idnits 2.17.1 draft-templin-intarea-6706bis-49.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 (May 3, 2020) is 1452 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 1041 -- Looks like a reference, but probably isn't: '2' on line 1047 == Missing Reference: 'N' is mentioned on line 1057, but not defined == Unused Reference: 'RFC2474' is defined on line 3295, but no explicit reference was found in the text == Unused Reference: 'RFC5175' is defined on line 3333, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-6man-segment-routing-header' is defined on line 3362, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-intarea-gue' is defined on line 3374, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-intarea-gue-extensions' is defined on line 3379, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-intarea-tunnels' is defined on line 3384, but no explicit reference was found in the text == Unused Reference: 'RFC1191' is defined on line 3434, but no explicit reference was found in the text == Unused Reference: 'RFC2764' is defined on line 3459, but no explicit reference was found in the text == Unused Reference: 'RFC2923' is defined on line 3473, but no explicit reference was found in the text == Unused Reference: 'RFC3819' is defined on line 3491, but no explicit reference was found in the text == Unused Reference: 'RFC4607' is defined on line 3546, but no explicit reference was found in the text == Unused Reference: 'RFC4963' is defined on line 3550, but no explicit reference was found in the text == Unused Reference: 'RFC5991' is defined on line 3592, but no explicit reference was found in the text == Unused Reference: 'RFC6864' is defined on line 3633, but no explicit reference was found in the text == Unused Reference: 'RFC7269' is defined on line 3647, but no explicit reference was found in the text == Unused Reference: 'RFC7421' is defined on line 3657, but no explicit reference was found in the text == Unused Reference: 'RFC8086' is defined on line 3669, but no explicit reference was found in the text == Outdated reference: A later version (-99) exists of draft-templin-6man-omni-interface-20 == 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 6106 (Obsoleted by RFC 8106) -- Obsolete informational reference (is this intentional?): RFC 6347 (Obsoleted by RFC 9147) Summary: 0 errors (**), 0 flaws (~~), 25 warnings (==), 5 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group F. Templin, Ed. 3 Internet-Draft Boeing Research & Technology 4 Obsoletes: rfc5320, rfc5558, rfc5720, May 3, 2020 5 rfc6179, rfc6706 (if 6 approved) 7 Intended status: Standards Track 8 Expires: November 4, 2020 10 Asymmetric Extended Route Optimization (AERO) 11 draft-templin-intarea-6706bis-49 13 Abstract 15 This document specifies the operation of IP over tunnel virtual links 16 using Asymmetric Extended Route Optimization (AERO). AERO interfaces 17 use an IPv6 link-local address format that supports operation of the 18 IPv6 Neighbor Discovery (ND) protocol and links ND to IP forwarding. 19 Prefix delegation/registration services are employed for network 20 admission and to manage the routing system. Multilink operation, 21 mobility management, quality of service (QoS) signaling and route 22 optimization are naturally supported through dynamic neighbor cache 23 updates. Standard IP multicasting services are also supported. AERO 24 is a widely-applicable mobile internetworking service especially 25 well-suited to aviation services, intelligent transportation systems, 26 mobile Virtual Private Networks (VPNs) and many other applications. 28 Status of This Memo 30 This Internet-Draft is submitted in full conformance with the 31 provisions of BCP 78 and BCP 79. 33 Internet-Drafts are working documents of the Internet Engineering 34 Task Force (IETF). Note that other groups may also distribute 35 working documents as Internet-Drafts. The list of current Internet- 36 Drafts is at https://datatracker.ietf.org/drafts/current/. 38 Internet-Drafts are draft documents valid for a maximum of six months 39 and may be updated, replaced, or obsoleted by other documents at any 40 time. It is inappropriate to use Internet-Drafts as reference 41 material or to cite them other than as "work in progress." 43 This Internet-Draft will expire on November 4, 2020. 45 Copyright Notice 47 Copyright (c) 2020 IETF Trust and the persons identified as the 48 document authors. All rights reserved. 50 This document is subject to BCP 78 and the IETF Trust's Legal 51 Provisions Relating to IETF Documents 52 (https://trustee.ietf.org/license-info) in effect on the date of 53 publication of this document. Please review these documents 54 carefully, as they describe your rights and restrictions with respect 55 to this document. Code Components extracted from this document must 56 include Simplified BSD License text as described in Section 4.e of 57 the Trust Legal Provisions and are provided without warranty as 58 described in the Simplified BSD License. 60 Table of Contents 62 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4 63 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 5 64 3. Asymmetric Extended Route Optimization (AERO) . . . . . . . . 11 65 3.1. AERO Link Reference Model . . . . . . . . . . . . . . . . 11 66 3.2. AERO Node Types . . . . . . . . . . . . . . . . . . . . . 12 67 3.3. AERO Routing System . . . . . . . . . . . . . . . . . . . 13 68 3.4. AERO Addresses . . . . . . . . . . . . . . . . . . . . . 15 69 3.5. Spanning Partitioned AERO Networks (SPAN) . . . . . . . . 16 70 3.5.1. SPAN Routing Topologies . . . . . . . . . . . . . . . 20 71 3.5.2. Segment Routing Over the SPAN . . . . . . . . . . . . 20 72 3.6. AERO Interface Characteristics . . . . . . . . . . . . . 21 73 3.7. AERO Interface Initialization . . . . . . . . . . . . . . 26 74 3.7.1. AERO Server/Relay Behavior . . . . . . . . . . . . . 26 75 3.7.2. AERO Proxy Behavior . . . . . . . . . . . . . . . . . 26 76 3.7.3. AERO Client Behavior . . . . . . . . . . . . . . . . 26 77 3.7.4. AERO Bridge Behavior . . . . . . . . . . . . . . . . 27 78 3.8. AERO Interface Neighbor Cache Maintenance . . . . . . . . 27 79 3.9. AERO Interface Encapsulation and Re-encapsulation . . . . 29 80 3.10. AERO Interface Decapsulation . . . . . . . . . . . . . . 30 81 3.11. AERO Interface Data Origin Authentication . . . . . . . . 30 82 3.12. AERO Interface MTU and Fragmentation . . . . . . . . . . 31 83 3.13. AERO Interface Forwarding Algorithm . . . . . . . . . . . 33 84 3.13.1. Client Forwarding Algorithm . . . . . . . . . . . . 33 85 3.13.2. Proxy Forwarding Algorithm . . . . . . . . . . . . . 34 86 3.13.3. Server/Relay Forwarding Algorithm . . . . . . . . . 35 87 3.13.4. Bridge Forwarding Algorithm . . . . . . . . . . . . 36 88 3.14. AERO Interface Error Handling . . . . . . . . . . . . . . 37 89 3.15. AERO Router Discovery, Prefix Delegation and 90 Autoconfiguration . . . . . . . . . . . . . . . . . . . . 39 91 3.15.1. AERO ND/PD Service Model . . . . . . . . . . . . . . 39 92 3.15.2. AERO Client Behavior . . . . . . . . . . . . . . . . 40 93 3.15.3. AERO Server Behavior . . . . . . . . . . . . . . . . 42 94 3.16. The AERO Proxy . . . . . . . . . . . . . . . . . . . . . 45 95 3.16.1. Detecting and Responding to Server Failures . . . . 47 96 3.16.2. Point-to-Multipoint Server Coordination . . . . . . 48 97 3.17. AERO Route Optimization . . . . . . . . . . . . . . . . . 48 98 3.17.1. Route Optimization Initiation . . . . . . . . . . . 49 99 3.17.2. Relaying the NS . . . . . . . . . . . . . . . . . . 49 100 3.17.3. Processing the NS and Sending the NA . . . . . . . . 49 101 3.17.4. Relaying the NA . . . . . . . . . . . . . . . . . . 51 102 3.17.5. Processing the NA . . . . . . . . . . . . . . . . . 51 103 3.17.6. Route Optimization Maintenance . . . . . . . . . . . 51 104 3.18. Neighbor Unreachability Detection (NUD) . . . . . . . . . 52 105 3.19. Mobility Management and Quality of Service (QoS) . . . . 53 106 3.19.1. Mobility Update Messaging . . . . . . . . . . . . . 54 107 3.19.2. Announcing Link-Layer Address and/or QoS Preference 108 Changes . . . . . . . . . . . . . . . . . . . . . . 55 109 3.19.3. Bringing New Links Into Service . . . . . . . . . . 55 110 3.19.4. Removing Existing Links from Service . . . . . . . . 55 111 3.19.5. Moving to a New Server . . . . . . . . . . . . . . . 56 112 3.20. Multicast . . . . . . . . . . . . . . . . . . . . . . . . 57 113 3.20.1. Source-Specific Multicast (SSM) . . . . . . . . . . 57 114 3.20.2. Any-Source Multicast (ASM) . . . . . . . . . . . . . 58 115 3.20.3. Bi-Directional PIM (BIDIR-PIM) . . . . . . . . . . . 59 116 3.21. Operation over Multiple AERO Links (VLANs) . . . . . . . 59 117 3.22. DNS Considerations . . . . . . . . . . . . . . . . . . . 60 118 3.23. Transition Considerations . . . . . . . . . . . . . . . . 61 119 3.24. Detecting and Reacting to Server and Bridge Failures . . 61 120 3.25. AERO Clients on the Open Internet . . . . . . . . . . . . 62 121 3.25.1. Use of SEND and CGA . . . . . . . . . . . . . . . . 64 122 3.26. Time-Varying MNPs . . . . . . . . . . . . . . . . . . . . 65 123 4. Implementation Status . . . . . . . . . . . . . . . . . . . . 66 124 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 66 125 6. Security Considerations . . . . . . . . . . . . . . . . . . . 66 126 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 68 127 8. References . . . . . . . . . . . . . . . . . . . . . . . . . 69 128 8.1. Normative References . . . . . . . . . . . . . . . . . . 70 129 8.2. Informative References . . . . . . . . . . . . . . . . . 71 130 Appendix A. AERO Alternate Encapsulations . . . . . . . . . . . 79 131 Appendix B. Non-Normative Considerations . . . . . . . . . . . . 80 132 B.1. Implementation Strategies for Route Optimization . . . . 81 133 B.2. Implicit Mobility Management . . . . . . . . . . . . . . 81 134 B.3. Direct Underlying Interfaces . . . . . . . . . . . . . . 82 135 B.4. Operation on AERO Links with /64 ASPs . . . . . . . . . . 82 136 B.5. AERO Critical Infrastructure Considerations . . . . . . . 83 137 B.6. AERO Server Failure Implications . . . . . . . . . . . . 83 138 B.7. AERO Client / Server Architecture . . . . . . . . . . . . 84 139 Appendix C. Change Log . . . . . . . . . . . . . . . . . . . . . 86 140 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 95 142 1. Introduction 144 Asymmetric Extended Route Optimization (AERO) fulfills the 145 requirements of Distributed Mobility Management (DMM) [RFC7333] and 146 route optimization [RFC5522] for aeronautical networking and other 147 network mobility use cases such as intelligent transportation 148 systems. AERO is based on a Non-Broadcast, Multiple Access (NBMA) 149 virtual link model known as the AERO link. The AERO link is a 150 virtual overlay configured over one or more underlying Internetworks, 151 and nodes on the link can exchange IP packets via tunneling. 152 Multilink operation allows for increased reliability, bandwidth 153 optimization and traffic path diversity. 155 The AERO service comprises Clients, Proxys, Servers and Relays that 156 are seen as AERO link neighbors as well as Bridges that interconnect 157 AERO link segments. Each node's AERO interface uses an IPv6 link- 158 local address format (known as the AERO address) that supports 159 operation of the IPv6 Neighbor Discovery (ND) protocol [RFC4861] and 160 links ND to IP forwarding. A node's AERO interface can be configured 161 over multiple underlying interfaces, and may therefore appear as a 162 single interface with multiple link-layer addresses. Each link-layer 163 address is subject to change due to mobility and/or QoS fluctuations, 164 and link-layer address changes are signaled by ND messaging the same 165 as for any IPv6 link. 167 AERO links provide a cloud-based service where mobile nodes may use 168 any Server acting as a Mobility Anchor Point (MAP) and fixed nodes 169 may use any Relay on the link for efficient communications. Fixed 170 nodes forward packets destined to other AERO nodes to the nearest 171 Relay, which forwards them through the cloud. A mobile node's 172 initial packets are forwarded through the Server, while direct 173 routing is supported through asymmetric extended route optimization 174 while data packets are flowing. Both unicast and multicast 175 communications are supported, and mobile nodes may efficiently move 176 between locations while maintaining continuous communications with 177 correspondents and without changing their IP Address. 