domains. They support 709 a bridging function by first establishing forwarding table entries 710 for their partiion prefixes either via standard BGP routing or static 711 routes. For example, if three Bridges ('A', 'B' and 'C') from 712 different segments serviced fc80::1000/116, fc80::2000/116 and 713 fc80::3000/116 respectively, then the forwarding tables in each 714 Bridge are as follows: 716 A: fc80::1000/116->local, fc80::2000/116->B, fc80::3000/116->C 718 B: fc80::1000/116->A, fc80::2000/116->local, fc80::3000/116->C 720 C: fc80::1000/116->A, fc80::2000/116->B, fc80::3000/116->local 722 These forwarding table entries are permanent and never change, since 723 they correspond to fixed infrastructure elements in their respective 724 segments. 726 ULA Client prefixes are instead dynamically advertised in the AERO 727 routing system by Servers and Relays that provide service for their 728 corresponding MNPs. For example, if three Servers ('D', 'E' and 'F') 729 service the MNPs 2001:db8:1000:2000::/56, 2001:db8:3000:4000::/56 and 730 2001:db8:5000:6000::/56 then the routing system would include: 732 D: fc80:2001:db8:1000:2000::/72 734 E: fc80:2001:db8:3000:4000::/72 736 F: fc80:2001:db8:5000:6000::/72 738 A full discussion of the BGP-based routing system used by AERO is 739 found in [I-D.ietf-rtgwg-atn-bgp]. 741 3.2.4. OMNI Link Encapsulation 743 With the Client and partition prefixes in place in each Bridge's 744 forwarding table, control and data packets sent between AERO nodes in 745 different segments can therefore be carried over the spanning treee 746 via mid-layer encapsulation using the OMNI Adaptation Layer (OAL) 747 header. For example, when an AERO service node forwards a packet 748 with IPv6 address 2001:db8:1:2::1 to a target AERO node with IPv6 749 address 2001:db8:1000:2000::1, it first encapsulates the packet in an 750 OAL header with source address set to its own ULA address (e.g., 751 fc80::1000:2000) and destination address set to 752 fc80:2001:db8:1000:2000::. Next, it encapsulates the resulting OAL 753 packet in an INET header with source address set to its own INET 754 address (e.g., 192.0.2.100) and destination set to the INET address 755 of a Bridge (e.g., 192.0.2.1). 757 OAL encapsulation is based on Generic Packet Tunneling in IPv6 758 [RFC2473]; the encapsulation format in the above example is shown in 759 Figure 3: 761 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 762 | INET Header | 763 | src = 192.0.2.100 | 764 | dst = 192.0.2.1 | 765 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 766 | OAL Header | 767 | src = fc80::1000:2000 | 768 | dst=fc80:2001:db8:1000:2000:: | 769 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 770 | Inner IP Header | 771 | src = 2001:db8:1:2::1 | 772 | dst = 2001:db8:1000:2000::1 | 773 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 774 | | 775 ~ ~ 776 ~ Inner Packet Body ~ 777 ~ ~ 778 | | 779 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 781 Figure 3: SPAN Encapsulation 783 In this format, the inner IP header and packet body are the original 784 IP packet, the OAL header is an IPv6 header prepared according to 785 [RFC2473], and the INET header is prepared as discussed in 786 Section 3.6. 788 This gives rise to a routing system that contains both Client prefix 789 routes that may change dynamically due to regional node mobility and 790 partion prefix routes that rarely if ever change. The Bridges can 791 therefore provide link-layer bridging by sending packets over the 792 spanning tree instead of network-layer routing according to MNP 793 routes. As a result, opportunities for packet loss due to node 794 mobility between different segments are mitigated. 796 In normal operations, IPv6 ND messages are conveyed over secured 797 paths between OMNI link neighbors so that specific Proxys, Servers or 798 Relays can be addressed without being subject to mobility events. 799 Conversely, only the first few packets destined to Clients need to 800 traverse secured paths until route optimization can determine a more 801 direct path. 803 3.2.5. Segment Routing Topologies (SRTs) 805 The 16-bit sub-prefixes of fc80::/10 identify up to 64 distinct 806 Segment Routing Topologies (SRTs). Each SRT is a mutually-exclusive 807 OMNI link overlay instance using a mutually-exclusive set of ULAs, 808 and emulates a Virtual LAN (VLAN) service for the OMNI link. In some 809 cases (e.g., when redundant topologies are needed for fault tolerance 810 and reliability) it may be beneficial to deploy multiple SRTs that 811 act as independent overlay instances. A communication failure in one 812 instance therefore will not affect communications in other instances. 814 Each SRT is identified by a distinct value in bits 10-15 of fc80::10, 815 i.e., as fc80::/16, fc81::/16, fc82::/16, etc. This document asserts 816 that up to four SRTs provide a level of safety sufficient for 817 critical communications such as civil aviation. Each SRT is 818 designated with a color that identifies a different OMNI link 819 instance as follows: 821 o Red - corresponds to fc80::/16 823 o Green - corresponds to fc81::/16 825 o Blue-1 - corresponds to fc82::/16 827 o Blue-2 - corresponds to fc83::/16 829 o fc84::/16 through fcbf::/16 are available for additional SRTs. 831 Each OMNI interface assigns an anycast ULA corresponding to its SRT 832 prefix. For example, the anycast ULA for the Green SRT is simply 833 fc81::. The anycast ULA is used for OMNI interface determination in 834 Safety-Based Multilink (SBM) as discussed in 835 [I-D.templin-6man-omni-interface]. Each OMNI interface further 836 applies Performance-Based Multilink (PBM) internally. 838 3.2.6. Segment Routing To the OMNI Link 840 An original IPv6 source can direct a packet to an OMNI link Client by 841 including a Segment Routing Header (SRH) with the anycast ULA for the 842 selected SRT as either the IPv6 destination or as an intermediate hop 843 within the SRH. This allows the original source to determine the 844 specific topology a packet will traverse when there may be multiple 845 alternatives to choose from. Since the SRH contains no useful 846 information for the destination, the Client may elect to delete the 847 SRH before forwarding in order to reduce overhead. This form of 848 Segment Routing supports Safety-Based Multilink (SBM), and can be 849 exercised through general-purpose SRH types such as [RFC8754]. 851 3.2.7. Segment Routing Within the OMNI Link 853 AERO nodes that insert an OAL header can use Segment Routing within 854 the OMNI link when necessary to influence the path of packets 855 destined to targets in remote segments without requiring all packets 856 to traverse strict spanning tree paths. 858 When a Client, Proxy or Server has a packet to send to a target 859 discovered through route optimization located in the same OMNI link 860 segment, it encapsulates the packet in an OAL header with the ULA of 861 the target as the destination address if fragmentation is necessary; 862 otherwise, it may omit the OAL header. The node then uses the 863 target's Link Layer Address (L2ADDR) information for INET 864 encapsulation without including an SRH. 866 When a Client, Proxy or Server has a packet to send to a route 867 optimization target located in a remote OMNI link segment, it 868 encapsulates the packet in an OAL header with its own ULA as the 869 source address. The node then SHOULD include an SRH [RFC8754] while 870 forwarding the packet to a Bridge. 872 When the SRH is omitted, the node sets the OAL header destination 873 address to the ULA of the target Client/Proxy/Server and packet 874 forwarding is via spanning tree paths. When the SRH is included, the 875 node first sets the destination address to the ULA Subnet Router 876 Anycast address of the remote segment and sets the lower 32 bits of 877 the ULA of the target's Proxy/Server as the Last Hop Segment (LHS). 878 The node also includes an AERO Route Optimization specification in 879 the SRH TLV section as shown in Figure 4: 881 0 1 2 3 882 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 1 883 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 884 | Type=TBD | Length | MNPlen|V| FMT | MNP[1] | 885 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 886 | MNP[2] | MNP[3] | ... | MNP[i] | 887 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 888 ~ Link Layer Address (L2ADDR) ~ 889 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 891 Figure 4: AERO Route Optimization SRH TLV 893 In this format: 895 o Type is TBD to be assigned according to the Segment Routing Header 896 TLV registry [RFC8754]. 898 o Length is the length of the body of the TLV in bytes, excluding 899 the Type and Length fields. 901 o MNPlen encodes a value between 0 and 15 that, when added to 1, 902 indicates the number of octets of the IPv4/IPv6 MNP prefix that 903 follows. (For example, when MNPlen encodes the value 2 the number 904 of octets 'i' is set to the value 3.) 906 o V indicates the IP protocol version of the MNP that follows. V is 907 set to 0 for IPv4 or 1 for IPv6. 909 o FMT is a three bit code that determines the context and format of 910 the L2ADDR exactly as specified in the OMNI Interface Attributes 911 sub-option [I-D.templin-6man-omni-interface]. 913 o MNP[1], MNP[2], etc. up to MNP[i] encode the leading 'i' octets of 914 the MNP, beginning with the most significant octet followed by the 915 next most significant octet, etc. The number of MNP octets to be 916 included is determined by the number of trailing zero octets in 917 the prefix. For example, for the IPv6 MNP 2001:db8:1:2::/64, 'i' 918 is set to 8 and only the leftmost 8 octets of the MNP are 919 included. In the same way, for the IPv4 MNP 192.0.2/24, 'i' is 920 set to 3 and only the leftmost 3 octets of the MNP are included. 922 o Link Layer Address (L2ADDR) is a UDP Port Number and IP address 923 encoded according to FMT exactly as specified in 924 [I-D.templin-6man-omni-interface]. 926 The node then forwards the packet via a local Bridge, which will 927 eventually direct it to a Bridge on the same segment as the target. 929 When a Bridge receives a packet with Segments Left=1 and with LHS on 930 a local segment, it checks to see if there is an AERO Route 931 Optimization TLV. If so, the Bridge creates a ULA destination 932 according to FMT. If FMT indicates that L2ADDR corresponds to a 933 target Proxy/Server, the Bridge concatenates the SRT fc*::/96 prefix 934 with the 32 bit LHS value to form the ULA destination. Otherwise, 935 the Bridge concatenates the SRT fc*::/16 prefix with the leading 936 MNPlen octets of the MNP and sets the remaining rightmost bits to 0 937 to form a Subnet Router Anycast ULA destination. The Bridge then 938 writes the ULA into the OAL header destination address and 939 encapsulates the packet in an INET header with the target's L2ADDR as 940 the destination then forwards the packet. Since the SRH contains no 941 useful information for the destination, the Bridge may elect to 942 delete the SRH before forwarding in order to reduce overhead. 944 In this way, the Bridge participates in route optimization to reduce 945 traffic load and suboptimal routing through strict spanning tree 946 paths. Note that if the Bridge does not recognize the AERO Route 947 Optimization TLV, it instead places the SRT fc*::/96 prefix 948 concatenated with the 32 bit LHS in the IPv6 destination address and 949 forwards according to the spanning tree. (Note that this is the same 950 behavior that would occur if the AERO Route Optimization TLV were not 951 present). 953 3.2.8. Segment Routing Header Compression 955 In the Segment Routing use cases discussed above, the segment routing 956 headers must be kept to a minimum size since source and target 957 Clients may be located behind low-end wireless links (e.g., 1Mbps or 958 less). The Compressed Routing Header (CRH) 959 [I-D.bonica-6man-comp-rtg-hdr] provides a compact form that reduces 960 the header size by omitting invariant information. The CRH Helper 961 option [I-D.bonica-6man-crh-helper-opt] can be used to encode the 962 AERO Route Optimization TLV, and the final hop Bridge that performs 963 route optimization may remove the CRH and its helper before 964 encapsulating and forwarding to the target. 966 The CRH and its companion helper option are therefore seen as 967 critical architectural elements that should be quickly progressed 968 through the standards process. Implementations SHOULD use the CRH 969 and its companion helper option instead of other Routing Header types 970 whenever possible to conserve bandwidth. 972 3.3. OMNI Interface Characteristics 974 OMNI interfaces are virtual interfaces configured over one or more 975 underlying interfaces classified as follows: 977 o INET interfaces connect to an INET either natively or through one 978 or several IPv4 Network Address Translators (NATs). Native INET 979 interfaces have global IP addresses that are reachable from any 980 INET correspondent. All Server, Relay and Bridge interfaces are 981 native interfaces, as are INET-facing interfaces of Proxys. NATed 982 INET interfaces connect to a private network behind one or more 983 NATs that provide INET access. Clients that are behind a NAT are 984 required to send periodic keepalive messages to keep NAT state 985 alive when there are no data packets flowing. 987 o ANET interfaces connect to an ANET that is separated from the open 988 INET by a Proxy. Proxys can actively issue control messages over 989 the INET on behalf of the Client to reduce ANET congestion. 991 o VPNed interfaces use security encapsulation over the INET to a 992 Virtual Private Network (VPN) server that also acts as a Server or 993 Proxy. Other than the link-layer encapsulation format, VPNed 994 interfaces behave the same as Direct interfaces. 996 o Direct interfaces connect a Client directly to a Server or Proxy 997 without crossing any ANET/INET paths. An example is a line-of- 998 sight link between a remote pilot and an unmanned aircraft. The 999 same Client considerations apply as for VPNed interfaces. 1001 OMNI interfaces use OAL encapsulation as necessary as discussed in 1002 Section 3.2.4. OMNI interfaces use link-layer encapsulation (see: 1003 Section 3.6) to exchange packets with OMNI link neighbors over INET 1004 or VPNed interfaces. OMNI interfaces do not use link-layer 1005 encapsulation over ANET and Direct underlying interfaces. 1007 OMNI interfaces maintain a neighbor cache for tracking per-neighbor 1008 state the same as for any interface. OMNI interfaces use ND messages 1009 including Router Solicitation (RS), Router Advertisement (RA), 1010 Neighbor Solicitation (NS) and Neighbor Advertisement (NA) for 1011 neighbor cache management. 1013 OMNI interfaces send ND messages with an OMNI option formatted as 1014 specified in [I-D.templin-6man-omni-interface]. The OMNI option 1015 includes prefix registration information and Interface Attributes 1016 containing link information parameters for the OMNI interface's 1017 underlying interfaces. Each OMNI option may include multiple 1018 Interface Attributes sub-options, each idenfiied by an ifIndex value. 1020 A Client's OMNI interface may be configured over multiple underlying 1021 interface connections. For example, common mobile handheld devices 1022 have both wireless local area network ("WLAN") and cellular wireless 1023 links. These links are often used "one at a time" with low-cost WLAN 1024 preferred and highly-available cellular wireless as a standby, but a 1025 simultaneous-use capability could provide benefits. In a more 1026 complex example, aircraft frequently have many wireless data link 1027 types (e.g. satellite-based, cellular, terrestrial, air-to-air 1028 directional, etc.) with diverse performance and cost properties. 1030 If a Client's multiple underlying interfaces are used "one at a time" 1031 (i.e., all other interfaces are in standby mode while one interface 1032 is active), then ND message OMNI options include only a single 1033 Interface Attribues sub-option set to constant values. In that case, 1034 the Client would appear to have a single interface but with a 1035 dynamically changing link-layer address. 1037 If the Client has multiple active underlying interfaces, then from 1038 the perspective of ND it would appear to have multiple link-layer 1039 addresses. In that case, ND message OMNI options MAY include 1040 multiple Interface Attributes sub-options - each with values that 1041 correspond to a specific interface. Every ND message need not 1042 include Interface Attributes for all underlying interfaces; for any 1043 attributes not included, the neighbor considers the status as 1044 unchanged. 1046 Bridge, Server and Proxy OMNI interfaces may be configured over one 1047 or more secured tunnel interfaces. The OMNI interface configures 1048 both an LLA and its corresponding ULA, while the underlying secured 1049 tunnel interfaces are either unnumbered or configure the same ULA. 1050 The OMNI interface encapsulates each IP packet in an OAL header and 1051 presents the packet to the underlying secured tunnel interface. 1052 Routing protocols such as BGP that run over the OMNI interface do not 1053 employ OAL encapsulation, but rather present the routing protocol 1054 messages directly to the underlying secured tunnels while using the 1055 ULA as the source address. This distinction must be honored 1056 consistently according to each node's configuration so that the IP 1057 forwarding table will associate discovered IP routes with the correct 1058 interface. 1060 3.4. OMNI Interface Initialization 1062 AERO Servers, Proxys and Clients configure OMNI interfaces as their 1063 point of attachment to the OMNI link. AERO nodes assign the MSPs for 1064 the link to their OMNI interfaces (i.e., as a "route-to-interface") 1065 to ensure that packets with destination addresses covered by an MNP 1066 not explicitly assigned to a non-OMNI interface are directed to the 1067 OMNI interface. 1069 OMNI interface initialization procedures for Servers, Proxys, Clients 1070 and Bridges are discussed in the following sections. 1072 3.4.1. AERO Server/Relay Behavior 1074 When a Server enables an OMNI interface, it assigns an LLA/ULA 1075 appropriate for the given OMNI link segment. The Server also 1076 configures secured tunnels with one or more neighboring Bridges and 1077 engages in a BGP routing protocol session with each Bridge. 1079 The OMNI interface provides a single interface abstraction to the IP 1080 layer, but internally comprises multiple secured tunnels as well as 1081 an NBMA nexus for sending encapsulated data packets to OMNI interface 1082 neighbors. The Server further configures a service to facilitate ND 1083 exchanges with AERO Clients and manages per-Client neighbor cache 1084 entries and IP forwarding table entries based on control message 1085 exchanges. 1087 Relays are simply Servers that run a dynamic routing protocol to 1088 redistribute routes between the OMNI interface and INET/EUN 1089 interfaces (see: Section 3.2.3). The Relay provisions MNPs to 1090 networks on the INET/EUN interfaces (i.e., the same as a Client would 1091 do) and advertises the MSP(s) for the OMNI link over the INET/EUN 1092 interfaces. The Relay further provides an attachment point of the 1093 OMNI link to a non-MNP-based global topology. 1095 3.4.2. AERO Proxy Behavior 1097 When a Proxy enables an OMNI interface, it assigns an LLA/ULA and 1098 configures permanent neighbor cache entries the same as for Servers. 