idnits 2.17.1 draft-templin-intarea-6706bis-58.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- == The 'Obsoletes: ' line in the draft header should list only the _numbers_ of the RFCs which will be obsoleted by this document (if approved); it should not include the word 'RFC' in the list. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (June 29, 2020) is 1396 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 1013 -- Looks like a reference, but probably isn't: '2' on line 1019 -- Looks like a reference, but probably isn't: '3' on line 869 == Missing Reference: 'N' is mentioned on line 1029, but not defined == Unused Reference: 'BGP' is defined on line 3266, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-intarea-frag-fragile' is defined on line 3279, but no explicit reference was found in the text == Outdated reference: A later version (-99) exists of draft-templin-6man-omni-interface-25 == Outdated reference: A later version (-31) exists of draft-bonica-6man-comp-rtg-hdr-22 == Outdated reference: A later version (-04) exists of draft-bonica-6man-crh-helper-opt-01 == Outdated reference: A later version (-13) exists of draft-ietf-intarea-tunnels-10 == Outdated reference: A later version (-26) exists of draft-ietf-rtgwg-atn-bgp-05 == Outdated reference: A later version (-11) exists of draft-templin-6man-dhcpv6-ndopt-09 == Outdated reference: A later version (-27) exists of draft-templin-v6ops-pdhost-25 -- Obsolete informational reference (is this intentional?): RFC 3330 (Obsoleted by RFC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, June 29, 2020 5 rfc6179, rfc6706 (if 6 approved) 7 Intended status: Standards Track 8 Expires: December 31, 2020 10 Asymmetric Extended Route Optimization (AERO) 11 draft-templin-intarea-6706bis-58 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 uses an IPv6 link-local address format that supports operation 19 of the IPv6 Neighbor Discovery (ND) protocol and links ND to IP 20 forwarding. Prefix delegation/registration services are employed for 21 network admission and to manage the routing system. Multilink 22 operation, mobility management, quality of service (QoS) signaling 23 and route optimization are naturally supported through dynamic 24 neighbor cache updates. Standard IP multicasting services are also 25 supported. AERO is a widely-applicable mobile internetworking 26 service especially well-suited to aviation services, intelligent 27 transportation systems, mobile Virtual Private Networks (VPNs) and 28 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 December 31, 2020. 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. AERO 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 . . . . . . . . . . . . . . 25 80 3.4.1. AERO Server/Relay Behavior . . . . . . . . . . . . . 25 81 3.4.2. AERO Proxy Behavior . . . . . . . . . . . . . . . . . 26 82 3.4.3. AERO Client Behavior . . . . . . . . . . . . . . . . 26 83 3.4.4. AERO Bridge Behavior . . . . . . . . . . . . . . . . 26 84 3.5. OMNI Interface Neighbor Cache Maintenance . . . . . . . . 26 85 3.6. OMNI Interface Encapsulation and Re-encapsulation . . . . 28 86 3.7. OMNI Interface Decapsulation . . . . . . . . . . . . . . 30 87 3.8. OMNI Interface Data Origin Authentication . . . . . . . . 30 88 3.9. OMNI Interface MTU and Fragmentation . . . . . . . . . . 30 89 3.10. OMNI Interface Forwarding Algorithm . . . . . . . . . . . 30 90 3.10.1. Client Forwarding Algorithm . . . . . . . . . . . . 31 91 3.10.2. Proxy Forwarding Algorithm . . . . . . . . . . . . . 32 92 3.10.3. Server/Relay Forwarding Algorithm . . . . . . . . . 33 93 3.10.4. Bridge Forwarding Algorithm . . . . . . . . . . . . 34 94 3.11. OMNI Interface Error Handling . . . . . . . . . . . . . . 35 95 3.12. AERO Router Discovery, Prefix Delegation and 96 Autoconfiguration . . . . . . . . . . . . . . . . . . . . 37 97 3.12.1. AERO ND/PD Service Model . . . . . . . . . . . . . . 37 98 3.12.2. AERO Client Behavior . . . . . . . . . . . . . . . . 38 99 3.12.3. AERO Server Behavior . . . . . . . . . . . . . . . . 40 100 3.13. The AERO Proxy . . . . . . . . . . . . . . . . . . . . . 43 101 3.13.1. Ancillary Servers Acting as Proxies . . . . . . . . 45 102 3.13.2. Detecting and Responding to Server Failures . . . . 45 103 3.13.3. Point-to-Multipoint Server Coordination . . . . . . 46 104 3.14. AERO Route Optimization / Address Resolution . . . . . . 47 105 3.14.1. Route Optimization Initiation . . . . . . . . . . . 47 106 3.14.2. Relaying the NS . . . . . . . . . . . . . . . . . . 48 107 3.14.3. Processing the NS and Sending the NA . . . . . . . . 48 108 3.14.4. Relaying the NA . . . . . . . . . . . . . . . . . . 49 109 3.14.5. Processing the NA . . . . . . . . . . . . . . . . . 49 110 3.14.6. Route Optimization Maintenance . . . . . . . . . . . 49 111 3.15. Neighbor Unreachability Detection (NUD) . . . . . . . . . 50 112 3.16. Mobility Management and Quality of Service (QoS) . . . . 52 113 3.16.1. Mobility Update Messaging . . . . . . . . . . . . . 52 114 3.16.2. Announcing Link-Layer Address and/or QoS Preference 115 Changes . . . . . . . . . . . . . . . . . . . . . . 53 116 3.16.3. Bringing New Links Into Service . . . . . . . . . . 54 117 3.16.4. Removing Existing Links from Service . . . . . . . . 54 118 3.16.5. Moving to a New Server . . . . . . . . . . . . . . . 54 119 3.17. Multicast . . . . . . . . . . . . . . . . . . . . . . . . 55 120 3.17.1. Source-Specific Multicast (SSM) . . . . . . . . . . 55 121 3.17.2. Any-Source Multicast (ASM) . . . . . . . . . . . . . 57 122 3.17.3. Bi-Directional PIM (BIDIR-PIM) . . . . . . . . . . . 58 123 3.18. Operation over Multiple OMNI Links . . . . . . . . . . . 58 124 3.19. DNS Considerations . . . . . . . . . . . . . . . . . . . 59 125 3.20. Transition Considerations . . . . . . . . . . . . . . . . 59 126 3.21. Detecting and Reacting to Server and Bridge Failures . . 60 127 3.22. AERO Clients on the Open Internet . . . . . . . . . . . . 60 128 3.22.1. Use of SEND and CGA . . . . . . . . . . . . . . . . 63 129 3.23. Time-Varying MNPs . . . . . . . . . . . . . . . . . . . . 64 130 4. Implementation Status . . . . . . . . . . . . . . . . . . . . 65 131 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 65 132 6. Security Considerations . . . . . . . . . . . . . . . . . . . 65 133 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 67 134 8. References . . . . . . . . . . . . . . . . . . . . . . . . . 69 135 8.1. Normative References . . . . . . . . . . . . . . . . . . 69 136 8.2. Informative References . . . . . . . . . . . . . . . . . 70 137 Appendix A. Non-Normative Considerations . . . . . . . . . . . . 76 138 A.1. Implementation Strategies for Route Optimization . . . . 76 139 A.2. Implicit Mobility Management . . . . . . . . . . . . . . 76 140 A.3. Direct Underlying Interfaces . . . . . . . . . . . . . . 77 141 A.4. AERO Critical Infrastructure Considerations . . . . . . . 77 142 A.5. AERO Server Failure Implications . . . . . . . . . . . . 78 143 A.6. AERO Client / Server Architecture . . . . . . . . . . . . 78 144 Appendix B. Change Log . . . . . . . . . . . . . . . . . . . . . 80 145 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 81 147 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. Multilink operation 159 allows for increased reliability, bandwidth optimization and traffic 160 path diversity. 162 The AERO service comprises Clients, Proxys, Servers and Relays that 163 are seen as OMNI link neighbors as well as Bridges that interconnect 164 OMNI link segments. Each node's OMNI interface uses an IPv6 link- 165 local address format that supports operation of the IPv6 Neighbor 166 Discovery (ND) protocol [RFC4861] and links ND to IP forwarding. A 167 node's OMNI interface can be configured over multiple underlying 168 interfaces, and may therefore appear as a single interface with 169 multiple link-layer addresses. Each link-layer address is subject to 170 change due to mobility and/or QoS fluctuations, and link-layer 171 address changes are signaled by ND messaging the same as for any IPv6 172 link. 174 AERO provides a cloud-based service where mobile nodes may use any 175 Server acting as a Mobility Anchor Point (MAP) and fixed nodes may 176 use any Relay on the link for efficient communications. Fixed nodes 177 forward packets destined to other AERO nodes to the nearest Relay, 178 which forwards them through the cloud. A mobile node's initial 179 packets are forwarded through the Server, while direct routing is 180 supported through asymmetric extended route optimization while data 181 packets are flowing. Both unicast and multicast communications are 182 supported, and mobile nodes may efficiently move between locations 183 while maintaining continuous communications with correspondents and 184 without changing their IP Address. 186 AERO Bridges are interconnected in a secured private BGP overlay 187 routing instance using encapsulation to provide a hybrid routing/ 188 bridging service that joins the underlying Internetworks of multiple 189 disjoint administrative domains into a single unified OMNI link. 190 Each OMNI link instance is characterized by the set of Mobility 191 Service Prefixes (MSPs) common to all mobile nodes. The link extends 192 to the point where a Relay/Server is on the optimal route from any 193 correspondent node on the link, and provides a conduit between the 194 underlying Internetwork and the OMNI link. To the underlying 195 Internetwork, the Relay/Server is the source of a route to the MSP, 196 and hence uplink traffic to the mobile node is naturally routed to 197 the nearest Relay/Server. 199 AERO assumes the use of PIM Sparse Mode in support of multicast 200 communication. In support of Source Specific Multicast (SSM) when a 201 Mobile Node is the source, AERO route optimization ensures that a 202 shortest-path multicast tree is established with provisions for 203 mobility and multilink operation. In all other multicast scenarios 204 there are no AERO dependencies. 206 AERO was designed for aeronautical networking for both manned and 207 unmanned aircraft, where the aircraft is treated as a mobile node 208 that can connect an Internet of Things (IoT). AERO is also 209 applicable to a wide variety of other use cases. For example, it can 210 be used to coordinate the Virtual Private Network (VPN) links of 211 mobile nodes (e.g., cellphones, tablets, laptop computers, etc.) that 212 connect into a home enterprise network via public access networks 213 using services such as OpenVPN [OVPN]. It can also be used to 214 facilitate vehicular and pedestrian communications services for 215 intelligent transportation systems. Other applicable use cases are 216 also in scope. 218 The following numbered sections present the AERO specification. The 219 appendices at the end of the document are non-normative. 221 2. Terminology 223 The terminology in the normative references applies; especially, the 224 terminology in the OMNI specification 225 [I-D.templin-6man-omni-interface] is used extensively throughout. 226 The following terms are defined within the scope of this document: 228 IPv6 Neighbor Discovery (ND) 229 an IPv6 control message service for coordinating neighbor 230 relationships between nodes connected to a common link. AERO uses 231 the ND service specified in [RFC4861]. 233 IPv6 Prefix Delegation (PD) 234 a networking service for delegating IPv6 prefixes to nodes on the 235 link. The nominal PD service is DHCPv6 [RFC8415], however 236 alternate services (e.g., based on ND messaging) are also in 237 scope. Most notably, a minimal form of PD known as "prefix 238 registration" can be used if the Client knows its prefix in 239 advance and can represent it in the IPv6 source address of an ND 240 message. 242 Access Network (ANET) 243 a node's first-hop data link service network (e.g., a radio access 244 network, cellular service provider network, corporate enterprise 245 network, etc.) that often provides link-layer security services 246 such as IEEE 802.1X and physical-layer security prevent 247 unauthorized access internally and with border network-layer 248 security services such as firewalls and proxies that prevent 249 unauthorized outside access. 251 ANET interface 252 a node's attachment to a link in an ANET. 254 Internetwork (INET) 255 a connected IP network topology with a coherent routing and 256 addressing plan and that provides a transit backbone service for 257 ANET end systems. INETs also provide an underlay service over 258 which the AERO virtual link is configured. Example INETs include 259 corporate enterprise networks, aviation networks, and the public 260 Internet itself. When there is no administrative boundary between 261 an ANET and the INET, the ANET and INET are one and the same. 263 INET Partition 264 frequently, INETs such as large corporate enterprise networks are 265 sub-divided internally into separate isolated partitions. Each 266 partition is fully connected internally but disconnected from 267 other partitions, and there is no requirement that separate 268 partitions maintain consistent Internet Protocol and/or addressing 269 plans. (Each INET partition is seen as a separate OMNI link 270 segment as discussed below.) 272 INET interface 273 a node's attachment to a link in an INET. 275 INET address 276 an IP address assigned to a node's interface connection to an 277 INET. 279 INET encapsulation 280 the encapsulation of a packet in an outer header or headers that 281 can be routed within the scope of the local INET partition. 283 OMNI link 284 the same as defined in [I-D.templin-6man-omni-interface], and 285 manifested by IPv6 encapsulation [RFC2473]. The OMNI link spans 286 underlying INET segments joined by virtual bridges in a spanning 287 tree the same as a bridged campus LAN. AERO nodes on the OMNI 288 link appear as single-hop neighbors even though they may be 289 separated by multiple underlying INET hops, and can use Segment 290 Routing [RFC8402] to cause packets to visit selected waypoints on 291 the link. 293 OMNI Interface 294 a node's attachment to an OMNI link. Since the addresses assigned 295 to an OMNI interface are managed for uniqueness, OMNI interfaces 296 do not require Duplicate Address Detection (DAD) and therefore set 297 the administrative variable 'DupAddrDetectTransmits' to zero 298 [RFC4862]. 300 OMNI Link-Local Address (LLA) 301 a link local IPv6 address per [RFC4291] constructed as specified 302 in Section 3.2.2. 304 OMNI Unique-Local Address (ULA) 305 a unique local IPv6 address per [RFC4193] constructed as specified 306 in Section 3.2.2. OMNI ULAs are statelessly derived from OMNI 307 LLAs, and vice-versa. 309 underlying interface 310 an ANET or INET interface over which an OMNI interface is 311 configured. 313 Mobility Service Prefix (MSP) 314 an IP prefix assigned to the OMNI link and from which more- 315 specific Mobile Network Prefixes (MNPs) are derived. 317 Mobile Network Prefix (MNP) 318 an IP prefix allocated from an MSP and delegated to an AERO Client 319 or Relay. 321 AERO node 322 a node that is connected to an OMNI link and participates in the 323 AERO internetworking and mobility service. 325 AERO Client ("Client") 326 an AERO node that connects over one or more underlying interfaces 327 and requests MNP PDs from AERO Servers. The Client assigns a 328 Client LLA to the OMNI interface for use in ND exchanges with 329 other AERO nodes and forwards packets to correspondents according 330 to OMNI interface neighbor cache state. 332 AERO Server ("Server") 333 an INET node that configures an OMNI interface to provide default 334 forwarding and mobility/multilink services for AERO Clients. The 335 Server assigns an administratively-provisioned LLA to its OMNI 336 interface to support the operation of the ND/PD services, and 337 advertises all of its associated MNPs via BGP peerings with 338 Bridges. 340 AERO Relay ("Relay") 341 an AERO Server that also provides forwarding services between 342 nodes reached via the OMNI link and correspondents on other links. 343 AERO Relays are provisioned with MNPs (i.e., the same as for an 344 AERO Client) and run a dynamic routing protocol to discover any 345 non-MNP IP routes. In both cases, the Relay advertises the MSP(s) 346 to its downstream networks, and distributes all of its associated 347 MNPs and non-MNP IP routes via BGP peerings with Bridges (i.e., 348 the same as for an AERO Server). 350 AERO Bridge ("Bridge") 351 a node that provides hybrid routing/bridging services (as well as 352 a security trust anchor) for nodes on an OMNI link. As a router, 353 the Bridge forwards packets using standard IP forwarding. As a 354 bridge, the Bridge forwards packets over the OMNI link without 355 decrementing the IPv6 Hop Limit. AERO Bridges peer with Servers 356 and other Bridges to discover the full set of MNPs for the link as 357 well as any non-MNPs that are reachable via Relays. 359 AERO Proxy ("Proxy") 360 a node that provides proxying services between Clients in an ANET 361 and Servers in external INETs. The AERO Proxy is a conduit 362 between the ANET and external INETs in the same manner as for 363 common web proxies, and behaves in a similar fashion as for ND 364 proxies [RFC4389]. 366 ingress tunnel endpoint (ITE) 367 an OMNI interface endpoint that injects encapsulated packets into 368 an OMNI link. 370 egress tunnel endpoint (ETE) 371 an OMNI interface endpoint that receives encapsulated packets from 372 an OMNI link. 374 link-layer address 375 an IP address used as an encapsulation header source or 376 destination address from the perspective of the OMNI interface. 377 When an upper layer protocol (e.g., UDP) is used as part of the 378 encapsulation, the port number is also considered as part of the 379 link-layer address. 381 network layer address 382 the source or destination address of an encapsulated IP packet 383 presented to the OMNI interface. 385 end user network (EUN) 386 an internal virtual or external edge IP network that an AERO 387 Client or Relay connects to the rest of the network via the OMNI 388 interface. The Client/Relay sees each EUN as a "downstream" 389 network, and sees the OMNI interface as the point of attachment to 390 the "upstream" network. 392 Mobile Node (MN) 393 an AERO Client and all of its downstream-attached networks that 394 move together as a single unit, i.e., an end system that connects 395 an Internet of Things. 397 Mobile Router (MR) 398 a MN's on-board router that forwards packets between any 399 downstream-attached networks and the OMNI link. 401 Route Optimization Source (ROS) 402 the AERO node nearest the source that initiates route 403 optimization. The ROS may be a Server or Proxy acting on behalf 404 of the source Client. 406 Route Optimization responder (ROR) 407 the AERO node nearest the target destination that responds to 408 route optimization requests. The ROR may be a Server acting on 409 behalf of a target MNP Client, or a Relay for a non-MNP 410 destination. 412 MAP List 413 a geographically and/or topologically referenced list of addresses 414 of all Servers within the same OMNI link. There is a single MAP 415 list for the entire OMNI link. 417 Distributed Mobility Management (DMM) 418 a BGP-based overlay routing service coordinated by Servers and 419 Bridges that tracks all Server-to-Client associations. 421 Mobility Service (MS) 422 the collective set of all Servers, Proxys, Bridges and Relays that 423 provide the AERO Service to Clients. 425 Mobility Service Endpoint MSE) 426 an individual Server, Proxy, Bridge or Relay in the Mobility 427 Service. 429 Throughout the document, the simple terms "Client", "Server", 430 "Bridge", "Proxy" and "Relay" refer to "AERO Client", "AERO Server", 431 "AERO Bridge", "AERO Proxy" and "AERO Relay", respectively. 432 Capitalization is used to distinguish these terms from other common 433 Internetworking uses in which they appear without capitalization. 435 The terminology of DHCPv6 [RFC8415] and IPv6 ND [RFC4861] (including 436 the names of node variables, messages and protocol constants) is used 437 throughout this document. The terms "All-Routers multicast", "All- 438 Nodes multicast", "Solicited-Node multicast" and "Subnet-Router 439 anycast" are defined in [RFC4291]. Also, the term "IP" is used to 440 generically refer to either Internet Protocol version, i.e., IPv4 441 [RFC0791] or IPv6 [RFC8200]. 443 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 444 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 445 "OPTIONAL" in this document are to be interpreted as described in BCP 446 14 [RFC2119][RFC8174] when, and only when, they appear in all 447 capitals, as shown here. 449 3. Asymmetric Extended Route Optimization (AERO) 451 The following sections specify the operation of IP over OMNI links 452 using the AERO service: 454 3.1. AERO Node Types 456 AERO Bridges provide hybrid routing/bridging services (as well as a 457 security trust anchor) for nodes on an OMNI link. Bridges use 458 standard IPv6 routing to forward packets both within the same INET 459 partitions and between disjoint INET partitions based on a mid-layer 460 IPv6 encapsulation per [RFC2473]. The inner IP layer experiences a 461 virtual bridging service since the inner IP TTL/Hop Limit is not 462 decremented during forwarding. Each Bridge also peers with Servers 463 and other Bridges in a dynamic routing protocol instance to provide a 464 Distributed Mobility Management (DMM) service for the list of active 465 MNPs (see Section 3.2.3). Bridges present the OMNI link as a set of 466 one or more Mobility Service Prefixes (MSPs) and configure secured 467 tunnels with Servers, Relays, Proxys and other Bridges; they further 468 maintain IP forwarding table entries for each Mobile Network Prefix 469 (MNP) and any other reachable non-MNP prefixes. 471 AERO Servers provide default forwarding and mobility/multilink 472 services for AERO Client Mobile Nodes (MNs). Each Server also peers 473 with Bridges in a dynamic routing protocol instance to advertise its 474 list of associated MNPs (see Section 3.2.3). Servers facilitate PD 475 exchanges with Clients, where each delegated prefix becomes an MNP 476 taken from an MSP. Servers forward packets between OMNI interface 477 neighbors and track each Client's mobility profiles. 479 AERO Clients register their MNPs through PD exchanges with AERO 480 Servers over the OMNI link, and distribute the MNPs to nodes on EUNs. 481 A Client may also be co-resident on the same physical or virtual 482 platform as a Server; in that case, the Client and Server behave as a 483 single functional unit. 485 AERO Proxys provide a conduit for ANET Clients to associate with 486 Servers in external INETs. Client and Servers exchange control plane 487 messages via the Proxy acting as a bridge between the ANET/INET 488 boundary. The Proxy forwards data packets between Clients and the 489 OMNI link according to forwarding information in the neighbor cache. 490 The Proxy function is specified in Section 3.13. 492 AERO Relays are Servers that provide forwarding services between the 493 OMNI interface and INET/EUN interfaces. Relays are provisioned with 494 MNPs the same as for an AERO Client, and also run a dynamic routing 495 protocol to discover any non-MNP IP routes. The Relay advertises the 496 MSP(s) to its connected networks, and distributes all of its 497 associated MNPs and non-MNP IP routes via BGP peerings with Bridges. 499 AERO Bridges, Servers, Proxys and Relays are critical infrastructure 500 elements in fixed (i.e., non-mobile) INET deployments and hence have 501 permanent and unchanging INET addresses. AERO Clients are MNs that 502 connect via underlying interfaces with addresses that may change when 503 the Client moves to a new network connection point. 