179 AERO Bridges are interconnected in a secured private BGP overlay 180 routing instance known as "The SPAN". The SPAN provides a hybrid 181 routing/bridging service to join the underlying Internetworks of 182 multiple disjoint administrative domains into a single unified AERO 183 link. Each AERO link instance is characterized by the set of 184 Mobility Service Prefixes (MSPs) common to all mobile nodes. The 185 link extends to the point where a Relay/Server is on the optimal 186 route from any correspondent node on the link, and provides a conduit 187 between the underlying Internetwork and the SPAN. To the underlying 188 Internetwork, the Relay/Server is the source of a route to the MSP, 189 and hence uplink traffic to the mobile node is naturally routed to 190 the nearest Relay/Server. 192 AERO assumes the use of PIM Sparse Mode in support of multicast 193 communication. In support of Source Specific Multicast (SSM) when a 194 Mobile Node is the source, AERO route optimization ensures that a 195 shortest-path multicast tree is established with provisions for 196 mobility and multilink operation. In all other multicast scenarios 197 there are no AERO dependencies. 199 AERO was designed for aeronautical networking for both manned and 200 unmanned aircraft, where the aircraft is treated as a mobile node 201 that can connect an Internet of Things (IoT). AERO is also 202 applicable to a wide variety of other use cases. For example, it can 203 be used to coordinate the Virtual Private Network (VPN) links of 204 mobile nodes (e.g., cellphones, tablets, laptop computers, etc.) that 205 connect into a home enterprise network via public access networks 206 using services such as OpenVPN [OVPN]. It can also be used to 207 facilitate vehicular and pedestrian communications services for 208 intelligent transportation systems. Other applicable use cases are 209 also in scope. 211 The following numbered sections present the AERO specification. The 212 appendices at the end of the document are non-normative. 214 2. Terminology 216 The terminology in the normative references applies; the following 217 terms are defined within the scope of this document: 219 IPv6 Neighbor Discovery (ND) 220 an IPv6 control message service for coordinating neighbor 221 relationships between nodes connected to a common link. AERO 222 interfaces use the ND service specified in [RFC4861]. 224 IPv6 Prefix Delegation (PD) 225 a networking service for delegating IPv6 prefixes to nodes on the 226 link. The nominal PD service is DHCPv6 [RFC8415], however 227 alternate services (e.g., based on ND messaging) are also in scope 228 [I-D.templin-v6ops-pdhost][I-D.templin-6man-dhcpv6-ndopt]. Most 229 notably, a minimal form of PD known as "prefix registration" can 230 be used if the Client knows its prefix in advance and can 231 represent it in the IPv6 source address of an ND message. 233 Access Network (ANET) 234 a node's first-hop data link service network, e.g., a radio access 235 network, cellular service provider network, corporate enterprise 236 network, or the public Internet itself. For secured ANETs, link- 237 layer security services such as IEEE 802.1X and physical-layer 238 security prevent unauthorized access internally while border 239 network-layer security services such as firewalls and proxies 240 prevent unauthorized outside access. 242 ANET interface 243 a node's attachment to a link in an ANET. 245 ANET address 246 an IP address assigned to a node's interface connection to an 247 ANET. 249 Internetwork (INET) 250 a connected IP network topology with a coherent routing and 251 addressing plan and that provides a transit backbone service for 252 ANET end systems. INETs also provide an underlay service over 253 which the AERO virtual link is configured. Example INETs include 254 corporate enterprise networks, aviation networks, and the public 255 Internet itself. When there is no administrative boundary between 256 an ANET and the INET, the ANET and INET are one and the same. 258 INET Partition 259 frequently, INETs such as large corporate enterprise networks are 260 sub-divided internally into separate isolated partitions. Each 261 partition is fully connected internally but disconnected from 262 other partitions, and there is no requirement that separate 263 partitions maintain consistent Internet Protocol and/or addressing 264 plans. (Each INET partition is seen as a separate SPAN segment as 265 discussed below.) 267 INET interface 268 a node's attachment to a link in an INET. 270 INET address 271 an IP address assigned to a node's interface connection to an 272 INET. 274 INET encapsulation 275 the encapsulation of a packet in an outer header or headers that 276 can be routed within the scope of the local INET partition. 278 AERO link 279 a Non-Broadcast, Multiple Access (NBMA) tunnel virtual overlay 280 configured over one or more underlying INETs. Nodes on the AERO 281 link appear as single-hop neighbors from the perspective of the 282 virtual overlay even though they may be separated by many 283 underlying INET hops. AERO links may be configured over multiple 284 underlying SPAN segments (see below). 286 AERO interface 287 a node's attachment to an AERO link. Since the addresses assigned 288 to an AERO interface are managed for uniqueness, AERO interfaces 289 do not require Duplicate Address Detection (DAD) and therefore set 290 the administrative variable 'DupAddrDetectTransmits' to zero 291 [RFC4862]. 293 underlying interface 294 an ANET or INET interface over which an AERO interface is 295 configured. 297 AERO address 298 an IPv6 link-local address assigned to an AERO interface and 299 constructed as specified in Section 3.4. 301 base AERO address 302 the lowest-numbered AERO address aggregated by the MNP (see 303 Section 3.4). 305 Mobility Service Prefix (MSP) 306 an IP prefix assigned to the AERO link and from which more- 307 specific Mobile Network Prefixes (MNPs) are derived. 309 Mobile Network Prefix (MNP) 310 an IP prefix allocated from an MSP and delegated to an AERO Client 311 or Relay. 313 AERO node 314 a node that is connected to an AERO link, or that provides 315 services to other nodes on an AERO link. 317 AERO Client ("Client") 318 an AERO node that connects over one or more underlying interfaces 319 and requests MNP PDs from AERO Servers. The Client assigns a 320 Client AERO address to the AERO interface for use in ND exchanges 321 with other AERO nodes and forwards packets to correspondents 322 according to AERO interface neighbor cache state. 324 AERO Server ("Server") 325 an INET node that configures an AERO interface to provide default 326 forwarding and mobility/multilink services for AERO Clients. The 327 Server assigns an administratively-provisioned AERO address to its 328 AERO interface to support the operation of the ND/PD services, and 329 advertises all of its associated MNPs via BGP peerings with 330 Bridges. 332 AERO Relay ("Relay") 333 an AERO Server that also provides forwarding services between 334 nodes reached via the AERO link and correspondents on other links. 335 AERO Relays are provisioned with MNPs (i.e., the same as for an 336 AERO Client) and run a dynamic routing protocol to discover any 337 non-MNP IP routes. In both cases, the Relay advertises the MSP(s) 338 to its downstream networks, and distributes all of its associated 339 MNPs and non-MNP IP routes via BGP peerings with Bridges (i.e., 340 the same as for an AERO Server). 342 AERO Bridge ("Bridge") 343 a node that provides hybrid routing/bridging services (as well as 344 a security trust anchor) for nodes on an AERO link. As a router, 345 the Bridge forwards packets using standard IP forwarding. As a 346 bridge, the Bridge forwards packets over the AERO link without 347 decrementing the IPv6 Hop Limit. AERO Bridges peer with Servers 348 and other Bridges to discover the full set of MNPs for the link as 349 well as any non-MNPs that are reachable via Relays. 351 AERO Proxy ("Proxy") 352 a node that provides proxying services between Clients in an ANET 353 and Servers in external INETs. The AERO Proxy is a conduit 354 between the ANET and external INETs in the same manner as for 355 common web proxies, and behaves in a similar fashion as for ND 356 proxies [RFC4389]. 358 Spanning Partitioned AERO Networks (SPAN) 359 a means for bridging disjoint INET partitions as segments of a 360 unified AERO link the same as for a bridged campus LAN. The SPAN 361 is a mid-layer IPv6 encapsulation service in the AERO routing 362 system that supports a unified AERO link view for all segments. 363 Each segment in the SPAN is a distinct INET partition, and 364 individual segments are joined by AERO Bridges. Segment Routing 365 [RFC8402][RFC8754] can be used to cause packets to visit selected 366 hops on the SPAN. 368 SPAN Service Prefix (SSP) 369 a global or unique local /96 IPv6 prefix assigned to the AERO link 370 to support SPAN services. 372 SPAN Partition Prefix (SPP) 373 a sub-prefix of the SPAN Service Prefix uniquely assigned to a 374 single SPAN segment. 376 SPAN Client Prefix (SCP) 377 a SPAN prefix formed from an AERO Client address. 379 SPAN Address 380 a unique local IPv6 address taken from a SPAN Client/Partition 381 Prefix and constructed as specified in Section 3.5. SPAN 382 addresses are statelessly derived from AERO addresses, and vice- 383 versa. 385 SPAN encapsulation 386 the addition of an IPv6 header with SPAN source and destinations 387 per [RFC2473]. SPAN encapsulation is often used in conjunction 388 with INET encapsulation, or in raw form over ANET interfaces. 390 ingress tunnel endpoint (ITE) 391 an AERO interface endpoint that injects encapsulated packets into 392 an AERO link. 394 egress tunnel endpoint (ETE) 395 an AERO interface endpoint that receives encapsulated packets from 396 an AERO link. 398 link-layer address 399 an IP address used as an encapsulation header source or 400 destination address from the perspective of the AERO interface. 401 When an upper layer protocol (e.g., UDP) is used as part of the 402 encapsulation, the port number is also considered as part of the 403 link-layer address. From the perspective of the AERO interface, 404 the link-layer address is either an INET address for intra-segment 405 encapsulation or a SPAN address for inter-segment encapsulation. 407 network layer address 408 the source or destination address of an encapsulated IP packet 409 presented to the AERO interface. 411 end user network (EUN) 412 an internal virtual or external edge IP network that an AERO 413 Client or Relay connects to the rest of the network via the AERO 414 interface. The Client/Relay sees each EUN as a "downstream" 415 network, and sees the AERO interface as the point of attachment to 416 the "upstream" network. 418 Mobile Node (MN) 419 an AERO Client and all of its downstream-attached networks that 420 move together as a single unit, i.e., an end system that connects 421 an Internet of Things. 423 Mobile Router (MR) 424 a MN's on-board router that forwards packets between any 425 downstream-attached networks and the AERO link. 427 Route Optimization Source (ROS) 428 the AERO node nearest the source that initiates route 429 optimization. The ROS may be a Server or Proxy acting on behalf 430 of the source Client. 432 Route Optimization responder (ROR) 433 the AERO node nearest the target destination that responds to 434 route optimization requests. The ROR may be a Server acting on 435 behalf of a target MNP Client, or a Relay for a non-MNP 436 destination. 438 MAP List 439 a geographically and/or topologically referenced list of AERO 440 addresses of all Servers within the same AERO link. There is a 441 single MAP list for the entire AERO link. 443 Distributed Mobility Management (DMM) 444 a BGP-based overlay routing service coordinated by Servers and 445 Bridges that tracks all Server-to-Client associations. 447 Mobility Service (MS) 448 the collective set of all Servers, Proxys, Bridges and Relays that 449 provide the AERO Service to Clients. 451 Mobility Service Endpoint MSE) 452 an individual Server, Proxy, Bridge or Relay in the Mobility 453 Service. 455 Throughout the document, the simple terms "Client", "Server", 456 "Bridge", "Proxy" and "Relay" refer to "AERO Client", "AERO Server", 457 "AERO Bridge", "AERO Proxy" and "AERO Relay", respectively. 458 Capitalization is used to distinguish these terms from other common 459 Internetworking uses in which they appear without capitalization. 