1099 The Proxy also configures secured tunnels with one or more 1100 neighboring Bridges and maintains per-Client neighbor cache entries 1101 based on control message exchanges. Importantly Proxys are often 1102 configured to act as Servers, and vice-versa. 1104 3.4.3. AERO Client Behavior 1106 When a Client enables an OMNI interface, it sends RS messages with ND 1107 parameters over its underlying interfaces to a Server, which returns 1108 an RA message with corresponding parameters. (The RS/RA messages may 1109 pass through a Proxy in the case of a Client's ANET interface, or 1110 through one or more NATs in the case of a Client's INET interface.) 1112 3.4.4. AERO Bridge Behavior 1114 AERO Bridges configure an OMNI interface and assign the ULA Subnet 1115 Router Anycast address for each OMNI link segment they connect to. 1116 Bridges configure secured tunnels with Servers, Proxys and other 1117 Bridges; they also configure LLAs/ULAs and permanent neighbor cache 1118 entries the same as Servers. Bridges engage in a BGP routing 1119 protocol session with a subset of the Servers and other Bridges on 1120 the spanning tree (see: Section 3.2.3). 1122 3.5. OMNI Interface Neighbor Cache Maintenance 1124 Each OMNI interface maintains a conceptual neighbor cache that 1125 includes an entry for each neighbor it communicates with on the OMNI 1126 link per [RFC4861]. OMNI interface neighbor cache entries are said 1127 to be one of "permanent", "symmetric", "asymmetric" or "proxy". 1129 Permanent neighbor cache entries are created through explicit 1130 administrative action; they have no timeout values and remain in 1131 place until explicitly deleted. AERO Bridges maintain permanent 1132 neighbor cache entries for their associated Proxys/Servers (and vice- 1133 versa). Each entry maintains the mapping between the neighbor's 1134 network-layer LLA and corresponding INET address. 1136 Symmetric neighbor cache entries are created and maintained through 1137 RS/RA exchanges as specified in Section 3.12, and remain in place for 1138 durations bounded by prefix lifetimes. AERO Servers maintain 1139 symmetric neighbor cache entries for each of their associated 1140 Clients, and AERO Clients maintain symmetric neighbor cache entries 1141 for each of their associated Servers. 1143 Asymmetric neighbor cache entries are created or updated based on 1144 route optimization messaging as specified in Section 3.14, and are 1145 garbage-collected when keepalive timers expire. AERO ROSs maintain 1146 asymmetric neighbor cache entries for active targets with lifetimes 1147 based on ND messaging constants. Asymmetric neighbor cache entries 1148 are unidirectional since only the ROS (and not the ROR) creates an 1149 entry. 1151 Proxy neighbor cache entries are created and maintained by AERO 1152 Proxys when they process Client/Server ND exchanges, and remain in 1153 place for durations bounded by ND and prefix lifetimes. AERO Proxys 1154 maintain proxy neighbor cache entries for each of their associated 1155 Clients. Proxy neighbor cache entries track the Client state and the 1156 address of the Client's associated Server(s). 1158 To the list of neighbor cache entry states in Section 7.3.2 of 1159 [RFC4861], Proxy and Server OMNI interfaces add an additional state 1160 DEPARTED that applies to symmetric and proxy neighbor cache entries 1161 for Clients that have recently departed. The interface sets a 1162 "DepartTime" variable for the neighbor cache entry to "DEPART_TIME" 1163 seconds. DepartTime is decremented unless a new ND message causes 1164 the state to return to REACHABLE. While a neighbor cache entry is in 1165 the DEPARTED state, packets destined to the target Client are 1166 forwarded to the Client's new location instead of being dropped. 1167 When DepartTime decrements to 0, the neighbor cache entry is deleted. 1168 It is RECOMMENDED that DEPART_TIME be set to the default constant 1169 value REACHABLE_TIME plus 10 seconds (40 seconds by default) to allow 1170 a window for packets in flight to be delivered while stale route 1171 optimization state may be present. 1173 When an ROR receives an authentic NS message used for route 1174 optimization, it searches for a symmetric neighbor cache entry for 1175 the target Client. The ROR then returns a solicited NA message 1176 without creating a neighbor cache entry for the ROS, but creates or 1177 updates a target Client "Report List" entry for the ROS and sets a 1178 "ReportTime" variable for the entry to REPORT_TIME seconds. The ROR 1179 resets ReportTime when it receives a new authentic NS message, and 1180 otherwise decrements ReportTime while no authentic NS messages have 1181 been received. It is RECOMMENDED that REPORT_TIME be set to the 1182 default constant value REACHABLE_TIME plus 10 seconds (40 seconds by 1183 default) to allow a window for route optimization to converge before 1184 ReportTime decrements below REACHABLE_TIME. 1186 When the ROS receives a solicited NA message response to its NS 1187 message used for route optimization, it creates or updates an 1188 asymmetric neighbor cache entry for the target network-layer and 1189 link-layer addresses. The ROS then (re)sets ReachableTime for the 1190 neighbor cache entry to REACHABLE_TIME seconds and uses this value to 1191 determine whether packets can be forwarded directly to the target, 1192 i.e., instead of via a default route. The ROS otherwise decrements 1193 ReachableTime while no further solicited NA messages arrive. It is 1194 RECOMMENDED that REACHABLE_TIME be set to the default constant value 1195 30 seconds as specified in [RFC4861]. 1197 AERO nodes also use the value MAX_UNICAST_SOLICIT to limit the number 1198 of NS keepalives sent when a correspondent may have gone unreachable, 1199 the value MAX_RTR_SOLICITATIONS to limit the number of RS messages 1200 sent without receiving an RA and the value MAX_NEIGHBOR_ADVERTISEMENT 1201 to limit the number of unsolicited NAs that can be sent based on a 1202 single event. It is RECOMMENDED that MAX_UNICAST_SOLICIT, 1203 MAX_RTR_SOLICITATIONS and MAX_NEIGHBOR_ADVERTISEMENT be set to 3 the 1204 same as specified in [RFC4861]. 1206 Different values for DEPART_TIME, REPORT_TIME, REACHABLE_TIME, 1207 MAX_UNICAST_SOLICIT, MAX_RTR_SOLCITATIONS and 1208 MAX_NEIGHBOR_ADVERTISEMENT MAY be administratively set; however, if 1209 different values are chosen, all nodes on the link MUST consistently 1210 configure the same values. Most importantly, DEPART_TIME and 1211 REPORT_TIME SHOULD be set to a value that is sufficiently longer than 1212 REACHABLE_TIME to avoid packet loss due to stale route optimization 1213 state. 1215 3.6. OMNI Interface Encapsulation and Re-encapsulation 1217 The OMNI Adaptation Layer (OAL) inserts a mid-layer IPv6 header known 1218 as the OAL header when necessary as discussed in the following 1219 sections. After either inserting or omitting the OAL header, the 1220 OMNI interface also inserts or omits an outer encapsulation header as 1221 discussed below. 1223 OMNI interfaces avoid outer encapsulation over Direct underlying 1224 interfaces and ANET underlying interfaces for which the first-hop 1225 access router is connected to the same underlying link. Otherwise, 1226 OMNI interfaces encapsulate packets according to whether they are 1227 entering the OMNI interface from the network layer or if they are 1228 being re-admitted into the same OMNI link they arrived on. This 1229 latter form of encapsulation is known as "re-encapsulation". 1231 For packets entering the OMNI interface from the network layer, the 1232 OMNI interface copies the "TTL/Hop Limit", "Type of Service/Traffic 1233 Class" [RFC2983], "Flow Label"[RFC6438] (for IPv6) and "Congestion 1234 Experienced" [RFC3168] values in the inner packet's IP header into 1235 the corresponding fields in the OAL and outer encapsulation 1236 header(s). 1238 For packets undergoing re-encapsulation, the OMNI interface instead 1239 copies these values from the original encapsulation header into the 1240 new encapsulation header, i.e., the values are transferred between 1241 encapsulation headers and *not* copied from the encapsulated packet's 1242 network-layer header. (Note especially that by copying the TTL/Hop 1243 Limit between encapsulation headers the value will eventually 1244 decrement to 0 if there is a (temporary) routing loop.) 1246 OMNI interfaces configured over ANET underlying interfaces which 1247 employ a different IP protocol version and/or may be located multiple 1248 IP hops from the nearest Proxy/Server use IP-in-IP encapsulation so 1249 that the inner packet can traverse the ANET. IPv6 underlying ANET 1250 interfaces use [RFC2473] encapsulation, while IPv4 interfaces use the 1251 appropriate encapsulation per one of [RFC2529][RFC5214][RFC2003]. 1253 OMNI interfaces configured over INET underlying interfaces 1254 encapsulate packets in INET headers according to the next hop 1255 determined in the forwarding algorithm in Section 3.10. If the next 1256 hop is reached via a secured tunnel, the OMNI interface uses an 1257 encapsulation format specific to the secured tunnel type (see: 1258 Section 6). If the next hop is reached via an unsecured INET 1259 interface, the OMNI interface instead uses UDP/IP encapsulation per 1260 [RFC4380] and as extended in [RFC6081]. 1262 When UDP/IP encapsulation is used, the OMNI interface next sets the 1263 UDP source port to a constant value that it will use in each 1264 successive packet it sends, and sets the UDP length field to the 1265 length of the encapsulated packet plus 8 bytes for the UDP header 1266 itself plus the length of any included extension headers or trailers. 1267 The encapsulated packet may be either IPv6 or IPv4, as distinguished 1268 by the version number found in the first four bits. 1270 For UDP/IP-encapsulated packets sent to a Server, Relay or Bridge, 1271 the OMNI interface sets the UDP destination port to 8060, i.e., the 1272 IANA-registered port number for AERO. For packets sent to a Client, 1273 the OMNI interface sets the UDP destination port to the port value 1274 stored in the neighbor cache entry for this Client. The OMNI 1275 interface finally includes/omits the UDP checksum according to 1276 [RFC6935][RFC6936]. 1278 3.7. OMNI Interface Decapsulation 1280 OMNI interfaces decapsulate packets destined either to the AERO node 1281 itself or to a destination reached via an interface other than the 1282 OMNI interface the packet was received on. When the encapsulated 1283 packet arrives in multiple OAL fragments, the OMNI interface 1284 reassembles as discussed in Section 3.9. Further decapsulation steps 1285 are performed according to the appropriate encapsulation format 1286 specification. 1288 3.8. OMNI Interface Data Origin Authentication 1290 AERO nodes employ simple data origin authentication procedures. In 1291 particular: 1293 o AERO Bridges, Servers and Proxys accept encapsulated data packets 1294 and control messages received from the (secured) spanning tree. 1296 o AERO Proxys and Clients accept packets that originate from within 1297 the same secured ANET. 1299 o AERO Clients and Relays accept packets from downstream network 1300 correspondents based on ingress filtering. 1302 o AERO Clients, Relays and Servers verify the outer UDP/IP 1303 encapsulation addresses according to [RFC4380]. 1305 AERO nodes silently drop any packets that do not satisfy the above 1306 data origin authentication procedures. Further security 1307 considerations are discussed in Section 6. 1309 3.9. OMNI Adaptation Layer (OAL) and OMNI Interface MTU 1311 The OMNI interface observes the link nature of tunnels, including the 1312 Maximum Transmission Unit (MTU), Maximum Reassembly Unit (MRU) and 1313 the role of fragmentation and reassembly[I-D.ietf-intarea-tunnels]. 1314 The OMNI interface employs an OMNI Adaptation Layer (OAL) for 1315 accommodating multiple underlying links with diverse MTUs. The 1316 functions of the OAL and the OMNI interface MTU/MRU are specified in 1317 Section 5 of [I-D.templin-6man-omni-interface], with MTU/MRU both set 1318 to the constant value 9180 bytes. 1320 3.10. OMNI Interface Forwarding Algorithm 1322 IP packets enter a node's OMNI interface either from the network 1323 layer (i.e., from a local application or the IP forwarding system) or 1324 from the link layer (i.e., from an OMNI interface neighbor). All 1325 packets entering a node's OMNI interface first undergo data origin 1326 authentication as discussed in Section 3.8. Packets that satisfy 1327 data origin authentication are processed further, while all others 1328 are dropped silently. The OMNI interface OAL wraps accepted packets 1329 in an OAL header and SRH if necessary as discussed above. 1331 Packets that enter the OMNI interface from the network layer are 1332 forwarded to an OMNI interface neighbor. Packets that enter the OMNI 1333 interface from the link layer are either re-admitted into the OMNI 1334 link or forwarded to the network layer where they are subject to 1335 either local delivery or IP forwarding. In all cases, the OMNI 1336 interface itself MUST NOT decrement the network layer TTL/Hop-count 1337 since its forwarding actions occur below the network layer. 1339 OMNI interfaces may have multiple underlying interfaces and/or 1340 neighbor cache entries for neighbors with multiple underlying 1341 interfaces (see Section 3.3). The OMNI interface uses traffic 1342 classifiers (e.g., DSCP value, port number, etc.) to select an 1343 outgoing underlying interface for each packet based on the node's own 1344 QoS preferences, and also to select a destination link-layer address 1345 based on the neighbor's underlying interface with the highest 1346 preference. AERO implementations SHOULD allow for QoS preference 1347 values to be modified at runtime through network management. 1349 If multiple outgoing interfaces and/or neighbor interfaces have a 1350 preference of "high", the AERO node replicates the packet and sends 1351 one copy via each of the (outgoing / neighbor) interface pairs; 1352 otherwise, the node sends a single copy of the packet via an 1353 interface with the highest preference. AERO nodes keep track of 1354 which underlying interfaces are currently "reachable" or 1355 "unreachable", and only use "reachable" interfaces for forwarding 1356 purposes. 1358 The following sections discuss the OMNI interface forwarding 1359 algorithms for Clients, Proxys, Servers and Bridges. In the 1360 following discussion, a packet's destination address is said to 1361 "match" if it is the same as a cached address, or if it is covered by 1362 a cached prefix (which may be encoded in an LLA). 1364 3.10.1. Client Forwarding Algorithm 1366 When an IP packet enters a Client's OMNI interface from the network 1367 layer the Client searches for an asymmetric neighbor cache entry that 1368 matches the destination. If there is a match, the Client uses one or 1369 more "reachable" neighbor interfaces in the entry for packet 1370 forwarding. If there is no asymmetric neighbor cache entry, the 1371 Client instead forwards the packet toward a Server (the packet is 1372 intercepted by a Proxy if there is a Proxy on the path). The Client 1373 encapsulates the packet in an OAL header and SRH if necessary and 1374 fragments according to MTU requirements (see: Section 3.9). 1376 When an IP packet enters a Client's OMNI interface from the link- 1377 layer, if the destination matches one of the Client's MNPs or link- 1378 local addresses the Client reassembles and decapsulates as necessary 1379 and delivers the inner packet to the network layer. Otherwise, the 1380 Client drops the packet and MAY return a network-layer ICMP 1381 Destination Unreachable message subject to rate limiting (see: 1382 Section 3.11). 1384 3.10.2. Proxy Forwarding Algorithm 1386 For control messages originating from or destined to a Client, the 1387 Proxy intercepts the message and updates its proxy neighbor cache 1388 entry for the Client. The Proxy then forwards a (proxyed) copy of 1389 the control message. (For example, the Proxy forwards a proxied 1390 version of a Client's NS/RS message to the target neighbor, and 1391 forwards a proxied version of the NA/RA reply to the Client.) 1393 When the Proxy receives a data packet from a Client within the ANET, 1394 the Proxy reassembles and re-fragments if necessary then searches for 1395 an asymmetric neighbor cache entry that matches the destination and 1396 forwards as follows: 1398 o if the destination matches an asymmetric neighbor cache entry, the 1399 Proxy uses one or more "reachable" neighbor interfaces in the 1400 entry for packet forwarding using OAL encapsulation and including 1401 a SRH if necessary according to the cached link-layer address 1402 information. If the neighbor interface is in the same OMNI link 1403 segment, the Proxy forwards the packet directly to the neighbor; 1404 otherwise, it forwards the packet to a Bridge. 1406 o else, the Proxy uses OAL encapsulation and forwards the packet to 1407 a Bridge while using the ULA corresponding to the packet's 1408 destination as the OAL destination address. 1410 When the Proxy receives an encapsulated data packet from an INET 1411 neighbor or from a secured tunnel from a Bridge, it accepts the 1412 packet only if data origin authentication succeeds and if there is a 1413 proxy neighbor cache entry that matches the inner destination. Next, 1414 the Proxy reassembles the packet (if necessary) and continues 1415 processing. If the reassembly is complete and the neighbor cache 1416 state is REACHABLE, the Proxy then returns a PTB if necessary (see: 1417 Section 3.9) then either drops or forwards the packet to the Client 1418 while performing OAL encapsulation and re-fragmentation if necessary. 1419 If the neighbor cache entry state is DEPARTED, the Proxy instead 1420 changes the OAL destination address to the address of the new Server 1421 and forwards it to a Bridge while performing OAL re-fragmentation if 1422 necessary. 1424 3.10.3. Server/Relay Forwarding Algorithm 1426 For control messages destined to a target Client's LLA that are 1427 received from a secured tunnel, the Server intercepts the message and 1428 sends a Proxyed response on behalf of the Client. (For example, the 1429 Server sends a Proxyed NA message reply in response to an NS message 1430 directed to one of its associated Clients.) If the Client's neighbor 1431 cache entry is in the DEPARTED state, however, the Server instead 1432 forwards the packet to the Client's new Server as discussed in 1433 Section 3.16. 1435 When the Server receives an encapsulated data packet from an INET 1436 neighbor or from a secured tunnel, it accepts the packet only if data 1437 origin authentication succeeds. The Server then continues processing 1438 as follows: 1440 o if the network layer destination matches a symmetric neighbor 1441 cache entry in the REACHABLE state the Server prepares the packet 1442 for forwarding to the destination Client. The Server first 1443 reassembles (if necessary) and forwards the packet (while re- 1444 fragmenting if necessary) as specified in Section 3.9. 1446 o else, if the destination matches a symmetric neighbor cache entry 1447 in the DEPARETED state the Server re-encapsulates the packet and 1448 forwards it using the ULA of the Client's new Server as the OAL 1449 header destination. 1451 o else, if the destination matches an asymmetric neighbor cache 1452 entry, the Server uses one or more "reachable" neighbor interfaces 1453 in the entry for packet forwarding via the local INET if the 1454 neighbor is in the same OMNI link segment or using OAL 1455 encapsulation and Segment Routing if necessary with the final 1456 destination set to the neighbor's ULA otherwise. 