505 3.2. The AERO Service over OMNI Links 507 3.2.1. AERO/OMNI Reference Model 509 Figure 1 presents the basic OMNI link reference model: 511 +----------------+ 512 | AERO Bridge B1 | 513 | Nbr: S1, S2, P1| 514 |(X1->S1; X2->S2)| 515 | MSP M1 | 516 +-+---------+--+-+ 517 +--------------+ | Secured | | +--------------+ 518 |AERO Server S1| | tunnels | | |AERO Server S2| 519 | Nbr: C1, B1 +-----+ | +-----+ Nbr: C2, B1 | 520 | default->B1 | | | default->B1 | 521 | X1->C1 | | | X2->C2 | 522 +-------+------+ | +------+-------+ 523 | OMNI link | | 524 X===+===+===================+==)===============+===+===X 525 | | | | 526 +-----+--------+ +--------+--+-----+ +--------+-----+ 527 |AERO Client C1| | AERO Proxy P1 | |AERO Client C2| 528 | Nbr: S1 | |(Proxy Nbr Cache)| | Nbr: S2 | 529 | default->S1 | +--------+--------+ | default->S2 | 530 | MNP X1 | | | MNP X2 | 531 +------+-------+ .--------+------. +-----+--------+ 532 | (- Proxyed Clients -) | 533 .-. `---------------' .-. 534 ,-( _)-. ,-( _)-. 535 .-(_ IP )-. +-------+ +-------+ .-(_ IP )-. 536 (__ EUN )--|Host H1| |Host H2|--(__ EUN ) 537 `-(______)-' +-------+ +-------+ `-(______)-' 539 Figure 1: AERO/OMNI Reference Model 541 In this model: 543 o the OMNI link is an overlay network service configured over one or 544 more underlying INET partitions which may be managed by different 545 administrative authorities and have incompatible protocols and/or 546 addressing plans. 548 o AERO Bridge B1 aggregates Mobility Service Prefix (MSP) M1, 549 discovers Mobile Network Prefixes (MNPs) X* and advertises the MSP 550 via BGP peerings over secured tunnels to Servers (S1, S2). 551 Bridges connect the disjoint segments of a partitioned OMNI link. 553 o AERO Servers/Relays S1 and S2 configure secured tunnels with 554 Bridge B1 and also provide mobility, multilink and default router 555 services for their associated Clients C1 and C2. 557 o AERO Clients C1 and C2 associate with Servers S1 and S2, 558 respectively. They receive Mobile Network Prefix (MNP) 559 delegations X1 and X2, and also act as default routers for their 560 associated physical or internal virtual EUNs. Simple hosts H1 and 561 H2 attach to the EUNs served by Clients C1 and C2, respectively. 563 o AERO Proxy P1 configures a secured tunnel with Bridge B1 and 564 provides proxy services for AERO Clients in secured enclaves that 565 cannot associate directly with other OMNI link neighbors. 567 An OMNI link configured over a single INET appears as a single 568 unified link with a consistent underlying network addressing plan. 569 In that case, all nodes on the link can exchange packets via simple 570 INET encapsulation, since the underlying INET is connected. In 571 common practice, however, an OMNI link may be partitioned into 572 multiple "segments", where each segment is a distinct INET 573 potentially managed under a different administrative authority (e.g., 574 as for worldwide aviation service providers such as ARINC, SITA, 575 Inmarsat, etc.). Individual INETs may also themselves be partitioned 576 internally, in which case each internal partition is seen as a 577 separate segment. 579 The addressing plan of each segment is consistent internally but will 580 often bear no relation to the addressing plans of other segments. 581 Each segment is also likely to be separated from others by network 582 security devices (e.g., firewalls, proxies, packet filtering 583 gateways, etc.), and in many cases disjoint segments may not even 584 have any common physical link connections. Therefore, nodes can only 585 be assured of exchanging packets directly with correspondents in the 586 same segment, and not with those in other segments. The only means 587 for joining the segments therefore is through inter-domain peerings 588 between AERO Bridges. 590 The same as for traditional campus LANs, multiple OMNI link segments 591 can be joined into a single unified link via a virtual bridging 592 service using a mid-layer IPv6 encpasulation per [RFC2473] known as 593 the "SPAN header" that supports inter-segment forwarding (i.e., 594 bridging) without decrementing the network-layer TTL/Hop Limit. This 595 bridging of OMNI link segments is shown in Figure 2: 597 . . . . . . . . . . . . . . . . . . . . . . . 598 . . 599 . .-(::::::::) . 600 . .-(::::::::::::)-. +-+ . 601 . (:::: Segment A :::)--|B|---+ . 602 . `-(::::::::::::)-' +-+ | . 603 . `-(::::::)-' | . 604 . | . 605 . .-(::::::::) | . 606 . .-(::::::::::::)-. +-+ | . 607 . (:::: Segment B :::)--|B|---+ . 608 . `-(::::::::::::)-' +-+ | . 609 . `-(::::::)-' | . 610 . | . 611 . .-(::::::::) | . 612 . .-(::::::::::::)-. +-+ | . 613 . (:::: Segment C :::)--|B|---+ . 614 . `-(::::::::::::)-' +-+ | . 615 . `-(::::::)-' | . 616 . | . 617 . ..(etc).. x . 618 . . 619 . . 620 . <- OMNI link Bridged by encapsulation -> . 621 . . . . . . . . . . . . . .. . . . . . . . . 623 Figure 2: Bridging OMNI Link Segments 625 Bridges, Servers, Relays and Proxys connect via secured INET tunnels 626 over their respecitve segments in a spanning tree topology rooted at 627 the Bridges. The secured spanning tree supports strong 628 authentication for IPv6 ND control messages and may also be used to 629 convey the initial data packets in a flow. Route optimization can 630 then be employed to cause data packets to take more direct paths 631 between OMNI link neighbors without having to strictly follow the 632 spanning tree. 634 3.2.2. Link-Local Addresses (LLAs) and Unique Local Addresses (ULAs) 636 AERO nodes on OMNI links use the Link-Local Address (LLA) prefix 637 fe80::/10 [RFC4193] to assign LLAs used for network-layer addresses 638 in IPv6 ND and data messages. They also use the Unique Local Address 639 (ULA) prefix fc80::/10 [RFC4193] to form ULAs used for SPAN header 640 source and desitnation addresses. See 641 [I-D.templin-6man-omni-interface] for a full specification of the 642 LLAs and ULAs used by AERO nodes on OMNI links. 644 For routing system organization (see Section 3.2.3), ULAs are 645 organized in partition prefixes, e.g., fc80::1000/116. For each such 646 partition prefix, the Bridge(s) that connect that segment assign the 647 all-zero's address of the prefix as a Subnet Router Anycast address. 648 For example, the Subnet Router Anycast address for fc80::1000/116 is 649 simply fc80::1000. 651 3.2.3. AERO Routing System 653 The AERO routing system comprises a private instance of the Border 654 Gateway Protocol (BGP) [RFC4271] that is coordinated between Bridges 655 and Servers and does not interact with either the public Internet BGP 656 routing system or any underlying INET routing systems. 658 In a reference deployment, each Server is configured as an Autonomous 659 System Border Router (ASBR) for a stub Autonomous System (AS) using 660 an AS Number (ASN) that is unique within the BGP instance, and each 661 Server further uses eBGP to peer with one or more Bridges but does 662 not peer with other Servers. Each INET of a multi-segment OMNI link 663 must include one or more Bridges, which peer with the Servers and 664 Proxys within that INET. All Bridges within the same INET are 665 members of the same hub AS using a common ASN, and use iBGP to 666 maintain a consistent view of all active MNPs currently in service. 667 The Bridges of different INETs peer with one another using eBGP. 669 Bridges advertise the OMNI link's MSPs and any non-MNP routes to each 670 of their Servers. This means that any aggregated non-MNPs (including 671 "default") are advertised to all Servers. Each Bridge configures a 672 black-hole route for each of its MSPs. By black-holing the MSPs, the 673 Bridge will maintain forwarding table entries only for the MNPs that 674 are currently active, and packets destined to all other MNPs will 675 correctly incur Destination Unreachable messages due to the black- 676 hole route. In this way, Servers have only partial topology 677 knowledge (i.e., they know only about the MNPs of their directly 678 associated Clients) and they forward all other packets to Bridges 679 which have full topology knowledge. 681 Each OMNI link segment assigns a unique sub-prefix of fc80::/96 known 682 as the ULA partition prefix. For example, a first segment could 683 assign fc80::1000/116, a second could assign fc80::2000/116, a third 684 could assign fc80::3000/116, etc. The administrative authorities for 685 each segment must therefore coordinate to assure mutually-exclusive 686 partiton prefix assignments, but internal provisioning of each prefix 687 is an independent local consideration for each administrative 688 authority. 690 ULA partition prefixes are statitcally represented in Bridge 691 forwarding tables. Bridges join multiple segments into a unified 692 OMNI link over multiple diverse administrative domains. They support 693 a bridging function by first establishing forwarding table entries 694 for their partiion prefixes either via standard BGP routing or static 695 routes. For example, if three Bridges ('A', 'B' and 'C') from 696 different segments serviced fc80::1000/116, fc80::2000/116 and 697 fc80::3000/116 respectively, then the forwarding tables in each 698 Bridge are as follows: 700 A: fc80::1000/116->local, fc80::2000/116->B, fc80::3000/116->C 702 B: fc80::1000/116->A, fc80::2000/116->local, fc80::3000/116->C 704 C: fc80::1000/116->A, fc80::2000/116->B, fc80::3000/116->local 706 These forwarding table entries are permanent and never change, since 707 they correspond to fixed infrastructure elements in their respective 708 segments. 710 ULA Client prefixes are instead dynamically advertised in the AERO 711 routing system by Servers and Relays that provide service for their 712 corresponding MNPs. For example, if three Servers ('D', 'E' and 'F') 713 service the MNPs 2001:db8:1000:2000::/56, 2001:db8:3000:4000::/56 and 714 2001:db8:5000:6000::/56 then the routing system would include: 716 D: fc80:2001:db8:1000:2000::/72 718 E: fc80:2001:db8:3000:4000::/72 720 F: fc80:2001:db8:5000:6000::/72 722 A full discussion of the BGP-based routing system used by AERO is 723 found in [I-D.ietf-rtgwg-atn-bgp]. 725 3.2.4. AERO Encapsulation 727 With the Client and partition prefixes in place in each Bridge's 728 forwarding table, control and data packets sent between AERO nodes in 729 different segments can therefore be carried over the via mid-layer 730 encapsulation using the SPAN header. For example, when a source AERO 731 node forwards a packet with IPv6 address 2001:db8:1:2::1 to a target 732 AERO node with IPv6 address 2001:db8:1000:2000::1, it first 733 encapsulates the packet in a SPAN header with source address set to 734 fc80:2001:db8:1:2:: and destination address set to 735 fc80:2001:db8:1000:2000::. Next, it encapsulates the resulting SPAN 736 packet in an INET header with source address set to its own INET 737 address (e.g., 192.0.2.100) and destination set to the INET address 738 of a Bridge (e.g., 192.0.2.1). 740 SPAN encapsulation is based on Generic Packet Tunneling in IPv6 741 [RFC2473]; the encapsulation format in the above example is shown in 742 Figure 3: 744 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 745 | INET Header | 746 | src = 192.0.2.100 | 747 | dst = 192.0.2.1 | 748 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 749 | SPAN Header | 750 | src = fc80:2001:db8:1:2:: | 751 | dst=fc80:2001:db8:1000:2000:: | 752 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 753 | Inner IP Header | 754 | src = 2001:db8:1:2::1 | 755 | dst = 2001:db8:1000:2000::1 | 756 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 757 | | 758 ~ ~ 759 ~ Inner Packet Body ~ 760 ~ ~ 761 | | 762 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 764 Figure 3: SPAN Encapsulation 766 In this format, the inner IP header and packet body are the original 767 IP packet, the SPAN header is an IPv6 header prepared according to 768 [RFC2473], and the INET header is prepared as discussed in 769 Section 3.6. 771 This gives rise to a routing system that contains both Client prefix 772 routes that may change dynamically due to regional node mobility and 773 partion prefix routes that never change. The Bridges can therefore 774 provide link-layer bridging by sending packets over the spanning tree 775 instead of network-layer routing according to MNP routes. As a 776 result, opportunities for packet loss due to node mobility between 777 different segments are mitigated. 779 In normal operations, IPv6 ND messages are conveyed over secured 780 paths between OMNI link neighbors so that specific Proxys, Servers or 781 Relays can be addressed without being subject to mobility events. 782 Conversely, only the first few packets destined to Clients need to 783 traverse secured paths until route optimization can determine a more 784 direct path. 786 3.2.5. Segment Routing Topologies (SRTs) 788 The 16-bit sub-prefixes of fc80::/10 identify up to 64 distinct 789 Segment Routing Topologies (SRTs). Each SRT is a mutually-exclusive 790 OMNI link overlay instance using a mutually-exclusive set of ULAs, 791 and emulates a Virtual LAN (VLAN) service for the OMNI link. In some 792 cases (e.g., when redundant topologies are needed for fault tolerance 793 and reliability) it may be beneficial to deploy multiple SRTs that 794 act as independent overlay instances. A communication failure in one 795 instance therefore will not affect communications in other instances. 797 Each SRT is identified by a distinct value in bits 10-15 of fc80::10, 798 i.e., as fc80::/16, fc81::/16, fc82::/16, etc. This document asserts 799 that up to four SRTs provide a level of safety sufficient for 800 critical communications such as civil aviation. Each SRT is 801 designated with a color that identifies a different OMNI link 802 instance as follows: 804 o Red - corresponds to fc80::/16 806 o Green - corresponds to fc81::/16 808 o Blue-1 - corresponds to fc82::/16 810 o Blue-2 - corresponds to fc83::/16 812 o fc84::/16 through fcbf::/16 are available for additional SRTs. 814 Each OMNI interface assigns an anycast ULA corresponding to its SRT 815 prefix. For example, the anycast ULA for the Green SRT is simply 816 fc81::. The anycast ULA is used for OMNI interface determination in 817 Safety-Based Multilink (SBM) as discussed in 818 [I-D.templin-6man-omni-interface]. Each OMNI interface further 819 applies Performance-Based Multilink (PBM) internally. 821 3.2.6. Segment Routing To the OMNI Link 823 An original IPv6 source can direct a packet to an OMNI link Client by 824 including a Segment Routing Header (SRH) with the anycast ULA for the 825 selected SRT as either the IPv6 destination or as an intermediate hop 826 within the SRH. This allows the original source to determine the 827 specific topology a packet will traverse when there may be multiple 828 alternatives to choose from. Since the SRH contains no useful 829 information for the destination, the Client may elect to delete the 830 SRH before forwarding in order to reduce overhead. This form of 831 Segment Routing supports Safety-Based Multilink (SBM), and can be 832 exercised through general-purpose SRH types such as [RFC8754]. 834 3.2.7. Segment Routing Within the OMNI Link 836 AERO nodes that insert a SPAN header can use Segment Routing within 837 the OMNI link when necessary to influence the path of packets 838 destined to targets in remote segments without requiring all packets 839 to traverse strict spanning tree paths. 841 When a Client, Proxy or Server has a packet to send to a target 842 discovered through route optimization located in the same OMNI link 843 segment, it encapsulates the packet in a SPAN header with the ULA of 844 the target as the destination address if fragmentation is necessary; 845 otherwise, it may omit the SPAN header. The node then uses the 846 target's Link Layer Address (L2ADDR) information for INET 847 encapsulation without including an SRH. 849 When a Client, Proxy or Server has a packet to send to a route 850 optimization target located in a remote OMNI link segment, it 851 encapsulates the packet in a SPAN header with its own ULA as the 852 source address. The node then SHOULD include an SRH [RFC8754] while 853 forwarding the packet to a Bridge. 855 When the SRH is omitted, the node sets the destination address to the 856 ULA of the target Client/Proxy/Server and packet forwarding is via 857 spanning tree paths. When the SRH is included, the node first sets 858 the destination address to the ULA Subnet Router Anycast address of 859 the remote segment and sets the ULA of the target's Proxy/Server as 860 the Last Hop Segment (LHS) ID. The node also includes an AERO Route 861 Optimization specification in the SRH TLV section as shown in 862 Figure 4: 864 0 1 2 3 865 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 1 866 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 867 | Type=TBD | Length | MNPlen|V| FMT | MNP[1] | 868 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 869 | MNP[2] | MNP[3] | ... | MNP[i] | 870 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 871 ~ Link Layer Address (L2ADDR) ~ 872 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 874 Figure 4: AERO Route Optimization SRH TLV 876 In this format: 878 o Type is TBD to be assigned according to the Segment Routing Header 879 TLV registry [RFC8754]. 881 o Length is the length of the body of the TLV in bytes, excluding 882 the Type and Length fields. 884 o MNPlen encodes a value 'i' (between 0 and 15) that indicates the 885 number of octets of the IPv4/IPv6 MNP prefix that follows. 887 o V indicates the IP protocol version of the MNP that follows. V is 888 set to 0 for IPv4 or 1 for IPv6. 890 o FMT is a three bit code that determines the context and format of 891 the L2ADDR exactly as specified in Figure 5. 893 o MNP{1], MNP[2], etc. up to MNP[i] encode the leading 'i' octets of 894 the MNP, beginning with the most significant octet followed by the 895 next most significant octet, etc. The number of MNP octets to be 896 included is determined by the number of trailing zero octets in 897 the prefix. For example, for the IPv6 MNP 2001:db8:1:2::/64, 'i' 898 is set to 8 and only the leftmost 8 octets of the MNP are 899 included. In the same way, for the IPv4 MNP 192.0.2/24, 'i' is 900 set to 3 and only the leftmost 3 octets of the MNP are included. 902 o Link Layer Address (L2ADDR) is a UDP Port Number and IP address 903 encoded according to FMT exactly as specified in Figure 5. 905 The node then forwards the packet via a local Bridge, which will 906 eventually direct it to a Bridge on the same segment as the target. 908 When a Bridge receives a packet with Segments Left=1 and with LHS ID 909 on a local segment, it checks to see if there is an AERO Route 910 Optimization TLV. If so, the Bridge creates a ULA destination 911 according to FMT. If FMT indicates that L2ADDR corresponds to a 912 target Proxy/Server, the Bridge concatenates the SRT ::/16 prefix 913 with the LHS ID to form the ULA destination. Otherwise, the Bridge 914 concatenates the SRT ::/16 prefix with the leading MNPlen octets of 915 the MNP and sets the remaining rightmost bits to 0 to form a Subnet 916 Router Anycast ULA destination. The Bridge then writes the ULA into 917 the SPAN header destination address and encapsulates the packet in an 918 INET header with the target's L2ADDR as the destination then forwards 919 the packet. Since the SRH contains no useful information for the 920 destination, the Bridge may elect to delete the SRH before forwarding 921 in order to reduce overhead. 923 In this way, the Bridge participates in route optimization to reduce 924 traffic load and suboptimal routing through strict spanning tree 925 paths. Note that if the Bridge does not recognize the AERO Route 926 Optimization TLV, it instead places the LHS ID concatentaed with the 927 SRT ::/16 prefix in the IPv6 destination address and forwards 928 according to the spanning tree. (Note that this is the same behavior 929 that would occur if the AERO Route Optimization TLV were not 930 present). 932 3.2.8. Segment Routing Header Compression 934 In the Segment Routing use cases discussed above, the segment routing 935 headers must be kept to a minimum size since source and target 936 Clients may be located behind low-end wireless links (e.g., 1Mbps or 937 less). The Compressed Routing Header (CRH) 938 [I-D.bonica-6man-comp-rtg-hdr] provides a compact form that reduces 939 the header size by omitting invariant information. The CRH Helper 940 option [I-D.bonica-6man-crh-helper-opt] can be used to encode the 941 AERO Route Optimization TLV, and the final hop Bridge that performs 942 route optimization may remove the CRH and its helper before 943 encapsulating and forwarding to the target. 945 The CRH and its companion helper option are therefore seen as 946 critical architectural elements that should be quickly progressed 947 through the standards process. Implementations SHOULD use the CRH 948 and its companion helper option instead of other Routing Header types 949 whenever possible to conserve bandwidth. 951 3.3. OMNI Interface Characteristics 953 OMNI interfaces are virtual interfaces configured over one or more 954 underlying interfaces classified as follows: 956 o INET interfaces connect to an INET either natively or through one 957 or several IPv4 Network Address Translators (NATs). Native INET 958 interfaces have global IP addresses that are reachable from any 959 INET correspondent. All Server, Relay and Bridge interfaces are 960 native interfaces, as are INET-facing interfaces of Proxys. NATed 961 INET interfaces connect to a private network behind one or more 962 NATs that provide INET access. Clients that are behind a NAT are 963 required to send periodic keepalive messages to keep NAT state 964 alive when there are no data packets flowing. 966 o Proxyed interfaces connect to an ANET that is separated from the 967 open INET by a Proxy. Proxys can actively issue control messages 968 over the INET on behalf of the Client to reduce ANET congestion. 970 o VPNed interfaces use security encapsulation over the INET to a 971 Virtual Private Network (VPN) server that also acts as a Server or 972 Proxy. Other than the link-layer encapsulation format, VPNed 973 interfaces behave the same as Direct interfaces. 975 o Direct interfaces connect a Client directly to a Server or Proxy 976 without crossing any ANET/INET paths. An example is a line-of- 977 sight link between a remote pilot and an unmanned aircraft. The 978 same Client considerations apply as for VPNed interfaces. 980 o In all cases, Clients examine the P flag in received RAs. If the 981 P flag is 1, the node that returned the RA message is acting as a 982 Proxy; otherwise, it is acting as a Server. 984 OMNI interfaces use SPAN encapsulation as necessary as discussed in 985 Section 3.2.4. OMNI interfaces use link-layer encapsulation (see: 986 Section 3.6) to exchange packets with OMNI link neighbors over INET 987 or VPNed interfaces. OMNI interfaces do not use link-layer 988 encapsulation over Proxyed and Direct underlying interfaces. 990 OMNI interfaces maintain a neighbor cache for tracking per-neighbor 991 state the same as for any interface. OMNI interfaces use ND messages 992 including Router Solicitation (RS), Router Advertisement (RA), 993 Neighbor Solicitation (NS) and Neighbor Advertisement (NA) for 994 neighbor cache management. 996 OMNI interfaces send ND messages with an OMNI option formatted as 997 specified in [I-D.templin-6man-omni-interface]. The OMNI option 998 includes prefix registration information and "ifIndex-tuples" 999 containing link information parameters for the OMNI interface's 1000 underlying interfaces. 1002 SPAN-encapsulated OMNI interface ND messages also include a Source/ 1003 Target Link-Layer Address Option (S/TLLAO) formatted as shown in 1004 Figure 5: 1006 0 1 2 3 1007 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 1008 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1009 | Type | Length | ifIndex[1] | SRT | FMT | 1010 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1011 | Last Hop Segment (LHS) ID [1] | 1012 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1013 ~ Link Layer Address (L2ADDR) [1] ~ 1014 ~ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1015 ~ | ifIndex[2] | SRT | FMT | 1016 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1017 | Last Hop Segment (LHS) ID [2] | 1018 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1019 ~ Link Layer Address (L2ADDR) [2] ~ 1020 ~ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1021 ~ | .... ~ 1022 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ~ 1023 ~ ... ~ 1024 ~ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1025 ~ | ifIndex[N] | SRT | FMT | 1026 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1027 | Last Hop Segment (LHS) ID [N] | 1028 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1029 ~ Link Layer Address (L2ADDR) [N] ~ 1030 ~ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1031 ~ | Zero Padding (if necessary) ... 