461 The terminology of DHCPv6 [RFC8415] and IPv6 ND [RFC4861] (including 462 the names of node variables, messages and protocol constants) is used 463 throughout this document. The terms "All-Routers multicast", "All- 464 Nodes multicast", "Solicited-Node multicast" and "Subnet-Router 465 anycast" are defined in [RFC4291] (with Link-Local scope assumed). 466 Also, the term "IP" is used to generically refer to either Internet 467 Protocol version, i.e., IPv4 [RFC0791] or IPv6 [RFC8200]. 469 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 470 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 471 "OPTIONAL" in this document are to be interpreted as described in BCP 472 14 [RFC2119][RFC8174] when, and only when, they appear in all 473 capitals, as shown here. 475 3. Asymmetric Extended Route Optimization (AERO) 477 The following sections specify the operation of IP over Asymmetric 478 Extended Route Optimization (AERO) links: 480 3.1. AERO Link Reference Model 482 +----------------+ 483 | AERO Bridge B1 | 484 | Nbr: S1, S2, P1| 485 |(X1->S1; X2->S2)| 486 | MSP M1 | 487 +-+---------+--+-+ 488 +--------------+ | Secured | | +--------------+ 489 |AERO Server S1| | tunnels | | |AERO Server S2| 490 | Nbr: C1, B1 +-----+ | +-----+ Nbr: C2, B1 | 491 | default->B1 | | | default->B1 | 492 | X1->C1 | | | X2->C2 | 493 +-------+------+ | +------+-------+ 494 | AERO Link | | 495 X===+===+===================+==)===============+===+===X 496 | | | | 497 +-----+--------+ +--------+--+-----+ +--------+-----+ 498 |AERO Client C1| | AERO Proxy P1 | |AERO Client C2| 499 | Nbr: S1 | |(Proxy Nbr Cache)| | Nbr: S2 | 500 | default->S1 | +--------+--------+ | default->S2 | 501 | MNP X1 | | | MNP X2 | 502 +------+-------+ .--------+------. +-----+--------+ 503 | (- Proxyed Clients -) | 504 .-. `---------------' .-. 505 ,-( _)-. ,-( _)-. 506 .-(_ IP )-. +-------+ +-------+ .-(_ IP )-. 507 (__ EUN )--|Host H1| |Host H2|--(__ EUN ) 508 `-(______)-' +-------+ +-------+ `-(______)-' 510 Figure 1: AERO Link Reference Model 512 Figure 1 presents the AERO link reference model. In this model: 514 o the AERO link is an overlay network service configured over one or 515 more underlying INET partitions which may be managed by different 516 administrative authorities and have incompatible protocols and/or 517 addressing plans. 519 o AERO Bridge B1 aggregates Mobility Service Prefix (MSP) M1, 520 discovers Mobile Network Prefixes (MNPs) X* and advertises the MSP 521 via BGP peerings over secured tunnels to Servers (S1, S2). 523 Bridges use the SPAN service to bridge disjoint segments of a 524 partitioned AERO link. 526 o AERO Servers S1 and S2 configure secured tunnels with Bridge B1 527 and also provide mobility, multilink and default router services 528 for their associated Clients C1 and C2. 530 o AERO Clients C1 and C2 associate with Servers S1 and S2, 531 respectively. They receive Mobile Network Prefix (MNP) 532 delegations X1 and X2, and also act as default routers for their 533 associated physical or internal virtual EUNs. Simple hosts H1 and 534 H2 attach to the EUNs served by Clients C1 and C2, respectively. 536 o AERO Proxy P1 configures a secured tunnel with Bridge B1 and 537 provides proxy services for AERO Clients in secured enclaves that 538 cannot associate directly with other AERO link neighbors. 540 Each node on the AERO link maintains an AERO interface neighbor cache 541 and an IP forwarding table the same as for any link. Although the 542 figure shows a limited deployment, in common operational practice 543 there will normally be many additional Bridges, Servers, Clients and 544 Proxys. 546 3.2. AERO Node Types 548 AERO Bridges provide hybrid routing/bridging services (as well as a 549 security trust anchor) for nodes on an AERO link. Bridges use 550 standard IPv6 routing to forward packets both within the same INET 551 partitions and between disjoint INET partitions based on a mid-layer 552 IPv6 encapsulation known as the SPAN header. The inner IP layer 553 experiences a virtual bridging service since the inner IP TTL/Hop 554 Limit is not decremented during forwarding. Each Bridge also peers 555 with Servers and other Bridges in a dynamic routing protocol instance 556 to provide a Distributed Mobility Management (DMM) service for the 557 list of active MNPs (see Section 3.3). Bridges present the AERO link 558 as a set of one or more Mobility Service Prefixes (MSPs) but as link- 559 layer devices need not connect directly to the AERO link themselves 560 unless an administrative interface is desired. Bridges configure 561 secured tunnels with Servers, Proxys and other Bridges; they further 562 maintain IP forwarding table entries for each Mobile Network Prefix 563 (MNP) and any other reachable non-MNP prefixes. 565 AERO Servers provide default forwarding and mobility/multilink 566 services for AERO Client Mobile Nodes (MNs). Each Server also peers 567 with Bridges in a dynamic routing protocol instance to advertise its 568 list of associated MNPs (see Section 3.3). Servers facilitate PD 569 exchanges with Clients, where each delegated prefix becomes an MNP 570 taken from an MSP. Servers forward packets between AERO interface 571 neighbors and track each Client's mobility profiles. 573 AERO Clients register their MNPs through PD exchanges with AERO 574 Servers over the AERO link, and distribute the MNPs to nodes on EUNs. 575 A Client may also be co-resident on the same physical or virtual 576 platform as a Server; in that case, the Client and Server behave as a 577 single functional unit. 579 AERO Proxys provide a conduit for ANET Clients to associate with 580 Servers in external INETs. Client and Servers exchange control plane 581 messages via the Proxy acting as a bridge between the ANET/INET 582 boundary. The Proxy forwards data packets between Clients and the 583 AERO link according to forwarding information in the neighbor cache. 584 The Proxy function is specified in Section 3.16. 586 AERO Relays are Servers that provide forwarding services between the 587 AERO interface and INET/EUN interfaces. Relays are provisioned with 588 MNPs the same as for an AERO Client, and also run a dynamic routing 589 protocol to discover any non-MNP IP routes. The Relay advertises the 590 MSP(s) to its connected networks, and distributes all of its 591 associated MNPs and non-MNP IP routes via BGP peerings with Bridges. 593 AERO Bridges, Servers, Proxys and Relays are critical infrastructure 594 elements in fixed (i.e., non-mobile) INET deployments and hence have 595 permanent and unchanging INET addresses. AERO Clients are MNs that 596 connect via underlying interfaces with addresses that may change when 597 the Client moves to a new network connection point. 599 3.3. AERO Routing System 601 The AERO routing system comprises a private instance of the Border 602 Gateway Protocol (BGP) [RFC4271] that is coordinated between Bridges 603 and Servers and does not interact with either the public Internet BGP 604 routing system or any underlying INET routing systems. 606 In a reference deployment, each Server is configured as an Autonomous 607 System Border Router (ASBR) for a stub Autonomous System (AS) using 608 an AS Number (ASN) that is unique within the BGP instance, and each 609 Server further uses eBGP to peer with one or more Bridges but does 610 not peer with other Servers. Each INET of a multi-segment AERO link 611 must include one or more Bridges, which peer with the Servers and 612 Proxys within that INET. All Bridges within the same INET are 613 members of the same hub AS using a common ASN, and use iBGP to 614 maintain a consistent view of all active MNPs currently in service. 615 The Bridges of different INETs peer with one another using eBGP. 617 Bridges advertise the AERO link's MSPs and any non-MNP routes to each 618 of their Servers. This means that any aggregated non-MNPs (including 619 "default") are advertised to all Servers. Each Bridge configures a 620 black-hole route for each of its MSPs. By black-holing the MSPs, the 621 Bridge will maintain forwarding table entries only for the MNPs that 622 are currently active, and packets destined to all other MNPs will 623 correctly incur Destination Unreachable messages due to the black- 624 hole route. In this way, Servers have only partial topology 625 knowledge (i.e., they know only about the MNPs of their directly 626 associated Clients) and they forward all other packets to Bridges 627 which have full topology knowledge. 629 Servers maintain a working set of associated MNPs, and dynamically 630 announce new MNPs and withdraw departed MNPs in eBGP updates to 631 Bridges. Servers that are configured as Relays also redistribute 632 non-MNP routes learned from non-AERO interfaces via their eBGP Bridge 633 peerings. 635 Clients are expected to remain associated with their current Servers 636 for extended timeframes, however Servers SHOULD selectively suppress 637 updates for impatient Clients that repeatedly associate and 638 disassociate with them in order to dampen routing churn. Servers 639 that are configured as Relays advertise the MSPs via INET/EUN 640 interfaces, and forward packets between INET/EUN interfaces and the 641 AERO interface using standard IP forwarding. 643 Scaling properties of the AERO routing system are limited by the 644 number of BGP routes that can be carried by Bridges. As of 2015, the 645 global public Internet BGP routing system manages more than 500K 646 routes with linear growth and no signs of router resource exhaustion 647 [BGP]. More recent network emulation studies have also shown that a 648 single Bridge can accommodate at least 1M dynamically changing BGP 649 routes even on a lightweight virtual machine, i.e., and without 650 requiring high-end dedicated router hardware. 652 Therefore, assuming each Bridge can carry 1M or more routes, this 653 means that at least 1M Clients can be serviced by a single set of 654 Bridges. A means of increasing scaling would be to assign a 655 different set of Bridges for each set of MSPs. In that case, each 656 Server still peers with one or more Bridges, but institutes route 657 filters so that BGP updates are only sent to the specific set of 658 Bridges that aggregate the MSP. For example, if the MSP for the AERO 659 link is 2001:db8::/32, a first set of Bridges could service the MSP 660 2001:db8::/40, a second set of Bridges could service 661 2001:db8:0100::/40, a third set could service 2001:db8:0200::/40, 662 etc. 664 Assuming up to 1K sets of Bridges, the AERO routing system can then 665 accommodate 1B or more MNPs with no additional overhead (for example, 666 it should be possible to service 1B /64 MNPs taken from a /34 MSP and 667 even more for shorter prefixes). In this way, each set of Bridges 668 services a specific set of MSPs that they advertise to the native 669 Internetwork routing system, and each Server configures MSP-specific 670 routes that list the correct set of Bridges as next hops. This 671 arrangement also allows for natural incremental deployment, and can 672 support small scale initial deployments followed by dynamic 673 deployment of additional Clients, Servers and Bridges without 674 disturbing the already-deployed base. 676 Server and Bridges can use the Bidirectional Forwarding Detection 677 (BFD) protocol [RFC5880] to quickly detect link failures that don't 678 result in interface state changes, BGP peer failures, and 679 administrative state changes. BFD is important in environments where 680 rapid response to failures is required for routing reconvergence and, 681 hence, communications continuity. 683 A full discussion of the BGP-based routing system used by AERO is 684 found in [I-D.ietf-rtgwg-atn-bgp]. The system provides for 685 Distributed Mobility Management (DMM) per the distributed mobility 686 anchoring architecture [I-D.ietf-dmm-distributed-mobility-anchoring]. 688 3.4. AERO Addresses 690 A Client's AERO address is an IPv6 link-local address formed from the 691 Client's delegated MNP. Bridge, Server, Relay and Proxy AERO 692 addresses are assigned from the range fe80::/96 and include an 693 administratively-provisioned value in the lower 32 bits. 695 IPv6 Client AERO addresses encode the Subnet-Router anycast address 696 of a MNP (or non-MNP globally routable prefix) within the least- 697 significant 112 bits of the IPv6 link-local prefix fe80::/16. For 698 example, for the MNP 2001:db8:1000:2000::/56 the corresponding AERO 699 address is fe80:2001:db8:1000:2000::/72. 