1458 o else, if the destination matches a non-MNP route in the IP 1459 forwarding table or an LLA assigned to the Server's OMNI 1460 interface, the Server reassembles if necessary, decapsulates the 1461 packet and releases it to the network layer for local delivery or 1462 IP forwarding. 1464 o else, the Server drops the packet. 1466 When the Server's OMNI interface receives a data packet from the 1467 network layer or from a VPNed or Direct Client, it performs OAL 1468 encapsulation and fragmentation if necessary, then processes the 1469 packet according to the network-layer destination address as follows: 1471 o if the destination matches a symmetric or asymmetric neighbor 1472 cache entry the Server processes the packet as above. 1474 o else, the Server encapsulates the packet in an OAL header and 1475 forwards it to a Bridge using its own ULA as the source and the 1476 ULA corresponding to the destination as the destination. 1478 3.10.4. Bridge Forwarding Algorithm 1480 Bridges forward OAL-encapsulated packets over secured tunnels the 1481 same as any IP router. When the Bridge receives an OAL-encapsulated 1482 packet via a secured tunnel, it removes the outer INET header and 1483 searches for a forwarding table entry that matches the OAL 1484 destination address. The Bridge then processes the packet as 1485 follows: 1487 o if the destination matches its ULA Subnet Router Anycast address, 1488 the Bridge checks for a SRH. If there is a SRH with Segments 1489 Left=1, with the ULA of a Proxy/Server on the local segment as the 1490 LHS, and with an AERO Route Optimization TLV, the Bridge examines 1491 the FMT to determine if the target is behind a NAT. If no NAT is 1492 indicated, the Bridge copies the MNP Subnet Router Anycast address 1493 if an MNP is included (otherwise copies the Proxy/Server ULA) into 1494 the destination address then forwards the packet directly to the 1495 L2ADDR using link-layer (UDP/IP) encapsulation. If a NAT is 1496 indicated, the Bridge MAY perform NAT traversal procedures by 1497 sending bubbles per [RFC4380]. The Bridge then either applies 1498 AERO route optimization if NAT traversal procedures have been 1499 successfully applied, or forwards the packet directly to the 1500 Server. 1502 o if the destination matches one of the Bridge's own addresses, the 1503 Bridge submits the packet for local delivery. 1505 o else, if the destination matches a forwarding table entry the 1506 Bridge forwards the packet via a secured tunnel to the next hop. 1507 If the destination matches an MSP without matching an MNP, 1508 however, the Bridge instead drops the packet and returns an ICMP 1509 Destination Unreachable message subject to rate limiting (see: 1510 Section 3.11). 1512 o else, the Bridge drops the packet and returns an ICMP Destination 1513 Unreachable as above. 1515 As for any IP router, the Bridge decrements the TTL/Hop Limit when it 1516 forwards the packet. Therefore, only the Hop Limit in the OAL header 1517 is decremented, and not the TTL/Hop Limit in the inner packet header. 1518 Bridges do not insert OAL headers themselves; instead, they act as 1519 IPv6 routers and forward packets based on the destination address 1520 found in the OAL headers of packets they receive. 1522 3.11. OMNI Interface Error Handling 1524 When an AERO node admits a packet into the OMNI interface, it may 1525 receive link-layer or network-layer error indications. 1527 A link-layer error indication is an ICMP error message generated by a 1528 router in the INET on the path to the neighbor or by the neighbor 1529 itself. The message includes an IP header with the address of the 1530 node that generated the error as the source address and with the 1531 link-layer address of the AERO node as the destination address. 1533 The IP header is followed by an ICMP header that includes an error 1534 Type, Code and Checksum. Valid type values include "Destination 1535 Unreachable", "Time Exceeded" and "Parameter Problem" 1536 [RFC0792][RFC4443]. (OMNI interfaces ignore all link-layer IPv4 1537 "Fragmentation Needed" and IPv6 "Packet Too Big" messages since they 1538 only emit packets that are guaranteed to be no larger than the IP 1539 minimum link MTU as discussed in Section 3.9.) 1541 The ICMP header is followed by the leading portion of the packet that 1542 generated the error, also known as the "packet-in-error". For 1543 ICMPv6, [RFC4443] specifies that the packet-in-error includes: "As 1544 much of invoking packet as possible without the ICMPv6 packet 1545 exceeding the minimum IPv6 MTU" (i.e., no more than 1280 bytes). For 1546 ICMPv4, [RFC0792] specifies that the packet-in-error includes: 1547 "Internet Header + 64 bits of Original Data Datagram", however 1548 [RFC1812] Section 4.3.2.3 updates this specification by stating: "the 1549 ICMP datagram SHOULD contain as much of the original datagram as 1550 possible without the length of the ICMP datagram exceeding 576 1551 bytes". 1553 The link-layer error message format is shown in Figure 5 (where, "L2" 1554 and "L3" refer to link-layer and network-layer, respectively): 1556 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1557 ~ ~ 1558 | L2 IP Header of | 1559 | error message | 1560 ~ ~ 1561 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1562 | L2 ICMP Header | 1563 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1564 ~ ~ P 1565 | IP and other encapsulation | a 1566 | headers of original L3 packet | c 1567 ~ ~ k 1568 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ e 1569 ~ ~ t 1570 | IP header of | 1571 | original L3 packet | i 1572 ~ ~ n 1573 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1574 ~ ~ e 1575 | Upper layer headers and | r 1576 | leading portion of body | r 1577 | of the original L3 packet | o 1578 ~ ~ r 1579 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1581 Figure 5: OMNI Interface Link-Layer Error Message Format 1583 The AERO node rules for processing these link-layer error messages 1584 are as follows: 1586 o When an AERO node receives a link-layer Parameter Problem message, 1587 it processes the message the same as described as for ordinary 1588 ICMP errors in the normative references [RFC0792][RFC4443]. 1590 o When an AERO node receives persistent link-layer Time Exceeded 1591 messages, the IP ID field may be wrapping before earlier fragments 1592 awaiting reassembly have been processed. In that case, the node 1593 should begin including integrity checks and/or institute rate 1594 limits for subsequent packets. 1596 o When an AERO node receives persistent link-layer Destination 1597 Unreachable messages in response to encapsulated packets that it 1598 sends to one of its asymmetric neighbor correspondents, the node 1599 should process the message as an indication that a path may be 1600 failing, and optionally initiate NUD over that path. If it 1601 receives Destination Unreachable messages over multiple paths, the 1602 node should allow future packets destined to the correspondent to 1603 flow through a default route and re-initiate route optimization. 1605 o When an AERO Client receives persistent link-layer Destination 1606 Unreachable messages in response to encapsulated packets that it 1607 sends to one of its symmetric neighbor Servers, the Client should 1608 mark the path as unusable and use another path. If it receives 1609 Destination Unreachable messages on many or all paths, the Client 1610 should associate with a new Server and release its association 1611 with the old Server as specified in Section 3.16.5. 1613 o When an AERO Server receives persistent link-layer Destination 1614 Unreachable messages in response to encapsulated packets that it 1615 sends to one of its symmetric neighbor Clients, the Server should 1616 mark the underlying path as unusable and use another underlying 1617 path. 1619 o When an AERO Server or Proxy receives link-layer Destination 1620 Unreachable messages in response to an encapsulated packet that it 1621 sends to one of its permanent neighbors, it treats the messages as 1622 an indication that the path to the neighbor may be failing. 1623 However, the dynamic routing protocol should soon reconverge and 1624 correct the temporary outage. 1626 When an AERO Bridge receives a packet for which the network-layer 1627 destination address is covered by an MSP, if there is no more- 1628 specific routing information for the destination the Bridge drops the 1629 packet and returns a network-layer Destination Unreachable message 1630 subject to rate limiting. The Bridge writes the network-layer source 1631 address of the original packet as the destination address and uses 1632 one of its non link-local addresses as the source address of the 1633 message. 1635 When an AERO node receives an encapsulated packet for which the 1636 reassembly buffer it too small, it drops the packet and returns a 1637 network-layer Packet Too Big (PTB) message. The node first writes 1638 the MRU value into the PTB message MTU field, writes the network- 1639 layer source address of the original packet as the destination 1640 address and writes one of its non link-local addresses as the source 1641 address. 1643 3.12. AERO Router Discovery, Prefix Delegation and Autoconfiguration 1645 AERO Router Discovery, Prefix Delegation and Autoconfiguration are 1646 coordinated as discussed in the following Sections. 1648 3.12.1. AERO Service Model 1650 Each AERO Server on the OMNI link is configured to facilitate Client 1651 prefix delegation/registration requests. Each Server is provisioned 1652 with a database of MNP-to-Client ID mappings for all Clients enrolled 1653 in the AERO service, as well as any information necessary to 1654 authenticate each Client. The Client database is maintained by a 1655 central administrative authority for the OMNI link and securely 1656 distributed to all Servers, e.g., via the Lightweight Directory 1657 Access Protocol (LDAP) [RFC4511], via static configuration, etc. 1658 Clients receive the same service regardless of the Servers they 1659 select. 1661 AERO Clients and Servers use ND messages to maintain neighbor cache 1662 entries. AERO Servers configure their OMNI interfaces as advertising 1663 NBMA interfaces, and therefore send unicast RA messages with a short 1664 Router Lifetime value (e.g., ReachableTime seconds) in response to a 1665 Client's RS message. Thereafter, Clients send additional RS messages 1666 to keep Server state alive. 1668 AERO Clients and Servers include prefix delegation and/or 1669 registration parameters in RS/RA messages (see 1670 [I-D.templin-6man-omni-interface]). The ND messages are exchanged 1671 between Client and Server according to the prefix management schedule 1672 required by the service. If the Client knows its MNP in advance, it 1673 can employ prefix registration by including its LLA as the source 1674 address of an RS message and with an OMNI option with valid prefix 1675 registration information for the MNP. If the Server (and Proxy) 1676 accept the Client's MNP assertion, they inject the prefix into the 1677 routing system and establish the necessary neighbor cache state. 1679 The following sections specify the Client and Server behavior. 1681 3.12.2. AERO Client Behavior 1683 AERO Clients discover the addresses of Servers in a similar manner as 1684 described in [RFC5214]. Discovery methods include static 1685 configuration (e.g., from a flat-file map of Server addresses and 1686 locations), or through an automated means such as Domain Name System 1687 (DNS) name resolution [RFC1035]. Alternatively, the Client can 1688 discover Server addresses through a layer 2 data link login exchange, 1689 or through a unicast RA response to a multicast/anycast RS as 1690 described below. In the absence of other information, the Client can 1691 resolve the DNS Fully-Qualified Domain Name (FQDN) 1692 "linkupnetworks.[domainname]" where "linkupnetworks" is a constant 1693 text string and "[domainname]" is a DNS suffix for the OMNI link 1694 (e.g., "example.com"). 1696 To associate with a Server, the Client acts as a requesting router to 1697 request MNPs. The Client prepares an RS message with prefix 1698 management parameters and includes a Nonce and Timestamp option if 1699 the Client needs to correlate RA replies. If the Client already 1700 knows the Server's LLA, it includes the LLA as the network-layer 1701 destination address; otherwise, it includes (link-local) All-Routers 1702 multicast as the network-layer destination. If the Client already 1703 knows its own LLA, it uses the LLA as the network-layer source 1704 address; otherwise, it uses the unspecified IPv6 address (::/128) as 1705 the network-layer source address. 1707 The Client next includes an OMNI option in the RS message to register 1708 its link-layer information with the Server. The Client sets the OMNI 1709 option prefix registration information according to the MNP, and 1710 includes Interface Attributes corresponding to the underlying 1711 interface over which the Client will send the RS message. The Client 1712 MAY include additional Interface Attributes specific to other 1713 underlying interfaces. 1715 The Client then sends the RS message (either directly via Direct 1716 interfaces, via a VPN for VPNed interfaces, via a Proxy for ANET 1717 interfaces or via INET encapsulation for INET interfaces) and waits 1718 for an RA message reply (see Section 3.12.3). The Client retries up 1719 to MAX_RTR_SOLICITATIONS times until an RA is received. If the 1720 Client receives no RAs, or if it receives an RA with Router Lifetime 1721 set to 0, the Client SHOULD abandon this Server and try another 1722 Server. Otherwise, the Client processes the prefix information found 1723 in the RA message. 1725 Next, the Client creates a symmetric neighbor cache entry with the 1726 Server's LLA as the network-layer address and the Server's 1727 encapsulation and/or link-layer addresses as the link-layer address. 1728 The Client records the RA Router Lifetime field value in the neighbor 1729 cache entry as the time for which the Server has committed to 1730 maintaining the MNP in the routing system via this underlying 1731 interface, and caches the other RA configuration information 1732 including Cur Hop Limit, M and O flags, Reachable Time and Retrans 1733 Timer. The Client then autoconfigures LLAs for each of the delegated 1734 MNPs and assigns them to the OMNI interface. The Client also caches 1735 any MSPs included in Route Information Options (RIOs) [RFC4191] as 1736 MSPs to associate with the OMNI link, and assigns the MTU value in 1737 the MTU option to the underlying interface. 1739 The Client then registers additional underlying interfaces with the 1740 Server by sending RS messages via each additional interface. The RS 1741 messages include the same parameters as for the initial RS/RA 1742 exchange, but with destination address set to the Server's LLA. 1744 Following autoconfiguration, the Client sub-delegates the MNPs to its 1745 attached EUNs and/or the Client's own internal virtual interfaces as 1746 described in [I-D.templin-v6ops-pdhost] to support the Client's 1747 downstream attached "Internet of Things (IoT)". The Client 1748 subsequently sends additional RS messages over each underlying 1749 interface before the Router Lifetime received for that interface 1750 expires. 1752 After the Client registers its underlying interfaces, it may wish to 1753 change one or more registrations, e.g., if an interface changes 1754 address or becomes unavailable, if QoS preferences change, etc. To 1755 do so, the Client prepares an RS message to send over any available 1756 underlying interface. The RS includes an OMNI option with prefix 1757 registration information specific to its MNP, with Interface 1758 Attributes specific to the selected underlying interface, and with 1759 any additional Interface Attributes specific to other underlying 1760 interfaces. When the Client receives the Server's RA response, it 1761 has assurance that the Server has been updated with the new 1762 information. 1764 If the Client wishes to discontinue use of a Server it issues an RS 1765 message over any underlying interface with an OMNI option with a 1766 prefix release indication. When the Server processes the message, it 1767 releases the MNP, sets the symmetric neighbor cache entry state for 1768 the Client to DEPARTED and returns an RA reply with Router Lifetime 1769 set to 0. After a short delay (e.g., 2 seconds), the Server 1770 withdraws the MNP from the routing system. 1772 3.12.3. AERO Server Behavior 1774 AERO Servers act as IP routers and support a prefix delegation/ 1775 registration service for Clients. Servers arrange to add their LLAs 1776 to a static map of Server addresses for the link and/or the DNS 1777 resource records for the FQDN "linkupnetworks.[domainname]" before 1778 entering service. Server addresses should be geographically and/or 1779 topologically referenced, and made available for discovery by Clients 1780 on the OMNI link. 1782 When a Server receives a prospective Client's RS message on its OMNI 1783 interface, it SHOULD return an immediate RA reply with Router 1784 Lifetime set to 0 if it is currently too busy or otherwise unable to 1785 service the Client. Otherwise, the Server authenticates the RS 1786 message and processes the prefix delegation/registration parameters. 1787 The Server first determines the correct MNPs to provide to the Client 1788 by searching the Client database. When the Server returns the MNPs, 1789 it also creates a forwarding table entry for the ULA corredponding to 1790 each MNP so that the MNPs are propagated into the routing system 1791 (see: Section 3.2.3). For IPv6, the Server creates an IPv6 1792 forwarding table entry for each MNP. For IPv4, the Server creates an 1793 IPv6 forwarding table entry with the IPv4-compatibility ULA prefix 1794 corresponding to the IPv4 address. 1796 The Server next creates a symmetric neighbor cache entry for the 1797 Client using the base LLA as the network-layer address and with 1798 lifetime set to no more than the smallest prefix lifetime. Next, the 1799 Server updates the neighbor cache entry by recording the information 1800 in each Interface Attributes sub-option in the RS OMNI option. The 1801 Server also records the actual OAL/INET addresses in the neighbor 1802 cache entry. 1804 Next, the Server prepares an RA message using its LLA as the network- 1805 layer source address and the network-layer source address of the RS 1806 message as the network-layer destination address. The Server sets 1807 the Router Lifetime to the time for which it will maintain both this 1808 underlying interface individually and the symmetric neighbor cache 1809 entry as a whole. The Server also sets Cur Hop Limit, M and O flags, 1810 Reachable Time and Retrans Timer to values appropriate for the OMNI 1811 link. The Server includes the MNPs, any other prefix management 1812 parameters and an OMNI option with no Interface Attributes. The 1813 Server then includes one or more RIOs that encode the MSPs for the 1814 OMNI link, plus an MTU option (see Section 3.9). The Server finally 1815 forwards the message to the Client using OAL/INET, INET, or NULL 1816 encapsulation as necessary. 