1032 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+- 1034 Figure 5: OMNI Source/Target Link-Layer Address Option (S/TLLAO) 1035 Format 1037 In this format, Type and Length are set the same as specified for S/ 1038 TLLAOs in [RFC4861], with trailing zero padding octets added as 1039 necessary to produce an integral number of 8 octet blocks. The S/ 1040 TLLAO includes N ifIndex-tuples corresponding to a proper subset of 1041 the ifIndex-tuples that appear in the OMNI option. Each ifIndex- 1042 tuple includes the following information: 1044 o ifIndex - the same value as in the corresponding ifIndex-tuple 1045 included in the OMNI option. 1047 o SRT - a 5-bit value that when added to 96 determines the prefix 1048 length to apply to the ULA formed from concatenating the SRT ::/16 1049 prefix with Last Hop Segment (LHS) ID. For example, the prefix 1050 length for the value 16 is 112. 1052 o FMT - a 3-bit "Framework/Mode/Type" code corresponding to the 1053 included Link Layer Address (L2ADDR) as follows: 1055 o 1057 * When the most significant bit (i.e., "Framework") is set to 0, 1058 L2ADDR is the INET encapsulation address of a Proxy/Server; 1059 otherwise, it is the addresss for the Source/Target itself 1061 * When the next most significant bit (i.e., "Mode") is set to 0, 1062 the Source/Target L2ADDR is on the open INET; otherwise, it is 1063 (likely) located behind a NAT. 1065 * When the least significant bit (i.e., "Type") is set to 0, 1066 L2ADDR is an IPv4 address; else, it is an IPv6 address. 1068 o Last Hop Segment (LHS) ID - Includes the least significant 32 bits 1069 of the last hop Proxy/Server ULA prior to encapsulation according 1070 to L2ADDR. When SRT and LHS are both set to 0, the last hop 1071 Proxy/Server ULA is considered unspecified in this IPv6 ND 1072 message. 1074 o Link Layer Address (L2ADDR) - Included according to FMT, and 1075 identifies the link-layer address (i.e., the encapsulation 1076 address) of the source/target. The Port Number and IP address are 1077 recorded in ones-compliment "obfuscated" form per [RFC4380]. 1079 If an S/TLLAO is included, any ifIndex-tuples correspond to a proper 1080 subset of the OMNI option ifIndex-tuples. Any S/TLLAO ifIndex-tuple 1081 with an ifIndex value that does not appear in an OMNI option ifindex- 1082 tuple is ignored. If the same ifIndex value appears in multiple 1083 ifIndex-tuples, the first tuple is processed and the remaining tuples 1084 are ignored. Any S/TLLAO ifIndex-tuples can therefore be viewed as 1085 extensions of their corresponding OMNI option ifIndex-tuples, i.e., 1086 the OMNI option and S/TLLAO are companions that are interpreted in 1087 conjunction with each other. 1089 A Client's OMNI interface may be configured over multiple underlying 1090 interface connections. For example, common mobile handheld devices 1091 have both wireless local area network ("WLAN") and cellular wireless 1092 links. These links are often used "one at a time" with low-cost WLAN 1093 preferred and highly-available cellular wireless as a standby, but a 1094 simultaneous-use capability could provide benefits. In a more 1095 complex example, aircraft frequently have many wireless data link 1096 types (e.g. satellite-based, cellular, terrestrial, air-to-air 1097 directional, etc.) with diverse performance and cost properties. 1099 If a Client's multiple underlying interfaces are used "one at a time" 1100 (i.e., all other interfaces are in standby mode while one interface 1101 is active), then ND message OMNI options include only a single 1102 ifIndex-tuple set to constant values. In that case, the Client would 1103 appear to have a single interface but with a dynamically changing 1104 link-layer address. 1106 If the Client has multiple active underlying interfaces, then from 1107 the perspective of ND it would appear to have multiple link-layer 1108 addresses. In that case, ND message OMNI options MAY include 1109 multiple ifIndex-tuples - each with values that correspond to a 1110 specific interface. Every ND message need not include all OMNI and/ 1111 or S/TLLAO ifIndex-tuples; for any ifIndex-tuple not included, the 1112 neighbor considers the status as unchanged. 1114 Bridge, Server and Proxy OMNI interfaces may be configured over one 1115 or more secured tunnel interfaces. The OMNI interface configures 1116 both an LLA and its corresponding ULA, while the underlying secured 1117 tunnel interfaces are either unnumbered or configure the same ULA. 1118 The OMNI interface encapsulates each IP packet in a SPAN header and 1119 presents the packet to the underlying secured tunnel interface. 1120 Routing protocols such as BGP that run over the OMNI interface do not 1121 employ SPAN encapsulation, but rather present the routing protocol 1122 messages directly to the underlying secured tunnels while using the 1123 ULA as the source address. This distinction must be honored 1124 consistently according to each node's configuration so that the IP 1125 forwarding table will associate discovered IP routes with the correct 1126 interface. 1128 3.4. OMNI Interface Initialization 1130 AERO Servers, Proxys and Clients configure OMNI interfaces as their 1131 point of attachment to the OMNI link. AERO nodes assign the MSPs for 1132 the link to their OMNI interfaces (i.e., as a "route-to-interface") 1133 to ensure that packets with destination addresses covered by an MNP 1134 not explicitly assigned to a non-OMNI interface are directed to the 1135 OMNI interface. 1137 OMNI interface initialization procedures for Servers, Proxys, Clients 1138 and Bridges are discussed in the following sections. 1140 3.4.1. AERO Server/Relay Behavior 1142 When a Server enables an OMNI interface, it assigns an LLA/ULA 1143 appropriate for the given OMNI link segment. The Server also 1144 configures secured tunnels with one or more neighboring Bridges and 1145 engages in a BGP routing protocol session with each Bridge. 1147 The OMNI interface provides a single interface abstraction to the IP 1148 layer, but internally comprises multiple secured tunnels as well as 1149 an NBMA nexus for sending encapsulated data packets to OMNI interface 1150 neighbors. The Server further configures a service to facilitate ND/ 1151 PD exchanges with AERO Clients and manages per-Client neighbor cache 1152 entries and IP forwarding table entries based on control message 1153 exchanges. 1155 Relays are simply Servers that run a dynamic routing protocol to 1156 redistribute routes between the OMNI interface and INET/EUN 1157 interfaces (see: Section 3.2.3). The Relay provisions MNPs to 1158 networks on the INET/EUN interfaces (i.e., the same as a Client would 1159 do) and advertises the MSP(s) for the OMNI link over the INET/EUN 1160 interfaces. The Relay further provides an attachment point of the 1161 OMNI link to a non-MNP-based global topology. 1163 3.4.2. AERO Proxy Behavior 1165 When a Proxy enables an OMNI interface, it assigns an LLA/ULA and 1166 configures permanent neighbor cache entries the same as for Servers. 1167 The Proxy also configures secured tunnels with one or more 1168 neighboring Bridges and maintains per-Client neighbor cache entries 1169 based on control message exchanges. 1171 3.4.3. AERO Client Behavior 1173 When a Client enables an OMNI interface, it sends RS messages with 1174 ND/PD parameters over its underlying interfaces to a Server in the 1175 MAP list, which returns an RA message with corresponding parameters. 1176 (The RS/RA messages may pass through a Proxy in the case of a 1177 Client's Proxyed interface, or through one or more NATs in the case 1178 of a Client's INET interface.) 1180 3.4.4. AERO Bridge Behavior 1182 AERO Bridges configure an OMNI interface and assign the ULA Subnet 1183 Router Anycast address for each OMNI link segment they connect to. 1184 Bridges configure secured tunnels with Servers, Proxys and other 1185 Bridges; they also configure LLAs/ULAs and permanent neighbor cache 1186 entries the same as Servers. Bridges engage in a BGP routing 1187 protocol session with a subset of the Servers and other Bridges on 1188 the spanning tree (see: Section 3.2.3). 1190 3.5. OMNI Interface Neighbor Cache Maintenance 1192 Each OMNI interface maintains a conceptual neighbor cache that 1193 includes an entry for each neighbor it communicates with on the OMNI 1194 link per [RFC4861]. OMNI interface neighbor cache entries are said 1195 to be one of "permanent", "symmetric", "asymmetric" or "proxy". 1197 Permanent neighbor cache entries are created through explicit 1198 administrative action; they have no timeout values and remain in 1199 place until explicitly deleted. AERO Bridges maintain permanent 1200 neighbor cache entries for their associated Proxys and Servers (and 1201 vice-versa). Each entry maintains the mapping between the neighbor's 1202 network-layer LLA and corresponding INET address. 1204 Symmetric neighbor cache entries are created and maintained through 1205 RS/RA exchanges as specified in Section 3.12, and remain in place for 1206 durations bounded by ND/PD lifetimes. AERO Servers maintain 1207 symmetric neighbor cache entries for each of their associated 1208 Clients, and AERO Clients maintain symmetric neighbor cache entries 1209 for each of their associated Servers. The list of all Servers on the 1210 OMNI link is maintained in the link's MAP list. 1212 Asymmetric neighbor cache entries are created or updated based on 1213 route optimization messaging as specified in Section 3.14, and are 1214 garbage-collected when keepalive timers expire. AERO ROSs maintain 1215 asymmetric neighbor cache entries for active targets with lifetimes 1216 based on ND messaging constants. Asymmetric neighbor cache entries 1217 are unidirectional since only the ROS (and not the ROR) creates an 1218 entry. 1220 Proxy neighbor cache entries are created and maintained by AERO 1221 Proxys when they process Client/Server ND/PD exchanges, and remain in 1222 place for durations bounded by ND/PD lifetimes. AERO Proxys maintain 1223 proxy neighbor cache entries for each of their associated Clients. 1224 Proxy neighbor cache entries track the Client state and the address 1225 of the Client's associated Server(s). 1227 To the list of neighbor cache entry states in Section 7.3.2 of 1228 [RFC4861], Proxy and Server OMNI interfaces add an additional state 1229 DEPARTED that applies to symmetric and proxy neighbor cache entries 1230 for Clients that have recently departed. The interface sets a 1231 "DepartTime" variable for the neighbor cache entry to "DEPART_TIME" 1232 seconds. DepartTime is decremented unless a new ND message causes 1233 the state to return to REACHABLE. While a neighbor cache entry is in 1234 the DEPARTED state, packets destined to the target Client are 1235 forwarded to the Client's new location instead of being dropped. 1236 When DepartTime decrements to 0, the neighbor cache entry is deleted. 1237 It is RECOMMENDED that DEPART_TIME be set to the default constant 1238 value REACHABLE_TIME plus 10 seconds (40 seconds by default) to allow 1239 a window for packets in flight to be delivered while stale route 1240 optimization state may be present. 1242 When an ROR receives an authentic NS message used for route 1243 optimization, it searches for a symmetric neighbor cache entry for 1244 the target Client. The ROR then returns a solicited NA message 1245 without creating a neighbor cache entry for the ROS, but creates or 1246 updates a target Client "Report List" entry for the ROS and sets a 1247 "ReportTime" variable for the entry to REPORT_TIME seconds. The ROR 1248 resets ReportTime when it receives a new authentic NS message, and 1249 otherwise decrements ReportTime while no authentic NS messages have 1250 been received. It is RECOMMENDED that REPORT_TIME be set to the 1251 default constant value REACHABLE_TIME plus 10 seconds (40 seconds by 1252 default) to allow a window for route optimization to converge before 1253 ReportTime decrements below REACHABLE_TIME. 1255 When the ROS receives a solicited NA message response to its NS 1256 message used for route optimization, it creates or updates an 1257 asymmetric neighbor cache entry for the target network-layer and 1258 link-layer addresses. The ROS then (re)sets ReachableTime for the 1259 neighbor cache entry to REACHABLE_TIME seconds and uses this value to 1260 determine whether packets can be forwarded directly to the target, 1261 i.e., instead of via a default route. The ROS otherwise decrements 1262 ReachableTime while no further solicited NA messages arrive. It is 1263 RECOMMENDED that REACHABLE_TIME be set to the default constant value 1264 30 seconds as specified in [RFC4861]. 1266 AERO nodes also use the value MAX_UNICAST_SOLICIT to limit the number 1267 of NS keepalives sent when a correspondent may have gone unreachable, 1268 the value MAX_RTR_SOLICITATIONS to limit the number of RS messages 1269 sent without receiving an RA and the value MAX_NEIGHBOR_ADVERTISEMENT 1270 to limit the number of unsolicited NAs that can be sent based on a 1271 single event. It is RECOMMENDED that MAX_UNICAST_SOLICIT, 1272 MAX_RTR_SOLICITATIONS and MAX_NEIGHBOR_ADVERTISEMENT be set to 3 the 1273 same as specified in [RFC4861]. 1275 Different values for DEPART_TIME, REPORT_TIME, REACHABLE_TIME, 1276 MAX_UNICAST_SOLICIT, MAX_RTR_SOLCITATIONS and 1277 MAX_NEIGHBOR_ADVERTISEMENT MAY be administratively set; however, if 1278 different values are chosen, all nodes on the link MUST consistently 1279 configure the same values. Most importantly, DEPART_TIME and 1280 REPORT_TIME SHOULD be set to a value that is sufficiently longer than 1281 REACHABLE_TIME to avoid packet loss due to stale route optimization 1282 state. 1284 3.6. OMNI Interface Encapsulation and Re-encapsulation 1286 OMNI interfaces insert a mid-layer IPv6 header known as the SPAN 1287 header when necessary as discussed in the following sections. After 1288 either inserting or omitting the SPAN header, the OMNI interface also 1289 inserts or omits an outer encapsulation header as discussed below. 1291 OMNI interfaces avoid outer encapsulation over Direct underlying 1292 interfaces and Proxyed underlying interfaces for which the first-hop 1293 access router is AERO-aware. Other OMNI interfaces encapsulate 1294 packets according to whether they are entering the OMNI interface 1295 from the network layer or if they are being re-admitted into the same 1296 OMNI link they arrived on. This latter form of encapsulation is 1297 known as "re-encapsulation". 1299 For packets entering the OMNI interface from the network layer, the 1300 OMNI interface copies the "TTL/Hop Limit", "Type of Service/Traffic 1301 Class" [RFC2983], "Flow Label"[RFC6438] (for IPv6) and "Congestion 1302 Experienced" [RFC3168] values in the inner packet's IP header into 1303 the corresponding fields in the SPAN and outer encapsulation 1304 header(s). 1306 For packets undergoing re-encapsulation, the OMNI interface instead 1307 copies these values from the original encapsulation header into the 1308 new encapsulation header, i.e., the values are transferred between 1309 encapsulation headers and *not* copied from the encapsulated packet's 1310 network-layer header. (Note especially that by copying the TTL/Hop 1311 Limit between encapsulation headers the value will eventually 1312 decrement to 0 if there is a (temporary) routing loop.) 1314 OMNI interfaces configured over INET underlying interfaces 1315 encapsulate packets in INET headers according to the next hop 1316 determined in the forwarding algorithm in Section 3.10. If the next 1317 hop is reached via a secured tunnel, the OMNI interface uses an 1318 encapsulation format specific to the secured tunnel type (see: 1319 Section 6). If the next hop is reached via an unsecured INET 1320 interface, the OMNI interface instead uses UDP/IP encapsulation per 1321 [RFC4380] and as extended in [RFC6081]. 1323 When UDP/IP encapsulation is used, the OMNI interface next sets the 1324 UDP source port to a constant value that it will use in each 1325 successive packet it sends, and sets the UDP length field to the 1326 length of the encapsulated packet plus 8 bytes for the UDP header 1327 itself plus the length of any included extension headers or trailers. 1328 The encapsulated packet may be either IPv6 or IPv4, as distinguished 1329 by the version number found in the first four bits. 1331 For UDP/IP-encapsulated packets sent to a Server, Relay or Bridge, 1332 the OMNI interface sets the UDP destination port to 8060, i.e., the 1333 IANA-registered port number for AERO. For packets sent to a Client, 1334 the OMNI interface sets the UDP destination port to the port value 1335 stored in the neighbor cache entry for this Client. The OMNI 1336 interface finally includes/omits the UDP checksum according to 1337 [RFC6935][RFC6936]. 1339 3.7. OMNI Interface Decapsulation 1341 OMNI interfaces decapsulate packets destined either to the AERO node 1342 itself or to a destination reached via an interface other than the 1343 OMNI interface the packet was received on. When the encapsulated 1344 packet arrives in multiple SPAN fragments, the OMNI interface 1345 reassembles as discussed in Section 3.9. Further decapsulation steps 1346 are performed according to the appropriate encapsulation format 1347 specification. 1349 3.8. OMNI Interface Data Origin Authentication 1351 AERO nodes employ simple data origin authentication procedures. In 1352 particular: 1354 o AERO Bridges, Servers and Proxys accept encapsulated data packets 1355 and control messages received from the (secured) spanning tree. 1357 o AERO Proxys and Clients accept packets that originate from within 1358 the same secured ANET. 1360 o AERO Clients and Relays accept packets from downstream network 1361 correspondents based on ingress filtering. 1363 o AERO Clients, Relays and Servers verify the outer UDP/IP 1364 encapsulation addresses according to [RFC4380]. 1366 AERO nodes silently drop any packets that do not satisfy the above 1367 data origin authentication procedures. Further security 1368 considerations are discussed in Section 6. 1370 3.9. OMNI Interface MTU and Fragmentation 1372 The OMNI interface observes the link nature of tunnels, including the 1373 Maximum Transmission Unit (MTU) and the role of fragmentation and 1374 reassembly[I-D.ietf-intarea-tunnels]. OMNI interface MTU and 1375 fragmentation/reassembly procedures are specified in 1376 [I-D.templin-6man-omni-interface]. 1378 3.10. OMNI Interface Forwarding Algorithm 1380 IP packets enter a node's OMNI interface either from the network 1381 layer (i.e., from a local application or the IP forwarding system) or 1382 from the link layer (i.e., from an OMNI interface neighbor). All 1383 packets entering a node's OMNI interface first undergo data origin 1384 authentication as discussed in Section 3.8. Packets that satisfy 1385 data origin authentication are processed further, while all others 1386 are dropped silently. OMNI interfaces wrap accepted packets in a 1387 SPAN header and SRH if necessary as discussed above. 1389 Packets that enter the OMNI interface from the network layer are 1390 forwarded to an OMNI interface neighbor. Packets that enter the OMNI 1391 interface from the link layer are either re-admitted into the OMNI 1392 link or forwarded to the network layer where they are subject to 1393 either local delivery or IP forwarding. In all cases, the OMNI 1394 interface itself MUST NOT decrement the network layer TTL/Hop-count 1395 since its forwarding actions occur below the network layer. 1397 OMNI interfaces may have multiple underlying interfaces and/or 1398 neighbor cache entries for neighbors with multiple ifIndex-tuple 1399 registrations (see Section 3.3). The OMNI interface uses traffic 1400 classifiers (e.g., DSCP value, port number, etc.) to select an 1401 outgoing underlying interface for each packet based on the node's own 1402 QoS preferences, and also to select a destination link-layer address 1403 based on the neighbor's underlying interface with the highest 1404 preference. AERO implementations SHOULD allow for QoS preference 1405 values to be modified at runtime through network management. 1407 If multiple outgoing interfaces and/or neighbor interfaces have a 1408 preference of "high", the AERO node replicates the packet and sends 1409 one copy via each of the (outgoing / neighbor) interface pairs; 1410 otherwise, the node sends a single copy of the packet via an 1411 interface with the highest preference. AERO nodes keep track of 1412 which underlying interfaces are currently "reachable" or 1413 "unreachable", and only use "reachable" interfaces for forwarding 1414 purposes. 1416 The following sections discuss the OMNI interface forwarding 1417 algorithms for Clients, Proxys, Servers and Bridges. In the 1418 following discussion, a packet's destination address is said to 1419 "match" if it is the same as a cached address, or if it is covered by 1420 a cached prefix (which may be encoded in an LLA). 1422 3.10.1. Client Forwarding Algorithm 1424 When an IP packet enters a Client's OMNI interface from the network 1425 layer the Client searches for an asymmetric neighbor cache entry that 1426 matches the destination. If there is a match, the Client uses one or 1427 more "reachable" neighbor interfaces in the entry for packet 1428 forwarding. If there is no asymmetric neighbor cache entry, the 1429 Client instead forwards the packet toward a Server (the packet is 1430 intercepted by a Proxy if there is a Proxy on the path). The Client 1431 encapsulates the packet in a SPAN header and SRH if necessary and 1432 fragments according to MTU requirements (see: Section 3.9). 1434 When an IP packet enters a Client's OMNI interface from the link- 1435 layer, if the destination matches one of the Client's MNPs or link- 1436 local addresses the Client reassembles and decapsulates as necessary 1437 and delivers the inner packet to the network layer. Otherwise, the 1438 Client drops the packet and MAY return a network-layer ICMP 1439 Destination Unreachable message subject to rate limiting (see: 1440 Section 3.11). 1442 3.10.2. Proxy Forwarding Algorithm 1444 For control messages originating from or destined to a Client, the 1445 Proxy intercepts the message and updates its proxy neighbor cache 1446 entry for the Client. The Proxy then forwards a (proxyed) copy of 1447 the control message. (For example, the Proxy forwards a proxied 1448 version of a Client's NS/RS message to the target neighbor, and 1449 forwards a proxied version of the NA/RA reply to the Client.) 1451 When the Proxy receives a data packet from a Client within the ANET, 1452 the Proxy reassembles and re-fragments if necessary then searches for 1453 an asymmetric neighbor cache entry that matches the destination and 1454 forwards as follows: 1456 o if the destination matches an asymmetric neighbor cache entry, the 1457 Proxy uses one or more "reachable" neighbor interfaces in the 1458 entry for packet forwarding using SPAN encapsulation and including 1459 a SRH if necessary according to the cached TLLAO information. If 1460 the neighbor interface is in the same SPAN segment, the Proxy 1461 forwards the packet directly to the neighbor; otherwise, it 1462 forwards the packet to a Bridge. 1464 o else, the Proxy uses SPAN encapsulation and forwards the packet to 1465 a Bridge while using the ULA corresponding to the packet's 1466 destination as the SPAN destination address. 