701 IPv4-compatible Client AERO addresses are based on an IPv4-mapped 702 IPv6 address [RFC4291] formed from an IPv4 MNP and with a prefix 703 length of 96 plus the MNP prefix length. For example, for the IPv4 704 MNP 192.0.2.16/28 the IPv4-mapped IPv6 MNP is: 706 0:0:0:0:0:ffff:192.0.2.16/124 (also written as 707 0:0:0:0:0:ffff:c000:0210/124) 709 The Client then constructs its AERO address with the prefix fe80::/64 710 and with the lower 64 bits of the IPv4-mapped IPv6 address in the 711 interface identifier as: fe80::ffff:192.0.2.16. 713 Mobility Service (MS) AERO addresses (used by Bridges, Servers, 714 Relays and Proxys) are allocated from the range fe80::/96, and MUST 715 be managed for uniqueness. The lower 32 bits of the AERO address 716 includes a unique integer value between 1 and 0xfeffffff (e.g., 717 fe80::1, fe80::2, fe80::3, etc., fe80::feff:ffff) as assigned by the 718 administrative authority for the link. The address fe80:: is the 719 IPv6 link-local Subnet-Router anycast address, and the address range 720 fe80::ff00:0000/104 is reserved for future use. 722 Finally, the address range fe80::/32 is used as the Teredo service 723 prefix for AERO according to the format in Section 4 of [RFC4380] 724 (see Section 3.25 for further discussion). 726 For a full discussion of the above address format and implications 727 for the /64 boundary, see: [I-D.templin-6man-omni-interface]. 729 3.5. Spanning Partitioned AERO Networks (SPAN) 731 An AERO link configured over a single INET appears as a single 732 unified link with a consistent underlying network addressing plan. 733 In that case, all nodes on the link can exchange packets via simple 734 INET encapsulation, since the underlying INET is connected. In 735 common practice, however, an AERO link may be partitioned into 736 multiple "segments", where each segment is a distinct INET 737 potentially managed under a different administrative authority (e.g., 738 as for worldwide aviation service providers such as ARINC, SITA, 739 Inmarsat, etc.). Individual INETs may also themselves be partitioned 740 internally, in which case each internal partition is seen as a 741 separate segment. 743 The addressing plan of each segment is consistent internally but will 744 often bear no relation to the addressing plans of other segments. 745 Each segment is also likely to be separated from others by network 746 security devices (e.g., firewalls, proxies, packet filtering 747 gateways, etc.), and in many cases disjoint segments may not even 748 have any common physical link connections. Therefore, nodes can only 749 be assured of exchanging packets directly with correspondents in the 750 same segment, and not with those in other segments. The only means 751 for joining the segments therefore is through inter-domain peerings 752 between AERO Bridges. 754 The same as for traditional campus LANs, multiple AERO link segments 755 can be joined into a single unified link via a virtual bridging 756 service termed "The SPAN". The SPAN performs link-layer packet 757 forwarding between segments (i.e., bridging) without decrementing the 758 network-layer TTL/Hop Limit. The SPAN model is depicted in Figure 2: 760 . . . . . . . . . . . . . . . . . . . . . . . 761 . . 762 . .-(::::::::) . 763 . .-(::::::::::::)-. +-+ . 764 . (:::: Segment A :::)--|B|---+ . 765 . `-(::::::::::::)-' +-+ | . 766 . `-(::::::)-' | . 767 . | . 768 . .-(::::::::) | . 769 . .-(::::::::::::)-. +-+ | . 770 . (:::: Segment B :::)--|B|---+ . 771 . `-(::::::::::::)-' +-+ | . 772 . `-(::::::)-' | . 773 . | . 774 . .-(::::::::) | . 775 . .-(::::::::::::)-. +-+ | . 776 . (:::: Segment C :::)--|B|---+ . 777 . `-(::::::::::::)-' +-+ | . 778 . `-(::::::)-' | . 779 . | . 780 . ..(etc).. x . 781 . . 782 . . 783 . <- AERO Link Bridged by the SPAN -> . 784 . . . . . . . . . . . . . .. . . . . . . . . 786 Figure 2: The SPAN 788 To support the SPAN, AERO links use the Unique Local Address (ULA) 789 prefix fd80::/10 [RFC4193] as the SPAN Service Prefix (SSP). The 790 prefix length intentionally matches the IPv6 link-local prefix 791 (fe80::/10), and enables a simple stateless translation between AERO 792 and SPAN addresses. Additional SSPs can be employed to identify 793 distinct SPAN Routing Topologies (SRTs) (see: Section 3.5.1). 795 Each segment in the SPAN assigns a unique sub-prefix of SSP::/96 796 termed the "SPAN Partition Prefix (SPP)". For example, for fd80::/96 797 a first segment could assign fd80::1000/116, a second could assign 798 fd80::2000/116, a third could assign fd80::3000/116, etc. The 799 administrative authorities for each segment must therefore coordinate 800 to assure mutually-exclusive SPP assignments, but internal 801 provisioning of the SPP is an independent local consideration for 802 each administrative authority. 804 SPAN addresses are formed by simply rewriting the upper 16 bits of 805 the corresponding AERO address. For example: 807 o the SPAN address formed from the IPv6 Client AERO address 808 fe80:2001:db8:1000:2000:: is simply fd80:2001:db8:1000:2000:: 810 o the SPAN address formed from the IPv4-compatible Client AERO 811 address fe80::ffff:192.0.2.1 is simply fd80::ffff:192.0.2.1 813 o the SPAN address formed from the administrative AERO address 814 fe80::1001 is simply fd80::1001. 816 AERO Bridges join multiple segments into a unified AERO link over 817 multiple diverse administrative domains. They support a bridging 818 function by first establishing forwarding table entries for their 819 SPPs either via standard BGP routing or static routes. For example, 820 if three Bridges ('A', 'B' and 'C') from different segments serviced 821 the SPPs fd80::1000/116, fd80::2000/116 and fd80::3000/116 822 respectively, then the forwarding tables in each Bridge are as 823 follows: 825 A: fd80::1000/116->local, fd80::2000/116->B, fd80::3000/116->C 827 B: fd80::1000/116->A, fd80::2000/116->local, fd80::3000/116->C 829 C: fd80::1000/116->A, fd80::2000/116->B, fd80::3000/116->local 831 These forwarding table entries are permanent and never change, since 832 they correspond to fixed infrastructure elements in their respective 833 segments. 835 SPAN Client Prefixes (SCPs) are instead dynamically advertised in the 836 AERO link routing system by Servers and Relays that provide service 837 for their corresponding MNPs. For example, if three Servers ('D', 838 'E' and 'F') service the MNPs 2001:db8:1000:2000::/56, 839 2001:db8:3000:4000::/56 and 2001:db8:5000:6000::/56 then the routing 840 system would include: 842 D: fd80:2001:db8:1000:2000::/72 844 E: fd80:2001:db8:3000:4000::/72 846 F: fd80:2001:db8:5000:6000::/72 848 With the SCPs and SPPs in place in each Bridge's forwarding table, 849 control and data packets sent between AERO nodes in different 850 segments can therefore be carried over the via encapsulation in a 851 mid-layer IPv6 header known as the "SPAN header". For example, when 852 a source AERO node forwards a packet with IPv6 address 853 2001:db8:1:2::1 to a target AERO node with IPv6 address 854 2001:db8:1000:2000::1, it first encapsulates the packet in a SPAN 855 header with source address set to fd80:2001:db8:1:2:: and destination 856 address set to fd80:2001:db8:1000:2000::. Next, it encapsulates the 857 resulting SPAN packet in an INET header with source address set to 858 its own INET address (e.g., 192.0.2.100) and destination set to the 859 INET address of a Bridge (e.g., 192.0.2.1). 861 SPAN encapsulation is based on Generic Packet Tunneling in IPv6 862 [RFC2473]; the encapsulation format in the above example is shown in 863 Figure 3: 865 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 866 | INET Header | 867 | src = 192.0.2.100 | 868 | dst = 192.0.2.1 | 869 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 870 | SPAN Header | 871 | src = fd80:2001:db8:1:2:: | 872 | dst=fd80:2001:db8:1000:2000:: | 873 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 874 | Inner IP Header | 875 | src = 2001:db8:1:2::1 | 876 | dst = 2001:db8:1000:2000::1 | 877 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 878 | | 879 ~ ~ 880 ~ Inner Packet Body ~ 881 ~ ~ 882 | | 883 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 885 Figure 3: SPAN Encapsulation 887 In this format, the inner IP header and packet body are the original 888 IP packet, the SPAN header is an IPv6 header prepared according to 889 [RFC2473], and the INET header is prepared as discussed in 890 Section 3.9. A packet is said to be "forwarded/sent into the SPAN" 891 when it is encapsulated as described above then forwarded via a 892 secured tunnel to a neighboring Bridge. 894 This gives rise to a routing system that contains both SCP routes 895 that may change dynamically due to regional node mobility and SPP 896 routes that never change. The Bridges can therefore provide link- 897 layer bridging by sending packets into the SPAN instead of network- 898 layer routing according to MNP routes. As a result, opportunities 899 for packet loss due to node mobility between different segments are 900 mitigated. 902 In normal operations, IPv6 ND messages are conveyed to SPP addresses 903 over the SPAN so that specific Proxys, Servers or Relays can be 904 addressed without being subject to mobility events. Conversely, only 905 the first few packets destined to SCP addresses traverse the SPAN 906 until route optimization can determine a more direct path. 908 3.5.1. SPAN Routing Topologies 910 In some cases (e.g., when redundant topologies are needed for fault 911 tolerance and reliability) it may be beneficial to deploy multiple 912 SPAN Routing Topologies (SRTs) that act as independent overlay 913 instances. A communication failure in one instance therefore will 914 not affect communications in other instances. 916 Each SRT is identified by a distinct value in bits 10-15 of he SSP 917 fd80::10, i.e., as fd80::/16, fd81::/16, fd82::/16, etc. This 918 document asserts that up to four SRTs provide a level of safety 919 sufficient for critical communications such as civil aviation. Each 920 SRT is designated with a color that identifies a different AERO link 921 instance as follows: 923 o Red (default) - corresponds to the SSP fd80::/16 925 o Green - corresponds to the SSP fd81::/16 927 o Blue-1 - corresponds to the SSP fd82::/16 929 o Blue-2 - corresponds to SSP fd83::/16 931 o SSPs fd84::/16 through fdbf::/16 are reserved for future use. 933 Each AERO interface assigns a SPAN Anycast address corresponding to 934 its SRT prefix. For example, the SPAN anycast address for the Green 935 SRT is simply fd81::. The SPAN anycast address is used for AERO 936 interface determination in Safety-Based Multilink (SBM) as discussed 937 in [I-D.templin-6man-omni-interface]. Each AERO interface further 938 applies Performance-Based Multilink (PBM) internally. 940 3.5.2. Segment Routing Over the SPAN 942 As discussed in the following sections, Segment Routing is used over 943 the SPAN to influence the path of packets destined to Clients on INET 944 interfaces without causing all packets to traverse the Client's 945 Server. When a Client, Proxy or Server has a packet to send to a 946 target discovered through route optimization located in the same SPAN 947 segment, it encapsulates the packet in a SPAN header with the SPAN 948 address of the target as the destination address, then uses the 949 target's Link Layer Address information for INET encapsulation. 951 When a Client, Proxy or Server has a packet to send to a route 952 optimization target located in a different SPAN segment, it 953 encapsulates the packet in a SPAN header with the SPAN address of the 954 target's Server as the destination. The node also includes a Segment 955 Routing Header (SRH) [RFC8754] with the SPAN address of the target as 956 the penultimate address and with the IP encapsulation address of the 957 target as the ultimate address. (When the encapsulation address is 958 an IPv6 address and a port number is included, the port number is 959 written into the SRH Tag field.) The node then forwards the packet 960 into the SPAN, which will eventually direct it to a Bridge on the 961 same segment as the target's Server. 