1818 After the initial RS/RA exchange, the Server maintains a 1819 ReachableTime timer for each of the Client's underlying interfaces 1820 individually (and for the Client's symmetric neighbor cache entry 1821 collectively) set to expire after ReachableTime seconds. If the 1822 Client (or Proxy) issues additional RS messages, the Server sends an 1823 RA response and resets ReachableTime. If the Server receives an ND 1824 message with a prefix release indication it sets the Client's 1825 symmetric neighbor cache entry to the DEPARTED state and withdraws 1826 the MNP from the routing system after a short delay (e.g., 2 1827 seconds). If ReachableTime expires before a new RS is received on an 1828 individual underlying interface, the Server marks the interface as 1829 DOWN. If ReachableTime expires before any new RS is received on any 1830 individual underlying interface, the Server sets the symmetric 1831 neighbor cache entry state to STALE and sets a 10 second timer. If 1832 the Server has not received a new RS or ND message with a prefix 1833 release indication before the 10 second timer expires, it deletes the 1834 neighbor cache entry and withdraws the MNP from the routing system. 1836 The Server processes any ND messages pertaining to the Client and 1837 returns an NA/RA reply in response to solicitations. The Server may 1838 also issue unsolicited RA messages, e.g., with reconfigure parameters 1839 to cause the Client to renegotiate its prefix delegation/ 1840 registrations, with Router Lifetime set to 0 if it can no longer 1841 service this Client, etc. Finally, If the symmetric neighbor cache 1842 entry is in the DEPARTED state, the Server deletes the entry after 1843 DepartTime expires. 1845 Note: Clients SHOULD notify former Servers of their departures, but 1846 Servers are responsible for expiring neighbor cache entries and 1847 withdrawing routes even if no departure notification is received 1848 (e.g., if the Client leaves the network unexpectedly). Servers 1849 SHOULD therefore set Router Lifetime to ReachableTime seconds in 1850 solicited RA messages to minimize persistent stale cache information 1851 in the absence of Client departure notifications. A short Router 1852 Lifetime also ensures that proactive Client/Server RS/RA messaging 1853 will keep any NAT state alive (see above). 1855 Note: All Servers on an OMNI link MUST advertise consistent values in 1856 the RA Cur Hop Limit, M and O flags, Reachable Time and Retrans Timer 1857 fields the same as for any link, since unpredictable behavior could 1858 result if different Servers on the same link advertised different 1859 values. 1861 3.12.3.1. Lightweight DHCPv6 Relay Agent (LDRA) 1863 When the OMNI option includes a DHCPv6 Unique Identifier (DUID) sub- 1864 option, the use of DHCPv6 Prefix Delegation is indicated. AERO 1865 Clients and Servers are always on the same link (i.e., the OMNI link) 1866 from the perspective of DHCPv6, however in some implementations the 1867 DHCPv6 server and ND function may be located in separate modules. In 1868 that case, the Server's OMNI interface module can act as a 1869 Lightweight DHCPv6 Relay Agent (LDRA)[RFC6221] to relay prefix 1870 delegation messages to and from the DHCPv6 server module. 1872 When the LDRA receives an authentic RS message, it constructs a 1873 corresponding IPv6/UDP/DHCPv6 message. It sets the IPv6 source 1874 address to the source address of the RS message, sets the IPv6 1875 destination address to 'All_DHCP_Relay_Agents_and_Servers' and sets 1876 the UDP fields to values that will be understood by the DHCPv6 1877 server. 1879 The LDRA then wraps the message in a DHCPv6 'Relay-Forward' message 1880 header and includes an 'Interface-Id' option that includes enough 1881 information to allow the LDRA to forward the resulting Reply message 1882 back to the Client (e.g., the Client's link-layer addresses, a 1883 security association identifier, etc.). The LDRA also wraps the OMNI 1884 option into the Interface-Id option, then forwards the message to the 1885 DHCPv6 server. 1887 When the DHCPv6 server prepares a Reply message, it wraps the message 1888 in a 'Relay-Reply' message and echoes the Interface-Id option. The 1889 DHCPv6 server then delivers the Relay-Reply message to the LDRA, 1890 which discards the Relay-Reply wrapper and IPv6/UDP headers, then 1891 uses the DHCPv6 message to construct an RA response to the Client. 1892 The Server uses the information in the Interface-Id option to prepare 1893 the RA message and to cache the link-layer addresses taken from the 1894 OMNI option echoed in the Interface-Id option. 1896 3.13. The AERO Proxy 1898 Clients may connect to protected-spectrum ANETs that employ physical 1899 and/or link-layer security services to facilitate communications to 1900 Servers in outside INETs. In that case, the ANET can employ an AERO 1901 Proxy. The Proxy is located at the ANET/INET border and listens for 1902 RS messages originating from or RA messages destined to ANET Clients. 1903 The Proxy acts on these control messages as follows: 1905 o when the Proxy receives an RS message from a new ANET Client, it 1906 first authenticates the message then examines the network-layer 1907 destination address. If the destination address is a Server's 1908 LLA, the Proxy proceeds to the next step. Otherwise, if the 1909 destination is (link-local) All-Routers multicast, the Proxy 1910 selects a "nearby" Server that is likely to be a good candidate to 1911 serve the Client and replaces the destination address with the 1912 Server's LLA. Next, the Proxy creates a proxy neighbor cache 1913 entry and caches the Client and Server link-layer addresses along 1914 with the OMNI option information and any other identifying 1915 information including Transaction IDs, Client Identifiers, Nonce 1916 values, etc. The Proxy finally encapsulates the (proxyed) RS 1917 message in an OAL header with source set to the Proxy's ULA and 1918 destination set to the Server's ULA then forwards the message into 1919 the OMNI link. 1921 o when the Server receives the RS, it authenticates the message then 1922 creates or updates a symmetric neighbor cache entry for the Client 1923 with the Proxy's ULA as the link-layer address. The Server then 1924 sends an RA message back to the Proxy via the spanning tree. 1926 o when the Proxy receives the RA, it authenticates the message and 1927 matches it with the proxy neighbor cache entry created by the RS. 1928 The Proxy then caches the prefix information as a mapping from the 1929 Client's MNPs to the Client's link-layer address, caches the 1930 Server's advertised Router Lifetime and sets the neighbor cache 1931 entry state to REACHABLE. The Proxy then optionally rewrites the 1932 Router Lifetime and forwards the (proxyed) message to the Client. 1933 The Proxy finally includes an MTU option (if necessary) with an 1934 MTU to use for the underlying ANET interface. 1936 After the initial RS/RA exchange, the Proxy forwards any Client data 1937 packets for which there is no matching asymmetric neighbor cache 1938 entry to a Bridge using OAL encapsulation with its own ULA as the 1939 source and the ULA corresponding to the Client as the destination. 1940 The Proxy instead forwards any Client data destined to an asymmetric 1941 neighbor cache target directly to the target according to the OAL/ 1942 link-layer information - the process of establishing asymmetric 1943 neighbor cache entries is specified in Section 3.14. 1945 While the Client is still attached to the ANET, the Proxy sends NS, 1946 RS and/or unsolicited NA messages to update the Server's symmetric 1947 neighbor cache entries on behalf of the Client and/or to convey QoS 1948 updates. This allows for higher-frequency Proxy-initiated RS/RA 1949 messaging over well-connected INET infrastructure supplemented by 1950 lower-frequency Client-initiated RS/RA messaging over constrained 1951 ANET data links. 1953 If the Server ceases to send solicited advertisements, the Proxy 1954 sends unsolicited RAs on the ANET interface with destination set to 1955 (link-local) All-Nodes multicast and with Router Lifetime set to zero 1956 to inform Clients that the Server has failed. Although the Proxy 1957 engages in ND exchanges on behalf of the Client, the Client can also 1958 send ND messages on its own behalf, e.g., if it is in a better 1959 position than the Proxy to convey QoS changes, etc. For this reason, 1960 the Proxy marks any Client-originated solicitation messages (e.g. by 1961 inserting a Nonce option) so that it can return the solicited 1962 advertisement to the Client instead of processing it locally. 1964 If the Client becomes unreachable, the Proxy sets the neighbor cache 1965 entry state to DEPARTED and retains the entry for DepartTime seconds. 1966 While the state is DEPARTED, the Proxy forwards any packets destined 1967 to the Client to a Bridge via OAL encapsulation with the Client's 1968 current Server as the destination. The Bridge in turn forwards the 1969 packets to the Client's current Server. When DepartTime expires, the 1970 Proxy deletes the neighbor cache entry and discards any further 1971 packets destined to this (now forgotten) Client. 1973 In some ANETs that employ a Proxy, the Client's MNP can be injected 1974 into the ANET routing system. In that case, the Client can send data 1975 messages without encapsulation so that the ANET routing system 1976 transports the unencapsulated packets to the Proxy. This can be very 1977 beneficial, e.g., if the Client connects to the ANET via low-end data 1978 links such as some aviation wireless links. 1980 If the first-hop ANET access router is on the same underlying link 1981 and recognizes the AERO/OMNI protocol, the Client can avoid 1982 encapsulation for both its control and data messages. When the 1983 Client connects to the link, it can send an unencapsulated RS message 1984 with source address set to its LLA and with destination address set 1985 to the LLA of the Client's selected Server or to (link-local) All- 1986 Routers multicast. The Client includes an OMNI option formatted as 1987 specified in [I-D.templin-6man-omni-interface]. 1989 The Client then sends the unencapsulated RS message, which will be 1990 intercepted by the AERO-Aware access router. The access router then 1991 encapsulates the RS message in an ANET header with its own address as 1992 the source address and the address of a Proxy as the destination 1993 address. The access router further remembers the address of the 1994 Proxy so that it can encapsulate future data packets from the Client 1995 via the same Proxy. If the access router needs to change to a new 1996 Proxy, it simply sends another RS message toward the Server via the 1997 new Proxy on behalf of the Client. 1999 In some cases, the access router and Proxy may be one and the same 2000 node. In that case, the node would be located on the same physical 2001 link as the Client, but its message exchanges with the Server would 2002 need to pass through a security gateway at the ANET/INET border. The 2003 method for deploying access routers and Proxys (i.e. as a single node 2004 or multiple nodes) is an ANET-local administrative consideration. 2006 3.13.1. Combined Proxy/Servers 2008 Clients may need to connect directly to Servers via INET, Direct and 2009 VPNed interfaces (i.e., non-ANET interfaces). If the Client's 2010 underlying interfaces all connect via the same INET partition, then 2011 it can connect to a single controlling Server via all interfaces. 2013 If some Client interfaces connect via different INET partitions, 2014 however, the Client still selects a set of controlling Servers and 2015 sends RS messages via their directly-connected Servers while using 2016 the LLA of the controlling Server as the destination. 2018 When a Server receives an RS with destination set to the LLA of a 2019 controlling Server, it acts as a Proxy to forward the message to the 2020 controlling Server while forwarding the corresponding RA reply to the 2021 Client. 2023 3.13.2. Detecting and Responding to Server Failures 2025 In environments where fast recovery from Server failure is required, 2026 Proxys SHOULD use proactive Neighbor Unreachability Detection (NUD) 2027 to track Server reachability in a similar fashion as for 2028 Bidirectional Forwarding Detection (BFD) [RFC5880]. Proxys can then 2029 quickly detect and react to failures so that cached information is 2030 re-established through alternate paths. The NUD control messaging is 2031 carried only over well-connected ground domain networks (i.e., and 2032 not low-end aeronautical radio links) and can therefore be tuned for 2033 rapid response. 2035 Proxys perform proactive NUD with Servers for which there are 2036 currently active ANET Clients by sending continuous NS messages in 2037 rapid succession, e.g., one message per second. The Proxy sends the 2038 NS message via the spanning tree with the Proxy's LLA as the source 2039 and the LLA of the Server as the destination. When the Proxy is also 2040 sending RS messages to the Server on behalf of ANET Clients, the 2041 resulting RA responses can be considered as equivalent hints of 2042 forward progress. This means that the Proxy need not also send a 2043 periodic NS if it has already sent an RS within the same period. If 2044 the Server fails (i.e., if the Proxy ceases to receive 2045 advertisements), the Proxy can quickly inform Clients by sending 2046 multicast RA messages on the ANET interface. 2048 The Proxy sends RA messages on the ANET interface with source address 2049 set to the Server's address, destination address set to (link-local) 2050 All-Nodes multicast, and Router Lifetime set to 0. The Proxy SHOULD 2051 send MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small 2052 delays [RFC4861]. Any Clients on the ANET that had been using the 2053 failed Server will receive the RA messages and associate with a new 2054 Server. 2056 3.13.3. Point-to-Multipoint Server Coordination 2058 In environments where Client messaging over ANETs is bandwidth- 2059 limited and/or expensive, Clients can enlist the services of the 2060 Proxy to coordinate with multiple Servers in a single RS/RA message 2061 exchange. The Client can send a single RS message to (link-local) 2062 All-Routers multicast that includes the ID's of multiple Servers in 2063 MS-Register sub-options of the OMNI option. 2065 When the Proxy receives the RS and processes the OMNI option, it 2066 sends a separate RS to each MS-Register Server ID. When the Proxy 2067 receives an RA, it can optionally return an immediate "singleton" RA 2068 to the Client or record the Server's ID for inclusion in a pending 2069 "aggregate" RA message. The Proxy can then return aggregate RA 2070 messages to the Client including multiple Server IDs in order to 2071 conserve bandwidth. Each RA includes a proper subset of the Server 2072 IDs from the original RS message, and the Proxy must ensure that the 2073 message contents of each RA are consistent with the information 2074 received from the (aggregated) Servers. 2076 Clients can thereafter employ efficient point-to-multipoint Server 2077 coordination under the assistance of the Proxy to reduce the number 2078 of messages sent over the ANET while enlisting the support of 2079 multiple Servers for fault tolerance. Clients can further include 2080 MS-Release suboptions in IPv6 ND messages to request the Proxy to 2081 release from former Servers via the procedures discussed in 2082 Section 3.16.5. 2084 The OMNI interface specification [I-D.templin-6man-omni-interface] 2085 provides further discussion of the Client/Proxy RS/RA messaging 2086 involved in point-to-multipoint coordination. 2088 3.14. AERO Route Optimization / Address Resolution 2090 While data packets are flowing between a source and target node, 2091 route optimization SHOULD be used. Route optimization is initiated 2092 by the first eligible Route Optimization Source (ROS) closest to the 2093 source as follows: 2095 o For Clients on VPNed and Direct interfaces, the Server is the ROS. 2097 o For Clients on ANET interfaces, the Proxy is the ROS. 2099 o For Clients on INET interfaces, the Client itself is the ROS. 2101 o For correspondent nodes on INET/EUN interfaces serviced by a 2102 Relay, the Relay is the ROS. 2104 The route optimization procedure is conducted between the ROS and the 2105 target Server/Relay acting as a Route Optimization Responder (ROR) in 2106 the same manner as for IPv6 ND Address Resolution and using the same 2107 NS/NA messaging. The target may either be a MNP Client serviced by a 2108 Server, or a non-MNP correspondent reachable via a Relay. 2110 The procedures are specified in the following sections. 2112 3.14.1. Route Optimization Initiation 2114 While data packets are flowing from the source node toward a target 2115 node, the ROS performs address resolution by sending an NS message 2116 for Address Resolution (NS(AR)) to receive a solicited NA message 2117 from the ROR. When the ROS sends an NS(AR), it includes: 2119 o the LLA of the ROS as the source address. 2121 o the data packet's destination as the Target Address. 2123 o the Solicited-Node multicast address [RFC4291] formed from the 2124 lower 24 bits of the data packet's destination as the destination 2125 address, e.g., for 2001:db8:1:2::10:2000 the NS destination 2126 address is ff02:0:0:0:0:1:ff10:2000. 2128 The NS(AR) message includes an OMNI option with no Interface 2129 Attributes, such that the target will not create a neighbor cache 2130 entry. The Prefix Length in the OMNI option is set to the Prefix 2131 Length associated with the ROS's LLA. 2133 The ROS then encapsulates the NS(AR) message in an OAL header with 2134 source set to its own ULA and destination set to the ULA 2135 corresponding to the packet's final destination, then sends the 2136 message into the spanning tree without decrementing the network-layer 2137 TTL/Hop Limit field. 2139 3.14.2. Relaying the NS 2141 When the Bridge receives the NS(AR) message from the ROS, it discards 2142 the INET header and determines that the ROR is the next hop by 2143 consulting its standard IPv6 forwarding table for the OAL header 2144 destination address. The Bridge then forwards the message toward the 2145 ROR via the spanning tree the same as for any IPv6 router. The 2146 final-hop Bridge in the spanning tree will deliver the message via a 2147 secured tunnel to the ROR. 2149 3.14.3. Processing the NS and Sending the NA 2151 When the ROR receives the NS(AR) message, it examines the Target 2152 Address to determine whether it has a neighbor cache entry and/or 2153 route that matches the target. If there is no match, the ROR drops 2154 the message. Otherwise, the ROR continues processing as follows: 2156 o if the target belongs to an MNP Client neighbor in the DEPARTED 2157 state the ROR changes the NS(AR) message OAL destination address 2158 to the ULA of the Client's new Server, forwards the message into 2159 the spanning tree and returns from processing. 2161 o If the target belongs to an MNP Client neighbor in the REACHABLE 2162 state, the ROR instead adds the AERO source address to the target 2163 Client's Report List with time set to ReportTime. 2165 o If the target belongs to a non-MNP route, the ROR continues 2166 processing without adding an entry to the Report List. 2168 The ROR then prepares a solicited NA message to send back to the ROS 2169 but does not create a neighbor cache entry. The ROR sets the NA 2170 source address to the LLA corresponding to the target, sets the 2171 Target Address to the target of the solicitation, and sets the 2172 destination address to the source of the solicitation. The ROR then 2173 includes an OMNI option with Prefix Length set to the length 2174 associated with the LLA. 2176 If the target is an MNP Client, the ROR next includes Interface 2177 Attributes in the OMNI option for each of the target Client's 2178 underlying interfaces with current information for each interface and 2179 with the ifIndex field in the OMNI header set to 0 to indicate that 2180 the message originated from the ROR and not the Client. 