1468 When the Proxy receives an encapsulated data packet from an INET 1469 neighbor or from a secured tunnel from a Bridge, it accepts the 1470 packet only if data origin authentication succeeds and if there is a 1471 proxy neighbor cache entry that matches the inner destination. Next, 1472 the Proxy reassembles the packet (if necessary) and continues 1473 processing. 1475 Next if reassembly is complete and the neighbor cache state is 1476 REACHABLE, the Proxy returns a PTB if necessary (see: Section 3.9) 1477 then either drops or forwards the packet to the Client while 1478 performing SPAN encapsulation and re-fragmentation to the ANET MTU 1479 size if necessary. If the neighbor cache entry state is DEPARTED, 1480 the Proxy instead changes the SPAN destination address to the address 1481 of the new Server and forwards it to a Bridge while performing re- 1482 fragmentation to 1280 bytes if necessary. 1484 3.10.3. Server/Relay Forwarding Algorithm 1486 For control messages destined to a target Client's LLA that are 1487 received from a secured tunnel, the Server intercepts the message and 1488 sends an appropriate response on behalf of the Client. (For example, 1489 the Server sends an NA message reply in response to an NS message 1490 directed to one of its associated Clients.) If the Client's neighbor 1491 cache entry is in the DEPARTED state, however, the Server instead 1492 forwards the packet to the Client's new Server as discussed in 1493 Section 3.16. 1495 When the Server receives an encapsulated data packet from an INET 1496 neighbor or from a secured tunnel, it accepts the packet only if data 1497 origin authentication succeeds. If the SPAN destination address is 1498 its own address, the Server continues processing as follows: 1500 o if the destination matches a symmetric neighbor cache entry in the 1501 REACHABLE state the Server prepares the packet for forwarding to 1502 the destination Client. The Server first reassembles (if 1503 necessary) and forwards the packet (while re-fragmenting if 1504 necessary) as specified in Section 3.9. 1506 o else, if the destination matches a symmetric neighbor cache entry 1507 in the DEPARETED state the Server re-encapsulates the packet and 1508 forwards it using the ULA of the Client's new Server as the 1509 destination. 1511 o else, if the destination matches an asymmetric neighbor cache 1512 entry, the Server uses one or more "reachable" neighbor interfaces 1513 in the entry for packet forwarding via the local INET if the 1514 neighbor is in the same OMNI link segment or using SPAN 1515 encapsulation and Segment Routing if necessary with the final 1516 destination set to the neighbor's ULA otherwise. 1518 o else, if the destination is an LLA that is not assigned on the 1519 OMNI interface the Server drops the packet. 1521 o else, the Server (acting as a Relay) reassembles if necessary, 1522 decapsulates the packet and releases it to the network layer for 1523 local delivery or IP forwarding. Based on the information in the 1524 forwarding table, the network layer may return the packet to the 1525 same OMNI interface in which case further processing occurs as 1526 below. (Note that this arrangement accommodates common 1527 implementations in which the IP forwarding table is not accessible 1528 from within the OMNI interface. If the OMNI interface can 1529 directly access the IP forwarding table (such as for in-kernel 1530 implementations) the forwarding table lookup can instead be 1531 performed internally from within the OMNI interface itself.) 1533 When the Server's OMNI interface receives a data packet from the 1534 network layer or from a VPNed or Direct Client, it performs SPAN 1535 encapsulation and fragmentation if necessary, then processes the 1536 packet according to the network-layer destination address as follows: 1538 o if the destination matches a symmetric or asymmetric neighbor 1539 cache entry the Server processes the packet as above. 1541 o else, the Server encapsulates the packet and forwards it to a 1542 Bridge using its own ULA as the source and the ULA corresponding 1543 to the destination as the destination. 1545 3.10.4. Bridge Forwarding Algorithm 1547 Bridges forward SPAN-encapsulated packets over secured tunnels the 1548 same as any IP router. When the Bridge receives a SPAN-encapsulated 1549 packet via a secured tunnel, it removes the outer INET header and 1550 searches for a forwarding table entry that matches the SPAN 1551 destination address. The Bridge then processes the packet as 1552 follows: 1554 o if the destination matches its ULA Subnet Router Anycast address, 1555 the Bridge checks for a SRH. If there is a SRH with Segments 1556 Left=1, with the ULA of a Proxy/Server on the local segment as the 1557 LHS ID, and with an AERO Route Optimization TLV, the Bridge 1558 examines the FMT to determine if the target is behind a NAT. If 1559 no NAT is indicated, the Bridge copies the MNP Subnet Router 1560 Anycast address if an MNP is included (otherwise copies the Proxy/ 1561 Server ULA) into the destination address then forwards the packet 1562 directly to the L2ADDR using link-layer (UDP/IP) encapsulation. 1563 If a NAT is indicated, the Bridge MAY perform NAT traversal 1564 procedures by sending bubbles per [RFC4380]. The Bridge then 1565 either applies AERO route optimization if NAT traversal procedures 1566 have been successfully applied, or forwards the packet directly to 1567 the Server. 1569 o if the destination matches one of the Bridge's own addresses, the 1570 Bridge submits the packet for local delivery. 1572 o else, if the destination matches a forwarding table entry the 1573 Bridge forwards the packet via a secured tunnel to the next hop. 1574 If the destination matches an MSP without matching an MNP, 1575 however, the Bridge instead drops the packet and returns an ICMP 1576 Destination Unreachable message subject to rate limiting (see: 1577 Section 3.11). 1579 o else, the Bridge drops the packet and returns an ICMP Destination 1580 Unreachable as above. 1582 As for any IP router, the Bridge decrements the TTL/Hop Limit when it 1583 forwards the packet. Therefore, only the Hop Limit in the SPAN 1584 header is decremented, and not the TTL/Hop Limit in the inner packet 1585 header. 1587 3.11. OMNI Interface Error Handling 1589 When an AERO node admits a packet into the OMNI interface, it may 1590 receive link-layer or network-layer error indications. 1592 A link-layer error indication is an ICMP error message generated by a 1593 router in the INET on the path to the neighbor or by the neighbor 1594 itself. The message includes an IP header with the address of the 1595 node that generated the error as the source address and with the 1596 link-layer address of the AERO node as the destination address. 1598 The IP header is followed by an ICMP header that includes an error 1599 Type, Code and Checksum. Valid type values include "Destination 1600 Unreachable", "Time Exceeded" and "Parameter Problem" 1601 [RFC0792][RFC4443]. (OMNI interfaces ignore all link-layer IPv4 1602 "Fragmentation Needed" and IPv6 "Packet Too Big" messages since they 1603 only emit packets that are guaranteed to be no larger than the IP 1604 minimum link MTU as discussed in Section 3.9.) 1606 The ICMP header is followed by the leading portion of the packet that 1607 generated the error, also known as the "packet-in-error". For 1608 ICMPv6, [RFC4443] specifies that the packet-in-error includes: "As 1609 much of invoking packet as possible without the ICMPv6 packet 1610 exceeding the minimum IPv6 MTU" (i.e., no more than 1280 bytes). For 1611 ICMPv4, [RFC0792] specifies that the packet-in-error includes: 1612 "Internet Header + 64 bits of Original Data Datagram", however 1613 [RFC1812] Section 4.3.2.3 updates this specification by stating: "the 1614 ICMP datagram SHOULD contain as much of the original datagram as 1615 possible without the length of the ICMP datagram exceeding 576 1616 bytes". 1618 The link-layer error message format is shown in Figure 6 (where, "L2" 1619 and "L3" refer to link-layer and network-layer, respectively): 1621 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1622 ~ ~ 1623 | L2 IP Header of | 1624 | error message | 1625 ~ ~ 1626 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1627 | L2 ICMP Header | 1628 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1629 ~ ~ P 1630 | IP and other encapsulation | a 1631 | headers of original L3 packet | c 1632 ~ ~ k 1633 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ e 1634 ~ ~ t 1635 | IP header of | 1636 | original L3 packet | i 1637 ~ ~ n 1638 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1639 ~ ~ e 1640 | Upper layer headers and | r 1641 | leading portion of body | r 1642 | of the original L3 packet | o 1643 ~ ~ r 1644 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1646 Figure 6: OMNI Interface Link-Layer Error Message Format 1648 The AERO node rules for processing these link-layer error messages 1649 are as follows: 1651 o When an AERO node receives a link-layer Parameter Problem message, 1652 it processes the message the same as described as for ordinary 1653 ICMP errors in the normative references [RFC0792][RFC4443]. 1655 o When an AERO node receives persistent link-layer Time Exceeded 1656 messages, the IP ID field may be wrapping before earlier fragments 1657 awaiting reassembly have been processed. In that case, the node 1658 should begin including integrity checks and/or institute rate 1659 limits for subsequent packets. 1661 o When an AERO node receives persistent link-layer Destination 1662 Unreachable messages in response to encapsulated packets that it 1663 sends to one of its asymmetric neighbor correspondents, the node 1664 should process the message as an indication that a path may be 1665 failing, and optionally initiate NUD over that path. If it 1666 receives Destination Unreachable messages over multiple paths, the 1667 node should allow future packets destined to the correspondent to 1668 flow through a default route and re-initiate route optimization. 1670 o When an AERO Client receives persistent link-layer Destination 1671 Unreachable messages in response to encapsulated packets that it 1672 sends to one of its symmetric neighbor Servers, the Client should 1673 mark the path as unusable and use another path. If it receives 1674 Destination Unreachable messages on many or all paths, the Client 1675 should associate with a new Server and release its association 1676 with the old Server as specified in Section 3.16.5. 1678 o When an AERO Server receives persistent link-layer Destination 1679 Unreachable messages in response to encapsulated packets that it 1680 sends to one of its symmetric neighbor Clients, the Server should 1681 mark the underlying path as unusable and use another underlying 1682 path. 1684 o When an AERO Server or Proxy receives link-layer Destination 1685 Unreachable messages in response to an encapsulated packet that it 1686 sends to one of its permanent neighbors, it treats the messages as 1687 an indication that the path to the neighbor may be failing. 1688 However, the dynamic routing protocol should soon reconverge and 1689 correct the temporary outage. 1691 When an AERO Bridge receives a packet for which the network-layer 1692 destination address is covered by an MSP, if there is no more- 1693 specific routing information for the destination the Bridge drops the 1694 packet and returns a network-layer Destination Unreachable message 1695 subject to rate limiting. The Bridge writes the network-layer source 1696 address of the original packet as the destination address and uses 1697 one of its non link-local addresses as the source address of the 1698 message. 1700 When an AERO node receives an encapsulated packet for which the 1701 reassembly buffer it too small, it drops the packet and returns a 1702 network-layer Packet Too Big (PTB) message. The node first writes 1703 the MRU value into the PTB message MTU field, writes the network- 1704 layer source address of the original packet as the destination 1705 address and writes one of its non link-local addresses as the source 1706 address. 1708 3.12. AERO Router Discovery, Prefix Delegation and Autoconfiguration 1710 AERO Router Discovery, Prefix Delegation and Autoconfiguration are 1711 coordinated as discussed in the following Sections. 1713 3.12.1. AERO ND/PD Service Model 1715 Each AERO Server on the OMNI link configures a PD service to 1716 facilitate Client requests. Each Server is provisioned with a 1717 database of MNP-to-Client ID mappings for all Clients enrolled in the 1718 AERO service, as well as any information necessary to authenticate 1719 each Client. The Client database is maintained by a central 1720 administrative authority for the OMNI link and securely distributed 1721 to all Servers, e.g., via the Lightweight Directory Access Protocol 1722 (LDAP) [RFC4511], via static configuration, etc. Clients receive the 1723 same service regardless of the Servers they select. 1725 AERO Clients and Servers use ND messages to maintain neighbor cache 1726 entries. AERO Servers configure their OMNI interfaces as advertising 1727 NBMA interfaces, and therefore send unicast RA messages with a short 1728 Router Lifetime value (e.g., ReachableTime seconds) in response to a 1729 Client's RS message. Thereafter, Clients send additional RS messages 1730 to keep Server state alive. 1732 AERO Clients and Servers include PD parameters in RS/RA messages (see 1733 [I-D.templin-6man-dhcpv6-ndopt] for ND/PD alternatives). The unified 1734 ND/PD messages are exchanged between Client and Server according to 1735 the prefix management schedule required by the PD service. If the 1736 Client knows its MNP in advance, it can instead employ prefix 1737 registration by including its LLA as the source address of an RS 1738 message and with an OMNI option with valid prefix registration 1739 information for the MNP. If the Server (and Proxy) accept the 1740 Client's MNP assertion, they inject the prefix into the routing 1741 system and establish the necessary neighbor cache state. 1743 The following sections specify the Client and Server behavior. 1745 3.12.2. AERO Client Behavior 1747 AERO Clients discover the addresses of Servers in a similar manner as 1748 described in [RFC5214]. Discovery methods include static 1749 configuration (e.g., from a flat-file map of Server addresses and 1750 locations), or through an automated means such as Domain Name System 1751 (DNS) name resolution [RFC1035]. Alternatively, the Client can 1752 discover Server addresses through a layer 2 data link login exchange, 1753 or through a unicast RA response to a multicast/anycast RS as 1754 described below. In the absence of other information, the Client can 1755 resolve the DNS Fully-Qualified Domain Name (FQDN) 1756 "linkupnetworks.[domainname]" where "linkupnetworks" is a constant 1757 text string and "[domainname]" is a DNS suffix for the OMNI link 1758 (e.g., "example.com"). 1760 To associate with a Server, the Client acts as a requesting router to 1761 request MNPs. The Client prepares an RS message with PD parameters 1762 and includes a Nonce and Timestamp option if the Client needs to 1763 correlate RA replies. If the Client already knows the Server's LLA, 1764 it includes the LLA as the network-layer destination address; 1765 otherwise, it includes (link-local) All-Routers multicast as the 1766 network-layer destination. If the Client already knows its own LLA, 1767 it uses the LLA as the network-layer source address; otherwise, it 1768 uses the unspecified IPv6 address (::/128) as the network-layer 1769 source address. 1771 The Client next includes an OMNI option in the RS message to register 1772 its link-layer information with the Server. The Client sets the OMNI 1773 option prefix registration information according to the MNP, and 1774 includes an ifIndex-tuple with S set to '1' corresponding to the 1775 underlying interface over which the Client will send the RS message. 1776 The Client MAY include additional ifIndex-tuples specific to other 1777 underlying interfaces. The Client MAY also include an SLLAO 1778 corresponding to the OMNI option ifIndex-tuple with S set to '1'. 1780 The Client then sends the RS message (either directly via Direct 1781 interfaces, via a VPN for VPNed interfaces, via a Proxy for proxyed 1782 interfaces or via INET encapsulation for INET interfaces) and waits 1783 for an RA message reply (see Section 3.12.3). The Client retries up 1784 to MAX_RTR_SOLICITATIONS times until an RA is received. If the 1785 Client receives no RAs, or if it receives an RA with Router Lifetime 1786 set to 0, the Client SHOULD abandon this Server and try another 1787 Server. Otherwise, the Client processes the PD information found in 1788 the RA message. 1790 Next, the Client creates a symmetric neighbor cache entry with the 1791 Server's LLA as the network-layer address and the Server's 1792 encapsulation and/or link-layer addresses as the link-layer address. 1793 The Client records the RA Router Lifetime field value in the neighbor 1794 cache entry as the time for which the Server has committed to 1795 maintaining the MNP in the routing system via this underlying 1796 interface, and caches the other RA configuration information 1797 including Cur Hop Limit, M and O flags, Reachable Time and Retrans 1798 Timer. The Client then autoconfigures LLAs for each of the delegated 1799 MNPs and assigns them to the OMNI interface. The Client also caches 1800 any MSPs included in Route Information Options (RIOs) [RFC4191] as 1801 MSPs to associate with the OMNI link, and assigns the MTU value in 1802 the MTU option to the underlying interface. 1804 The Client then registers additional underlying interfaces with the 1805 Server by sending RS messages via each additional interface. The RS 1806 messages include the same parameters as for the initial RS/RA 1807 exchange, but with destination address set to the Server's LLA. 1809 Following autoconfiguration, the Client sub-delegates the MNPs to its 1810 attached EUNs and/or the Client's own internal virtual interfaces as 1811 described in [I-D.templin-v6ops-pdhost] to support the Client's 1812 downstream attached "Internet of Things (IoT)". The Client 1813 subsequently sends additional RS messages over each underlying 1814 interface before the Router Lifetime received for that interface 1815 expires. 1817 After the Client registers its underlying interfaces, it may wish to 1818 change one or more registrations, e.g., if an interface changes 1819 address or becomes unavailable, if QoS preferences change, etc. To 1820 do so, the Client prepares an RS message to send over any available 1821 underlying interface. The RS includes an OMNI option with prefix 1822 registration information specific to its MNP, with an ifIndex-tuple 1823 specific to the selected underlying interface with S set to '1', and 1824 with any additional ifIndex-tuples specific to other underlying 1825 interfaces. The Client includes fresh ifIndex-tuple values to update 1826 the Server's neighbor cache entry. When the Client receives the 1827 Server's RA response, it has assurance that the Server has been 1828 updated with the new information. 1830 If the Client wishes to discontinue use of a Server it issues an RS 1831 message over any underlying interface with an OMNI option with a 1832 prefix release indication. When the Server processes the message, it 1833 releases the MNP, sets the symmetric neighbor cache entry state for 1834 the Client to DEPARTED and returns an RA reply with Router Lifetime 1835 set to 0. After a short delay (e.g., 2 seconds), the Server 1836 withdraws the MNP from the routing system. 1838 3.12.3. AERO Server Behavior 1840 AERO Servers act as IP routers and support a PD service for Clients. 1841 Servers arrange to add their LLAs to a static map of Server addresses 1842 for the link and/or the DNS resource records for the FQDN 1843 "linkupnetworks.[domainname]" before entering service. Server 1844 addresses should be geographically and/or topologically referenced, 1845 and made available for discovery by Clients on the OMNI link. 1847 When a Server receives a prospective Client's RS message on its OMNI 1848 interface, it SHOULD return an immediate RA reply with Router 1849 Lifetime set to 0 if it is currently too busy or otherwise unable to 1850 service the Client. Otherwise, the Server authenticates the RS 1851 message and processes the PD parameters. The Server first determines 1852 the correct MNPs to delegate to the Client by searching the Client 1853 database. When the Server delegates the MNPs, it also creates a 1854 forwarding table entry for each MNP so that the MNPs are propagated 1855 into the routing system (see: Section 3.2.3). For IPv6, the Server 1856 creates an IPv6 forwarding table entry for each MNP. For IPv4, the 1857 Server creates an IPv6 forwarding table entry with the SPAN 1858 Compatibility Prefix (SCP) corresponding to the IPv4 address. 1860 The Server next creates a symmetric neighbor cache entry for the 1861 Client using the base LLA as the network-layer address and with 1862 lifetime set to no more than the smallest PD lifetime. Next, the 1863 Server updates the neighbor cache entry by recording the information 1864 in each ifIndex-tuple in the RS OMNI option. The Server also records 1865 the actual SPAN/INET addresses in the neighbor cache entry. 1867 Next, the Server prepares an RA message using its LLA as the network- 1868 layer source address and the network-layer source address of the RS 1869 message as the network-layer destination address. The Server sets 1870 the Router Lifetime to the time for which it will maintain both this 1871 underlying interface individually and the symmetric neighbor cache 1872 entry as a whole. The Server also sets Cur Hop Limit, M and O flags, 1873 Reachable Time and Retrans Timer to values appropriate for the OMNI 1874 link. The Server includes the delegated MNPs, any other PD 1875 parameters and an OMNI option with no ifIndex-tuples. The Server 1876 then includes one or more RIOs that encode the MSPs for the OMNI 1877 link, plus an MTU option (see Section 3.9). The Server finally 1878 forwards the message to the Client using SPAN/INET, INET, or NULL 1879 encapsulation as necessary. 1881 After the initial RS/RA exchange, the Server maintains a 1882 ReachableTime timer for each of the Client's underlying interfaces 1883 individually (and for the Client's symmetric neighbor cache entry 1884 collectively) set to expire after ReachableTime seconds. If the 1885 Client (or Proxy) issues additional RS messages, the Server sends an 1886 RA response and resets ReachableTime. If the Server receives an ND 1887 message with PD release indication it sets the Client's symmetric 1888 neighbor cache entry to the DEPARTED state and withdraws the MNP from 1889 the routing system after a short delay (e.g., 2 seconds). If 1890 ReachableTime expires before a new RS is received on an individual 1891 underlying interface, the Server marks the interface as DOWN. If 1892 ReachableTime expires before any new RS is received on any individual 1893 underlying interface, the Server sets the symmetric neighbor cache 1894 entry state to STALE and sets a 10 second timer. If the Server has 1895 not received a new RS or ND message with PD release indication before 1896 the 10 second timer expires, it deletes the neighbor cache entry and 1897 withdraws the MNP from the routing system. 1899 The Server processes any ND/PD messages pertaining to the Client and 1900 returns an NA/RA reply in response to solicitations. The Server may 1901 also issue unsolicited RA messages, e.g., with PD reconfigure 1902 parameters to cause the Client to renegotiate its PDs, with Router 1903 Lifetime set to 0 if it can no longer service this Client, etc. 1904 Finally, If the symmetric neighbor cache entry is in the DEPARTED 1905 state, the Server deletes the entry after DepartTime expires. 