963 When a Bridge on the same segment as the target's Server receives a 964 SPAN-encapsulated packet destined to the target Server, it looks 965 ahead into the Segment Routing List to determine that the penultimate 966 destination is set to the target's SPAN address and the ultimate 967 destination is set to the target's Link Layer Address. The Bridge 968 then advances the SPAN destination address to the target's SPAN 969 address and encapsulates the SPAN packet in an INET header based on 970 the target's Link Layer Address, then forwards the packet to the 971 target directly while bypassing the target's Server. In this way, 972 the Bridge participates in route optimization to greatly reduce 973 traffic load and suboptimal routing through the target's Server. 975 3.6. AERO Interface Characteristics 977 AERO interfaces are virtual interfaces configured over one or more 978 underlying interfaces classified as follows: 980 o INET interfaces connect to an INET either natively or through one 981 or several IPv4 Network Address Translators (NATs). Native INET 982 interfaces have global IP addresses that are reachable from any 983 INET correspondent. All Server, Relay and Bridge interfaces are 984 native interfaces, as are INET-facing interfaces of Proxys. NATed 985 INET interfaces connect to a private network behind one or more 986 NATs that provide INET access. Clients that are behind a NAT are 987 required to send periodic keepalive messages to keep NAT state 988 alive when there are no data packets flowing. 990 o Proxyed interfaces connect to an ANET that is separated from the 991 open INET by an AERO Proxy. Proxys can actively issue control 992 messages over the INET on behalf of the Client to reduce ANET 993 congestion. Clients connected to Proxyed interfaces receive RAs 994 with the P flag set to 1. 996 o VPNed interfaces use security encapsulation over the INET to a 997 Virtual Private Network (VPN) server that also acts as an AERO 998 Server. Clients connected to VPNed interfaces receive RAs with 999 the P flag set to 1 the same as for Proxyed interfaces. Other 1000 than the link-layer encapsulation format, VPNed interfaces behave 1001 the same as Direct interfaces. 1003 o Direct interfaces connect a Client directly to a Server without 1004 crossing any ANET/INET paths. An example is a line-of-sight link 1005 between a remote pilot and an unmanned aircraft. The same Client 1006 considerations apply as for VPNed interfaces above, and the Client 1007 receives RA messages with the P flag set to 1. 1009 AERO interfaces use SPAN-layer encapsulation as necessary as 1010 discussed in Section 3.5. AERO interfaces use link-layer 1011 encapsulation (see: Section 3.9) to exchange packets with AERO link 1012 neighbors over INET or VPNed interfaces. AERO interfaces do not use 1013 encapsulation over Proxyed and Direct underlying interfaces. 1015 AERO interfaces maintain a neighbor cache for tracking per-neighbor 1016 state the same as for any interface. AERO interfaces use ND messages 1017 including Router Solicitation (RS), Router Advertisement (RA), 1018 Neighbor Solicitation (NS) and Neighbor Advertisement (NA) for 1019 neighbor cache management. 1021 AERO interfaces send ND messages with an Overlay Multilink Network 1022 Interface (OMNI) option formatted as specified in 1023 [I-D.templin-6man-omni-interface]. The OMNI option includes prefix 1024 registration information and "ifIndex-tuples" containing link 1025 information parameters for the AERO interface's underlying 1026 interfaces. 1028 When encapsulation is used, AERO interface ND messages MAY also 1029 include an AERO Source/Target Link-Layer Address Option (S/TLLAO) 1030 formatted as shown in Figure 4: 1032 0 1 2 3 1033 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 1034 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1035 | Type | Length | ifIndex[1] | SRT | LHS |FMT| 1036 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1037 ~ Segment Routing List [1] ~ 1038 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1039 ~ Link Layer Address [1] ~ 1040 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1041 | Port Number [1] | ifIndex[2] | SRT | LHS |FMT| 1042 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1043 ~ Segment Routing List [2] ~ 1044 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1045 ~ Link Layer Address [2] ~ 1046 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1047 | Port Number [2] | .... ~ 1048 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ~ 1049 ~ ... ~ 1050 ~ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1051 ~ | ifIndex[N] | SRT | LHS |FMT| 1052 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1053 ~ Segment Routing List [N] ~ 1054 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1055 ~ Link Layer Address [N] ~ 1056 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1057 | Port Number [N] | Zero Padding (if necessary) ... 1058 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+--+-+-+-+-+-+-+-+-+-+-+-+-+-+- 1060 Figure 4: AERO Source/Target Link-Layer Address Option (S/TLLAO) 1061 Format 1063 In this format, Type and Length are set the same as specified for S/ 1064 TLLAOs in [RFC4861], with trailing zero padding octets added as 1065 necessary to produce an integral number of 8 octet blocks. The S/ 1066 TLLAO includes N ifIndex-tuples in correspondence to ifIndex-tuples 1067 that appear in the OMNI option. Each ifIndex-tuple includes the 1068 following information: 1070 o ifIndex[i] - the same value as in the corresponding ifIndex-tuple 1071 included in the OMNI option. 1073 o SRT[i] - a 2-bit "SPAN Routing Topology" value (see: 1074 Section 3.5.1) coded as follows: 1076 * 000 - Red 1078 * 001 - Green 1079 * 010 - Blue-1 1081 * 011 - Blue-2 1083 * 100 - 111 - Reserved 1085 o LHS[i] - a 3-bit "LookaHead Segments" value that encodes the 1086 number (from 0 to 7) of entries in Segment Routing List [i]. 1088 o FMT[i] - a 2-bit "Format" code. Determines the format of the Link 1089 Layer Address [i] field as follows: 1091 * 00 - Link Layer Address [i] encodes a Teredo-format AERO 1092 address for a node behind a NAT. 1094 * 01 - Link Layer Address [i] encodes a Teredo-format AERO 1095 address for a node on the open INET. 1097 * 10 - Link Layer Address [i] encodes a native IPv6 address. 1099 * 11 - Link Layer Address [i] encodes a native IPv6 address with 1100 Port Number [i] field included. 1102 o Segment Routing List [i] - Includes LHS[i]-many 16 byte SPAN 1103 addresses corresponding to the Segment IDs (SIDs) that must be 1104 visited prior to forwarding to Link Layer Address [i]. The 1105 ultimate SID appears first, followed by the penultimate SID 1106 second, etc. 1108 o Link Layer Address [i] - Included according to FMT[i], and 1109 identifies the link-layer address of the source/target. 1111 o Port Number [i] - Present only when FMT[i] is 11. When present, 1112 the field is 2 bytes in length and immediately follows Link Layer 1113 Address [i]. Encodes the upper layer protocol port number to be 1114 used as the encapsulation source port. 1116 If an S/TLLAO is included, any ifIndex-tuples correspond to a proper 1117 subset of the OMNI option ifIndex-tuples. Any S/TLLAO ifIndex-tuple 1118 having an ifIndex value that does not appear in an OMNI option 1119 ifindex-tuple is ignored. If the same ifIndex value appears in 1120 multiple ifIndex-tuples, the first tuple is processed and the 1121 remaining tuples are ignored. Any S/TLLAO ifIndex-tuples can 1122 therefore be viewed as inter-dependent extensions of their 1123 corresponding OMNI option ifIndex-tuples, i.e., the OMNI option and 1124 S/TLLAO are companions that are interpreted in conjunction with each 1125 other. 1127 A Client's AERO interface may be configured over multiple underlying 1128 interface connections. For example, common mobile handheld devices 1129 have both wireless local area network ("WLAN") and cellular wireless 1130 links. These links are often used "one at a time" with low-cost WLAN 1131 preferred and highly-available cellular wireless as a standby, but a 1132 simultaneous-use capability could provide benefits. In a more 1133 complex example, aircraft frequently have many wireless data link 1134 types (e.g. satellite-based, cellular, terrestrial, air-to-air 1135 directional, etc.) with diverse performance and cost properties. 1137 If a Client's multiple underlying interfaces are used "one at a time" 1138 (i.e., all other interfaces are in standby mode while one interface 1139 is active), then ND message OMNI options include only a single 1140 ifIndex-tuple set to constant values. In that case, the Client would 1141 appear to have a single interface but with a dynamically changing 1142 link-layer address. 1144 If the Client has multiple active underlying interfaces, then from 1145 the perspective of ND it would appear to have multiple link-layer 1146 addresses. In that case, ND message OMNI options MAY include 1147 multiple ifIndex-tuples - each with values that correspond to a 1148 specific interface. Every ND message need not include all OMNI and/ 1149 or S/TLLAO ifIndex-tuples; for any ifIndex-tuple not included, the 1150 neighbor considers the status as unchanged. 1152 Bridge, Server and Proxy AERO interfaces may be configured over one 1153 or more secured tunnel interfaces. The AERO interface configures 1154 both an AERO address and its corresponding SPAN address, while the 1155 underlying secured tunnel interfaces are either unnumbered or 1156 configure the same SPAN address. The AERO interface encapsulates 1157 each IP packet in a SPAN header and presents the packet to the 1158 underlying secured tunnel interface. For Bridges that do not 1159 configure an AERO interface, the secured tunnel interfaces themselves 1160 are exposed to the IP layer with each interface configuring the 1161 Bridge's SPAN address. Routing protocols such as BGP therefore run 1162 directly over the Bridge's secured tunnel interfaces. For nodes that 1163 configure an AERO interface, routing protocols such as BGP run over 1164 the AERO interface but do not employ SPAN encapsulation. Instead, 1165 the AERO interface presents the routing protocol messages directly to 1166 the underlying secured tunnels without applying encapsulation and 1167 while using the SPAN address as the source address. This distinction 1168 must be honored consistently according to each node's configuration 1169 so that the IP forwarding table will associate discovered IP routes 1170 with the correct interface. 1172 3.7. AERO Interface Initialization 1174 AERO Servers, Proxys and Clients configure AERO interfaces as their 1175 point of attachment to the AERO link. AERO nodes assign the MSPs for 1176 the link to their AERO interfaces (i.e., as a "route-to-interface") 1177 to ensure that packets with destination addresses covered by an MNP 1178 not explicitly assigned to a non-AERO interface are directed to the 1179 AERO interface. 1181 AERO interface initialization procedures for Servers, Proxys, Clients 1182 and Bridges are discussed in the following sections. 1184 3.7.1. AERO Server/Relay Behavior 1186 When a Server enables an AERO interface, it assigns AERO/SPAN 1187 addresses appropriate for the given SPAN segment. The Server also 1188 configures secured tunnels with one or more neighboring Bridges and 1189 engages in a BGP routing protocol session with each Bridge. 1191 The AERO interface provides a single interface abstraction to the IP 1192 layer, but internally comprises multiple secured tunnels as well as 1193 an NBMA nexus for sending encapsulated data packets to AERO interface 1194 neighbors. The Server further configures a service to facilitate ND/ 1195 PD exchanges with AERO Clients and manages per-Client neighbor cache 1196 entries and IP forwarding table entries based on control message 1197 exchanges. 1199 Relays are simply Servers that run a dynamic routing protocol to 1200 redistribute routes between the AERO interface and INET/EUN 1201 interfaces (see: Section 3.3). The Relay provisions MNPs to networks 1202 on the INET/EUN interfaces (i.e., the same as a Client would do) and 1203 advertises the MSP(s) for the AERO link over the INET/EUN interfaces. 1204 The Relay further provides an attachment point of the AERO link to a 1205 non-MNP-based global topology. 1207 3.7.2. AERO Proxy Behavior 1209 When a Proxy enables an AERO interface, it assigns AERO/SPAN 1210 addresses and configures permanent neighbor cache entries the same as 1211 for Servers. The Proxy also configures secured tunnels with one or 1212 more neighboring Bridges and maintains per-Client neighbor cache 1213 entries based on control message exchanges. 1215 3.7.3. AERO Client Behavior 1217 When a Client enables an AERO interface, it sends RS messages with 1218 ND/PD parameters over its underlying interfaces to a Server in the 1219 MAP list, which returns an RA message with corresponding parameters. 