2182 For each Interface Attributes sub-option, the ROR sets the L2ADDR 2183 according to its own INET address for VPNed or Direct interfaces, to 2184 the INET address of the Proxy or to the Client's INET address for 2185 INET interfaces. The ROR then includes the lower 32 bits of its own 2186 ULA (or the ULA of the Proxy) as the LHS, encodes the ULA prefix 2187 length code in the SRT field and sets the FMT code accordingly as 2188 specified in Section 3.3. 2190 The ROR then sets the NA message R flag to 1 (as a router), S flag to 2191 1 (as a response to a solicitation), and O flag to 0 (as a proxy). 2192 The ROR finally encapsulates the NA message in an OAL header with 2193 source set to its own ULA and destination set to the source ULA of 2194 the NS(AR) message, then forwards the message into the spanning tree 2195 without decrementing the network-layer TTL/Hop Limit field. 2197 3.14.4. Relaying the NA 2199 When the Bridge receives the NA message from the ROR, it discards the 2200 INET header and determines that the ROS is the next hop by consulting 2201 its standard IPv6 forwarding table for the OAL header destination 2202 address. The Bridge then forwards the OAL-encapsulated NA message 2203 toward the ROS the same as for any IPv6 router. The final-hop Bridge 2204 in the spanning tree will deliver the message via a secured tunnel to 2205 the ROS. 2207 3.14.5. Processing the NA 2209 When the ROS receives the solicited NA message, it processes the 2210 message the same as for standard IPv6 Address Resolution [RFC4861]. 2211 In the process, it caches the source ULA then creates an asymmetric 2212 neighbor cache entry for the target and caches all information found 2213 in the OMNI option. The ROS finally sets the asymmetric neighbor 2214 cache entry lifetime to ReachableTime seconds. 2216 3.14.6. Route Optimization Maintenance 2218 Following route optimization, the ROS forwards future data packets 2219 destined to the target via the addresses found in the cached link- 2220 layer information. The route optimization is shared by all sources 2221 that send packets to the target via the ROS, i.e., and not just the 2222 source on behalf of which the route optimization was initiated. 2224 While new data packets destined to the target are flowing through the 2225 ROS, it sends additional NS(AR) messages to the ROR before 2226 ReachableTime expires to receive a fresh solicited NA message the 2227 same as described in the previous sections (route optimization 2228 refreshment strategies are an implementation matter, with a non- 2229 normative example given in Appendix A.1). The ROS uses the cached 2230 ULA of the ROR as the NS(AR) OAL destination address, and sends up to 2231 MAX_MULTICAST_SOLICIT NS(AR) messages separated by 1 second until an 2232 NA is received. If no NA is received, the ROS assumes that the 2233 current ROR has become unreachable and deletes the target neighbor 2234 cache entry. Subsequent data packets will trigger a new route 2235 optimization per Section 3.14.1 to discover a new ROR while initial 2236 data packets travel over a suboptimal route. 2238 If an NA is received, the ROS then updates the asymmetric neighbor 2239 cache entry to refresh ReachableTime, while (for MNP destinations) 2240 the ROR adds or updates the ROS address to the target's Report List 2241 and with time set to ReportTime. While no data packets are flowing, 2242 the ROS instead allows ReachableTime for the asymmetric neighbor 2243 cache entry to expire. When ReachableTime expires, the ROS deletes 2244 the asymmetric neighbor cache entry. Any future data packets flowing 2245 through the ROS will again trigger a new route optimization. 2247 The ROS may also receive unsolicited NA messages from the ROR at any 2248 time (see: Section 3.16). If there is an asymmetric neighbor cache 2249 entry for the target, the ROS updates the link-layer information but 2250 does not update ReachableTime since the receipt of an unsolicited NA 2251 does not confirm that any forward paths are working. If there is no 2252 asymmetric neighbor cache entry, the ROS simply discards the 2253 unsolicited NA. 2255 In this arrangement, the ROS holds an asymmetric neighbor cache entry 2256 for the target via the ROR, but the ROR does not hold an asymmetric 2257 neighbor cache entry for the ROS. The route optimization neighbor 2258 relationship is therefore asymmetric and unidirectional. If the 2259 target node also has packets to send back to the source node, then a 2260 separate route optimization procedure is performed in the reverse 2261 direction. But, there is no requirement that the forward and reverse 2262 paths be symmetric. 2264 3.15. Neighbor Unreachability Detection (NUD) 2266 AERO nodes perform Neighbor Unreachability Detection (NUD) per 2267 [RFC4861] either reactively in response to persistent link-layer 2268 errors (see Section 3.11) or proactively to confirm reachability. 2269 The NUD algorithm is based on periodic control message exchanges. 2270 The algorithm may further be seeded by ND hints of forward progress, 2271 but care must be taken to avoid inferring reachability based on 2272 spoofed information. For example, authentic IPv6 ND message 2273 exchanges may be considered as acceptable hints of forward progress, 2274 while spurious data packets should not be. 2276 AERO Servers, Proxys and Relays can use (OAL-encapsulated) standard 2277 NS/NA NUD exchanges sent over the OMNI link spanning tree to securely 2278 test reachability without risk of DoS attacks from nodes pretending 2279 to be a neighbor; Proxys can further perform NUD to securely verify 2280 Server reachability on behalf of their proxyed Clients. However, a 2281 means for an ROS to test the unsecured forward directions of target 2282 route optimized paths is also necessary. 2284 When an ROR directs an ROS to a neighbor with one or more target 2285 link-layer addresses, the ROS can proactively test each such 2286 unsecured route optimized path by sending "loopback" NS(NUD) 2287 messages. While testing the paths, the ROS can optionally continue 2288 to send packets via the spanning tree, maintain a small queue of 2289 packets until target reachability is confirmed, or (optimistically) 2290 allow packets to flow via the route optimized paths. 2292 When the ROS sends a loopback NS(NUD) message, it uses its LLA as 2293 both the IPv6 source and destination address, and the MNP Subnet- 2294 Router anycast address as the Target Address. The ROS includes a 2295 Nonce and Timestamp option, then encapsulates the message in OAL/INET 2296 headers with its own ULA as the source and the ULA of the route 2297 optimization target as the destination. The ROS then forwards the 2298 message to the target (either directly to the L2ADDR of the target if 2299 the target is in the same OMNI link segment, or via a Bridge if the 2300 target is in a different OMNI link segment). 2302 When the route optimization target receives the NS(NUD) message, it 2303 notices that the IPv6 destination address is the same as the source 2304 address. It then reverses the OAL header source and destination 2305 addresses and returns the message to the ROS (either directly or via 2306 the spanning tree). The route optimization target does not decrement 2307 the NS(NUD) message IPv6 Hop-Limit in the process, since the message 2308 has not exited the OMNI link. 2310 When the ROS receives the NS(NUD) message, it can determine from the 2311 Nonce, Timestamp and Target Address that the message originated from 2312 itself and that it transited the forward path. The ROS need not 2313 prepare an NA response, since the destination of the response would 2314 be itself and testing the route optimization path again would be 2315 redundant. 2317 The ROS marks route optimization target paths that pass these NUD 2318 tests as "reachable", and those that do not as "unreachable". These 2319 markings inform the OMNI interface forwarding algorithm specified in 2320 Section 3.10. 2322 Note that to avoid a DoS vector nodes MUST NOT return loopback 2323 NS(NUD) messages received from an unsecured link-layer source via the 2324 spanning tree. 2326 3.16. Mobility Management and Quality of Service (QoS) 2328 AERO is a Distributed Mobility Management (DMM) service. Each Server 2329 is responsible for only a subset of the Clients on the OMNI link, as 2330 opposed to a Centralized Mobility Management (CMM) service where 2331 there is a single network mobility collective entity for all Clients. 2332 Clients coordinate with their associated Servers via RS/RA exchanges 2333 to maintain the DMM profile, and the AERO routing system tracks all 2334 current Client/Server peering relationships. 2336 Servers provide default routing and mobility/multilink services for 2337 their dependent Clients. Clients are responsible for maintaining 2338 neighbor relationships with their Servers through periodic RS/RA 2339 exchanges, which also serves to confirm neighbor reachability. When 2340 a Client's underlying interface address and/or QoS information 2341 changes, the Client is responsible for updating the Server with this 2342 new information. Note that when there is a Proxy in the path, the 2343 Proxy can also perform some RS/RA exchanges on the Client's behalf. 2345 Mobility management considerations are specified in the following 2346 sections. 2348 3.16.1. Mobility Update Messaging 2350 Servers accommodate Client mobility/multilink and/or QoS change 2351 events by sending unsolicited NA (uNA) messages to each ROS in the 2352 target Client's Report List. When a Server sends a uNA message, it 2353 sets the IPv6 source address to the Client's LLA, sets the 2354 destination address to (link-local) All-Nodes multicast and sets the 2355 Target Address to the Client's Subnet-Router anycast address. The 2356 Server also includes an OMNI option with Prefix Length set to the 2357 length associated with the Client's LLA, with Interface Attributes 2358 for the target Client's underlying interfaces and with ifIndex in the 2359 OMNI header set to 0. The Server sets the NA R flag to 1, the S flag 2360 to 0 and the O flag to 0, then encapsulates the message in an OAL 2361 header with source set to its own ULA and destination set to the ULA 2362 of the ROS and sends the message into the spanning tree. 2364 As discussed in Section 7.2.6 of [RFC4861], the transmission and 2365 reception of uNA messages is unreliable but provides a useful 2366 optimization. In well-connected Internetworks with robust data links 2367 uNA messages will be delivered with high probability, but in any case 2368 the Server can optionally send up to MAX_NEIGHBOR_ADVERTISEMENT uNAs 2369 to each ROS to increase the likelihood that at least one will be 2370 received. 2372 When the ROS receives a uNA message, it ignores the message if there 2373 is no existing neighbor cache entry for the Client. Otherwise, it 2374 uses the included OMNI option information to update the neighbor 2375 cache entry, but does not reset ReachableTime since the receipt of an 2376 unsolicited NA message from the target Server does not provide 2377 confirmation that any forward paths to the target Client are working. 2379 If uNA messages are lost, the ROS may be left with stale address and/ 2380 or QoS information for the Client for up to ReachableTime seconds. 2381 During this time, the ROS can continue sending packets according to 2382 its stale neighbor cache information. When ReachableTime is close to 2383 expiring, the ROS will re-initiate route optimization and receive 2384 fresh link-layer address information. 2386 In addition to sending uNA messages to the current set of ROSs for 2387 the Client, the Server also sends uNAs to the former link-layer 2388 address for any underlying interface for which the link-layer address 2389 has changed. The uNA messages update Proxys that cannot easily 2390 detect (e.g., without active probing) when a formerly-active Client 2391 has departed. 2393 3.16.2. Announcing Link-Layer Address and/or QoS Preference Changes 2395 When a Client needs to change its underlying interface addresses and/ 2396 or QoS preferences (e.g., due to a mobility event), either the Client 2397 or its Proxys send RS messages to the Server via the spanning tree 2398 with an OMNI option that includes Interface attributes with the new 2399 link quality and address information. 2401 Up to MAX_RTR_SOLICITATIONS RS messages MAY be sent in parallel with 2402 sending actual data packets in case one or more RAs are lost. If all 2403 RAs are lost, the Client SHOULD re-associate with a new Server. 2405 When the Server receives the Client's changes, it sends uNA messages 2406 to all nodes in the Report List the same as described in the previous 2407 section. 2409 3.16.3. Bringing New Links Into Service 2411 When a Client needs to bring new underlying interfaces into service 2412 (e.g., when it activates a new data link), it sends an RS message to 2413 the Server via the underlying interface with an OMNI option that 2414 includes Interface Attributes with appropriate link quality values 2415 and with link-layer address information for the new link. 2417 3.16.4. Removing Existing Links from Service 2419 When a Client needs to remove existing underlying interfaces from 2420 service (e.g., when it de-activates an existing data link), it sends 2421 an RS or uNA message to its Server with an OMNI option with 2422 appropriate link quality values. 2424 If the Client needs to send RS/uNA messages over an underlying 2425 interface other than the one being removed from service, it MUST 2426 include Interface Attributes with appropriate link quality values for 2427 any underlying interfaces being removed from service. 2429 3.16.5. Moving to a New Server 2431 When a Client associates with a new Server, it performs the Client 2432 procedures specified in Section 3.12.2. The Client also includes MS- 2433 Release identifiers in the RS message OMNI option per 2434 [I-D.templin-6man-omni-interface] if it wants the new Server to 2435 notify any old Servers from which the Client is departing. 2437 When the new Server receives the Client's RS message, it returns an 2438 RA as specified in Section 3.12.3 and sends up to 2439 MAX_NEIGHBOR_ADVERTIISEMENT uNA messages to any old Servers listed in 2440 OMNI option MS-Release identifiers. Each uNA message includes the 2441 Client's LLA as the source address, the old Server's LLA as the 2442 destination address, and an OMNI option with the old Server's ID in 2443 the MS-Release list. The new Server wraps the uNA in an OAL header 2444 with its own ULA as the source and the old Server's ULA as the 2445 destination, then sends the message into the spanning tree. 2447 When an old Server receives the uNA, it changes the Client's neighbor 2448 cache entry state to DEPARTED, sets the link-layer address of the 2449 Client to the new Server's ULA, and resets DepartTime. After a short 2450 delay (e.g., 2 seconds) the old Server withdraws the Client's MNP 2451 from the routing system. After DepartTime expires, the old Server 2452 deletes the Client's neighbor cache entry. 2454 The old Server also sends unsolicited NA messages to all ROSs in the 2455 Client's Report List with an OMNI option with a single Interface 2456 Attributes sub-option with ifIndex set to 0, and with the Link Layer 2457 address information of the new Server. When the ROS receives the NA, 2458 it caches the address of the new Server in the existing asymmetric 2459 neighbor cache entry and marks the entry as STALE for a period of 10 2460 seconds after which the cache entry is deleted. While in the STALE 2461 state, subsequent data packets flow according to any existing cached 2462 link-layer information and trigger a new NS(AR)/NA exchange via the 2463 new Server. 2465 Clients SHOULD NOT move rapidly between Servers in order to avoid 2466 causing excessive oscillations in the AERO routing system. Examples 2467 of when a Client might wish to change to a different Server include a 2468 Server that has gone unreachable, topological movements of 2469 significant distance, movement to a new geographic region, movement 2470 to a new OMNI link segment, etc. 2472 When a Client moves to a new Server, some of the fragments of a 2473 multiple fragment packet may have already arrived at the old Server 2474 while others are en route to the new Server, however no special 2475 attention in the reassembly algorithm is necessary when re-routed 2476 fragments are simply treated as loss. 2478 3.17. Multicast 2480 The AERO Client provides an IGMP (IPv4) [RFC2236] or MLD (IPv6) 2481 [RFC3810] proxy service for its EUNs and/or hosted applications 2482 [RFC4605]. The Client forwards IGMP/MLD messages over any of its 2483 underlying interfaces for which group membership is required. The 2484 IGMP/MLD messages may be further forwarded by a first-hop ANET access 2485 router acting as an IGMP/MLD-snooping switch [RFC4541], then 2486 ultimately delivered to an AERO Proxy/Server acting as a Protocol 2487 Independent Multicast - Sparse-Mode (PIM-SM, or simply "PIM") 2488 Designated Router (DR) [RFC7761]. AERO Relays also act as PIM 2489 routers (i.e., the same as AERO Proxys/Servers) on behalf of nodes on 2490 INET/EUN networks. The behaviors identified in the following 2491 sections correspond to Source-Specific Multicast (SSM) and Any-Source 2492 Multicast (ASM) operational modes. 2494 3.17.1. Source-Specific Multicast (SSM) 2496 When an ROS (i.e., an AERO Proxy/Server/Relay) "X" acting as PIM 2497 router receives a Join/Prune message from a node on its downstream 2498 interfaces containing one or more ((S)ource, (G)roup) pairs, it 2499 updates its Multicast Routing Information Base (MRIB) accordingly. 2500 For each S belonging to a prefix reachable via X's non-OMNI 2501 interfaces, X then forwards the (S, G) Join/Prune to any PIM routers 2502 on those interfaces per [RFC7761]. 2504 For each S belonging to a prefix reachable via X's OMNI interface, X 2505 originates a separate copy of the Join/Prune for each (S,G) in the 2506 message using its own LLA as the source address and ALL-PIM-ROUTERS 2507 as the destination address. X then encapsulates each message in an 2508 OAL header with source address set to the ULA of X and destination 2509 address set to S then forwards the message into the spanning tree, 2510 which delivers it to AERO Server/Relay "Y" that services S. At the 2511 same time, if the message was a Join, X sends a route-optimization NS 2512 message toward each S the same as discussed in Section 3.14. The 2513 resulting NAs will return the LLA for the prefix that matches S as 2514 the network-layer source address and with an OMNI option with the ULA 2515 corresponding to any underlying interfaces that are currently 2516 servicing S. 2518 When Y processes the Join/Prune message, if S located behind any 2519 INET, Direct, or VPNed interfaces Y acts as a PIM router and updates 2520 its MRIB to list X as the next hop in the reverse path. If S is 2521 located behind any Proxys "Z"*, Y also forwards the message to each 2522 Z* over the spanning tree while continuing to use the LLA of X as the 2523 source address. Each Z* then updates its MRIB accordingly and 2524 maintains the LLA of X as the next hop in the reverse path. Since 2525 the Bridges do not examine network layer control messages, this means 2526 that the (reverse) multicast tree path is simply from each Z* (and/or 2527 Y) to X with no other multicast-aware routers in the path. If any Z* 2528 (and/or Y) is located on the same OMNI link segment as X, the 2529 multicast data traffic sent to X directly using OAL/INET 2530 encapsulation instead of via a Bridge. 