1907 Note: Clients SHOULD notify former Servers of their departures, but 1908 Servers are responsible for expiring neighbor cache entries and 1909 withdrawing routes even if no departure notification is received 1910 (e.g., if the Client leaves the network unexpectedly). Servers 1911 SHOULD therefore set Router Lifetime to ReachableTime seconds in 1912 solicited RA messages to minimize persistent stale cache information 1913 in the absence of Client departure notifications. A short Router 1914 Lifetime also ensures that proactive Client/Server RS/RA messaging 1915 will keep any NAT state alive (see above). 1917 Note: All Servers on an OMNI link MUST advertise consistent values in 1918 the RA Cur Hop Limit, M and O flags, Reachable Time and Retrans Timer 1919 fields the same as for any link, since unpredictable behavior could 1920 result if different Servers on the same link advertised different 1921 values. 1923 3.12.3.1. Lightweight DHCPv6 Relay Agent (LDRA) 1925 When DHCPv6 is used as the ND/PD service back end, AERO Clients and 1926 Servers are always on the same link (i.e., the OMNI link) from the 1927 perspective of DHCPv6. However, in some implementations the DHCPv6 1928 server and ND function may be located in separate modules. In that 1929 case, the Server's OMNI interface module can act as a Lightweight 1930 DHCPv6 Relay Agent (LDRA)[RFC6221] to relay PD messages to and from 1931 the DHCPv6 server module. 1933 When the LDRA receives an authentic RS message, it extracts the PD 1934 message parameters and uses them to construct an IPv6/UDP/DHCPv6 1935 message. It sets the IPv6 source address to the source address of 1936 the RS message, sets the IPv6 destination address to 1937 'All_DHCP_Relay_Agents_and_Servers' and sets the UDP fields to values 1938 that will be understood by the DHCPv6 server. 1940 The LDRA then wraps the message in a DHCPv6 'Relay-Forward' message 1941 header and includes an 'Interface-Id' option that includes enough 1942 information to allow the LDRA to forward the resulting Reply message 1943 back to the Client (e.g., the Client's link-layer addresses, a 1944 security association identifier, etc.). The LDRA also wraps the OMNI 1945 option and SLLAO into the Interface-Id option, then forwards the 1946 message to the DHCPv6 server. 1948 When the DHCPv6 server prepares a Reply message, it wraps the message 1949 in a 'Relay-Reply' message and echoes the Interface-Id option. The 1950 DHCPv6 server then delivers the Relay-Reply message to the LDRA, 1951 which discards the Relay-Reply wrapper and IPv6/UDP headers, then 1952 uses the DHCPv6 message to construct an RA response to the Client. 1953 The Server uses the information in the Interface-Id option to prepare 1954 the RA message and to cache the link-layer addresses taken from the 1955 OMNI option and SLLAO echoed in the Interface-Id option. 1957 3.13. The AERO Proxy 1959 Clients may connect to protected-spectrum ANETs that deploy physical 1960 and/or link-layer security services to facilitate communications to 1961 Servers in outside INETs. In that case, the ANET can employ an AERO 1962 Proxy. The Proxy is located at the ANET/INET border and listens for 1963 RS messages originating from or RA messages destined to ANET Clients. 1964 The Proxy acts on these control messages as follows: 1966 o when the Proxy receives an RS message from a new ANET Client, it 1967 first authenticates the message then examines the network-layer 1968 destination address. If the destination address is a Server's 1969 LLA, the Proxy proceeds to the next step. Otherwise, if the 1970 destination is (link-local) All-Routers multicast, the Proxy 1971 selects a "nearby" Server that is likely to be a good candidate to 1972 serve the Client and replaces the destination address with the 1973 Server's LLA. Next, the Proxy creates a proxy neighbor cache 1974 entry and caches the Client and Server link-layer addresses along 1975 with the OMNI option information and any other identifying 1976 information including Transaction IDs, Client Identifiers, Nonce 1977 values, etc. The Proxy finally encapsulates the (proxyed) RS 1978 message in a SPAN header with source set to the Proxy's ULA and 1979 destination set to the Server's ULA then forwards the message into 1980 the SPAN. 1982 o when the Server receives the RS, it authenticates the message then 1983 creates or updates a symmetric neighbor cache entry for the Client 1984 with the Proxy's ULA as the link-layer address. The Server then 1985 sends an RA message back to the Proxy via the spanning tree. 1987 o when the Proxy receives the RA, it authenticates the message and 1988 matches it with the proxy neighbor cache entry created by the RS. 1989 The Proxy then caches the PD route information as a mapping from 1990 the Client's MNPs to the Client's link-layer address, caches the 1991 Server's advertised Router Lifetime and sets the neighbor cache 1992 entry state to REACHABLE. The Proxy then sets the P bit in the RA 1993 flags field, optionally rewrites the Router Lifetime and forwards 1994 the (proxyed) message to the Client. The Proxy finally includes 1995 an MTU option (if necessary) with an MTU to use for the underlying 1996 ANET interface. 1998 After the initial RS/RA exchange, the Proxy forwards any Client data 1999 packets for which there is no matching asymmetric neighbor cache 2000 entry to a Bridge using SPAN encapsulation with its own ULA as the 2001 source and the ULA corresponding to the Client as the destination. 2002 The Proxy instead forwards any Client data destined to an asymmetric 2003 neighbor cache target directly to the target according to the SPAN/ 2004 link-layer information - the process of establishing asymmetric 2005 neighbor cache entries is specified in Section 3.14. 2007 While the Client is still attached to the ANET, the Proxy sends NS, 2008 RS and/or unsolicited NA messages to update the Server's symmetric 2009 neighbor cache entries on behalf of the Client and/or to convey QoS 2010 updates. This allows for higher-frequency Proxy-initiated RS/RA 2011 messaging over well-connected INET infrastructure supplemented by 2012 lower-frequency Client-initiated RS/RA messaging over constrained 2013 ANET data links. 2015 If the Server ceases to send solicited advertisements, the Proxy 2016 sends unsolicited RAs on the ANET interface with destination set to 2017 (link-local) All-Nodes multicast and with Router Lifetime set to zero 2018 to inform Clients that the Server has failed. Although the Proxy 2019 engages in ND exchanges on behalf of the Client, the Client can also 2020 send ND messages on its own behalf, e.g., if it is in a better 2021 position than the Proxy to convey QoS changes, etc. For this reason, 2022 the Proxy marks any Client-originated solicitation messages (e.g. by 2023 inserting a Nonce option) so that it can return the solicited 2024 advertisement to the Client instead of processing it locally. 2026 If the Client becomes unreachable, the Proxy sets the neighbor cache 2027 entry state to DEPARTED and retains the entry for DepartTime seconds. 2028 While the state is DEPARTED, the Proxy forwards any packets destined 2029 to the Client to a Bridge via SPAN encapsulation with the Client's 2030 current Server as the destination. The Bridge in turn forwards the 2031 packets to the Client's current Server. When DepartTime expires, the 2032 Proxy deletes the neighbor cache entry and discards any further 2033 packets destined to this (now forgotten) Client. 2035 In some ANETs that employ a Proxy, the Client's MNP can be injected 2036 into the ANET routing system. In that case, the Client can send data 2037 messages without encapsulation so that the ANET routing system 2038 transports the unencapsulated packets to the Proxy. This can be very 2039 beneficial, e.g., if the Client connects to the ANET via low-end data 2040 links such as some aviation wireless links. 2042 If the first-hop ANET access router is AERO-aware, the Client can 2043 avoid encapsulation for both its control and data messages. When the 2044 Client connects to the link, it can send an unencapsulated RS message 2045 with source address set to its LLA and with destination address set 2046 to the LLA of the Client's selected Server or to (link-local) All- 2047 Routers multicast. The Client includes an OMNI option formatted as 2048 specified in [I-D.templin-6man-omni-interface]. 2050 The Client then sends the unencapsulated RS message, which will be 2051 intercepted by the AERO-Aware access router. The access router then 2052 encapsulates the RS message in an ANET header with its own address as 2053 the source address and the address of a Proxy as the destination 2054 address. The access router further remembers the address of the 2055 Proxy so that it can encapsulate future data packets from the Client 2056 via the same Proxy. If the access router needs to change to a new 2057 Proxy, it simply sends another RS message toward the Server via the 2058 new Proxy on behalf of the Client. 2060 In some cases, the access router and Proxy may be one and the same 2061 node. In that case, the node would be located on the same physical 2062 link as the Client, but its message exchanges with the Server would 2063 need to pass through a security gateway at the ANET/INET border. The 2064 method for deploying access routers and Proxys (i.e. as a single node 2065 or multiple nodes) is an ANET-local administrative consideration. 2067 3.13.1. Ancillary Servers Acting as Proxies 2069 Clients may need to connect directly to Servers via INET, Direct and 2070 VPNed interfaces (i.e., non-ANET interfaces). If the Client's 2071 underlying interfaces all connect via the same INET partition, then 2072 it can connect to a single controlling Server via all interfaces. 2074 If some Client interfaces connect via different INET partitions, 2075 however, the Client still selects a single controlling Server and 2076 sends RS messages over interfaces that connect via ancillary Servers 2077 while using the LLA of the controlling Server as the destination. 2079 When an ancillary Server receives an RS with destination set to the 2080 LLA of the controlling Server, it acts as a Proxy to forward the 2081 message to the controlling Server while forwarding the corresponding 2082 RA reply to the Client. When the ancillary Server forwards the RA 2083 reply, it sets the P bit in the RA flags field to indicate that it is 2084 acting in Proxy mode on behalf of this Client. 2086 3.13.2. Detecting and Responding to Server Failures 2088 In environments where fast recovery from Server failure is required, 2089 Proxys SHOULD use proactive Neighbor Unreachability Detection (NUD) 2090 to track Server reachability in a similar fashion as for 2091 Bidirectional Forwarding Detection (BFD) [RFC5880]. Proxys can then 2092 quickly detect and react to failures so that cached information is 2093 re-established through alternate paths. The NUD control messaging is 2094 carried only over well-connected ground domain networks (i.e., and 2095 not low-end aeronautical radio links) and can therefore be tuned for 2096 rapid response. 2098 Proxys perform proactive NUD with Servers for which there are 2099 currently active ANET Clients by sending continuous NS messages in 2100 rapid succession, e.g., one message per second. The Proxy sends the 2101 NS message via the spanning tree with the Proxy's LLA as the source 2102 and the LLA of the Server as the destination. When the Proxy is also 2103 sending RS messages to the Server on behalf of ANET Clients, the 2104 resulting RA responses can be considered as equivalent hints of 2105 forward progress. This means that the Proxy need not also send a 2106 periodic NS if it has already sent an RS within the same period. If 2107 the Server fails (i.e., if the Proxy ceases to receive 2108 advertisements), the Proxy can quickly inform Clients by sending 2109 multicast RA messages on the ANET interface. 2111 The Proxy sends RA messages on the ANET interface with source address 2112 set to the Server's address, destination address set to (link-local) 2113 All-Nodes multicast, and Router Lifetime set to 0. The Proxy SHOULD 2114 send MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small 2115 delays [RFC4861]. Any Clients on the ANET that had been using the 2116 failed Server will receive the RA messages and associate with a new 2117 Server. 2119 3.13.3. Point-to-Multipoint Server Coordination 2121 In environments where Client messaging over ANETs is bandwidth- 2122 limited and/or expensive, Clients can enlist the services of the 2123 Proxy to coordinate with multiple Servers in a single RS/RA message 2124 exchange. The Client can send a single RS message to (link-local) 2125 All-Routers multicast that includes the ID's of multiple Servers in 2126 MS-Register sub-options of the OMNI option. 2128 When the Proxy receives the RS and processes the OMNI option, it 2129 sends a separate RS to each MS-Register Server ID. When the Proxy 2130 receives an RA, it can optionally return an immediate "singleton" RA 2131 to the Client or record the Server's ID for inclusion in a pending 2132 "aggregate" RA message. The Proxy can then return aggregate RA 2133 messages to the Client including multiple Server IDs in order to 2134 conserve bandwidth. Each RA includes a proper subset of the Server 2135 IDs from the original RS message, and the Proxy must ensure that the 2136 message contents of each RA are consistent with the information 2137 received from the (aggregated) Servers. 2139 Clients can thereafter employ efficient point-to-multipoint Server 2140 coordination under the assistance of the Proxy to reduce the number 2141 of messages sent over the ANET while enlisting the support of 2142 multiple Servers for fault tolerance. Clients can further include 2143 MS-Release suboptions in IPv6 ND messages to request the Proxy to 2144 release from former Servers via the procedures discussed in 2145 Section 3.16.5. 2147 The OMNI interface specification [I-D.templin-6man-omni-interface] 2148 provides further discussion of the Client/Proxy RS/RA messaging 2149 involved in point-to-multipoint coordination. 2151 3.14. AERO Route Optimization / Address Resolution 2153 While data packets are flowing between a source and target node, 2154 route optimization SHOULD be used. Route optimization is initiated 2155 by the first eligible Route Optimization Source (ROS) closest to the 2156 source as follows: 2158 o For Clients on VPNed and Direct interfaces, the Server is the ROS. 2160 o For Clients on Proxyed interfaces, the Proxy is the ROS. 2162 o For Clients on INET interfaces, the Client itself is the ROS. 2164 o For correspondent nodes on INET/EUN interfaces serviced by a 2165 Relay, the Relay is the ROS. 2167 The route optimization procedure is conducted between the ROS and the 2168 target Server/Relay acting as a Route Optimization Responder (ROR) in 2169 the same manner as for IPv6 ND Address Resolution and using the same 2170 NS/NA messaging. The target may either be a MNP Client serviced by a 2171 Server, or a non-MNP correspondent reachable via a Relay. 2173 The procedures are specified in the following sections. 2175 3.14.1. Route Optimization Initiation 2177 While data packets are flowing from the source node toward a target 2178 node, the ROS performs address resolution by sending an NS message 2179 for Address Resolution (NS(AR)) to receive a solicited NA message 2180 from the ROR. When the ROS sends an NS(AR), it includes: 2182 o the LLA of the ROS as the source address. 2184 o the data packet's destination as the Target Address. 2186 o the Solicited-Node multicast address [RFC4291] formed from the 2187 lower 24 bits of the data packet's destination as the destination 2188 address, e.g., for 2001:db8:1:2::10:2000 the NS destination 2189 address is ff02:0:0:0:0:1:ff10:2000. 2191 The NS(AR) message includes an OMNI option with no ifIndex-tuples and 2192 no SLLAO, such that the target will not create a neighbor cache 2193 entry. 2195 The ROS then encapsulates the NS(AR) message in a SPAN header with 2196 source set to its own ULA and destination set to the ULA 2197 corresponding to the packet's final destination, then sends the 2198 message into the spanning tree without decrementing the network-layer 2199 TTL/Hop Limit field. 2201 3.14.2. Relaying the NS 2203 When the Bridge receives the NS(AR) message from the ROS, it discards 2204 the INET header and determines that the ROR is the next hop by 2205 consulting its standard IPv6 forwarding table for the SPAN header 2206 destination address. The Bridge then forwards the message toward the 2207 ROR via the spanning tree the same as for any IPv6 router. The 2208 final-hop Bridge in the spanning tree will deliver the message via a 2209 secured tunnel to the ROR. 2211 3.14.3. Processing the NS and Sending the NA 2213 When the ROR receives the NS(AR) message, it examines the Target 2214 Address to determine whether it has a neighbor cache entry and/or 2215 route that matches the target. If there is no match, the ROR drops 2216 the message. Otherwise, the ROR continues processing as follows: 2218 o if the target belongs to an MNP Client neighbor in the DEPARTED 2219 state the ROR changes the NS(AR) message SPAN destination address 2220 to the ULA of the Client's new Server, forwards the message into 2221 the spanning tree and returns from processing. 2223 o If the target belongs to an MNP Client neighbor in the REACHABLE 2224 state, the ROR instead adds the AERO source address to the target 2225 Client's Report List with time set to ReportTime. 2227 o If the target belongs to a non-MNP route, the ROR continues 2228 processing without adding an entry to the Report List. 2230 The ROR then prepares a solicited NA message to send back to the ROS 2231 but does not create a neighbor cache entry. The ROR sets the NA 2232 source address to the LLA corresponding to the target, sets the 2233 Target Address to the target of the solicitation, and sets the 2234 destination address to the source of the solicitation. 2236 The ROR then includes an OMNI option with prefix registration length 2237 set to the length of the MNP if the target is an MNP Client; 2238 otherwise, set to the maximum of the non-MNP prefix length and 64. 2239 (Note that a /64 limit is imposed to avoid causing the ROS to set 2240 short prefixes (e.g., "default") that would match destinations for 2241 which the routing system includes more-specific prefixes.) 2242 If the target is an MNP Client, the ROR next includes ifIndex-tuples 2243 in the OMNI option for each of the target Client's underlying 2244 interfaces with current information for each interface and with the S 2245 flag set to 0. The ROR then includes a TLLAO with ifIndex-tuples in 2246 one-to-one correspondence with the tuples that appear in the OMNI 2247 option. 2249 The ROR sets L2ADDR to its own INET address for VPNed or Direct 2250 interfaces, to the INET address of the Proxy for Proxyed interfaces 2251 or to the Client's INET address for INET interfaces. The ROR then 2252 includes the lower 32 bits of its own ULA (or the ULA of the Proxy, 2253 for Proxyed interfaces) as the LHS ID, encodes the ULA prefix length 2254 code in the SRT field and sets the FMT code accordingly as specified 2255 in Section 3.3. 2257 The ROR then sets the NA message R flag to 1 (as a router), S flag to 2258 1 (as a response to a solicitation), and O flag to 0 (as a proxy). 2259 The ROR finally encapsulates the NA message in a SPAN header with 2260 source set to its own ULA and destination set to the source ULA of 2261 the NS(AR) message, then forwards the message into the spanning tree 2262 without decrementing the network-layer TTL/Hop Limit field. 2264 3.14.4. Relaying the NA 2266 When the Bridge receives the NA message from the ROR, it discards the 2267 INET header and determines that the ROS is the next hop by consulting 2268 its standard IPv6 forwarding table for the SPAN header destination 2269 address. The Bridge then forwards the SPAN-encapsulated NA message 2270 toward the ROS the same as for any IPv6 router. The final-hop Bridge 2271 in the spanning tree will deliver the message via a secured tunnel to 2272 the ROS. 2274 3.14.5. Processing the NA 2276 When the ROS receives the solicited NA message, it processes the 2277 message the same as for standard IPv6 Address Resolution [RFC4861]. 2278 In the process, it caches the source ULA then creates an asymmetric 2279 neighbor cache entry for the ROR and caches all information found in 2280 the OMNI and TLLAO options. The ROS finally sets the asymmetric 2281 neighbor cache entry lifetime to ReachableTime seconds. 2283 3.14.6. Route Optimization Maintenance 2285 Following route optimization, the ROS forwards future data packets 2286 destined to the target via the addresses found in the cached link- 2287 layer information. The route optimization is shared by all sources 2288 that send packets to the target via the ROS, i.e., and not just the 2289 source on behalf of which the route optimization was initiated. 2291 While new data packets destined to the target are flowing through the 2292 ROS, it sends additional NS(AR) messages to the ROR before 2293 ReachableTime expires to receive a fresh solicited NA message the 2294 same as described in the previous sections (route optimization 2295 refreshment strategies are an implementation matter, with a non- 2296 normative example given in Appendix A.1). The ROS uses the cached 2297 ULA of the ROR as the NS(AR) SPAN destination address, and sends up 2298 to MAX_MULTICAST_SOLICIT NS(AR) messages separated by 1 second until 2299 an NA is received. If no NA is received, the ROS assumes that the 2300 current ROR has become unreachable and deletes the neighbor cache 2301 entry. Subsequent data packets will trigger a new route optimization 2302 per Section 3.14.1 to discover a new ROR while initial data packets 2303 travel over a suboptimal route. 2305 If an NA is received, the ROS then updates the asymmetric neighbor 2306 cache entry to refresh ReachableTime, while (for MNP destinations) 2307 the ROR adds or updates the ROS address to the target Client's Report 2308 List and with time set to ReportTime. While no data packets are 2309 flowing, the ROS instead allows ReachableTime for the asymmetric 2310 neighbor cache entry to expire. When ReachableTime expires, the ROS 2311 deletes the asymmetric neighbor cache entry. Any future data packets 2312 flowing through the ROS will again trigger a new route optimization. 2314 The ROS may also receive unsolicited NA messages from the ROR at any 2315 time (see: Section 3.16). If there is an asymmetric neighbor cache 2316 entry for the target, the ROS updates the link-layer information but 2317 does not update ReachableTime since the receipt of an unsolicited NA 2318 does not confirm that any forward paths are working. If there is no 2319 asymmetric neighbor cache entry, the ROS simply discards the 2320 unsolicited NA. 2322 In this arrangement, the ROS holds an asymmetric neighbor cache entry 2323 for the ROR, but the ROR does not hold an asymmetric neighbor cache 2324 entry for the ROS. The route optimization neighbor relationship is 2325 therefore asymmetric and unidirectional. If the target node also has 2326 packets to send back to the source node, then a separate route 2327 optimization procedure is performed in the reverse direction. But, 2328 there is no requirement that the forward and reverse paths be 2329 symmetric. 2331 3.15. Neighbor Unreachability Detection (NUD) 2333 AERO nodes perform Neighbor Unreachability Detection (NUD) per 2334 [RFC4861] either reactively in response to persistent link-layer 2335 errors (see Section 3.11) or proactively to confirm reachability. 2336 The NUD algorithm is based on periodic control message exchanges. 2337 The algorithm may further be seeded by ND hints of forward progress, 2338 but care must be taken to avoid inferring reachability based on 2339 spoofed information. For example, authentic IPv6 ND message 2340 exchanges may be considered as acceptable hints of forward progress, 2341 while spurious data packets should not be. 2343 AERO Servers, Proxys and Relays can use standard NS/NA NUD exchanges 2344 sent over the spanning tree to securely test reachability without 2345 risk of DoS attacks from nodes pretending to be a neighbor; Proxys 2346 can further perform NUD to securely verify Server reachability on 2347 behalf of their proxyed Clients. However, a means for a ROS to test 2348 the unsecured forward directions of target route optimized paths is 2349 also necessary. 