1221 (The RS/RA messages may pass through a Proxy in the case of a 1222 Client's Proxyed interface, or through one or more NATs in the case 1223 of a Client's INET interface.) 1225 3.7.4. AERO Bridge Behavior 1227 AERO Bridges need not connect directly to the AERO link, since they 1228 operate as link-layer forwarding devices instead of network layer 1229 routers. Configuration of AERO interfaces on Bridges is therefore 1230 OPTIONAL, e.g., if an administrative interface is needed. Bridges 1231 configure secured tunnels with Servers, Proxys and other Bridges; 1232 they also configure AERO/SPAN addresses and permanent neighbor cache 1233 entries the same as Servers. Bridges engage in a BGP routing 1234 protocol session with a subset of the Servers on the local SPAN 1235 segment, and with other Bridges on the SPAN (see: Section 3.3). 1237 3.8. AERO Interface Neighbor Cache Maintenance 1239 Each AERO interface maintains a conceptual neighbor cache that 1240 includes an entry for each neighbor it communicates with on the AERO 1241 link per [RFC4861]. AERO interface neighbor cache entries are said 1242 to be one of "permanent", "symmetric", "asymmetric" or "proxy". 1244 Permanent neighbor cache entries are created through explicit 1245 administrative action; they have no timeout values and remain in 1246 place until explicitly deleted. AERO Bridges maintain permanent 1247 neighbor cache entries for their associated Proxys and Servers (and 1248 vice-versa). Each entry maintains the mapping between the neighbor's 1249 network-layer AERO address and corresponding INET address. 1251 Symmetric neighbor cache entries are created and maintained through 1252 RS/RA exchanges as specified in Section 3.15, and remain in place for 1253 durations bounded by ND/PD lifetimes. AERO Servers maintain 1254 symmetric neighbor cache entries for each of their associated 1255 Clients, and AERO Clients maintain symmetric neighbor cache entries 1256 for each of their associated Servers. The list of all Servers on the 1257 AERO link is maintained in the link's MAP list. 1259 Asymmetric neighbor cache entries are created or updated based on 1260 route optimization messaging as specified in Section 3.17, and are 1261 garbage-collected when keepalive timers expire. AERO ROSs maintain 1262 asymmetric neighbor cache entries for active targets with lifetimes 1263 based on ND messaging constants. Asymmetric neighbor cache entries 1264 are unidirectional since only the ROS (and not the ROR) creates an 1265 entry. 1267 Proxy neighbor cache entries are created and maintained by AERO 1268 Proxys when they process Client/Server ND/PD exchanges, and remain in 1269 place for durations bounded by ND/PD lifetimes. AERO Proxys maintain 1270 proxy neighbor cache entries for each of their associated Clients. 1271 Proxy neighbor cache entries track the Client state and the address 1272 of the Client's associated Server(s). 1274 To the list of neighbor cache entry states in Section 7.3.2 of 1275 [RFC4861], Proxy and Server AERO interfaces add an additional state 1276 DEPARTED that applies to symmetric and proxy neighbor cache entries 1277 for Clients that have recently departed. The interface sets a 1278 "DepartTime" variable for the neighbor cache entry to "DEPART_TIME" 1279 seconds. DepartTime is decremented unless a new ND message causes 1280 the state to return to REACHABLE. While a neighbor cache entry is in 1281 the DEPARTED state, packets destined to the target Client are 1282 forwarded to the Client's new location instead of being dropped. 1283 When DepartTime decrements to 0, the neighbor cache entry is deleted. 1284 It is RECOMMENDED that DEPART_TIME be set to the default constant 1285 value REACHABLE_TIME plus 10 seconds (40 seconds by default) to allow 1286 a window for packets in flight to be delivered while stale route 1287 optimization state may be present. 1289 When an ROR receives an authentic NS message used for route 1290 optimization, it searches for a symmetric neighbor cache entry for 1291 the target Client. The ROR then returns a solicited NA message 1292 without creating a neighbor cache entry for the ROS, but creates or 1293 updates a target Client "Report List" entry for the ROS and sets a 1294 "ReportTime" variable for the entry to REPORT_TIME seconds. The ROR 1295 resets ReportTime when it receives a new authentic NS message, and 1296 otherwise decrements ReportTime while no authentic NS messages have 1297 been received. It is RECOMMENDED that REPORT_TIME be set to the 1298 default constant value REACHABLE_TIME plus 10 seconds (40 seconds by 1299 default) to allow a window for route optimization to converge before 1300 ReportTime decrements below REACHABLE_TIME. 1302 When the ROS receives a solicited NA message response to its NS 1303 message used for route optimization, it creates or updates an 1304 asymmetric neighbor cache entry for the target network-layer and 1305 link-layer addresses. The ROS then (re)sets ReachableTime for the 1306 neighbor cache entry to REACHABLE_TIME seconds and uses this value to 1307 determine whether packets can be forwarded directly to the target, 1308 i.e., instead of via a default route. The ROS otherwise decrements 1309 ReachableTime while no further solicited NA messages arrive. It is 1310 RECOMMENDED that REACHABLE_TIME be set to the default constant value 1311 30 seconds as specified in [RFC4861]. 1313 AERO nodes also use the value MAX_UNICAST_SOLICIT to limit the number 1314 of NS keepalives sent when a correspondent may have gone unreachable, 1315 the value MAX_RTR_SOLICITATIONS to limit the number of RS messages 1316 sent without receiving an RA and the value MAX_NEIGHBOR_ADVERTISEMENT 1317 to limit the number of unsolicited NAs that can be sent based on a 1318 single event. It is RECOMMENDED that MAX_UNICAST_SOLICIT, 1319 MAX_RTR_SOLICITATIONS and MAX_NEIGHBOR_ADVERTISEMENT be set to 3 the 1320 same as specified in [RFC4861]. 1322 Different values for DEPART_TIME, REPORT_TIME, REACHABLE_TIME, 1323 MAX_UNICAST_SOLICIT, MAX_RTR_SOLCITATIONS and 1324 MAX_NEIGHBOR_ADVERTISEMENT MAY be administratively set; however, if 1325 different values are chosen, all nodes on the link MUST consistently 1326 configure the same values. Most importantly, DEPART_TIME and 1327 REPORT_TIME SHOULD be set to a value that is sufficiently longer than 1328 REACHABLE_TIME to avoid packet loss due to stale route optimization 1329 state. 1331 3.9. AERO Interface Encapsulation and Re-encapsulation 1333 In some instances, AERO interfaces insert a mid-layer IPv6 header 1334 known as the SPAN header as discussed in the following sections. 1335 After either inserting or omitting the SPAN header, the AERO 1336 interface inserts an outer encapsulation header as discussed below. 1338 AERO interfaces avoid outer encapsulation over Direct underlying 1339 interfaces and Proxyed underlying interfaces for which the first-hop 1340 access router is AERO-aware. Other AERO interfaces encapsulate 1341 packets according to whether they are entering the AERO interface 1342 from the network layer or if they are being re-admitted into the same 1343 AERO link they arrived on. This latter form of encapsulation is 1344 known as "re-encapsulation". 1346 For packets entering the AERO interface from the network layer, the 1347 AERO interface copies the "TTL/Hop Limit", "Type of Service/Traffic 1348 Class" [RFC2983], "Flow Label"[RFC6438] (for IPv6) and "Congestion 1349 Experienced" [RFC3168] values in the inner packet's IP header into 1350 the corresponding fields in the SPAN and outer encapsulation 1351 header(s). 1353 For packets undergoing re-encapsulation, the AERO interface instead 1354 copies these values from the original encapsulation header into the 1355 new encapsulation header, i.e., the values are transferred between 1356 encapsulation headers and *not* copied from the encapsulated packet's 1357 network-layer header. (Note especially that by copying the TTL/Hop 1358 Limit between encapsulation headers the value will eventually 1359 decrement to 0 if there is a (temporary) routing loop.) For IPv4 1360 encapsulation/re-encapsulation, the AERO interface sets the DF bit as 1361 discussed in Section 3.12. 1363 AERO interfaces configured over INET underlying interfaces 1364 encapsulate packets in INET headers according to the next hop 1365 determined in the forwarding algorithm in Section 3.13. If the next 1366 hop is reached via a secured tunnel, the AERO interface uses an 1367 encapsulation format specific to the secured tunnel type (see: 1368 Section 6). If the next hop is reached via an unsecured INET 1369 interface, the AERO interface instead uses UDP/IP encapsulation 1370 according to the Teredo format specified in [RFC4380] and as extended 1371 in [RFC6081]. 1373 When Teredo encapsulation is used, the AERO interface next sets the 1374 UDP source port to a constant value that it will use in each 1375 successive packet it sends, and sets the UDP length field to the 1376 length of the encapsulated packet plus 8 bytes for the UDP header 1377 itself plus the length of any included Teredo extension headers or 1378 trailers. The encapsulated packet may be either IPv6 or IPv4, as 1379 distinguished by the version number found in the first four bits. 1381 For Teredo-encapsulated packets sent to a Server, Relay or Bridge, 1382 the AERO interface sets the UDP destination port to 8060, i.e., the 1383 IANA-registered port number for AERO. For packets sent to a Client, 1384 the AERO interface sets the UDP destination port to the port value 1385 stored in the neighbor cache entry for this Client. The AERO 1386 interface finally includes/omits the UDP checksum according to 1387 [RFC6935][RFC6936]. 1389 AERO interfaces observe the packet sizing and fragmentation 1390 considerations found in Section 3.12. 1392 3.10. AERO Interface Decapsulation 1394 AERO interfaces decapsulate packets destined either to the AERO node 1395 itself or to a destination reached via an interface other than the 1396 AERO interface the packet was received on. When the encapsulated 1397 packet arrives in multiple SPAN fragments, the AERO interface 1398 reassembles as discussed in Section 3.12. Further decapsulation 1399 steps are performed according to the appropriate encapsulation format 1400 specification. 1402 3.11. AERO Interface Data Origin Authentication 1404 AERO nodes employ simple data origin authentication procedures. In 1405 particular: 1407 o AERO Bridges, Servers and Proxys accept encapsulated data packets 1408 and control messages received from secured tunnels via the SPAN. 1410 o AERO Proxys and Clients accept packets that originate from within 1411 the same secured ANET. 1413 o AERO Clients and Relays accept packets from downstream network 1414 correspondents based on ingress filtering. 1416 o AERO Clients, Relays and Servers verify the outer UDP/IP 1417 encapsulation addresses according to the Teredo specification 1418 [RFC4380]. 1420 AERO nodes silently drop any packets that do not satisfy the above 1421 data origin authentication procedures. Further security 1422 considerations are discussed in Section 6. 1424 3.12. AERO Interface MTU and Fragmentation 1426 IPv6 underlying interfaces are REQUIRED to configure a minimum 1427 Maximum Transmission Unit (MTU) of 1280 bytes [RFC8200]. The minimum 1428 MTU for IPv4 underlying interfaces is only 68 bytes [RFC1122], 1429 meaning that a packet smaller than the IPv6 MTU may require 1430 fragmentation when IPv4 encapsulation is used. Therefore, the Don't 1431 Fragment (DF) bit in the IPv4 encapsulation header MUST be set to 0. 1433 The AERO interface configures an MTU of 9180 bytes [RFC2492]; the 1434 size is therefore not a reflection of the underlying interface MTUs, 1435 but rather determines the largest packet the AERO interface can 1436 forward or reassemble. The AERO interface therefore accommodates IP 1437 packets up to 9180 bytes while generating IPv6 Path MTU Discovery 1438 (PMTUD) Packet Too Big (PTB) messages [RFC8201] as necessary (see 1439 below). 