2532 Following the initial Join/Prune and NS/NA messaging, X maintains an 2533 asymmetric neighbor cache entry for each S the same as if X was 2534 sending unicast data traffic to S. In particular, X performs 2535 additional NS/NA exchanges to keep the neighbor cache entry alive for 2536 up to t_periodic seconds [RFC7761]. If no new Joins are received 2537 within t_periodic seconds, X allows the neighbor cache entry to 2538 expire. Finally, if X receives any additional Join/Prune messages 2539 for (S,G) it forwards the messages to each Y and Z* in the neighbor 2540 cache entry over the spanning tree. 2542 At some later time, Client C that holds an MNP for source S may 2543 depart from a first Proxy Z1 and/or connect via a new Proxy Z2. In 2544 that case, Y sends an unsolicited NA message to X the same as 2545 specified for unicast mobility in Section 3.16. When X receives the 2546 unsolicited NA message, it updates its asymmetric neighbor cache 2547 entry for the LLA for source S and sends new Join messages to any new 2548 Proxys Z2. There is no requirement to send any Prune messages to old 2549 Proxys Z1 since source S will no longer source any multicast data 2550 traffic via Z1. Instead, the multicast state for (S,G) in Proxy Z1 2551 will soon time out since no new Joins will arrive. 2553 After some later time, C may move to a new Server Y2 and depart from 2554 old Sever Y1. In that case, Y1 sends Join messages for any of C's 2555 active (S,G) groups to Y2 while including its own LLA as the source 2556 address. This causes Y2 to include Y1 in the multicast forwarding 2557 tree during the interim time that Y1's symmetric neighbor cache entry 2558 for C is in the DEPARTED state. At the same time, Y1 sends an 2559 unsolicited NA message to X with an OMNI option with ifIndex in the 2560 header set to 0 and a release indication to cause X to release its 2561 asymmetric neighbor cache entry. X then sends a new Join message to 2562 S via the spanning tree and re-initiates route optimization the same 2563 as if it were receiving a fresh Join message from a node on a 2564 downstream link. 2566 3.17.2. Any-Source Multicast (ASM) 2568 When an ROS X acting as a PIM router receives a Join/Prune from a 2569 node on its downstream interfaces containing one or more (*,G) pairs, 2570 it updates its Multicast Routing Information Base (MRIB) accordingly. 2571 X then forwards a copy of the message to the Rendezvous Point (RP) R 2572 for each G over the spanning tree. X uses its own LLA as the source 2573 address and ALL-PIM-ROUTERS as the destination address, then 2574 encapsulates each message in an OAL header with source address set to 2575 the ULA of X and destination address set to R, then sends the message 2576 into the spanning tree. At the same time, if the message was a Join 2577 X initiates NS/NA route optimization the same as for the SSM case 2578 discussed in Section 3.17.1. 2580 For each source S that sends multicast traffic to group G via R, the 2581 Proxy/Server Z* for the Client that aggregates S encapsulates the 2582 packets in PIM Register messages and forwards them to R via the 2583 spanning tree, which may then elect to send a PIM Join to Z*. This 2584 will result in an (S,G) tree rooted at Z* with R as the next hop so 2585 that R will begin to receive two copies of the packet; one native 2586 copy from the (S, G) tree and a second copy from the pre-existing (*, 2587 G) tree that still uses PIM Register encapsulation. R can then issue 2588 a PIM Register-stop message to suppress the Register-encapsulated 2589 stream. At some later time, if C moves to a new Proxy/Server Z*, it 2590 resumes sending packets via PIM Register encapsulation via the new 2591 Z*. 2593 At the same time, as multicast listeners discover individual S's for 2594 a given G, they can initiate an (S,G) Join for each S under the same 2595 procedures discussed in Section 3.17.1. Once the (S,G) tree is 2596 established, the listeners can send (S, G) Prune messages to R so 2597 that multicast packets for group G sourced by S will only be 2598 delivered via the (S, G) tree and not from the (*, G) tree rooted at 2599 R. All mobility considerations discussed for SSM apply. 2601 3.17.3. Bi-Directional PIM (BIDIR-PIM) 2603 Bi-Directional PIM (BIDIR-PIM) [RFC5015] provides an alternate 2604 approach to ASM that treats the Rendezvous Point (RP) as a Designated 2605 Forwarder (DF). Further considerations for BIDIR-PIM are out of 2606 scope. 2608 3.18. Operation over Multiple OMNI Links 2610 An AERO Client can connect to multiple OMNI links the same as for any 2611 data link service. In that case, the Client maintains a distinct 2612 OMNI interface for each link, e.g., 'omni0' for the first link, 2613 'omni1' for the second, 'omni2' for the third, etc. Each OMNI link 2614 would include its own distinct set of Bridges, Servers and Proxys, 2615 thereby providing redundancy in case of failures. 2617 Each OMNI link could utilize the same or different ANET connections. 2618 The links can be distinguished at the link-layer via the SRT prefix 2619 in a similar fashion as for Virtual Local Area Network (VLAN) tagging 2620 (e.g., IEEE 802.1Q) and/or through assignment of distinct sets of 2621 MSPs on each link. This gives rise to the opportunity for supporting 2622 multiple redundant networked paths, with each VLAN distinguished by a 2623 different SRT "color" (see: Section 3.2.5). 2625 The Client's IP layer can select the outgoing OMNI interface 2626 appropriate for a given traffic profile while (in the reverse 2627 direction) correspondent nodes must have some way of steering their 2628 packets destined to a target via the correct OMNI link. 2630 In a first alternative, if each OMNI link services different MSPs, 2631 then the Client can receive a distinct MNP from each of the links. 2632 IP routing will therefore assure that the correct Red/Green/Blue/etc. 2633 network is used for both outbound and inbound traffic. This can be 2634 accomplished using existing technologies and approaches, and without 2635 requiring any special supporting code in correspondent nodes or 2636 Bridges. 2638 In a second alternative, if each OMNI link services the same MSP(s) 2639 then each link could assign a distinct "OMNI link Anycast" address 2640 that is configured by all Bridges on the link. Correspondent nodes 2641 can then perform Segment Routing to select the correct SRT, which 2642 will then direct the packet over multiple hops to the target. 2644 3.19. DNS Considerations 2646 AERO Client MNs and INET correspondent nodes consult the Domain Name 2647 System (DNS) the same as for any Internetworking node. When 2648 correspondent nodes and Client MNs use different IP protocol versions 2649 (e.g., IPv4 correspondents and IPv6 MNs), the INET DNS must maintain 2650 A records for IPv4 address mappings to MNs which must then be 2651 populated in Relay NAT64 mapping caches. In that way, an IPv4 2652 correspondent node can send packets to the IPv4 address mapping of 2653 the target MN, and the Relay will translate the IPv4 header and 2654 destination address into an IPv6 header and IPv6 destination address 2655 of the MN. 2657 When an AERO Client registers with an AERO Server, the Server can 2658 return the address(es) of DNS servers in RDNSS options [RFC6106]. 2659 The DNS server provides the IP addresses of other MNs and 2660 correspondent nodes in AAAA records for IPv6 or A records for IPv4. 2662 3.20. Transition Considerations 2664 OAL encapsulation ensures that dissimilar INET partitions can be 2665 joined into a single unified OMNI link, even though the partitions 2666 themselves may have differing protocol versions and/or incompatible 2667 addressing plans. However, a commonality can be achieved by 2668 incrementally distributing globally routable (i.e., native) IP 2669 prefixes to eventually reach all nodes (both mobile and fixed) in all 2670 OMNI link segments. This can be accomplished by incrementally 2671 deploying AERO Relays on each INET partition, with each Relay 2672 distributing its MNPs and/or discovering non-MNP prefixes on its INET 2673 links. 2675 This gives rise to the opportunity to eventually distribute native IP 2676 addresses to all nodes, and to present a unified OMNI link view even 2677 if the INET partitions remain in their current protocol and 2678 addressing plans. In that way, the OMNI link can serve the dual 2679 purpose of providing a mobility/multilink service and a transition 2680 service. Or, if an INET partition is transitioned to a native IP 2681 protocol version and addressing scheme that is compatible with the 2682 OMNI link MNP-based addressing scheme, the partition and OMNI link 2683 can be joined by Relays. 2685 Relays that connect INETs/EUNs with dissimilar IP protocol versions 2686 may need to employ a network address and protocol translation 2687 function such as NAT64[RFC6146]. 2689 3.21. Detecting and Reacting to Server and Bridge Failures 2691 In environments where rapid failure recovery is required, Servers and 2692 Bridges SHOULD use Bidirectional Forwarding Detection (BFD) 2693 [RFC5880]. Nodes that use BFD can quickly detect and react to 2694 failures so that cached information is re-established through 2695 alternate nodes. BFD control messaging is carried only over well- 2696 connected ground domain networks (i.e., and not low-end radio links) 2697 and can therefore be tuned for rapid response. 2699 Servers and Bridges maintain BFD sessions in parallel with their BGP 2700 peerings. If a Server or Bridge fails, BGP peers will quickly re- 2701 establish routes through alternate paths the same as for common BGP 2702 deployments. Similarly, Proxys maintain BFD sessions with their 2703 associated Bridges even though they do not establish BGP peerings 2704 with them. 2706 Proxys SHOULD use proactive NUD for Servers for which there are 2707 currently active ANET Clients in a manner that parallels BFD, i.e., 2708 by sending unicast NS messages in rapid succession to receive 2709 solicited NA messages. When the Proxy is also sending RS messages on 2710 behalf of ANET Clients, the RS/RA messaging can be considered as 2711 equivalent hints of forward progress. This means that the Proxy need 2712 not also send a periodic NS if it has already sent an RS within the 2713 same period. If a Server fails, the Proxy will cease to receive 2714 advertisements and can quickly inform Clients of the outage by 2715 sending multicast RA messages on the ANET interface. 2717 The Proxy sends multicast RA messages with source address set to the 2718 Server's address, destination address set to (link-local) All-Nodes 2719 multicast, and Router Lifetime set to 0. The Proxy SHOULD send 2720 MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small delays 2721 [RFC4861]. Any Clients on the ANET interface that have been using 2722 the (now defunct) Server will receive the RA messages and associate 2723 with a new Server. 2725 3.22. AERO Clients on the Open Internet 2727 AERO Clients that connect to the open Internet via INET interfaces 2728 can establish a VPN or direct link to securely connect to a Server in 2729 a "tethered" arrangement with all of the Client's traffic transiting 2730 the Server. Alternatively, the Client can associate with an INET 2731 Server using UDP/IP encapsulation and asymmetric securing services as 2732 discussed in the following sections. 2734 When a Client's OMNI interface enables an INET underlying interface, 2735 it first determines whether the interface is likely to be behind a 2736 NAT. For IPv4, the Client assumes it is on the open Internet if the 2737 INET address is not a special-use IPv4 address per [RFC3330]. 2738 Similarly for IPv6, the Client assumes it is on the open Internet if 2739 the INET address is not a link-local [RFC4291] or unique-local 2740 [RFC4193] IPv6 address. 2742 The Client then prepares a UDP/IP-encapsulated RS message with IPv6 2743 source address set to its LLA, with IPv6 destination set to (link- 2744 local) All-Routers multicast and with an OMNI option with underlying 2745 interface parameters. If the Client believes that it is on the open 2746 Internet, it SHOULD include Interface Attributes with the L2ADDR used 2747 for INET encapsulation (otherwise, it MAY omit L2ADDR). If the 2748 underlying address is IPv4, the Client includes the Port Number and 2749 IPv4 address written in obfuscated form [RFC4380] as discussed in 2750 Section 3.3. If the underlying interface address is IPv6, the Client 2751 instead includes the Port Number and IPv6 address in obfuscated form. 2752 The Client finally includes an Authentication option per [RFC4380] to 2753 provide message authentication, sets the UDP/IP source to its INET 2754 address and UDP port, sets the UDP/IP destination to the Server's 2755 INET address and the AERO service port number (8060), then sends the 2756 message to the Server. 2758 When the Server receives the RS, it authenticates the message and 2759 registers the Client's MNP and INET interface information according 2760 to the OMNI option parameters. If the RS message includes an L2ADDR 2761 in the OMNI option, the Server compares the encapsulation IP address 2762 and UDP port number with the (unobfuscated) values. If the values 2763 are the same, the Server caches the Client's information as "INET" 2764 addresses meaning that the Client is likely to accept direct messages 2765 without requiring NAT traversal exchanges. If the values are 2766 different (or, if the OMNI option did not include an L2ADDR) the 2767 Server instead caches the Client's information as "NAT" addresses 2768 meaning that NAT traversal exchanges may be necessary. 2770 The Server then returns an RA message with IPv6 source and 2771 destination set corresponding to the addresses in the RS, and with an 2772 Authentication option per [RFC4380]. For IPv4, the Server also 2773 includes an Origin option per [RFC4380] with the mapped and 2774 obfuscated Port Number and IPv4 address observed in the encapsulation 2775 headers. For IPv6, the Server instead includes an IPv6 Origin option 2776 per Figure 6 with the mapped and obfuscated observed Port Number and 2777 IPv6 address (note that the value 0x02 in the second octet 2778 differentiates from other [RFC4380] option types). 2780 +--------+--------+-----------------+ 2781 | 0x00 | 0x02 | Origin port # | 2782 +--------+--------+-----------------+ 2783 ~ Origin IPv6 address ~ 2784 +-----------------------------------+ 2786 Figure 6: IPv6 Origin Option 2788 When the Client receives the RA message, it compares the mapped Port 2789 Number and IP address from the Origin option with its own address. 2790 If the addresses are the same, the Client assumes the open Internet / 2791 Cone NAT principle; if the addresses are different, the Client 2792 instead assumes that further qualification procedures are necessary 2793 to detect the type of NAT and proceeds according to standard 2794 [RFC4380] procedures. 2796 After the Client has registered its INET interfaces in such RS/RA 2797 exchanges it sends periodic RS messages to receive fresh RA messages 2798 before the Router Lifetime received on each INET interface expires. 2799 The Client also maintains default routes via its Servers, i.e., the 2800 same as described in earlier sections. 2802 When the Client sends messages to target IP addresses, it also 2803 invokes route optimization per Section 3.14 using IPv6 ND address 2804 resolution messaging. The Client sends the NS(AR) message to the 2805 Server wrapped in a UDP/IP header with an Authentication option with 2806 the NS source address set to the Client's LLA and destination address 2807 set to the target solicited node multicast address. The Server 2808 authenticates the message and sends a corresponding NS(AR) message 2809 over the spanning tree the same as if it were the ROS, but with the 2810 OAL source address set to the Server's ULA and destination set to the 2811 ULA of the target. When the ROR receives the NS(AR), it adds the 2812 Server's ULA and Client's LLA to the target's Report List, and 2813 returns an NA with OMNI option information for the target. The 2814 Server then returns a UDP/IP encapsulated NA message with an 2815 Authentication option to the Client. 2817 Following route optimization for targets in the same OMNI link 2818 segment, if the target's L2ADDR is on the open INET, the Client 2819 forwards data packets directly to the target INET address. If the 2820 target is behind a NAT, the Client first establishes NAT state for 2821 the L2ADDR using the "bubble" mechanisms specified in 2822 [RFC6081][RFC4380]. The Client continues to send data packets via 2823 its Server until NAT state is populated, then begins forwarding 2824 packets via the direct path through the NAT to the target. For 2825 targets in different OMNI link segments, the Client inserts an SRH 2826 and forwards data packets to the Bridge that returned the NA message. 2828 The ROR may return uNAs via the Server if the target moves, and the 2829 Server will send corresponding Authentication-protected uNAs to the 2830 Client. The Client can also send "loopback" NS(NUD) messages to test 2831 forward path reachability even though there is no security 2832 association between the Client and the target. 2834 The Client sends UDP/IP encapsulated IPv6 packets no larger than 1280 2835 bytes in one piece. In order to accommodate larger IPv6 packets (up 2836 to the OMNI interface MTU), the Client inserts an OAL header with 2837 source set to its own ULA and destination set to the ULA of the 2838 target and uses IPv6 fragmentation according to Section 3.9. The 2839 Client then encapsulates each fragment in a UDP/IP header and sends 2840 the fragments to the next hop. 2842 3.22.1. Use of SEND and CGA 2844 In some environments, use of the [RFC4380] Authentication option 2845 alone may be sufficient for assuring IPv6 ND message authentication 2846 between Clients and Servers. When additional protection is 2847 necessary, nodes should employ SEcure Neighbor Discovery (SEND) 2848 [RFC3971] with Cryptographically-Generated Addresses (CGA) [RFC3972]. 2850 When SEND/CGA are used, the Client prepares RS messages with its 2851 link-local CGA as the IPv6 source and (link-local) All-Routers 2852 multicast as the IPv6 Destination, includes any SEND options and 2853 wraps the message in an OAL header. The Client sets the OAL source 2854 address to its own ULA and sets the OAL destination address to (site- 2855 local) All-Routers multicast. The Client then wraps the RS message 2856 in UDP/IP headers according to [RFC4380] and sends the message to the 2857 Server. 2859 When the Server receives the message, it first verifies the 2860 Authentication option (if present) then uses the OAL source address 2861 to determine the MNP of the Client. The Server then processes the 2862 SEND options to authenticate the RS message and prepares an RA 2863 message response. The Server prepares the RA with its own link-local 2864 CGA as the IPv6 source and the CGA of the Client as the IPv6 2865 destination, includes any SEND options and wraps the message in an 2866 OAL header. The Server sets the OAL source address to its own ULA 2867 and sets the OAL destination address to the Client's ULA. The Server 2868 then wraps the RA message in UDP/IP headers according to [RFC4380] 2869 and sends the message to the Client. Thereafter, the Client/Server 2870 send additional RS/RA messages to maintain their association and any 2871 NAT state. 