2351 When an ROR directs an ROS to a neighbor with one or more target 2352 link-layer addresses, the ROS can proactively test each such 2353 unsecured route optimized path by sending "loopback" NS(NUD) 2354 messages. While testing the paths, the ROS can optionally continue 2355 to send packets via the spanning tree, maintain a small queue of 2356 packets until target reachability is confirmed, or (optimistically) 2357 allow packets to flow via the route optimized paths. 2359 When the ROS sends a loopback NS(NUD) message, it uses its LLA as 2360 both the IPv6 source and destination address, and the MNP Subnet- 2361 Router anycast address as the Target Address. The ROS includes a 2362 Nonce and Timestamp option, then encapsulates the message in SPAN/ 2363 INET headers with its own ULA as the source and the ULA of the route 2364 optimization target as the destination. The ROS then forwards the 2365 message to the target (either directly to the L2ADDR of the target if 2366 the target is in the same OMNI link segment, or via a Bridge if the 2367 target is in a different OMNI link segment). 2369 When the route optimization target receives the NS(NUD) message, it 2370 notices that the IPv6 destination address is the same as the source 2371 address. It then reverses the SPAN source and destination addresses 2372 and returns the message to the ROS (either directly or via the 2373 spanning tree). The route optimization target does not decrement the 2374 NS(NUD) message IPv6 Hop-Limit in the process, since the message has 2375 not exited the OMNI link. 2377 When the ROS receives the NS(NUD) message, it can determine from the 2378 Nonce, Timestamp and Target Address that the message originated from 2379 itself and that it transited the forward path. The ROS need not 2380 prepare an NA response, since the destination of the response would 2381 be itself and testing the route optimization path again would be 2382 redundant. 2384 The ROS marks route optimization target paths that pass these NUD 2385 tests as "reachable", and those that do not as "unreachable". These 2386 markings inform the OMNI interface forwarding algorithm specified in 2387 Section 3.10. 2389 Note that to avoid a DoS vector nodes MUST NOT return loopback 2390 NS(NUD) messages received from an unsecured link-layer source via the 2391 spanning tree. 2393 3.16. Mobility Management and Quality of Service (QoS) 2395 AERO is a Distributed Mobility Management (DMM) service. Each Server 2396 is responsible for only a subset of the Clients on the OMNI link, as 2397 opposed to a Centralized Mobility Management (CMM) service where 2398 there is a single network mobility collective entity for all Clients. 2399 Clients coordinate with their associated Servers via RS/RA exchanges 2400 to maintain the DMM profile, and the AERO routing system tracks all 2401 current Client/Server peering relationships. 2403 Servers provide default routing and mobility/multilink services for 2404 their dependent Clients. Clients are responsible for maintaining 2405 neighbor relationships with their Servers through periodic RS/RA 2406 exchanges, which also serves to confirm neighbor reachability. When 2407 a Client's underlying interface address and/or QoS information 2408 changes, the Client is responsible for updating the Server with this 2409 new information. Note that for Proxyed interfaces, however, the 2410 Proxy can also perform some RS/RA exchanges on the Client's behalf. 2412 Mobility management considerations are specified in the following 2413 sections. 2415 3.16.1. Mobility Update Messaging 2417 Servers accommodate Client mobility/multilink and/or QoS change 2418 events by sending unsolicited NA (uNA) messages to each ROS in the 2419 target Client's Report List. When a Server sends a uNA message, it 2420 sets the IPv6 source address to the Client's LLA, sets the 2421 destination address to (link-local) All-Nodes multicast and sets the 2422 Target Address to the Client's Subnet-Router anycast address. The 2423 Server also includes an OMNI option with prefix registration 2424 information and with ifIndex-tuples for the target Client's remaining 2425 interfaces. The Server then includes a TLLAO with corresponding 2426 ifIndex-tuples prepared the same as for the initial route 2427 optimization event. The Server sets the NA R flag to 1, the S flag 2428 to 0 and the O flag to 0, then encapsulates the message in a SPAN 2429 header with source set to its own ULA and destination set to the ULA 2430 of the ROS and sends the message into the spanning tree. 2432 As discussed in Section 7.2.6 of [RFC4861], the transmission and 2433 reception of uNA messages is unreliable but provides a useful 2434 optimization. In well-connected Internetworks with robust data links 2435 uNA messages will be delivered with high probability, but in any case 2436 the Server can optionally send up to MAX_NEIGHBOR_ADVERTISEMENT uNAs 2437 to each ROS to increase the likelihood that at least one will be 2438 received. 2440 When the ROS receives a uNA message, it ignores the message if there 2441 is no existing neighbor cache entry for the Client. Otherwise, it 2442 uses the included OMNI option and TLLAO information to update the 2443 neighbor cache entry, but does not reset ReachableTime since the 2444 receipt of an unsolicited NA message from the target Server does not 2445 provide confirmation that any forward paths to the target Client are 2446 working. 2448 If uNA messages are lost, the ROS may be left with stale address and/ 2449 or QoS information for the Client for up to ReachableTime seconds. 2450 During this time, the ROS can continue sending packets according to 2451 its stale neighbor cache information. When ReachableTime is close to 2452 expiring, the ROS will re-initiate route optimization and receive 2453 fresh link-layer address information. 2455 In addition to sending uNA messages to the current set of ROSs for 2456 the Client, the Server also sends uNAs to the former link-layer 2457 address for any ifIndex-tuple for which the link-layer address has 2458 changed. The uNA messages update Proxys that cannot easily detect 2459 (e.g., without active probing) when a formerly-active Client has 2460 departed. 2462 3.16.2. Announcing Link-Layer Address and/or QoS Preference Changes 2464 When a Client needs to change its underlying interface addresses and/ 2465 or QoS preferences (e.g., due to a mobility event), either the Client 2466 or its Proxys send RS messages to the Server via the spanning tree 2467 with an OMNI option that includes an ifIndex-tuple with the new link 2468 quality and address information. 2470 Up to MAX_RTR_SOLICITATIONS RS messages MAY be sent in parallel with 2471 sending actual data packets in case one or more RAs are lost. If all 2472 RAs are lost, the Client SHOULD re-associate with a new Server. 2474 When the Server receives the Client's changes, it sends uNA messages 2475 to all nodes in the Report List the same as described in the previous 2476 section. 2478 3.16.3. Bringing New Links Into Service 2480 When a Client needs to bring new underlying interfaces into service 2481 (e.g., when it activates a new data link), it sends an RS message to 2482 the Server via the underlying interface with an OMNI option that 2483 includes an ifIndex-tuple with appropriate link quality values and 2484 with link-layer address information for the new link. 2486 3.16.4. Removing Existing Links from Service 2488 When a Client needs to remove existing underlying interfaces from 2489 service (e.g., when it de-activates an existing data link), it sends 2490 an RS or uNA message to its Server with an OMNI option with 2491 appropriate link quality values. 2493 If the Client needs to send RS/uNA messages over an underlying 2494 interface other than the one being removed from service, it MUST 2495 include ifIndex-tuples with appropriate link quality values for any 2496 underlying interfaces being removed from service. 2498 3.16.5. Moving to a New Server 2500 When a Client associates with a new Server, it performs the Client 2501 procedures specified in Section 3.12.2. The Client also includes MS- 2502 Release identifiers in the RS message OMNI option per 2503 [I-D.templin-6man-omni-interface] if it wants the new Server to 2504 notify any old Servers from which the Client is departing. 2506 When the new Server receives the Client's RS message, it returns an 2507 RA as specified in Section 3.12.3 and sends up to 2508 MAX_NEIGHBOR_ADVERTIISEMENT uNA messages to any old Servers listed in 2509 OMNI option MS-Release identifiers. Each uNA message includes the 2510 Client's LLA as the source address, the old Server's LLA as the 2511 destination address, and an OMNI option with the Register/Release bit 2512 set to 0. The new Server wraps the uNA in a SPAN header with its own 2513 ULA as the source and the old Server's ULA as the destination, then 2514 sends the message into the spanning tree. 2516 When an old Server receives the uNA, it changes the Client's neighbor 2517 cache entry state to DEPARTED, sets the link-layer address of the 2518 Client to the new Server's ULA, and resets DepartTime. After a short 2519 delay (e.g., 2 seconds) the old Server withdraws the Client's MNP 2520 from the routing system. After DepartTime expires, the old Server 2521 deletes the Client's neighbor cache entry. 2523 The old Server also sends unsolicited NA messages to all ROSs in the 2524 Client's Report List with an OMNI option with a single ifIndex-tuple 2525 with ifIndex set to 0, and with the ULA of the new Server in a 2526 companion TLLAO. When the ROS receives the NA, it caches the address 2527 of the new Server in the existing asymmetric neighbor cache entry and 2528 marks the entry as STALE for a period of 10 seconds after which the 2529 cache entry is deleted. While in the STALE state, subsequent data 2530 packets flow according to any existing cached link-layer information 2531 and trigger a new NS(AR)/NA exchange via the new Server. 2533 Clients SHOULD NOT move rapidly between Servers in order to avoid 2534 causing excessive oscillations in the AERO routing system. Examples 2535 of when a Client might wish to change to a different Server include a 2536 Server that has gone unreachable, topological movements of 2537 significant distance, movement to a new geographic region, movement 2538 to a new OMNI link segment, etc. 2540 When a Client moves to a new Server, some of the fragments of a 2541 multiple fragment packet may have already arrived at the old Server 2542 while others are en route to the new Server, however no special 2543 attention in the reassembly algorithm is necessary when re-routed 2544 fragments are simply treated as loss. 2546 3.17. Multicast 2548 The AERO Client provides an IGMP (IPv4) [RFC2236] or MLD (IPv6) 2549 [RFC3810] proxy service for its EUNs and/or hosted applications 2550 [RFC4605]. The Client forwards IGMP/MLD messages over any of its 2551 underlying interfaces for which group membership is required. The 2552 IGMP/MLD messages may be further forwarded by a first-hop ANET access 2553 router acting as an IGMP/MLD-snooping switch [RFC4541], then 2554 ultimately delivered to an AERO Proxy/Server acting as a Protocol 2555 Independent Multicast - Sparse-Mode (PIM-SM, or simply "PIM") 2556 Designated Router (DR) [RFC7761]. AERO Relays also act as PIM 2557 routers (i.e., the same as AERO Proxys/Servers) on behalf of nodes on 2558 INET/EUN networks. The behaviors identified in the following 2559 sections correspond to Source-Specific Multicast (SSM) and Any-Source 2560 Multicast (ASM) operational modes. 2562 3.17.1. Source-Specific Multicast (SSM) 2564 When an ROS (i.e., an AERO Proxy/Server/Relay) "X" acting as PIM 2565 router receives a Join/Prune message from a node on its downstream 2566 interfaces containing one or more ((S)ource, (G)roup) pairs, it 2567 updates its Multicast Routing Information Base (MRIB) accordingly. 2568 For each S belonging to a prefix reachable via X's non-OMNI 2569 interfaces, X then forwards the (S, G) Join/Prune to any PIM routers 2570 on those interfaces per [RFC7761]. 2572 For each S belonging to a prefix reachable via X's OMNI interface, X 2573 originates a separate copy of the Join/Prune for each (S,G) in the 2574 message using its own LLA as the source address and ALL-PIM-ROUTERS 2575 as the destination address. X then encapsulates each message in a 2576 SPAN header with source address set to the ULA of X and destination 2577 address set to S then forwards the message into the spanning tree, 2578 which delivers it to AERO Server/Relay "Y" that services S. At the 2579 same time, if the message was a Join, X sends a route-optimization NS 2580 message toward each S the same as discussed in Section 3.14. The 2581 resulting NAs will return the LLA for the prefix that matches S as 2582 the network-layer source address and TLLAOs with the ULA 2583 corresponding to any ifIndex-tuples that are currently servicing S. 2585 When Y processes the Join/Prune message, if S located behind any 2586 INET, Direct, or VPNed interfaces Y acts as a PIM router and updates 2587 its MRIB to list X as the next hop in the reverse path. If S is 2588 located behind any Proxys "Z"*, Y also forwards the message to each 2589 Z* over the spanning tree while continuing to use the LLA of X as the 2590 source address. Each Z* then updates its MRIB accordingly and 2591 maintains the LLA of X as the next hop in the reverse path. Since 2592 the Bridges do not examine network layer control messages, this means 2593 that the (reverse) multicast tree path is simply from each Z* (and/or 2594 Y) to X with no other multicast-aware routers in the path. If any Z* 2595 (and/or Y) is located on the same OMNI link segment as X, the 2596 multicast data traffic sent to X directly using SPAN/INET 2597 encapsulation instead of via a Bridge. 2599 Following the initial Join/Prune and NS/NA messaging, X maintains an 2600 asymmetric neighbor cache entry for each S the same as if X was 2601 sending unicast data traffic to S. In particular, X performs 2602 additional NS/NA exchanges to keep the neighbor cache entry alive for 2603 up to t_periodic seconds [RFC7761]. If no new Joins are received 2604 within t_periodic seconds, X allows the neighbor cache entry to 2605 expire. Finally, if X receives any additional Join/Prune messages 2606 for (S,G) it forwards the messages to each Y and Z* in the neighbor 2607 cache entry over the spanning tree. 2609 At some later time, Client C that holds an MNP for source S may 2610 depart from a first Proxy Z1 and/or connect via a new Proxy Z2. In 2611 that case, Y sends an unsolicited NA message to X the same as 2612 specified for unicast mobility in Section 3.16. When X receives the 2613 unsolicited NA message, it updates its asymmetric neighbor cache 2614 entry for the LLA for source S and sends new Join messages to any new 2615 Proxys Z2. There is no requirement to send any Prune messages to old 2616 Proxys Z1 since source S will no longer source any multicast data 2617 traffic via Z1. Instead, the multicast state for (S,G) in Proxy Z1 2618 will soon time out since no new Joins will arrive. 2620 After some later time, C may move to a new Server Y2 and depart from 2621 old Sever Y1. In that case, Y1 sends Join messages for any of C's 2622 active (S,G) groups to Y2 while including its own LLA as the source 2623 address. This causes Y2 to include Y1 in the multicast forwarding 2624 tree during the interim time that Y1's symmetric neighbor cache entry 2625 for C is in the DEPARTED state. At the same time, Y1 sends an 2626 unsolicited NA message to X with an OMNI option and TLLAO with 2627 ifIndex-tuple set to 0 and a release indication to cause X to release 2628 its asymmetric neighbor cache entry. X then sends a new Join message 2629 to S via the spanning tree and re-initiates route optimization the 2630 same as if it were receiving a fresh Join message from a node on a 2631 downstream link. 2633 3.17.2. Any-Source Multicast (ASM) 2635 When an ROS X acting as a PIM router receives a Join/Prune from a 2636 node on its downstream interfaces containing one or more (*,G) pairs, 2637 it updates its Multicast Routing Information Base (MRIB) accordingly. 2638 X then forwards a copy of the message to the Rendezvous Point (RP) R 2639 for each G over the spanning tree. X uses its own LLA as the source 2640 address and ALL-PIM-ROUTERS as the destination address, then 2641 encapsulates each message in a SPAN header with source address set to 2642 the ULA of X and destination address set to R, then sends the message 2643 into the spanning tree. At the same time, if the message was a Join 2644 X initiates NS/NA route optimization the same as for the SSM case 2645 discussed in Section 3.17.1. 2647 For each source S that sends multicast traffic to group G via R, the 2648 Proxy/Server Z* for the Client that aggregates S encapsulates the 2649 packets in PIM Register messages and forwards them to R via the 2650 spanning tree, which may then elect to send a PIM Join to Z*. This 2651 will result in an (S,G) tree rooted at Z* with R as the next hop so 2652 that R will begin to receive two copies of the packet; one native 2653 copy from the (S, G) tree and a second copy from the pre-existing (*, 2654 G) tree that still uses PIM Register encapsulation. R can then issue 2655 a PIM Register-stop message to suppress the Register-encapsulated 2656 stream. At some later time, if C moves to a new Proxy/Server Z*, it 2657 resumes sending packets via PIM Register encapsulation via the new 2658 Z*. 2660 At the same time, as multicast listeners discover individual S's for 2661 a given G, they can initiate an (S,G) Join for each S under the same 2662 procedures discussed in Section 3.17.1. Once the (S,G) tree is 2663 established, the listeners can send (S, G) Prune messages to R so 2664 that multicast packets for group G sourced by S will only be 2665 delivered via the (S, G) tree and not from the (*, G) tree rooted at 2666 R. All mobility considerations discussed for SSM apply. 2668 3.17.3. Bi-Directional PIM (BIDIR-PIM) 2670 Bi-Directional PIM (BIDIR-PIM) [RFC5015] provides an alternate 2671 approach to ASM that treats the Rendezvous Point (RP) as a Designated 2672 Forwarder (DF). Further considerations for BIDIR-PIM are out of 2673 scope. 2675 3.18. Operation over Multiple OMNI Links 2677 An AERO Client can connect to multiple OMNI links the same as for any 2678 data link service. In that case, the Client maintains a distinct 2679 OMNI interface for each link, e.g., 'omni0' for the first link, 2680 'omni1' for the second, 'omni2' for the third, etc. Each OMNI link 2681 would include its own distinct set of Bridges, Servers and Proxys, 2682 thereby providing redundancy in case of failures. 2684 The Bridges, Servers and Proxys on each OMNI link can assign AERO and 2685 ULAs that use the same or different numberings from those on other 2686 links. Since the links are mutually independent there is no 2687 requirement for avoiding inter-link address duplication, e.g., the 2688 same LLA such as fe80::1000 could be used to number distinct nodes 2689 that connect to different OMNI links. 2691 Each OMNI link could utilize the same or different ANET connections. 2692 The links can be distinguished at the link-layer via the SRT prefix 2693 in a similar fashion as for Virtual Local Area Network (VLAN) tagging 2694 (e.g., IEEE 802.1Q) and/or through assignment of distinct sets of 2695 MSPs on each link. This gives rise to the opportunity for supporting 2696 multiple redundant networked paths, with each VLAN distinguished by a 2697 different SRT "color" (see: Section 3.2.5). 2699 The Client's IP layer can select the outgoing OMNI interface 2700 appropriate for a given traffic profile while (in the reverse 2701 direction) correspondent nodes must have some way of steering their 2702 packets destined to a target via the correct OMNI link. 2704 In a first alternative, if each OMNI link services different MSPs, 2705 then the Client can receive a distinct MNP from each of the links. 2706 IP routing will therefore assure that the correct Red/Green/Blue/etc. 2707 network is used for both outbound and inbound traffic. This can be 2708 accomplished using existing technologies and approaches, and without 2709 requiring any special supporting code in correspondent nodes or 2710 Bridges. 2712 In a second alternative, if each OMNI link services the same MSP(s) 2713 then each link could assign a distinct "OMNI link Anycast" address 2714 that is configured by all Bridges on the link. Correspondent nodes 2715 can then perform Segment Routing to select the correct SRT, which 2716 will then direct the packet over multiple hops to the target. 2718 3.19. DNS Considerations 2720 AERO Client MNs and INET correspondent nodes consult the Domain Name 2721 System (DNS) the same as for any Internetworking node. When 2722 correspondent nodes and Client MNs use different IP protocol versions 2723 (e.g., IPv4 correspondents and IPv6 MNs), the INET DNS must maintain 2724 A records for IPv4 address mappings to MNs which must then be 2725 populated in Relay NAT64 mapping caches. In that way, an IPv4 2726 correspondent node can send packets to the IPv4 address mapping of 2727 the target MN, and the Relay will translate the IPv4 header and 2728 destination address into an IPv6 header and IPv6 destination address 2729 of the MN. 2731 When an AERO Client registers with an AERO Server, the Server can 2732 return the address(es) of DNS servers in RDNSS options [RFC6106]. 2733 The DNS server provides the IP addresses of other MNs and 2734 correspondent nodes in AAAA records for IPv6 or A records for IPv4. 2736 3.20. Transition Considerations 2738 SPAN encapsulation ensures that dissimilar INET partitions can be 2739 joined into a single unified OMNI link, even though the partitions 2740 themselves may have differing protocol versions and/or incompatible 2741 addressing plans. However, a commonality can be achieved by 2742 incrementally distributing globally routable (i.e., native) IP 2743 prefixes to eventually reach all nodes (both mobile and fixed) in all 2744 OMNI link segments. This can be accomplished by incrementally 2745 deploying AERO Relays on each INET partition, with each Relay 2746 distributing its MNPs and/or discovering non-MNP prefixes on its INET 2747 links. 2749 This gives rise to the opportunity to eventually distribute native IP 2750 addresses to all nodes, and to present a unified OMNI link view even 2751 if the INET partitions remain in their current protocol and 2752 addressing plans. In that way, the OMNI link can serve the dual 2753 purpose of providing a mobility/multilink service and a transition 2754 service. Or, if an INET partition is transitioned to a native IP 2755 protocol version and addressing scheme that is compatible with the 2756 OMNI link MNP-based addressing scheme, the partition and OMNI link 2757 can be joined by Relays. 2759 Relays that connect INETs/EUNs with dissimilar IP protocol versions 2760 may need to employ a network address and protocol translation 2761 function such as NAT64[RFC6146]. 2763 3.21. Detecting and Reacting to Server and Bridge Failures 2765 In environments where rapid failure recovery is required, Servers and 2766 Bridges SHOULD use Bidirectional Forwarding Detection (BFD) 2767 [RFC5880]. Nodes that use BFD can quickly detect and react to 2768 failures so that cached information is re-established through 2769 alternate nodes. BFD control messaging is carried only over well- 2770 connected ground domain networks (i.e., and not low-end radio links) 2771 and can therefore be tuned for rapid response. 