1441 AERO interfaces employ mid-layer IPv6 encapsulation and 1442 fragmentation/reassembly per [RFC2473] (aka "SPAN encapsulation") to 1443 accommodate the 9180 byte MTU. The AERO interface returns 1444 internally-generated PTB messages for packets admitted into the 1445 interface that it deems too large (e.g., according to link 1446 performance characteristics, reassembly cost, etc.) while either 1447 dropping or forwarding the packet as necessary. The AERO interface 1448 performs PMTUD even if the destination appears to be on the same link 1449 since intermediate AERO link nodes may return a PTB. This ensures 1450 that the path MTU is adaptive and reflects the current path used for 1451 a given data flow. 1453 AERO nodes perform SPAN encapsulation and fragmentation/reassembly as 1454 follows: 1456 o When a node's AERO interface sends a packet over a Proxyed, VPNed 1457 or Direct underlying interface, it sends without SPAN 1458 encapsulation if the packet is no larger than the underlying 1459 interface MTU. Otherwise, it inserts a SPAN header with source 1460 address set to the node's own SPAN address and destination set to 1461 the SPAN address of the link-layer peer Proxy, Server or Client on 1462 the underlying interface. The AERO interface then uses IPv6 1463 fragmentation to break the packet into a minimum number of non- 1464 overlapping fragments, where the largest fragment size is 1465 determined by the underlying interface MTU and the smallest 1466 fragment is no smaller than 640 bytes. The AERO interface then 1467 sends the fragments to the link-layer peer, which reassembles 1468 before forwarding toward the final destination. 1470 o When a node's AERO interface sends a packet over an INET 1471 underlying interface, it sends encapsulated packets no larger than 1472 1280 bytes without a SPAN header if the destination is reached via 1473 an INET address within the same SPAN segment. Otherwise, it 1474 inserts a SPAN header with source address set to the node's SPAN 1475 address, destination set to the SPAN address of the next hop AERO 1476 node toward the final destination and (if necessary) with a SRH 1477 with the remaining Segment IDs on the path to the final 1478 destination. The AERO interface then uses IPv6 fragmentation to 1479 break the encapsulated packet into a minimum number of non- 1480 overlapping fragments, where the largest fragment size (including 1481 both SPAN and INET encapsulation) is 1280 bytes and the smallest 1482 fragment is no smaller than 640 bytes. The AERO interface then 1483 encapsulates the SPAN fragments in INET headers and sends them to 1484 the SPAN destination, which reassembles before forwarding toward 1485 the final destination. 1487 In order to avoid a "tiny fragment" attack, AERO interfaces 1488 unconditionally drop all SPAN fragments smaller than 640 bytes. In 1489 order to set the correct context for reassembly, the AERO interface 1490 that inserts a SPAN header MUST also be the one that inserts the IPv6 1491 Fragment Header Identification value. Although all fragments of the 1492 same fragmented SPAN packet are typically sent via the same 1493 underlying interface, this is not strictly required since all 1494 fragments will arrive at the AERO interface that performs reassembly 1495 even if they travel over different paths. 1497 Note that the AERO interface can forward large packets via 1498 encapsulation and fragmentation while at the same time returning 1499 advisory PTB messages, e.g., subject to rate limiting. The receiving 1500 node that performs reassembly can also send advisory PTB messages if 1501 reassembly conditions become unfavorable. The AERO interface can 1502 therefore continuously forward large packets without loss while 1503 returning advisory messages recommending a smaller size (but no 1504 smaller than 1280). Advisory PTB messages are differentiated from 1505 PTB messages that report loss by setting the Code field in the ICMPv6 1506 message header to the value 1. This document therefore updates 1507 [RFC4443] and [RFC8201]. 1509 3.13. AERO Interface Forwarding Algorithm 1511 IP packets enter a node's AERO interface either from the network 1512 layer (i.e., from a local application or the IP forwarding system) or 1513 from the link layer (i.e., from an AERO interface neighbor). All 1514 packets entering a node's AERO interface first undergo data origin 1515 authentication as discussed in Section 3.11. Packets that satisfy 1516 data origin authentication are processed further, while all others 1517 are dropped silently. 1519 Packets that enter the AERO interface from the network layer are 1520 forwarded to an AERO interface neighbor. Packets that enter the AERO 1521 interface from the link layer are either re-admitted into the AERO 1522 link or forwarded to the network layer where they are subject to 1523 either local delivery or IP forwarding. In all cases, the AERO 1524 interface itself MUST NOT decrement the network layer TTL/Hop-count 1525 since its forwarding actions occur below the network layer. 1527 AERO interfaces may have multiple underlying interfaces and/or 1528 neighbor cache entries for neighbors with multiple ifIndex-tuple 1529 registrations (see Section 3.6). The AERO interface uses traffic 1530 classifiers (e.g., DSCP value, port number, etc.) to select an 1531 outgoing underlying interface for each packet based on the node's own 1532 QoS preferences, and also to select a destination link-layer address 1533 based on the neighbor's underlying interface with the highest 1534 preference. AERO implementations SHOULD allow for QoS preference 1535 values to be modified at runtime through network management. 1537 If multiple outgoing interfaces and/or neighbor interfaces have a 1538 preference of "high", the AERO node replicates the packet and sends 1539 one copy via each of the (outgoing / neighbor) interface pairs; 1540 otherwise, the node sends a single copy of the packet via an 1541 interface with the highest preference. AERO nodes keep track of 1542 which underlying interfaces are currently "reachable" or 1543 "unreachable", and only use "reachable" interfaces for forwarding 1544 purposes. 1546 The following sections discuss the AERO interface forwarding 1547 algorithms for Clients, Proxys, Servers and Bridges. In the 1548 following discussion, a packet's destination address is said to 1549 "match" if it is the same as a cached address, or if it is covered by 1550 a cached prefix (which may be encoded in an AERO address). 1552 3.13.1. Client Forwarding Algorithm 1554 When an IP packet enters a Client's AERO interface from the network 1555 layer the Client searches for an asymmetric neighbor cache entry that 1556 matches the destination. If there is a match, the Client uses one or 1557 more "reachable" neighbor interfaces in the entry for packet 1558 forwarding. If there is no asymmetric neighbor cache entry, the 1559 Client instead forwards the packet toward a Server (the packet is 1560 intercepted by a Proxy if there is a Proxy on the path). The Client 1561 encapsulates the packet in a SPAN header and fragments if necessary 1562 according to MTU requirements (see: Section 3.12). 1564 When an IP packet enters a Client's AERO interface from the link- 1565 layer, if the destination matches one of the Client's MNPs or link- 1566 local addresses the Client reassembles and decapsulates as necessary 1567 and delivers the inner packet to the network layer. Otherwise, the 1568 Client drops the packet and MAY return a network-layer ICMP 1569 Destination Unreachable message subject to rate limiting (see: 1570 Section 3.14). 1572 3.13.2. Proxy Forwarding Algorithm 1574 For control messages originating from or destined to a Client, the 1575 Proxy intercepts the message and updates its proxy neighbor cache 1576 entry for the Client. The Proxy then forwards a (proxyed) copy of 1577 the control message. (For example, the Proxy forwards a proxied 1578 version of a Client's NS/RS message to the target neighbor, and 1579 forwards a proxied version of the NA/RA reply to the Client.) 1581 When the Proxy receives a data packet from a Client within the ANET, 1582 the Proxy reassembles and re-fragments if necessary then searches for 1583 an asymmetric neighbor cache entry that matches the destination and 1584 forwards as follows: 1586 o if the destination matches an asymmetric neighbor cache entry, the 1587 Proxy uses one or more "reachable" neighbor interfaces in the 1588 entry for packet forwarding using SPAN encapsulation and including 1589 a SRH if necessary according to the cached TLLAO information. If 1590 the neighbor interface is in the same SPAN segment, the Proxy 1591 forwards the packet directly to the neighbor; otherwise, it 1592 forwards the packet to a Bridge. 1594 o else, the Proxy uses SPAN encapsulation and forwards the packet to 1595 a Bridge while using the SPAN address corresponding to the 1596 packet's destination as the SPAN destination address. 1598 When the Proxy receives an encapsulated data packet from an INET 1599 neighbor or from a secured tunnel from a Bridge, it accepts the 1600 packet only if data origin authentication succeeds and if there is a 1601 proxy neighbor cache entry that matches the inner destination. Next, 1602 the Proxy reassembles the packet (if necessary) and continues 1603 processing. 1605 Next if reassembly is complete and the neighbor cache state is 1606 REACHABLE, the Proxy returns a PTB if necessary (see: Section 3.12) 1607 then either drops or forwards the packet to the Client while 1608 performing SPAN encapsulation and re-fragmentation to the ANET MTU 1609 size if necessary. If the neighbor cache entry state is DEPARTED, 1610 the Proxy instead changes the SPAN destination address to the address 1611 of the new Server and forwards it to a Bridge while performing re- 1612 fragmentation to 1280 bytes if necessary. 1614 3.13.3. Server/Relay Forwarding Algorithm 1616 For control messages destined to a target Client's AERO address that 1617 are received from a secured tunnel, the Server intercepts the message 1618 and sends an appropriate response on behalf of the Client. (For 1619 example, the Server sends an NA message reply in response to an NS 1620 message directed to one of its associated Clients.) If the Client's 1621 neighbor cache entry is in the DEPARTED state, however, the Server 1622 instead forwards the packet to the Client's new Server as discussed 1623 in Section 3.19. 1625 When the Server receives an encapsulated data packet from an INET 1626 neighbor or from a secured tunnel, it accepts the packet only if data 1627 origin authentication succeeds. If the SPAN destination address is 1628 its own address, the Server continues processing as follows: 1630 o if the destination matches a symmetric neighbor cache entry in the 1631 REACHABLE state the Server prepares the packet for forwarding to 1632 the destination Client. The Server first reassembles (if 1633 necessary) and forwards the packet (while re-fragmenting if 1634 necessary) as specified in Section 3.12. 1636 o else, if the destination matches a symmetric neighbor cache entry 1637 in the DEPARETED state the Server re-encapsulates the packet and 1638 forwards it using the SPAN address of the Client's new Server as 1639 the destination. 1641 o else, if the destination matches an asymmetric neighbor cache 1642 entry, the Server uses one or more "reachable" neighbor interfaces 1643 in the entry for packet forwarding via the local INET if the 1644 neighbor is in the same SPAN segment or using SPAN encapsulation 1645 and Segment Routing if necessary with the final destination set to 1646 the neighbor's SPAN address otherwise. 1648 o else, if the destination is an AERO address that is not assigned 1649 on the AERO interface the Server drops the packet. 1651 o else, the Server (acting as a Relay) reassembles if necessary, 1652 decapsulates the packet and releases it to the network layer for 1653 local delivery or IP forwarding. Based on the information in the 1654 forwarding table, the network layer may return the packet to the 1655 same AERO interface in which case further processing occurs as 1656 below. (Note that this arrangement accommodates common 1657 implementations in which the IP forwarding table is not accessible 1658 from within the AERO interface. If the AERO interface can 1659 directly access the IP forwarding table (such as for in-kernel 1660 implementations) the forwarding table lookup can instead be 1661 performed internally from within the AERO interface itself.) 1663 When the Server's AERO interface receives a data packet from the 1664 network layer or from a VPNed or Direct Client, it performs SPAN 1665 encapsulation and fragmentation if necessary, then processes the 1666 packet according to the network-layer destination address as follows: 1668 o if the destination matches a symmetric or asymmetric neighbor 1669 cache entry the Server processes the packet as above. 