2873 The Client and Server also may exchange NS/NA messages using their 2874 own CGA as the source and with OAL encapsulation as above. When a 2875 Client sends an NS(AR), it sets the IPv6 source to its CGA and sets 2876 the IPv6 destination to the Solicited-Node Multicast address of the 2877 target. The Client then wraps the message in an OAL header with its 2878 own ULA as the source and the ULA of the target as the destination 2879 and sends it to the Server. The Server authenticates the message, 2880 then changes the IPv6 source address to the Client's LLA, removes the 2881 SEND options, and sends a corresponding NS(AR) into the spanning 2882 tree. When the Server receives the corresponding OAL-encapsulated 2883 NA, it changes the IPv6 destination address to the Client's CGA, 2884 inserts SEND options, then wraps the message in UDP/IP headers and 2885 sends it to the Client. 2887 When a Client sends a uNA, it sets the IPv6 source address to its own 2888 CGA and sets the IPv6 destination address to (link-local) All-Nodes 2889 multicast, includes SEND options, wraps the message in OAL and UDP/IP 2890 headers and sends the message to the Server. The Server 2891 authenticates the message, then changes the IPv6 address to the 2892 Client's LLA, removes the SEND options and forwards the message the 2893 same as discussed in Section 3.16.1. In the reverse direction, when 2894 the Server forwards a uNA to the Client, it changes the IPv6 address 2895 to its own CGA and inserts SEND options then forwards the message to 2896 the Client. 2898 When a Client sends an NS(NUD), it sets both the IPv6 source and 2899 destination address to its own LLA, wraps the message in an OAL 2900 header and UDP/IP headers, then sends the message directly to the 2901 peer which will loop the message back. In this case alone, the 2902 Client does not use the Server as a trust broker for forwarding the 2903 ND message. 2905 3.23. Time-Varying MNPs 2907 In some use cases, it is desirable, beneficial and efficient for the 2908 Client to receive a constant MNP that travels with the Client 2909 wherever it moves. For example, this would allow air traffic 2910 controllers to easily track aircraft, etc. In other cases, however 2911 (e.g., intelligent transportation systems), the MN may be willing to 2912 sacrifice a modicum of efficiency in order to have time-varying MNPs 2913 that can be changed every so often to defeat adversarial tracking. 2915 The DHCPv6 service offers a way for Clients that desire time-varying 2916 MNPs to obtain short-lived prefixes (e.g., on the order of a small 2917 number of minutes). In that case, the identity of the Client would 2918 not be bound to the MNP but rather the Client's identity would be 2919 bound to the DHCPv6 Device Unique Identifier (DUID) and used as the 2920 seed for Prefix Delegation. The Client would then be obligated to 2921 renumber its internal networks whenever its MNP (and therefore also 2922 its LLA) changes. This should not present a challenge for Clients 2923 with automated network renumbering services, however presents limits 2924 for the durations of ongoing sessions that would prefer to use a 2925 constant address. 2927 4. Implementation Status 2929 An early AERO implementation based on OpenVPN (https://openvpn.net/) 2930 was announced on the v6ops mailing list on January 10, 2018 and an 2931 initial public release of the AERO proof-of-concept source code was 2932 announced on the intarea mailing list on August 21, 2015. 2934 AERO Release-3.0.1 was tagged on September 11, 2020, and is 2935 undergoing internal testing. Additional releases expected Q42020, 2936 with first public release expected before year-end. 2938 5. IANA Considerations 2940 The IANA has assigned a 4-octet Private Enterprise Number "45282" for 2941 AERO in the "enterprise-numbers" registry. 2943 The IANA has assigned the UDP port number "8060" for an earlier 2944 experimental version of AERO [RFC6706]. This document obsoletes 2945 [RFC6706] and claims the UDP port number "8060" for all future use. 2947 The IANA is instructed to assign a new type value TBD in the Segment 2948 Routing Header TLV registry [RFC8754]. 2950 No further IANA actions are required. 2952 6. Security Considerations 2954 AERO Bridges configure secured tunnels with AERO Servers, Realys and 2955 Proxys within their local OMNI link segments. Applicable secured 2956 tunnel alternatives include IPsec [RFC4301], TLS/SSL [RFC8446], DTLS 2957 [RFC6347], WireGuard [WG], etc. The AERO Bridges of all OMNI link 2958 segments in turn configure secured tunnels for their neighboring AERO 2959 Bridges in a spanning tree topology. Therefore, control messages 2960 exchanged between any pair of OMNI link neighbors on the spanning 2961 tree are already secured. 2963 AERO Servers, Relays and Proxys targeted by a route optimization may 2964 also receive data packets directly from arbitrary nodes in INET 2965 partitions instead of via the spanning tree. For INET partitions 2966 that apply effective ingress filtering to defeat source address 2967 spoofing, the simple data origin authentication procedures in 2968 Section 3.8 can be applied. 2970 For INET partitions that require strong security in the data plane, 2971 two options for securing communications include 1) disable route 2972 optimization so that all traffic is conveyed over secured tunnels, or 2973 2) enable on-demand secure tunnel creation between INET partition 2974 neighbors. Option 1) would result in longer routes than necessary 2975 and traffic concentration on critical infrastructure elements. 2976 Option 2) could be coordinated by establishing a secured tunnel on- 2977 demand instead of performing an NS/NA exchange in the route 2978 optimization procedures. Procedures for establishing on-demand 2979 secured tunnels are out of scope. 2981 AERO Clients that connect to secured ANETs need not apply security to 2982 their ND messages, since the messages will be intercepted by a 2983 perimeter Proxy that applies security on its INET-facing interface as 2984 part of the spanning tree (see above). AERO Clients connected to the 2985 open INET can use symmetric network and/or transport layer security 2986 services such as VPNs or can by some other means establish a direct 2987 link. When a VPN or direct link may be impractical, however, an 2988 asymmetric security service such as SEcure Neighbor Discovery (SEND) 2989 [RFC3971] with Cryptographically Generated Addresses (CGAs) [RFC3972] 2990 and/or the Authentication option [RFC4380] can be applied. 2992 Application endpoints SHOULD use application-layer security services 2993 such as TLS/SSL, DTLS or SSH [RFC4251] to assure the same level of 2994 protection as for critical secured Internet services. AERO Clients 2995 that require host-based VPN services SHOULD use symmetric network 2996 and/or transport layer security services such as IPsec, TLS/SSL, 2997 DTLS, etc. AERO Proxys and Servers can also provide a network-based 2998 VPN service on behalf of the Client, e.g., if the Client is located 2999 within a secured enclave and cannot establish a VPN on its own 3000 behalf. 3002 AERO Servers and Bridges present targets for traffic amplification 3003 Denial of Service (DoS) attacks. This concern is no different than 3004 for widely-deployed VPN security gateways in the Internet, where 3005 attackers could send spoofed packets to the gateways at high data 3006 rates. This can be mitigated by connecting Servers and Bridges over 3007 dedicated links with no connections to the Internet and/or when 3008 connections to the Internet are only permitted through well-managed 3009 firewalls. Traffic amplification DoS attacks can also target an AERO 3010 Client's low data rate links. This is a concern not only for Clients 3011 located on the open Internet but also for Clients in secured 3012 enclaves. AERO Servers and Proxys can institute rate limits that 3013 protect Clients from receiving packet floods that could DoS low data 3014 rate links. 3016 AERO Relays must implement ingress filtering to avoid a spoofing 3017 attack in which spurious messages with ULA addresses are injected 3018 into an OMNI link from an outside attacker. AERO Clients MUST ensure 3019 that their connectivity is not used by unauthorized nodes on their 3020 EUNs to gain access to a protected network, i.e., AERO Clients that 3021 act as routers MUST NOT provide routing services for unauthorized 3022 nodes. (This concern is no different than for ordinary hosts that 3023 receive an IP address delegation but then "share" the address with 3024 other nodes via some form of Internet connection sharing such as 3025 tethering.) 3027 The MAP list MUST be well-managed and secured from unauthorized 3028 tampering, even though the list contains only public information. 3029 The MAP list can be conveyed to the Client in a similar fashion as in 3030 [RFC5214] (e.g., through layer 2 data link login messaging, secure 3031 upload of a static file, DNS lookups, etc.). 3033 Although public domain and commercial SEND implementations exist, 3034 concerns regarding the strength of the cryptographic hash algorithm 3035 have been documented [RFC6273] [RFC4982]. 3037 SRH authentication facilities are specified in [RFC8754]. 3039 Security considerations for accepting link-layer ICMP messages and 3040 reflected packets are discussed throughout the document. 3042 Security considerations for IPv6 fragmentation and reassembly are 3043 discussed in [I-D.templin-6man-omni-interface]. 3045 7. Acknowledgements 3047 Discussions in the IETF, aviation standards communities and private 3048 exchanges helped shape some of the concepts in this work. 3049 Individuals who contributed insights include Mikael Abrahamsson, Mark 3050 Andrews, Fred Baker, Bob Braden, Stewart Bryant, Brian Carpenter, 3051 Wojciech Dec, Pavel Drasil, Ralph Droms, Adrian Farrel, Nick Green, 3052 Sri Gundavelli, Brian Haberman, Bernhard Haindl, Joel Halpern, Tom 3053 Herbert, Sascha Hlusiak, Lee Howard, Zdenek Jaron, Andre Kostur, 3054 Hubert Kuenig, Ted Lemon, Andy Malis, Satoru Matsushima, Tomek 3055 Mrugalski, Madhu Niraula, Alexandru Petrescu, Behcet Saikaya, Michal 3056 Skorepa, Joe Touch, Bernie Volz, Ryuji Wakikawa, Tony Whyman, Lloyd 3057 Wood and James Woodyatt. Members of the IESG also provided valuable 3058 input during their review process that greatly improved the document. 3059 Special thanks go to Stewart Bryant, Joel Halpern and Brian Haberman 3060 for their shepherding guidance during the publication of the AERO 3061 first edition. 3063 This work has further been encouraged and supported by Boeing 3064 colleagues including Kyle Bae, M. Wayne Benson, Dave Bernhardt, Cam 3065 Brodie, John Bush, Balaguruna Chidambaram, Irene Chin, Bruce Cornish, 3066 Claudiu Danilov, Don Dillenburg, Joe Dudkowski, Wen Fang, Samad 3067 Farooqui, Anthony Gregory, Jeff Holland, Seth Jahne, Brian Jaury, 3068 Greg Kimberly, Ed King, Madhuri Madhava Badgandi, Laurel Matthew, 3069 Gene MacLean III, Rob Muszkiewicz, Sean O'Sullivan, Vijay 3070 Rajagopalan, Greg Saccone, Rod Santiago, Kent Shuey, Brian Skeen, 3071 Mike Slane, Carrie Spiker, Katie Tran, Brendan Williams, Amelia 3072 Wilson, Julie Wulff, Yueli Yang, Eric Yeh and other members of the 3073 Boeing mobility, networking and autonomy teams. Kyle Bae, Wayne 3074 Benson, Katie Tran and Eric Yeh are especially acknowledged for 3075 implementing the AERO functions as extensions to the public domain 3076 OpenVPN distribution. 3078 Earlier works on NBMA tunneling approaches are found in 3079 [RFC2529][RFC5214][RFC5569]. 3081 Many of the constructs presented in this second edition of AERO are 3082 based on the author's earlier works, including: 3084 o The Internet Routing Overlay Network (IRON) 3085 [RFC6179][I-D.templin-ironbis] 3087 o Virtual Enterprise Traversal (VET) 3088 [RFC5558][I-D.templin-intarea-vet] 3090 o The Subnetwork Encapsulation and Adaptation Layer (SEAL) 3091 [RFC5320][I-D.templin-intarea-seal] 3093 o AERO, First Edition [RFC6706] 3095 Note that these works cite numerous earlier efforts that are not also 3096 cited here due to space limitations. The authors of those earlier 3097 works are acknowledged for their insights. 3099 This work is aligned with the NASA Safe Autonomous Systems Operation 3100 (SASO) program under NASA contract number NNA16BD84C. 3102 This work is aligned with the FAA as per the SE2025 contract number 3103 DTFAWA-15-D-00030. 3105 This work is aligned with the Boeing Commercial Airplanes (BCA) 3106 Internet of Things (IoT) and autonomy programs. 3108 This work is aligned with the Boeing Information Technology (BIT) 3109 MobileNet program. 3111 8. References 3113 8.1. Normative References 3115 [I-D.templin-6man-omni-interface] 3116 Templin, F. and T. Whyman, "Transmission of IPv6 Packets 3117 over Overlay Multilink Network (OMNI) Interfaces", draft- 3118 templin-6man-omni-interface-34 (work in progress), 3119 September 2020. 3121 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 3122 DOI 10.17487/RFC0791, September 1981, 3123 . 3125 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 3126 RFC 792, DOI 10.17487/RFC0792, September 1981, 3127 . 3129 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 3130 Requirement Levels", BCP 14, RFC 2119, 3131 DOI 10.17487/RFC2119, March 1997, 3132 . 3134 [RFC2473] Conta, A. and S. Deering, "Generic Packet Tunneling in 3135 IPv6 Specification", RFC 2473, DOI 10.17487/RFC2473, 3136 December 1998, . 3138 [RFC3971] Arkko, J., Ed., Kempf, J., Zill, B., and P. Nikander, 3139 "SEcure Neighbor Discovery (SEND)", RFC 3971, 3140 DOI 10.17487/RFC3971, March 2005, 3141 . 3143 [RFC3972] Aura, T., "Cryptographically Generated Addresses (CGA)", 3144 RFC 3972, DOI 10.17487/RFC3972, March 2005, 3145 . 3147 [RFC4191] Draves, R. and D. Thaler, "Default Router Preferences and 3148 More-Specific Routes", RFC 4191, DOI 10.17487/RFC4191, 3149 November 2005, . 3151 [RFC4193] Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast 3152 Addresses", RFC 4193, DOI 10.17487/RFC4193, October 2005, 3153 . 3155 [RFC4380] Huitema, C., "Teredo: Tunneling IPv6 over UDP through 3156 Network Address Translations (NATs)", RFC 4380, 3157 DOI 10.17487/RFC4380, February 2006, 3158 . 3160 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 3161 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 3162 DOI 10.17487/RFC4861, September 2007, 3163 . 3165 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 3166 Address Autoconfiguration", RFC 4862, 3167 DOI 10.17487/RFC4862, September 2007, 3168 . 3170 [RFC6081] Thaler, D., "Teredo Extensions", RFC 6081, 3171 DOI 10.17487/RFC6081, January 2011, 3172 . 3174 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 3175 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 3176 May 2017, . 3178 [RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6 3179 (IPv6) Specification", STD 86, RFC 8200, 3180 DOI 10.17487/RFC8200, July 2017, 3181 . 3183 [RFC8415] Mrugalski, T., Siodelski, M., Volz, B., Yourtchenko, A., 3184 Richardson, M., Jiang, S., Lemon, T., and T. Winters, 3185 "Dynamic Host Configuration Protocol for IPv6 (DHCPv6)", 3186 RFC 8415, DOI 10.17487/RFC8415, November 2018, 3187 . 3189 8.2. Informative References 3191 [BGP] Huston, G., "BGP in 2015, http://potaroo.net", January 3192 2016. 3194 [I-D.bonica-6man-comp-rtg-hdr] 3195 Bonica, R., Kamite, Y., Niwa, T., Alston, A., and L. 3196 Jalil, "The IPv6 Compact Routing Header (CRH)", draft- 3197 bonica-6man-comp-rtg-hdr-22 (work in progress), May 2020. 3199 [I-D.bonica-6man-crh-helper-opt] 3200 Li, X., Bao, C., Ruan, E., and R. Bonica, "Compressed 3201 Routing Header (CRH) Helper Option", draft-bonica-6man- 3202 crh-helper-opt-01 (work in progress), May 2020. 3204 [I-D.ietf-intarea-frag-fragile] 3205 Bonica, R., Baker, F., Huston, G., Hinden, R., Troan, O., 3206 and F. Gont, "IP Fragmentation Considered Fragile", draft- 3207 ietf-intarea-frag-fragile-17 (work in progress), September 3208 2019. 3210 [I-D.ietf-intarea-tunnels] 3211 Touch, J. and M. Townsley, "IP Tunnels in the Internet 3212 Architecture", draft-ietf-intarea-tunnels-10 (work in 3213 progress), September 2019. 3215 [I-D.ietf-rtgwg-atn-bgp] 3216 Templin, F., Saccone, G., Dawra, G., Lindem, A., and V. 3217 Moreno, "A Simple BGP-based Mobile Routing System for the 3218 Aeronautical Telecommunications Network", draft-ietf- 3219 rtgwg-atn-bgp-06 (work in progress), June 2020. 3221 [I-D.templin-6man-dhcpv6-ndopt] 3222 Templin, F., "A Unified Stateful/Stateless Configuration 3223 Service for IPv6", draft-templin-6man-dhcpv6-ndopt-10 3224 (work in progress), June 2020. 3226 [I-D.templin-intarea-seal] 3227 Templin, F., "The Subnetwork Encapsulation and Adaptation 3228 Layer (SEAL)", draft-templin-intarea-seal-68 (work in 3229 progress), January 2014. 3231 [I-D.templin-intarea-vet] 3232 Templin, F., "Virtual Enterprise Traversal (VET)", draft- 3233 templin-intarea-vet-40 (work in progress), May 2013. 3235 [I-D.templin-ironbis] 3236 Templin, F., "The Interior Routing Overlay Network 3237 (IRON)", draft-templin-ironbis-16 (work in progress), 3238 March 2014. 3240 [I-D.templin-v6ops-pdhost] 3241 Templin, F., "IPv6 Prefix Delegation and Multi-Addressing 3242 Models", draft-templin-v6ops-pdhost-26 (work in progress), 3243 June 2020. 3245 [OVPN] OpenVPN, O., "http://openvpn.net", October 2016. 3247 [RFC1035] Mockapetris, P., "Domain names - implementation and 3248 specification", STD 13, RFC 1035, DOI 10.17487/RFC1035, 3249 November 1987, . 3251 [RFC1812] Baker, F., Ed., "Requirements for IP Version 4 Routers", 3252 RFC 1812, DOI 10.17487/RFC1812, June 1995, 3253 . 3255 [RFC2003] Perkins, C., "IP Encapsulation within IP", RFC 2003, 3256 DOI 10.17487/RFC2003, October 1996, 3257 . 3259 [RFC2236] Fenner, W., "Internet Group Management Protocol, Version 3260 2", RFC 2236, DOI 10.17487/RFC2236, November 1997, 3261 . 3263 [RFC2529] Carpenter, B. and C. Jung, "Transmission of IPv6 over IPv4 3264 Domains without Explicit Tunnels", RFC 2529, 3265 DOI 10.17487/RFC2529, March 1999, 3266 . 3268 [RFC2983] Black, D., "Differentiated Services and Tunnels", 3269 RFC 2983, DOI 10.17487/RFC2983, October 2000, 3270 . 3272 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 3273 of Explicit Congestion Notification (ECN) to IP", 3274 RFC 3168, DOI 10.17487/RFC3168, September 2001, 3275 . 3277 [RFC3330] IANA, "Special-Use IPv4 Addresses", RFC 3330, 3278 DOI 10.17487/RFC3330, September 2002, 3279 . 3281 [RFC3810] Vida, R., Ed. and L. Costa, Ed., "Multicast Listener 3282 Discovery Version 2 (MLDv2) for IPv6", RFC 3810, 3283 DOI 10.17487/RFC3810, June 2004, 3284 . 3286 [RFC4251] Ylonen, T. and C. Lonvick, Ed., "The Secure Shell (SSH) 3287 Protocol Architecture", RFC 4251, DOI 10.17487/RFC4251, 3288 January 2006, . 3290 [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A 3291 Border Gateway Protocol 4 (BGP-4)", RFC 4271, 3292 DOI 10.17487/RFC4271, January 2006, 3293 . 3295 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 3296 Architecture", RFC 4291, DOI 10.17487/RFC4291, February 3297 2006, . 3299 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 3300 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 3301 December 2005, . 3303 [RFC4389] Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery 3304 Proxies (ND Proxy)", RFC 4389, DOI 10.17487/RFC4389, April 3305 2006, . 3307 [RFC4443] Conta, A., Deering, S., and M. Gupta, Ed., "Internet 3308 Control Message Protocol (ICMPv6) for the Internet 3309 Protocol Version 6 (IPv6) Specification", STD 89, 3310 RFC 4443, DOI 10.17487/RFC4443, March 2006, 3311 . 