2773 Servers and Bridges maintain BFD sessions in parallel with their BGP 2774 peerings. If a Server or Bridge fails, BGP peers will quickly re- 2775 establish routes through alternate paths the same as for common BGP 2776 deployments. Similarly, Proxys maintain BFD sessions with their 2777 associated Bridges even though they do not establish BGP peerings 2778 with them. 2780 Proxys SHOULD use proactive NUD for Servers for which there are 2781 currently active ANET Clients in a manner that parallels BFD, i.e., 2782 by sending unicast NS messages in rapid succession to receive 2783 solicited NA messages. When the Proxy is also sending RS messages on 2784 behalf of ANET Clients, the RS/RA messaging can be considered as 2785 equivalent hints of forward progress. This means that the Proxy need 2786 not also send a periodic NS if it has already sent an RS within the 2787 same period. If a Server fails, the Proxy will cease to receive 2788 advertisements and can quickly inform Clients of the outage by 2789 sending multicast RA messages on the ANET interface. 2791 The Proxy sends multicast RA messages with source address set to the 2792 Server's address, destination address set to (link-local) All-Nodes 2793 multicast, and Router Lifetime set to 0. The Proxy SHOULD send 2794 MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small delays 2795 [RFC4861]. Any Clients on the ANET interface that have been using 2796 the (now defunct) Server will receive the RA messages and associate 2797 with a new Server. 2799 3.22. AERO Clients on the Open Internet 2801 AERO Clients that connect to the open Internet via INET interfaces 2802 can establish a VPN or direct link to securely connect to a Server in 2803 a "tethered" arrangement with all of the Client's traffic transiting 2804 the Server. Alternatively, the Client can associate with an INET 2805 Server using UDP/IP encapsulation and asymmetric securing services as 2806 discussed in the following sections. 2808 When a Client's OMNI interface enables an INET underlying interface, 2809 it first determines whether the interface is likely to be behind a 2810 NAT. For IPv4, the Client assumes it is on the open Internet if the 2811 INET address is not a special-use IPv4 address per [RFC3330]. 2812 Similarly for IPv6, the Client assumes it is on the open Internet if 2813 the INET address is not a link-local [RFC4291] or unique-local 2814 [RFC4193] IPv6 address. 2816 The Client then prepares a UDP/IP-encapsulated RS message with IPv6 2817 source address set to its LLA, with IPv6 destination set to (link- 2818 local) All-Routers multicast and with an OMNI option with underlying 2819 interface parameters. If the Client believes that it is on the open 2820 Internet, it SHOULD also include an SLLAO set according to the 2821 address used for INET encapsulation (otherwise, it MAY omit the 2822 SLLAO). If the underlying address is IPv4, the Client includes the 2823 Port Number and IPv4 address written in obfuscated form [RFC4380] as 2824 discussed in Section 3.3. If the underlying interface address is 2825 IPv6, the Client instead includes the Port Number and IPv6 address in 2826 obfuscated form. The Client finally includes an Authentication 2827 option per [RFC4380] to provide message authentication, sets the UDP/ 2828 IP source to its INET address and UDP port, sets the UDP/IP 2829 destination to the Server's INET address and the AERO service port 2830 number (8060), then sends the message to the Server. 2832 When the Server receives the RS, it authenticates the message and 2833 registers the Client's MNP and INET interface information according 2834 to the OMNI option parameters. If the RS message includes an SLLAO, 2835 the Server compares the encapsulation IP address and UDP port number 2836 with the (unobfuscated) SLLAO values. If the values are the same, 2837 the Server caches the Client's information as "INET" addresses 2838 meaning that the Client is likely to accept direct messages without 2839 requiring NAT traversal exchanges. If the values are different (or, 2840 if there was no SLLAO) the Server instead caches the Client's 2841 information as "NAT" addresses meaning that NAT traversal exchanges 2842 may be necessary. 2844 The Server then returns an RA message with IPv6 source and 2845 destination set corresponding to the addresses in the RS, and with an 2846 Authentication option per [RFC4380]. For IPv4, the Server also 2847 includes an Origin option per [RFC4380] with the mapped and 2848 obfuscated Port Number and IPv4 address observed in the encapsulation 2849 headers. For IPv6, the Server instead includes an IPv6 Origin option 2850 per Figure 7 with the mapped and obfuscated observed Port Number and 2851 IPv6 address (note that the value 0x02 in the second octet 2852 differentiates from other [RFC4380] option types). 2854 +--------+--------+-----------------+ 2855 | 0x00 | 0x02 | Origin port # | 2856 +--------+--------+-----------------+ 2857 ~ Origin IPv6 address ~ 2858 +-----------------------------------+ 2860 Figure 7: IPv6 Origin Option 2862 When the Client receives the RA message, it compares the mapped Port 2863 Number and IP address from the Origin option with its own address. 2864 If the addresses are the same, the Client assumes the open Internet / 2865 Cone NAT principle; if the addresses are different, the Client 2866 instead assumes that further qualification procedures are necessary 2867 to detect the type of NAT and proceeds according to standard 2868 [RFC4380] procedures. 2870 After the Client has registered its INET interfaces in such RS/RA 2871 exchanges it sends periodic RS messages to receive fresh RA messages 2872 before the Router Lifetime received on each INET interface expires. 2873 The Client also maintains default routes via its Servers, i.e., the 2874 same as described in earlier sections. 2876 When the Client sends messages to target IP addresses, it also 2877 invokes route optimization per Section 3.14 using IPv6 ND address 2878 resolution messaging. The Client sends the NS(AR) message to the 2879 Server wrapped in a UDP/IP header with an Authentication option with 2880 the NS source address set to the Client's LLA and destination address 2881 set to the target solicited node multicast address. The Server 2882 authenticates the message and sends a corresponding NS(AR) message 2883 over the spanning tree the same as if it were the ROS, but with the 2884 SPAN source address set to the Server's ULA and destination set to 2885 the ULA of the target. When the ROR receives the NS(AR), it adds the 2886 Server's ULA and Client's LLA to the target's Report List, and 2887 returns an NA with OMNI and TLLAO information for the target. The 2888 Server then returns a UDP/IP encapsulated NA message with an 2889 Authentication option to the Client. 2891 Following route optimization, for targets in the same OMNI link 2892 segment if the target's TLLAO addresss is on the open INET, the 2893 Client forwards data packets directly to the target INET address. If 2894 the target's TLLAO address is behind a NAT, the Client first 2895 establishes NAT state for the L2ADDR using the "bubble" mechanisms 2896 specified in [RFC6081][RFC4380]. The Client continues to send data 2897 packets via its Server until NAT state is populated, then begins 2898 forwarding packets via the direct path through the NAT to the target. 2899 For targets in different OMNI link segments, the Client inserts an 2900 SRH and forwards data packets to the Bridge that returned the NA 2901 message. 2903 The ROR may return uNAs via the Server if the target moves, and the 2904 Server will send corresponding Authentication-protected uNAs to the 2905 Client. The Client can also send "loopback" NS(NUD) messages to test 2906 forward path reachability even though there is no security 2907 association between the Client and the target. 2909 The Client sends UDP/IP encapsulated IPv6 packets no larger than 1280 2910 bytes in one piece. In order to accommodate larger IPv6 packets (up 2911 to the OMNI interface MTU), the Client inserts a SPAN header with 2912 source set to its own ULA and destination set to the ULA of the 2913 target and uses IPv6 fragmentation according to Section 3.9. The 2914 Client then encapsulates each fragment in a UDP/IP header and sends 2915 the fragments to the next hop. 2917 3.22.1. Use of SEND and CGA 2919 In some environments, use of the [RFC4380] Authentication option 2920 alone may be sufficient for assuring IPv6 ND message authentication 2921 between Clients and Servers. When additional protection is 2922 necessary, nodes should employ SEcure Neighbor Discovery (SEND) 2923 [RFC3971] with Cryptographically-Generated Addresses (CGA) [RFC3972]. 2925 When SEND/CGA are used, the Client prepares RS messages with its 2926 link-local CGA as the IPv6 source and (link-local) All-Routers 2927 multicast as the IPv6 Destination, includes any SEND options and 2928 wraps the message in a SPAN header. The Client sets the SPAN source 2929 address to its own ULA and sets the SPAN destination address to 2930 (site-local) All-Routers multicast. The Client then wraps the RS 2931 message in UDP/IP headers according to [RFC4380] and sends the 2932 message to the Server. 2934 When the Server receives the message, it first verifies the 2935 Authentication option (if present) then uses the SPAN source address 2936 to determine the MNP of the Client. The Server then processes the 2937 SEND options to authenticate the RS message and prepares an RA 2938 message response. The Server prepares the RA with its own link-local 2939 CGA as the IPv6 source and the CGA of the Client as the IPv6 2940 destination, includes any SEND options and wraps the message in a 2941 SPAN header. The Server sets the SPAN source address to its own ULA 2942 and sets the SPAN destination address to the Client's ULA. The 2943 Server then wraps the RA message in UDP/IP headers according to 2944 [RFC4380] and sends the message to the Client. Thereafter, the 2945 Client/Server send additional RS/RA messages to maintain their 2946 association and any NAT state. 2948 The Client and Server also may exchange NS/NA messages using their 2949 own CGA as the source and with SPAN encapsulation as above. When a 2950 Client sends an NS(AR), it sets the IPv6 source to its CGA and sets 2951 the IPv6 destination to the Solicited-Node Multicast address of the 2952 target. The Client then wraps the message in a SPAN header with its 2953 own ULA as the source and the ULA of the target as the destination 2954 and sends it to the Server. The Server authenticates the message, 2955 then changes the IPv6 source address to the Client's LLA, removes the 2956 SEND options, and sends a corresponding NS(AR) into the spanning 2957 tree. When the Server receives the corresponding SPAN-encapsulated 2958 NA, it changes the IPv6 destination address to the Client's CGA, 2959 inserts SEND options, then wraps the message in UDP/IP headers and 2960 sends it to the Client. 2962 When a Client sends a uNA, it sets the IPv6 source address to its own 2963 CGA and sets the IPv6 destination address to (link-local) All-Nodes 2964 multicast, includes SEND options, wraps the message in SPAN and UDP/ 2965 IP headers and sends the message to the Server. The Server 2966 authenticates the message, then changes the IPv6 address to the 2967 Client's LLA, removes the SEND options and forwards the message the 2968 same as discussed in Section 3.16.1. In the reverse direction, when 2969 the Server forwards a uNA to the Client, it changes the IPv6 address 2970 to its own CGA and inserts SEND options then forwards the message to 2971 the Client. 2973 When a Client sends an NS(NUD), it sets both the IPv6 source and 2974 destination address to its own LLA, wraps the message in a SPAN 2975 header and UDP/IP headers, then sends the message directly to the 2976 peer which will loop the message back. In this case alone, the 2977 Client does not use the Server as a trust broker for forwarding the 2978 ND message. 2980 3.23. Time-Varying MNPs 2982 In some use cases, it is desirable, beneficial and efficient for the 2983 Client to receive a constant MNP that travels with the Client 2984 wherever it moves. For example, this would allow air traffic 2985 controllers to easily track aircraft, etc. In other cases, however 2986 (e.g., intelligent transportation systems), the MN may be willing to 2987 sacrifice a modicum of efficiency in order to have time-varying MNPs 2988 that can be changed every so often to defeat adversarial tracking. 2990 The DHCPv6-PD service offers a way for Clients that desire time- 2991 varying MNPs to obtain short-lived prefixes (e.g., on the order of a 2992 small number of minutes). In that case, the identity of the Client 2993 would not be bound to the MNP but rather the Client's identity would 2994 be bound to the DHCPv6 Device Unique Identifier (DUID) and used as 2995 the seed for Prefix Delegation. The Client would then be obligated 2996 to renumber its internal networks whenever its MNP (and therefore 2997 also its LLA) changes. This should not present a challenge for 2998 Clients with automated network renumbering services, however presents 2999 limits for the durations of ongoing sessions that would prefer to use 3000 a constant address. 3002 4. Implementation Status 3004 An AERO implementation based on OpenVPN (https://openvpn.net/) was 3005 announced on the v6ops mailing list on January 10, 2018 and an 3006 initial public release of the AERO proof-of-concept source code was 3007 announced on the intarea mailing list on August 21, 2015. 3009 As of 4/1/2020, more recent updated implementations are under 3010 internal development and testing with plans to release in the near 3011 future. 3013 5. IANA Considerations 3015 The IANA has assigned a 4-octet Private Enterprise Number "45282" for 3016 AERO in the "enterprise-numbers" registry. 3018 The IANA has assigned the UDP port number "8060" for an earlier 3019 experimental version of AERO [RFC6706]. This document obsoletes 3020 [RFC6706] and claims the UDP port number "8060" for all future use. 3022 The IANA is instructed to assign a new type value TBD in the Segment 3023 Routing Header TLV registry [RFC8754]. 3025 No further IANA actions are required. 3027 6. Security Considerations 3029 AERO Bridges configure secured tunnels with AERO Servers, Realys and 3030 Proxys within their local OMNI link segments. Applicable secured 3031 tunnel alternatives include IPsec [RFC4301], TLS/SSL [RFC8446], DTLS 3032 [RFC6347], WireGuard, etc. The AERO Bridges of all OMNI link 3033 segments in turn configure secured tunnels for their neighboring AERO 3034 Bridges in a spanning tree topology. Therefore, control messages 3035 exchanged between any pair of OMNI link neighbors on the spanning 3036 tree are already secured. 3038 AERO Servers, Relays and Proxys targeted by a route optimization may 3039 also receive data packets directly from arbitrary nodes in INET 3040 partitions instead of via the spanning tree. For INET partitions 3041 that apply effective ingress filtering to defeat source address 3042 spoofing, the simple data origin authentication procedures in 3043 Section 3.8 can be applied. 3045 For INET partitions that require strong security in the data plane, 3046 two options for securing communications include 1) disable route 3047 optimization so that all traffic is conveyed over secured tunnels, or 3048 2) enable on-demand secure tunnel creation between INET partition 3049 neighbors. Option 1) would result in longer routes than necessary 3050 and traffic concentration on critical infrastructure elements. 3051 Option 2) could be coordinated by establishing a secured tunnel on- 3052 demand instead of performing an NS/NA exchange in the route 3053 optimization procedures. Procedures for establishing on-demand 3054 secured tunnels are out of scope. 3056 AERO Clients that connect to secured ANETs need not apply security to 3057 their ND messages, since the messages will be intercepted by a 3058 perimeter Proxy that applies security on its INET-facing interface as 3059 part of the spanning tree (see above). AERO Clients connected to the 3060 open INET can use symmetric network and/or transport layer security 3061 services such as VPNs or can by some other means establish a direct 3062 link. When a VPN or direct link may be impractical, however, an 3063 asymmetric security service such as SEcure Neighbor Discovery (SEND) 3064 [RFC3971] with Cryptographically Generated Addresses (CGAs) [RFC3972] 3065 and/or the Authentication option [RFC4380] can be applied. 3067 Application endpoints SHOULD use application-layer security services 3068 such as TLS/SSL, DTLS or SSH [RFC4251] to assure the same level of 3069 protection as for critical secured Internet services. AERO Clients 3070 that require host-based VPN services SHOULD use symmetric network 3071 and/or transport layer security services such as IPsec, TLS/SSL, 3072 DTLS, etc. AERO Proxys and Servers can also provide a network-based 3073 VPN service on behalf of the Client, e.g., if the Client is located 3074 within a secured enclave and cannot establish a VPN on its own 3075 behalf. 3077 AERO Servers and Bridges present targets for traffic amplification 3078 Denial of Service (DoS) attacks. This concern is no different than 3079 for widely-deployed VPN security gateways in the Internet, where 3080 attackers could send spoofed packets to the gateways at high data 3081 rates. This can be mitigated by connecting Servers and Bridges over 3082 dedicated links with no connections to the Internet and/or when 3083 connections to the Internet are only permitted through well-managed 3084 firewalls. Traffic amplification DoS attacks can also target an AERO 3085 Client's low data rate links. This is a concern not only for Clients 3086 located on the open Internet but also for Clients in secured 3087 enclaves. AERO Servers and Proxys can institute rate limits that 3088 protect Clients from receiving packet floods that could DoS low data 3089 rate links. 3091 AERO Relays must implement ingress filtering to avoid a spoofing 3092 attack in which spurious messages with ULA addresses are injected 3093 into an OMNI link from an outside attacker. AERO Clients MUST ensure 3094 that their connectivity is not used by unauthorized nodes on their 3095 EUNs to gain access to a protected network, i.e., AERO Clients that 3096 act as routers MUST NOT provide routing services for unauthorized 3097 nodes. (This concern is no different than for ordinary hosts that 3098 receive an IP address delegation but then "share" the address with 3099 other nodes via some form of Internet connection sharing such as 3100 tethering.) 3102 The MAP list MUST be well-managed and secured from unauthorized 3103 tampering, even though the list contains only public information. 3104 The MAP list can be conveyed to the Client in a similar fashion as in 3105 [RFC5214] (e.g., through layer 2 data link login messaging, secure 3106 upload of a static file, DNS lookups, etc.). 3108 Although public domain and commercial SEND implementations exist, 3109 concerns regarding the strength of the cryptographic hash algorithm 3110 have been documented [RFC6273] [RFC4982]. 3112 SRH authentication facilities are specified in [RFC8754]. 3114 Security considerations for accepting link-layer ICMP messages and 3115 reflected packets are discussed throughout the document. 3117 Security considerations for IPv6 fragmentation and reassembly are 3118 discussed in [I-D.templin-6man-omni-interface]. 3120 7. Acknowledgements 3122 Discussions in the IETF, aviation standards communities and private 3123 exchanges helped shape some of the concepts in this work. 3124 Individuals who contributed insights include Mikael Abrahamsson, Mark 3125 Andrews, Fred Baker, Bob Braden, Stewart Bryant, Brian Carpenter, 3126 Wojciech Dec, Pavel Drasil, Ralph Droms, Adrian Farrel, Nick Green, 3127 Sri Gundavelli, Brian Haberman, Bernhard Haindl, Joel Halpern, Tom 3128 Herbert, Sascha Hlusiak, Lee Howard, Zdenek Jaron, Andre Kostur, 3129 Hubert Kuenig, Ted Lemon, Andy Malis, Satoru Matsushima, Tomek 3130 Mrugalski, Madhu Niraula, Alexandru Petrescu, Behcet Saikaya, Michal 3131 Skorepa, Joe Touch, Bernie Volz, Ryuji Wakikawa, Tony Whyman, Lloyd 3132 Wood and James Woodyatt. Members of the IESG also provided valuable 3133 input during their review process that greatly improved the document. 3134 Special thanks go to Stewart Bryant, Joel Halpern and Brian Haberman 3135 for their shepherding guidance during the publication of the AERO 3136 first edition. 3138 This work has further been encouraged and supported by Boeing 3139 colleagues including Kyle Bae, M. Wayne Benson, Dave Bernhardt, Cam 3140 Brodie, John Bush, Balaguruna Chidambaram, Irene Chin, Bruce Cornish, 3141 Claudiu Danilov, Don Dillenburg, Joe Dudkowski, Wen Fang, Samad 3142 Farooqui, Anthony Gregory, Jeff Holland, Seth Jahne, Brian Jaury, 3143 Greg Kimberly, Ed King, Madhuri Madhava Badgandi, Laurel Matthew, 3144 Gene MacLean III, Rob Muszkiewicz, Sean O'Sullivan, Vijay 3145 Rajagopalan, Greg Saccone, Rod Santiago, Kent Shuey, Brian Skeen, 3146 Mike Slane, Carrie Spiker, Katie Tran, Brendan Williams, Amelia 3147 Wilson, Julie Wulff, Yueli Yang, Eric Yeh and other members of the 3148 Boeing mobility, networking and autonomy teams. Kyle Bae, Wayne 3149 Benson, Katie Tran and Eric Yeh are especially acknowledged for 3150 implementing the AERO functions as extensions to the public domain 3151 OpenVPN distribution. 3153 Earlier works on NBMA tunneling approaches are found in 3154 [RFC2529][RFC5214][RFC5569]. 3156 Many of the constructs presented in this second edition of AERO are 3157 based on the author's earlier works, including: 3159 o The Internet Routing Overlay Network (IRON) 3160 [RFC6179][I-D.templin-ironbis] 3162 o Virtual Enterprise Traversal (VET) 3163 [RFC5558][I-D.templin-intarea-vet] 3165 o The Subnetwork Encapsulation and Adaptation Layer (SEAL) 3166 [RFC5320][I-D.templin-intarea-seal] 3168 o AERO, First Edition [RFC6706] 3170 Note that these works cite numerous earlier efforts that are not also 3171 cited here due to space limitations. The authors of those earlier 3172 works are acknowledged for their insights. 3174 This work is aligned with the NASA Safe Autonomous Systems Operation 3175 (SASO) program under NASA contract number NNA16BD84C. 3177 This work is aligned with the FAA as per the SE2025 contract number 3178 DTFAWA-15-D-00030. 3180 This work is aligned with the Boeing Commercial Airplanes (BCA) 3181 Internet of Things (IoT) and autonomy programs. 3183 This work is aligned with the Boeing Information Technology (BIT) 3184 MobileNet program. 3186 8. References 3188 8.1. Normative References 3190 [I-D.templin-6man-omni-interface] 3191 Templin, F. and T. Whyman, "Transmission of IPv6 Packets 3192 over Overlay Multilink Network (OMNI) Interfaces", draft- 3193 templin-6man-omni-interface-25 (work in progress), June 3194 2020. 3196 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 3197 DOI 10.17487/RFC0791, September 1981, 3198 . 3200 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 3201 RFC 792, DOI 10.17487/RFC0792, September 1981, 3202 . 3204 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 3205 Requirement Levels", BCP 14, RFC 2119, 3206 DOI 10.17487/RFC2119, March 1997, 3207 . 3209 [RFC2473] Conta, A. and S. Deering, "Generic Packet Tunneling in 3210 IPv6 Specification", RFC 2473, DOI 10.17487/RFC2473, 3211 December 1998, . 3213 [RFC3971] Arkko, J., Ed., Kempf, J., Zill, B., and P. Nikander, 3214 "SEcure Neighbor Discovery (SEND)", RFC 3971, 3215 DOI 10.17487/RFC3971, March 2005, 3216 . 3218 [RFC3972] Aura, T., "Cryptographically Generated Addresses (CGA)", 3219 RFC 3972, DOI 10.17487/RFC3972, March 2005, 3220 . 3222 [RFC4191] Draves, R. and D. Thaler, "Default Router Preferences and 3223 More-Specific Routes", RFC 4191, DOI 10.17487/RFC4191, 3224 November 2005, . 3226 [RFC4193] Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast 3227 Addresses", RFC 4193, DOI 10.17487/RFC4193, October 2005, 3228 . 3230 [RFC4380] Huitema, C., "Teredo: Tunneling IPv6 over UDP through 3231 Network Address Translations (NATs)", RFC 4380, 3232 DOI 10.17487/RFC4380, February 2006, 3233 . 3235 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 3236 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 3237 DOI 10.17487/RFC4861, September 2007, 3238 . 3240 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 3241 Address Autoconfiguration", RFC 4862, 3242 DOI 10.17487/RFC4862, September 2007, 3243 . 3245 [RFC6081] Thaler, D., "Teredo Extensions", RFC 6081, 3246 DOI 10.17487/RFC6081, January 2011, 3247 . 3249 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 3250 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 3251 May 2017, . 3253 [RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6 3254 (IPv6) Specification", STD 86, RFC 8200, 3255 DOI 10.17487/RFC8200, July 2017, 3256 . 3258 [RFC8415] Mrugalski, T., Siodelski, M., Volz, B., Yourtchenko, A., 3259 Richardson, M., Jiang, S., Lemon, T., and T. Winters, 3260 "Dynamic Host Configuration Protocol for IPv6 (DHCPv6)", 3261 RFC 8415, DOI 10.17487/RFC8415, November 2018, 3262 . 3264 8.2. Informative References 3266 [BGP] Huston, G., "BGP in 2015, http://potaroo.net", January 3267 2016. 3269 [I-D.bonica-6man-comp-rtg-hdr] 3270 Bonica, R., Kamite, Y., Niwa, T., Alston, A., and L. 3271 Jalil, "The IPv6 Compact Routing Header (CRH)", draft- 3272 bonica-6man-comp-rtg-hdr-22 (work in progress), May 2020. 3274 [I-D.bonica-6man-crh-helper-opt] 3275 Li, X., Bao, C., Ruan, E., and R. Bonica, "Compressed 3276 Routing Header (CRH) Helper Option", draft-bonica-6man- 3277 crh-helper-opt-01 (work in progress), May 2020. 3279 [I-D.ietf-intarea-frag-fragile] 3280 Bonica, R., Baker, F., Huston, G., Hinden, R., Troan, O., 3281 and F. Gont, "IP Fragmentation Considered Fragile", draft- 3282 ietf-intarea-frag-fragile-17 (work in progress), September 3283 2019. 3285 [I-D.ietf-intarea-tunnels] 3286 Touch, J. and M. Townsley, "IP Tunnels in the Internet 3287 Architecture", draft-ietf-intarea-tunnels-10 (work in 3288 progress), September 2019. 3290 [I-D.ietf-rtgwg-atn-bgp] 3291 Templin, F., Saccone, G., Dawra, G., Lindem, A., and V. 3292 Moreno, "A Simple BGP-based Mobile Routing System for the 3293 Aeronautical Telecommunications Network", draft-ietf- 3294 rtgwg-atn-bgp-05 (work in progress), January 2020. 3296 [I-D.templin-6man-dhcpv6-ndopt] 3297 Templin, F., "A Unified Stateful/Stateless Configuration 3298 Service for IPv6", draft-templin-6man-dhcpv6-ndopt-09 3299 (work in progress), January 2020. 3301 [I-D.templin-intarea-seal] 3302 Templin, F., "The Subnetwork Encapsulation and Adaptation 3303 Layer (SEAL)", draft-templin-intarea-seal-68 (work in 3304 progress), January 2014. 3306 [I-D.templin-intarea-vet] 3307 Templin, F., "Virtual Enterprise Traversal (VET)", draft- 3308 templin-intarea-vet-40 (work in progress), May 2013. 3310 [I-D.templin-ironbis] 3311 Templin, F., "The Interior Routing Overlay Network 3312 (IRON)", draft-templin-ironbis-16 (work in progress), 3313 March 2014. 3315 [I-D.templin-v6ops-pdhost] 3316 Templin, F., "IPv6 Prefix Delegation and Multi-Addressing 3317 Models", draft-templin-v6ops-pdhost-25 (work in progress), 3318 January 2020. 3320 [OVPN] OpenVPN, O., "http://openvpn.net", October 2016. 3322 [RFC1035] Mockapetris, P., "Domain names - implementation and 3323 specification", STD 13, RFC 1035, DOI 10.17487/RFC1035, 3324 November 1987, . 3326 [RFC1812] Baker, F., Ed., "Requirements for IP Version 4 Routers", 3327 RFC 1812, DOI 10.17487/RFC1812, June 1995, 3328 . 3330 [RFC2236] Fenner, W., "Internet Group Management Protocol, Version 3331 2", RFC 2236, DOI 10.17487/RFC2236, November 1997, 3332 . 3334 [RFC2529] Carpenter, B. and C. Jung, "Transmission of IPv6 over IPv4 3335 Domains without Explicit Tunnels", RFC 2529, 3336 DOI 10.17487/RFC2529, March 1999, 3337 . 3339 [RFC2983] Black, D., "Differentiated Services and Tunnels", 3340 RFC 2983, DOI 10.17487/RFC2983, October 2000, 3341 . 3343 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 3344 of Explicit Congestion Notification (ECN) to IP", 3345 RFC 3168, DOI 10.17487/RFC3168, September 2001, 3346 . 3348 [RFC3330] IANA, "Special-Use IPv4 Addresses", RFC 3330, 3349 DOI 10.17487/RFC3330, September 2002, 3350 . 3352 [RFC3810] Vida, R., Ed. and L. Costa, Ed., "Multicast Listener 3353 Discovery Version 2 (MLDv2) for IPv6", RFC 3810, 3354 DOI 10.17487/RFC3810, June 2004, 3355 . 3357 [RFC4251] Ylonen, T. and C. Lonvick, Ed., "The Secure Shell (SSH) 3358 Protocol Architecture", RFC 4251, DOI 10.17487/RFC4251, 3359 January 2006, . 3361 [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A 3362 Border Gateway Protocol 4 (BGP-4)", RFC 4271, 3363 DOI 10.17487/RFC4271, January 2006, 3364 . 3366 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 3367 Architecture", RFC 4291, DOI 10.17487/RFC4291, February 3368 2006, . 3370 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 3371 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 3372 December 2005, . 3374 [RFC4389] Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery 3375 Proxies (ND Proxy)", RFC 4389, DOI 10.17487/RFC4389, April 3376 2006, . 3378 [RFC4443] Conta, A., Deering, S., and M. Gupta, Ed., "Internet 3379 Control Message Protocol (ICMPv6) for the Internet 3380 Protocol Version 6 (IPv6) Specification", STD 89, 3381 RFC 4443, DOI 10.17487/RFC4443, March 2006, 3382 . 3384 [RFC4511] Sermersheim, J., Ed., "Lightweight Directory Access 3385 Protocol (LDAP): The Protocol", RFC 4511, 3386 DOI 10.17487/RFC4511, June 2006, 3387 . 3389 [RFC4541] Christensen, M., Kimball, K., and F. Solensky, 3390 "Considerations for Internet Group Management Protocol 3391 (IGMP) and Multicast Listener Discovery (MLD) Snooping 3392 Switches", RFC 4541, DOI 10.17487/RFC4541, May 2006, 3393 . 3395 [RFC4605] Fenner, B., He, H., Haberman, B., and H. Sandick, 3396 "Internet Group Management Protocol (IGMP) / Multicast 3397 Listener Discovery (MLD)-Based Multicast Forwarding 3398 ("IGMP/MLD Proxying")", RFC 4605, DOI 10.17487/RFC4605, 3399 August 2006, . 3401 [RFC4982] Bagnulo, M. and J. Arkko, "Support for Multiple Hash 3402 Algorithms in Cryptographically Generated Addresses 3403 (CGAs)", RFC 4982, DOI 10.17487/RFC4982, July 2007, 3404 . 3406 [RFC5015] Handley, M., Kouvelas, I., Speakman, T., and L. Vicisano, 3407 "Bidirectional Protocol Independent Multicast (BIDIR- 3408 PIM)", RFC 5015, DOI 10.17487/RFC5015, October 2007, 3409 . 3411 [RFC5214] Templin, F., Gleeson, T., and D. Thaler, "Intra-Site 3412 Automatic Tunnel Addressing Protocol (ISATAP)", RFC 5214, 3413 DOI 10.17487/RFC5214, March 2008, 3414 . 3416 [RFC5320] Templin, F., Ed., "The Subnetwork Encapsulation and 3417 Adaptation Layer (SEAL)", RFC 5320, DOI 10.17487/RFC5320, 3418 February 2010, . 3420 [RFC5522] Eddy, W., Ivancic, W., and T. Davis, "Network Mobility 3421 Route Optimization Requirements for Operational Use in 3422 Aeronautics and Space Exploration Mobile Networks", 3423 RFC 5522, DOI 10.17487/RFC5522, October 2009, 3424 . 3426 [RFC5558] Templin, F., Ed., "Virtual Enterprise Traversal (VET)", 3427 RFC 5558, DOI 10.17487/RFC5558, February 2010, 3428 . 3430 [RFC5569] Despres, R., "IPv6 Rapid Deployment on IPv4 3431 Infrastructures (6rd)", RFC 5569, DOI 10.17487/RFC5569, 3432 January 2010, . 3434 [RFC5880] Katz, D. and D. Ward, "Bidirectional Forwarding Detection 3435 (BFD)", RFC 5880, DOI 10.17487/RFC5880, June 2010, 3436 . 3438 [RFC6106] Jeong, J., Park, S., Beloeil, L., and S. Madanapalli, 3439 "IPv6 Router Advertisement Options for DNS Configuration", 3440 RFC 6106, DOI 10.17487/RFC6106, November 2010, 3441 . 3443 [RFC6146] Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful 3444 NAT64: Network Address and Protocol Translation from IPv6 3445 Clients to IPv4 Servers", RFC 6146, DOI 10.17487/RFC6146, 3446 April 2011, . 3448 [RFC6179] Templin, F., Ed., "The Internet Routing Overlay Network 3449 (IRON)", RFC 6179, DOI 10.17487/RFC6179, March 2011, 3450 . 3452 [RFC6221] Miles, D., Ed., Ooghe, S., Dec, W., Krishnan, S., and A. 3453 Kavanagh, "Lightweight DHCPv6 Relay Agent", RFC 6221, 3454 DOI 10.17487/RFC6221, May 2011, 3455 . 3457 [RFC6273] Kukec, A., Krishnan, S., and S. Jiang, "The Secure 3458 Neighbor Discovery (SEND) Hash Threat Analysis", RFC 6273, 3459 DOI 10.17487/RFC6273, June 2011, 3460 . 3462 [RFC6347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer 3463 Security Version 1.2", RFC 6347, DOI 10.17487/RFC6347, 3464 January 2012, . 3466 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 3467 for Equal Cost Multipath Routing and Link Aggregation in 3468 Tunnels", RFC 6438, DOI 10.17487/RFC6438, November 2011, 3469 . 3471 [RFC6706] Templin, F., Ed., "Asymmetric Extended Route Optimization 3472 (AERO)", RFC 6706, DOI 10.17487/RFC6706, August 2012, 3473 . 3475 [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and 3476 UDP Checksums for Tunneled Packets", RFC 6935, 3477 DOI 10.17487/RFC6935, April 2013, 3478 . 3480 [RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement 3481 for the Use of IPv6 UDP Datagrams with Zero Checksums", 3482 RFC 6936, DOI 10.17487/RFC6936, April 2013, 3483 . 3485 [RFC7333] Chan, H., Ed., Liu, D., Seite, P., Yokota, H., and J. 3486 Korhonen, "Requirements for Distributed Mobility 3487 Management", RFC 7333, DOI 10.17487/RFC7333, August 2014, 3488 . 3490 [RFC7761] Fenner, B., Handley, M., Holbrook, H., Kouvelas, I., 3491 Parekh, R., Zhang, Z., and L. Zheng, "Protocol Independent 3492 Multicast - Sparse Mode (PIM-SM): Protocol Specification 3493 (Revised)", STD 83, RFC 7761, DOI 10.17487/RFC7761, March 3494 2016, . 3496 [RFC8402] Filsfils, C., Ed., Previdi, S., Ed., Ginsberg, L., 3497 Decraene, B., Litkowski, S., and R. Shakir, "Segment 3498 Routing Architecture", RFC 8402, DOI 10.17487/RFC8402, 3499 July 2018, . 3501 [RFC8446] Rescorla, E., "The Transport Layer Security (TLS) Protocol 3502 Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018, 3503 . 3505 [RFC8754] Filsfils, C., Ed., Dukes, D., Ed., Previdi, S., Leddy, J., 3506 Matsushima, S., and D. Voyer, "IPv6 Segment Routing Header 3507 (SRH)", RFC 8754, DOI 10.17487/RFC8754, March 2020, 3508 . 3510 Appendix A. Non-Normative Considerations 3512 AERO can be applied to a multitude of Internetworking scenarios, with 3513 each having its own adaptations. The following considerations are 3514 provided as non-normative guidance: 3516 A.1. Implementation Strategies for Route Optimization 3518 Route optimization as discussed in Section 3.14 results in the route 3519 optimization source (ROS) creating an asymmetric neighbor cache entry 3520 for the target neighbor. The neighbor cache entry is maintained for 3521 at most ReachableTime seconds and then deleted unless updated. In 3522 order to refresh the neighbor cache entry lifetime before the 3523 ReachableTime timer expires, the specification requires 3524 implementations to issue a new NS/NA exchange to reset ReachableTime 3525 while data packets are still flowing. However, the decision of when 3526 to initiate a new NS/NA exchange and to perpetuate the process is 3527 left as an implementation detail. 3529 One possible strategy may be to monitor the neighbor cache entry 3530 watching for data packets for (ReachableTime - 5) seconds. If any 3531 data packets have been sent to the neighbor within this timeframe, 3532 then send an NS to receive a new NA. If no data packets have been 3533 sent, wait for 5 additional seconds and send an immediate NS if any 3534 data packets are sent within this "expiration pending" 5 second 3535 window. If no additional data packets are sent within the 5 second 3536 window, delete the neighbor cache entry. 3538 The monitoring of the neighbor data packet traffic therefore becomes 3539 an asymmetric ongoing process during the neighbor cache entry 3540 lifetime. If the neighbor cache entry expires, future data packets 3541 will trigger a new NS/NA exchange while the packets themselves are 3542 delivered over a longer path until route optimization state is re- 3543 established. 3545 A.2. Implicit Mobility Management 3547 OMNI interface neighbors MAY provide a configuration option that 3548 allows them to perform implicit mobility management in which no ND 3549 messaging is used. In that case, the Client only transmits packets 3550 over a single interface at a time, and the neighbor always observes 3551 packets arriving from the Client from the same link-layer source 3552 address. 3554 If the Client's underlying interface address changes (either due to a 3555 readdressing of the original interface or switching to a new 3556 interface) the neighbor immediately updates the neighbor cache entry 3557 for the Client and begins accepting and sending packets according to 3558 the Client's new address. This implicit mobility method applies to 3559 use cases such as cellphones with both WiFi and Cellular interfaces 3560 where only one of the interfaces is active at a given time, and the 3561 Client automatically switches over to the backup interface if the 3562 primary interface fails. 3564 A.3. Direct Underlying Interfaces 3566 When a Client's OMNI interface is configured over a Direct interface, 3567 the neighbor at the other end of the Direct link can receive packets 3568 without any encapsulation. In that case, the Client sends packets 3569 over the Direct link according to QoS preferences. If the Direct 3570 interface has the highest QoS preference, then the Client's IP 3571 packets are transmitted directly to the peer without going through an 3572 ANET/INET. If other interfaces have higher QoS preferences, then the 3573 Client's IP packets are transmitted via a different interface, which 3574 may result in the inclusion of Proxys, Servers and Bridges in the 3575 communications path. Direct interfaces must be tested periodically 3576 for reachability, e.g., via NUD. 3578 A.4. AERO Critical Infrastructure Considerations 3580 AERO Bridges can be either Commercial off-the Shelf (COTS) standard 3581 IP routers or virtual machines in the cloud. Bridges must be 3582 provisioned, supported and managed by the INET administrative 3583 authority, and connected to the Bridges of other INETs via inter- 3584 domain peerings. Cost for purchasing, configuring and managing 3585 Bridges is nominal even for very large OMNI links. 3587 AERO Servers can be standard dedicated server platforms, but most 3588 often will be deployed as virtual machines in the cloud. The only 3589 requirements for Servers are that they can run the AERO user-level 3590 code and have at least one network interface connection to the INET. 3591 As with Bridges, Servers must be provisioned, supported and managed 3592 by the INET administrative authority. Cost for purchasing, 3593 configuring and managing Servers is nominal especially for virtual 3594 Servers hosted in the cloud. 3596 AERO Proxys are most often standard dedicated server platforms with 3597 one network interface connected to the ANET and a second interface 3598 connected to an INET. As with Servers, the only requirements are 3599 that they can run the AERO user-level code and have at least one 3600 interface connection to the INET. Proxys must be provisioned, 3601 supported and managed by the ANET administrative authority. Cost for 3602 purchasing, configuring and managing Proxys is nominal, and borne by 3603 the ANET administrative authority. 3605 AERO Relays can be any dedicated server or COTS router platform 3606 connected to INETs and/or EUNs. The Relay connects to the OMNI link 3607 and engages in eBGP peering with one or more Bridges as a stub AS. 3608 The Relay then injects its MNPs and/or non-MNP prefixes into the BGP 3609 routing system, and provisions the prefixes to its downstream- 3610 attached networks. The Relay can perform ROS/ROR services the same 3611 as for any Server, and can route between the MNP and non-MNP address 3612 spaces. 3614 A.5. AERO Server Failure Implications 3616 AERO Servers may appear as a single point of failure in the 3617 architecture, but such is not the case since all Servers on the link 3618 provide identical services and loss of a Server does not imply 3619 immediate and/or comprehensive communication failures. Although 3620 Clients typically associate with a single Server at a time, Server 3621 failure is quickly detected and conveyed by Bidirectional Forward 3622 Detection (BFD) and/or proactive NUD allowing Clients to migrate to 3623 new Servers. 3625 If a Server fails, ongoing packet forwarding to Clients will continue 3626 by virtue of the asymmetric neighbor cache entries that have already 3627 been established in route optimization sources (ROSs). If a Client 3628 also experiences mobility events at roughly the same time the Server 3629 fails, unsolicited NA messages may be lost but proxy neighbor cache 3630 entries in the DEPARTED state will ensure that packet forwarding to 3631 the Client's new locations will continue for up to DepartTime 3632 seconds. 3634 If a Client is left without a Server for an extended timeframe (e.g., 3635 greater than ReachableTime seconds) then existing asymmetric neighbor 3636 cache entries will eventually expire and both ongoing and new 3637 communications will fail. The original source will continue to 3638 retransmit until the Client has established a new Server 3639 relationship, after which time continuous communications will resume. 3641 Therefore, providing many Servers on the link with high availability 3642 profiles provides resilience against loss of individual Servers and 3643 assurance that Clients can establish new Server relationships quickly 3644 in event of a Server failure. 3646 A.6. AERO Client / Server Architecture 3648 The AERO architectural model is client / server in the control plane, 3649 with route optimization in the data plane. The same as for common 3650 Internet services, the AERO Client discovers the addresses of AERO 3651 Servers and selects one Server to connect to. The AERO service is 3652 analogous to common Internet services such as google.com, yahoo.com, 3653 cnn.com, etc. However, there is only one AERO service for the link 3654 and all Servers provide identical services. 3656 Common Internet services provide differing strategies for advertising 3657 server addresses to clients. The strategy is conveyed through the 3658 DNS resource records returned in response to name resolution queries. 3659 As of January 2020 Internet-based 'nslookup' services were used to 3660 determine the following: 3662 o When a client resolves the domainname "google.com", the DNS always 3663 returns one A record (i.e., an IPv4 address) and one AAAA record 3664 (i.e., an IPv6 address). The client receives the same addresses 3665 each time it resolves the domainname via the same DNS resolver, 3666 but may receive different addresses when it resolves the 3667 domainname via different DNS resolvers. But, in each case, 3668 exactly one A and one AAAA record are returned. 3670 o When a client resolves the domainname "ietf.org", the DNS always 3671 returns one A record and one AAAA record with the same addresses 3672 regardless of which DNS resolver is used. 3674 o When a client resolves the domainname "yahoo.com", the DNS always 3675 returns a list of 4 A records and 4 AAAA records. Each time the 3676 client resolves the domainname via the same DNS resolver, the same 3677 list of addresses are returned but in randomized order (i.e., 3678 consistent with a DNS round-robin strategy). But, interestingly, 3679 the same addresses are returned (albeit in randomized order) when 3680 the domainname is resolved via different DNS resolvers. 3682 o When a client resolves the domainname "amazon.com", the DNS always 3683 returns a list of 3 A records and no AAAA records. As with 3684 "yahoo.com", the same three A records are returned from any 3685 worldwide Internet connection point in randomized order. 3687 The above example strategies show differing approaches to Internet 3688 resilience and service distribution offered by major Internet 3689 services. The Google approach exposes only a single IPv4 and a 3690 single IPv6 address to clients. Clients can then select whichever IP 3691 protocol version offers the best response, but will always use the 3692 same IP address according to the current Internet connection point. 3693 This means that the IP address offered by the network must lead to a 3694 highly-available server and/or service distribution point. In other 3695 words, resilience is predicated on high availability within the 3696 network and with no client-initiated failovers expected (i.e., it is 3697 all-or-nothing from the client's perspective). However, Google does 3698 provide for worldwide distributed service distribution by virtue of 3699 the fact that each Internet connection point responds with a 3700 different IPv6 and IPv4 address. The IETF approach is like google 3701 (all-or-nothing from the client's perspective), but provides only a 3702 single IPv4 or IPv6 address on a worldwide basis. This means that 3703 the addresses must be made highly-available at the network level with 3704 no client failover possibility, and if there is any worldwide service 3705 distribution it would need to be conducted by a network element that 3706 is reached via the IP address acting as a service distribution point. 3708 In contrast to the Google and IETF philosophies, Yahoo and Amazon 3709 both provide clients with a (short) list of IP addresses with Yahoo 3710 providing both IP protocol versions and Amazon as IPv4-only. The 3711 order of the list is randomized with each name service query 3712 response, with the effect of round-robin load balancing for service 3713 distribution. With a short list of addresses, there is still 3714 expectation that the network will implement high availability for 3715 each address but in case any single address fails the client can 3716 switch over to using a different address. The balance then becomes 3717 one of function in the network vs function in the end system. 3719 The same implications observed for common highly-available services 3720 in the Internet apply also to the AERO client/server architecture. 3721 When an AERO Client connects to one or more ANETs, it discovers one 3722 or more AERO Server addresses through the mechanisms discussed in 3723 earlier sections. Each Server address presumably leads to a fault- 3724 tolerant clustering arrangement such as supported by Linux-HA, 3725 Extended Virtual Synchrony or Paxos. Such an arrangement has 3726 precedence in common Internet service deployments in lightweight 3727 virtual machines without requiring expensive hardware deployment. 3728 Similarly, common Internet service deployments set service IP 3729 addresses on service distribution points that may relay requests to 3730 many different servers. 3732 For AERO, the expectation is that a combination of the Google/IETF 3733 and Yahoo/Amazon philosophies would be employed. The AERO Client 3734 connects to different ANET access points and can receive 1-2 Server 3735 LLAs at each point. It then selects one AERO Server address, and 3736 engages in RS/RA exchanges with the same Server from all ANET 3737 connections. The Client remains with this Server unless or until the 3738 Server fails, in which case it can switch over to an alternate 3739 Server. The Client can likewise switch over to a different Server at 3740 any time if there is some reason for it to do so. So, the AERO 3741 expectation is for a balance of function in the network and end 3742 system, with fault tolerance and resilience at both levels. 3744 Appendix B. Change Log 3746 << RFC Editor - remove prior to publication >> 3747 Changes from draft-templin-intarea-6706bis-54 to draft-templin- 3748 intrea-6706bis-55: 3750 o Updates on Segment Routing and S/TLLAO contents. 3752 o Various editorials and addressing cleanups. 3754 Changes from draft-templin-intarea-6706bis-52 to draft-templin- 3755 intrea-6706bis-53: 3757 o Normative reference to the OMNI spec, and remove portions that are 3758 already specified in OMNI. 3760 o Renamed "AERO interface/link" to "OMIN interface/link" throughout 3761 the document. 3763 o Truncated obsolete back section matter. 3765 Author's Address 3767 Fred L. Templin (editor) 3768 Boeing Research & Technology 3769 P.O. Box 3707 3770 Seattle, WA 98124 3771 USA 3773 Email: fltemplin@acm.org