3313 [RFC4511] Sermersheim, J., Ed., "Lightweight Directory Access 3314 Protocol (LDAP): The Protocol", RFC 4511, 3315 DOI 10.17487/RFC4511, June 2006, 3316 . 3318 [RFC4541] Christensen, M., Kimball, K., and F. Solensky, 3319 "Considerations for Internet Group Management Protocol 3320 (IGMP) and Multicast Listener Discovery (MLD) Snooping 3321 Switches", RFC 4541, DOI 10.17487/RFC4541, May 2006, 3322 . 3324 [RFC4605] Fenner, B., He, H., Haberman, B., and H. Sandick, 3325 "Internet Group Management Protocol (IGMP) / Multicast 3326 Listener Discovery (MLD)-Based Multicast Forwarding 3327 ("IGMP/MLD Proxying")", RFC 4605, DOI 10.17487/RFC4605, 3328 August 2006, . 3330 [RFC4982] Bagnulo, M. and J. Arkko, "Support for Multiple Hash 3331 Algorithms in Cryptographically Generated Addresses 3332 (CGAs)", RFC 4982, DOI 10.17487/RFC4982, July 2007, 3333 . 3335 [RFC5015] Handley, M., Kouvelas, I., Speakman, T., and L. Vicisano, 3336 "Bidirectional Protocol Independent Multicast (BIDIR- 3337 PIM)", RFC 5015, DOI 10.17487/RFC5015, October 2007, 3338 . 3340 [RFC5214] Templin, F., Gleeson, T., and D. Thaler, "Intra-Site 3341 Automatic Tunnel Addressing Protocol (ISATAP)", RFC 5214, 3342 DOI 10.17487/RFC5214, March 2008, 3343 . 3345 [RFC5320] Templin, F., Ed., "The Subnetwork Encapsulation and 3346 Adaptation Layer (SEAL)", RFC 5320, DOI 10.17487/RFC5320, 3347 February 2010, . 3349 [RFC5522] Eddy, W., Ivancic, W., and T. Davis, "Network Mobility 3350 Route Optimization Requirements for Operational Use in 3351 Aeronautics and Space Exploration Mobile Networks", 3352 RFC 5522, DOI 10.17487/RFC5522, October 2009, 3353 . 3355 [RFC5558] Templin, F., Ed., "Virtual Enterprise Traversal (VET)", 3356 RFC 5558, DOI 10.17487/RFC5558, February 2010, 3357 . 3359 [RFC5569] Despres, R., "IPv6 Rapid Deployment on IPv4 3360 Infrastructures (6rd)", RFC 5569, DOI 10.17487/RFC5569, 3361 January 2010, . 3363 [RFC5880] Katz, D. and D. Ward, "Bidirectional Forwarding Detection 3364 (BFD)", RFC 5880, DOI 10.17487/RFC5880, June 2010, 3365 . 3367 [RFC6106] Jeong, J., Park, S., Beloeil, L., and S. Madanapalli, 3368 "IPv6 Router Advertisement Options for DNS Configuration", 3369 RFC 6106, DOI 10.17487/RFC6106, November 2010, 3370 . 3372 [RFC6146] Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful 3373 NAT64: Network Address and Protocol Translation from IPv6 3374 Clients to IPv4 Servers", RFC 6146, DOI 10.17487/RFC6146, 3375 April 2011, . 3377 [RFC6179] Templin, F., Ed., "The Internet Routing Overlay Network 3378 (IRON)", RFC 6179, DOI 10.17487/RFC6179, March 2011, 3379 . 3381 [RFC6221] Miles, D., Ed., Ooghe, S., Dec, W., Krishnan, S., and A. 3382 Kavanagh, "Lightweight DHCPv6 Relay Agent", RFC 6221, 3383 DOI 10.17487/RFC6221, May 2011, 3384 . 3386 [RFC6273] Kukec, A., Krishnan, S., and S. Jiang, "The Secure 3387 Neighbor Discovery (SEND) Hash Threat Analysis", RFC 6273, 3388 DOI 10.17487/RFC6273, June 2011, 3389 . 3391 [RFC6347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer 3392 Security Version 1.2", RFC 6347, DOI 10.17487/RFC6347, 3393 January 2012, . 3395 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 3396 for Equal Cost Multipath Routing and Link Aggregation in 3397 Tunnels", RFC 6438, DOI 10.17487/RFC6438, November 2011, 3398 . 3400 [RFC6706] Templin, F., Ed., "Asymmetric Extended Route Optimization 3401 (AERO)", RFC 6706, DOI 10.17487/RFC6706, August 2012, 3402 . 3404 [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and 3405 UDP Checksums for Tunneled Packets", RFC 6935, 3406 DOI 10.17487/RFC6935, April 2013, 3407 . 3409 [RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement 3410 for the Use of IPv6 UDP Datagrams with Zero Checksums", 3411 RFC 6936, DOI 10.17487/RFC6936, April 2013, 3412 . 3414 [RFC7333] Chan, H., Ed., Liu, D., Seite, P., Yokota, H., and J. 3415 Korhonen, "Requirements for Distributed Mobility 3416 Management", RFC 7333, DOI 10.17487/RFC7333, August 2014, 3417 . 3419 [RFC7761] Fenner, B., Handley, M., Holbrook, H., Kouvelas, I., 3420 Parekh, R., Zhang, Z., and L. Zheng, "Protocol Independent 3421 Multicast - Sparse Mode (PIM-SM): Protocol Specification 3422 (Revised)", STD 83, RFC 7761, DOI 10.17487/RFC7761, March 3423 2016, . 3425 [RFC8402] Filsfils, C., Ed., Previdi, S., Ed., Ginsberg, L., 3426 Decraene, B., Litkowski, S., and R. Shakir, "Segment 3427 Routing Architecture", RFC 8402, DOI 10.17487/RFC8402, 3428 July 2018, . 3430 [RFC8446] Rescorla, E., "The Transport Layer Security (TLS) Protocol 3431 Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018, 3432 . 3434 [RFC8754] Filsfils, C., Ed., Dukes, D., Ed., Previdi, S., Leddy, J., 3435 Matsushima, S., and D. Voyer, "IPv6 Segment Routing Header 3436 (SRH)", RFC 8754, DOI 10.17487/RFC8754, March 2020, 3437 . 3439 [WG] Wireguard, "Wireguard, https://www.wireguard.com", August 3440 2020. 3442 Appendix A. Non-Normative Considerations 3444 AERO can be applied to a multitude of Internetworking scenarios, with 3445 each having its own adaptations. The following considerations are 3446 provided as non-normative guidance: 3448 A.1. Implementation Strategies for Route Optimization 3450 Route optimization as discussed in Section 3.14 results in the route 3451 optimization source (ROS) creating an asymmetric neighbor cache entry 3452 for the target neighbor. The neighbor cache entry is maintained for 3453 at most ReachableTime seconds and then deleted unless updated. In 3454 order to refresh the neighbor cache entry lifetime before the 3455 ReachableTime timer expires, the specification requires 3456 implementations to issue a new NS/NA exchange to reset ReachableTime 3457 while data packets are still flowing. However, the decision of when 3458 to initiate a new NS/NA exchange and to perpetuate the process is 3459 left as an implementation detail. 3461 One possible strategy may be to monitor the neighbor cache entry 3462 watching for data packets for (ReachableTime - 5) seconds. If any 3463 data packets have been sent to the neighbor within this timeframe, 3464 then send an NS to receive a new NA. If no data packets have been 3465 sent, wait for 5 additional seconds and send an immediate NS if any 3466 data packets are sent within this "expiration pending" 5 second 3467 window. If no additional data packets are sent within the 5 second 3468 window, delete the neighbor cache entry. 3470 The monitoring of the neighbor data packet traffic therefore becomes 3471 an asymmetric ongoing process during the neighbor cache entry 3472 lifetime. If the neighbor cache entry expires, future data packets 3473 will trigger a new NS/NA exchange while the packets themselves are 3474 delivered over a longer path until route optimization state is re- 3475 established. 3477 A.2. Implicit Mobility Management 3479 OMNI interface neighbors MAY provide a configuration option that 3480 allows them to perform implicit mobility management in which no ND 3481 messaging is used. In that case, the Client only transmits packets 3482 over a single interface at a time, and the neighbor always observes 3483 packets arriving from the Client from the same link-layer source 3484 address. 3486 If the Client's underlying interface address changes (either due to a 3487 readdressing of the original interface or switching to a new 3488 interface) the neighbor immediately updates the neighbor cache entry 3489 for the Client and begins accepting and sending packets according to 3490 the Client's new address. This implicit mobility method applies to 3491 use cases such as cellphones with both WiFi and Cellular interfaces 3492 where only one of the interfaces is active at a given time, and the 3493 Client automatically switches over to the backup interface if the 3494 primary interface fails. 3496 A.3. Direct Underlying Interfaces 3498 When a Client's OMNI interface is configured over a Direct interface, 3499 the neighbor at the other end of the Direct link can receive packets 3500 without any encapsulation. In that case, the Client sends packets 3501 over the Direct link according to QoS preferences. If the Direct 3502 interface has the highest QoS preference, then the Client's IP 3503 packets are transmitted directly to the peer without going through an 3504 ANET/INET. If other interfaces have higher QoS preferences, then the 3505 Client's IP packets are transmitted via a different interface, which 3506 may result in the inclusion of Proxys, Servers and Bridges in the 3507 communications path. Direct interfaces must be tested periodically 3508 for reachability, e.g., via NUD. 3510 A.4. AERO Critical Infrastructure Considerations 3512 AERO Bridges can be either Commercial off-the Shelf (COTS) standard 3513 IP routers or virtual machines in the cloud. Bridges must be 3514 provisioned, supported and managed by the INET administrative 3515 authority, and connected to the Bridges of other INETs via inter- 3516 domain peerings. Cost for purchasing, configuring and managing 3517 Bridges is nominal even for very large OMNI links. 3519 AERO Servers can be standard dedicated server platforms, but most 3520 often will be deployed as virtual machines in the cloud. The only 3521 requirements for Servers are that they can run the AERO user-level 3522 code and have at least one network interface connection to the INET. 3523 As with Bridges, Servers must be provisioned, supported and managed 3524 by the INET administrative authority. Cost for purchasing, 3525 configuring and managing Servers is nominal especially for virtual 3526 Servers hosted in the cloud. 3528 AERO Proxys are most often standard dedicated server platforms with 3529 one network interface connected to the ANET and a second interface 3530 connected to an INET. As with Servers, the only requirements are 3531 that they can run the AERO user-level code and have at least one 3532 interface connection to the INET. Proxys must be provisioned, 3533 supported and managed by the ANET administrative authority. Cost for 3534 purchasing, configuring and managing Proxys is nominal, and borne by 3535 the ANET administrative authority. 3537 AERO Relays can be any dedicated server or COTS router platform 3538 connected to INETs and/or EUNs. The Relay connects to the OMNI link 3539 and engages in eBGP peering with one or more Bridges as a stub AS. 3540 The Relay then injects its MNPs and/or non-MNP prefixes into the BGP 3541 routing system, and provisions the prefixes to its downstream- 3542 attached networks. The Relay can perform ROS/ROR services the same 3543 as for any Server, and can route between the MNP and non-MNP address 3544 spaces. 3546 A.5. AERO Server Failure Implications 3548 AERO Servers may appear as a single point of failure in the 3549 architecture, but such is not the case since all Servers on the link 3550 provide identical services and loss of a Server does not imply 3551 immediate and/or comprehensive communication failures. Although 3552 Clients typically associate with a single Server at a time, Server 3553 failure is quickly detected and conveyed by Bidirectional Forward 3554 Detection (BFD) and/or proactive NUD allowing Clients to migrate to 3555 new Servers. 3557 If a Server fails, ongoing packet forwarding to Clients will continue 3558 by virtue of the asymmetric neighbor cache entries that have already 3559 been established in route optimization sources (ROSs). If a Client 3560 also experiences mobility events at roughly the same time the Server 3561 fails, unsolicited NA messages may be lost but proxy neighbor cache 3562 entries in the DEPARTED state will ensure that packet forwarding to 3563 the Client's new locations will continue for up to DepartTime 3564 seconds. 3566 If a Client is left without a Server for an extended timeframe (e.g., 3567 greater than ReachableTime seconds) then existing asymmetric neighbor 3568 cache entries will eventually expire and both ongoing and new 3569 communications will fail. The original source will continue to 3570 retransmit until the Client has established a new Server 3571 relationship, after which time continuous communications will resume. 3573 Therefore, providing many Servers on the link with high availability 3574 profiles provides resilience against loss of individual Servers and 3575 assurance that Clients can establish new Server relationships quickly 3576 in event of a Server failure. 3578 A.6. AERO Client / Server Architecture 3580 The AERO architectural model is client / server in the control plane, 3581 with route optimization in the data plane. The same as for common 3582 Internet services, the AERO Client discovers the addresses of AERO 3583 Servers and selects one Server to connect to. The AERO service is 3584 analogous to common Internet services such as google.com, yahoo.com, 3585 cnn.com, etc. However, there is only one AERO service for the link 3586 and all Servers provide identical services. 3588 Common Internet services provide differing strategies for advertising 3589 server addresses to clients. The strategy is conveyed through the 3590 DNS resource records returned in response to name resolution queries. 3591 As of January 2020 Internet-based 'nslookup' services were used to 3592 determine the following: 3594 o When a client resolves the domainname "google.com", the DNS always 3595 returns one A record (i.e., an IPv4 address) and one AAAA record 3596 (i.e., an IPv6 address). The client receives the same addresses 3597 each time it resolves the domainname via the same DNS resolver, 3598 but may receive different addresses when it resolves the 3599 domainname via different DNS resolvers. But, in each case, 3600 exactly one A and one AAAA record are returned. 3602 o When a client resolves the domainname "ietf.org", the DNS always 3603 returns one A record and one AAAA record with the same addresses 3604 regardless of which DNS resolver is used. 3606 o When a client resolves the domainname "yahoo.com", the DNS always 3607 returns a list of 4 A records and 4 AAAA records. Each time the 3608 client resolves the domainname via the same DNS resolver, the same 3609 list of addresses are returned but in randomized order (i.e., 3610 consistent with a DNS round-robin strategy). But, interestingly, 3611 the same addresses are returned (albeit in randomized order) when 3612 the domainname is resolved via different DNS resolvers. 3614 o When a client resolves the domainname "amazon.com", the DNS always 3615 returns a list of 3 A records and no AAAA records. As with 3616 "yahoo.com", the same three A records are returned from any 3617 worldwide Internet connection point in randomized order. 3619 The above example strategies show differing approaches to Internet 3620 resilience and service distribution offered by major Internet 3621 services. The Google approach exposes only a single IPv4 and a 3622 single IPv6 address to clients. Clients can then select whichever IP 3623 protocol version offers the best response, but will always use the 3624 same IP address according to the current Internet connection point. 3625 This means that the IP address offered by the network must lead to a 3626 highly-available server and/or service distribution point. In other 3627 words, resilience is predicated on high availability within the 3628 network and with no client-initiated failovers expected (i.e., it is 3629 all-or-nothing from the client's perspective). However, Google does 3630 provide for worldwide distributed service distribution by virtue of 3631 the fact that each Internet connection point responds with a 3632 different IPv6 and IPv4 address. The IETF approach is like google 3633 (all-or-nothing from the client's perspective), but provides only a 3634 single IPv4 or IPv6 address on a worldwide basis. This means that 3635 the addresses must be made highly-available at the network level with 3636 no client failover possibility, and if there is any worldwide service 3637 distribution it would need to be conducted by a network element that 3638 is reached via the IP address acting as a service distribution point. 3640 In contrast to the Google and IETF philosophies, Yahoo and Amazon 3641 both provide clients with a (short) list of IP addresses with Yahoo 3642 providing both IP protocol versions and Amazon as IPv4-only. The 3643 order of the list is randomized with each name service query 3644 response, with the effect of round-robin load balancing for service 3645 distribution. With a short list of addresses, there is still 3646 expectation that the network will implement high availability for 3647 each address but in case any single address fails the client can 3648 switch over to using a different address. The balance then becomes 3649 one of function in the network vs function in the end system. 3651 The same implications observed for common highly-available services 3652 in the Internet apply also to the AERO client/server architecture. 3653 When an AERO Client connects to one or more ANETs, it discovers one 3654 or more AERO Server addresses through the mechanisms discussed in 3655 earlier sections. Each Server address presumably leads to a fault- 3656 tolerant clustering arrangement such as supported by Linux-HA, 3657 Extended Virtual Synchrony or Paxos. Such an arrangement has 3658 precedence in common Internet service deployments in lightweight 3659 virtual machines without requiring expensive hardware deployment. 3660 Similarly, common Internet service deployments set service IP 3661 addresses on service distribution points that may relay requests to 3662 many different servers. 3664 For AERO, the expectation is that a combination of the Google/IETF 3665 and Yahoo/Amazon philosophies would be employed. The AERO Client 3666 connects to different ANET access points and can receive 1-2 Server 3667 LLAs at each point. It then selects one AERO Server address, and 3668 engages in RS/RA exchanges with the same Server from all ANET 3669 connections. The Client remains with this Server unless or until the 3670 Server fails, in which case it can switch over to an alternate 3671 Server. The Client can likewise switch over to a different Server at 3672 any time if there is some reason for it to do so. So, the AERO 3673 expectation is for a balance of function in the network and end 3674 system, with fault tolerance and resilience at both levels. 3676 Appendix B. Change Log 3678 << RFC Editor - remove prior to publication >> 3680 Changes from draft-templin-intarea-6706bis-59 to draft-templin- 3681 intrea-6706bis-60: 3683 o Removed all references to S/TLLAO - all Interface Attributes are 3684 now maintained completely in the OMNI option. 3686 Changes from draft-templin-intarea-6706bis-58 to draft-templin- 3687 intrea-6706bis-59: 3689 o The term "Relay"used in older draft versions is now "Bridge". 3690 "Relay" now refers to what was formally called: "Gateway". 3692 o Fine-grained cleanup of Forwarding Algorithm; IPv6 ND message 3693 addressing; OMNI Prefix Lengths, etc. 3695 Changes from draft-templin-intarea-6706bis-54 to draft-templin- 3696 intrea-6706bis-55: 3698 o Updates on Segment Routing and S/TLLAO contents. 3700 o Various editorials and addressing cleanups. 3702 Changes from draft-templin-intarea-6706bis-52 to draft-templin- 3703 intrea-6706bis-53: 3705 o Normative reference to the OMNI spec, and remove portions that are 3706 already specified in OMNI. 3708 o Renamed "AERO interface/link" to "OMIN interface/link" throughout 3709 the document. 3711 o Truncated obsolete back section matter. 3713 Author's Address 3715 Fred L. Templin (editor) 3716 Boeing Research & Technology 3717 P.O. Box 3707 3718 Seattle, WA 98124 3719 USA 3721 Email: fltemplin@acm.org