1671 o else, the Server encapsulates the packet and forwards it to a 1672 Bridge using its own SPAN address as the source and the SPAN 1673 address corresponding to the destination as the destination. 1675 3.13.4. Bridge Forwarding Algorithm 1677 Bridges forward SPAN-encapsulated packets over secured tunnels the 1678 same as any IP router. When the Bridge receives a SPAN-encapsulated 1679 packet via a secured tunnel, it removes the outer INET header and 1680 searches for a forwarding table entry that matches the SPAN 1681 destination address. The Bridge then processes the packet as 1682 follows: 1684 o if the destination is the SPAN address of a Server located in the 1685 local SPAN partition, the Bridge checks for a SRH. If there is a 1686 SRH with the SPAN address of the final destination as the 1687 penultimate ID and with a Link Layer Address of the final 1688 destination as the ultimate ID, the Bridge copies the SPAN address 1689 of the final destination into the destination address. If the 1690 Link Layer Address does not indicate the presence of a NAT, the 1691 Bridge then forwards the packet directly to the Link Layer Address 1692 using link-layer (UDP/IP) encapsulation. Otherwise, the Bridge 1693 forwards the packet directly to the Server. 1695 o if the destination matches one of the Bridge's own addresses, the 1696 Bridge submits the packet for local delivery. 1698 o else, if the destination matches a forwarding table entry the 1699 Bridge forwards the packet via a secured tunnel to the next hop. 1700 If the destination matches an MSP without matching an MNP, 1701 however, the Bridge instead drops the packet and returns an ICMP 1702 Destination Unreachable message subject to rate limiting (see: 1703 Section 3.14). 1705 o else, the Bridge drops the packet and returns an ICMP Destination 1706 Unreachable as above. 1708 As for any IP router, the Bridge decrements the TTL/Hop Limit when it 1709 forwards the packet. Therefore, only the Hop Limit in the SPAN 1710 header is decremented, and not the TTL/Hop Limit in the inner packet 1711 header. 1713 3.14. AERO Interface Error Handling 1715 When an AERO node admits a packet into the AERO interface, it may 1716 receive link-layer or network-layer error indications. 1718 A link-layer error indication is an ICMP error message generated by a 1719 router in the INET on the path to the neighbor or by the neighbor 1720 itself. The message includes an IP header with the address of the 1721 node that generated the error as the source address and with the 1722 link-layer address of the AERO node as the destination address. 1724 The IP header is followed by an ICMP header that includes an error 1725 Type, Code and Checksum. Valid type values include "Destination 1726 Unreachable", "Time Exceeded" and "Parameter Problem" 1727 [RFC0792][RFC4443]. (AERO interfaces ignore all link-layer IPv4 1728 "Fragmentation Needed" and IPv6 "Packet Too Big" messages since they 1729 only emit packets that are guaranteed to be no larger than the IP 1730 minimum link MTU as discussed in Section 3.12.) 1732 The ICMP header is followed by the leading portion of the packet that 1733 generated the error, also known as the "packet-in-error". For 1734 ICMPv6, [RFC4443] specifies that the packet-in-error includes: "As 1735 much of invoking packet as possible without the ICMPv6 packet 1736 exceeding the minimum IPv6 MTU" (i.e., no more than 1280 bytes). For 1737 ICMPv4, [RFC0792] specifies that the packet-in-error includes: 1738 "Internet Header + 64 bits of Original Data Datagram", however 1739 [RFC1812] Section 4.3.2.3 updates this specification by stating: "the 1740 ICMP datagram SHOULD contain as much of the original datagram as 1741 possible without the length of the ICMP datagram exceeding 576 1742 bytes". 1744 The link-layer error message format is shown in Figure 5 (where, "L2" 1745 and "L3" refer to link-layer and network-layer, respectively): 1747 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1748 ~ ~ 1749 | L2 IP Header of | 1750 | error message | 1751 ~ ~ 1752 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1753 | L2 ICMP Header | 1754 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1755 ~ ~ P 1756 | IP and other encapsulation | a 1757 | headers of original L3 packet | c 1758 ~ ~ k 1759 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ e 1760 ~ ~ t 1761 | IP header of | 1762 | original L3 packet | i 1763 ~ ~ n 1764 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1765 ~ ~ e 1766 | Upper layer headers and | r 1767 | leading portion of body | r 1768 | of the original L3 packet | o 1769 ~ ~ r 1770 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ --- 1772 Figure 5: AERO Interface Link-Layer Error Message Format 1774 The AERO node rules for processing these link-layer error messages 1775 are as follows: 1777 o When an AERO node receives a link-layer Parameter Problem message, 1778 it processes the message the same as described as for ordinary 1779 ICMP errors in the normative references [RFC0792][RFC4443]. 1781 o When an AERO node receives persistent link-layer Time Exceeded 1782 messages, the IP ID field may be wrapping before earlier fragments 1783 awaiting reassembly have been processed. In that case, the node 1784 should begin including integrity checks and/or institute rate 1785 limits for subsequent packets. 1787 o When an AERO node receives persistent link-layer Destination 1788 Unreachable messages in response to encapsulated packets that it 1789 sends to one of its asymmetric neighbor correspondents, the node 1790 should process the message as an indication that a path may be 1791 failing, and optionally initiate NUD over that path. If it 1792 receives Destination Unreachable messages over multiple paths, the 1793 node should allow future packets destined to the correspondent to 1794 flow through a default route and re-initiate route optimization. 1796 o When an AERO Client receives persistent link-layer Destination 1797 Unreachable messages in response to encapsulated packets that it 1798 sends to one of its symmetric neighbor Servers, the Client should 1799 mark the path as unusable and use another path. If it receives 1800 Destination Unreachable messages on many or all paths, the Client 1801 should associate with a new Server and release its association 1802 with the old Server as specified in Section 3.19.5. 1804 o When an AERO Server receives persistent link-layer Destination 1805 Unreachable messages in response to encapsulated packets that it 1806 sends to one of its symmetric neighbor Clients, the Server should 1807 mark the underlying path as unusable and use another underlying 1808 path. 1810 o When an AERO Server or Proxy receives link-layer Destination 1811 Unreachable messages in response to an encapsulated packet that it 1812 sends to one of its permanent neighbors, it treats the messages as 1813 an indication that the path to the neighbor may be failing. 1814 However, the dynamic routing protocol should soon reconverge and 1815 correct the temporary outage. 1817 When an AERO Bridge receives a packet for which the network-layer 1818 destination address is covered by an MSP, if there is no more- 1819 specific routing information for the destination the Bridge drops the 1820 packet and returns a network-layer Destination Unreachable message 1821 subject to rate limiting. The Bridge writes the network-layer source 1822 address of the original packet as the destination address and uses 1823 one of its non link-local addresses as the source address of the 1824 message. 1826 When an AERO node receives an encapsulated packet for which the 1827 reassembly buffer it too small, it drops the packet and returns a 1828 network-layer Packet Too Big (PTB) message. The node first writes 1829 the MRU value into the PTB message MTU field, writes the network- 1830 layer source address of the original packet as the destination 1831 address and writes one of its non link-local addresses as the source 1832 address. 1834 3.15. AERO Router Discovery, Prefix Delegation and Autoconfiguration 1836 AERO Router Discovery, Prefix Delegation and Autoconfiguration are 1837 coordinated as discussed in the following Sections. 1839 3.15.1. AERO ND/PD Service Model 1841 Each AERO Server on the link configures a PD service to facilitate 1842 Client requests. Each Server is provisioned with a database of MNP- 1843 to-Client ID mappings for all Clients enrolled in the AERO service, 1844 as well as any information necessary to authenticate each Client. 1845 The Client database is maintained by a central administrative 1846 authority for the AERO link and securely distributed to all Servers, 1847 e.g., via the Lightweight Directory Access Protocol (LDAP) [RFC4511], 1848 via static configuration, etc. Clients receive the same service 1849 regardless of the Servers they select. 1851 AERO Clients and Servers use ND messages to maintain neighbor cache 1852 entries. AERO Servers configure their AERO interfaces as advertising 1853 NBMA interfaces, and therefore send unicast RA messages with a short 1854 Router Lifetime value (e.g., REACHABLE_TIME seconds) in response to a 1855 Client's RS message. Thereafter, Clients send additional RS messages 1856 to keep Server state alive. 1858 AERO Clients and Servers include PD parameters in RS/RA messages (see 1859 [I-D.templin-6man-dhcpv6-ndopt] for ND/PD alternatives). The unified 1860 ND/PD messages are exchanged between Client and Server according to 1861 the prefix management schedule required by the PD service. If the 1862 Client knows its MNP in advance, it can instead employ prefix 1863 registration by including its AERO address as the source address of 1864 an RS message and with an OMNI option with valid prefix registration 1865 information for the MNP. If the Server (and Proxy) accept the 1866 Client's MNP assertion, they inject the prefix into the routing 1867 system and establish the necessary neighbor cache state. 1869 The following sections specify the Client and Server behavior. 1871 3.15.2. AERO Client Behavior 1873 AERO Clients discover the addresses of Servers in a similar manner as 1874 described in [RFC5214]. Discovery methods include static 1875 configuration (e.g., from a flat-file map of Server addresses and 1876 locations), or through an automated means such as Domain Name System 1877 (DNS) name resolution [RFC1035]. Alternatively, the Client can 1878 discover Server addresses through a layer 2 data link login exchange, 1879 or through a unicast RA response to a multicast/anycast RS as 1880 described below. In the absence of other information, the Client can 1881 resolve the DNS Fully-Qualified Domain Name (FQDN) 1882 "linkupnetworks.[domainname]" where "linkupnetworks" is a constant 1883 text string and "[domainname]" is a DNS suffix for the AERO link 1884 (e.g., "example.com"). 1886 To associate with a Server, the Client acts as a requesting router to 1887 request MNPs. The Client prepares an RS message with PD parameters 1888 and includes a Nonce and Timestamp option if the Client needs to 1889 correlate RA replies. If the Client already knows the Server's AERO 1890 address, it includes the AERO address as the network-layer 1891 destination address; otherwise, it includes the link-scoped All- 1892 Routers multicast (ff02::2) or Subnet-Router anycast (fe80::) address 1893 as the network-layer destination. If the Client already knows its 1894 own AERO address, it uses the AERO address as the network-layer 1895 source address; otherwise, it uses the unspecified IPv6 address 1896 (::/128) as the network-layer source address. 1898 The Client next includes an OMNI option in the RS message to register 1899 its link-layer information with the Server. The Client sets the OMNI 1900 option prefix registration information according to the MNP, and 1901 includes an ifIndex-tuple with S set to '1' corresponding to the 1902 underlying interface over which the Client will send the RS message. 1903 The Client MAY include additional ifIndex-tuples specific to other 1904 underlying interfaces. The Client MAY also include an SLLAO 1905 corresponding to the OMNI option ifIndex-tuple with S set to '1'. 1907 The Client then sends the RS message (either directly via Direct 1908 interfaces, via a VPN for VPNed interfaces, via a Proxy for proxyed 1909 interfaces or via INET encapsulation for INET interfaces) and waits 1910 for an RA message reply (see Section 3.15.3). The Client retries up 1911 to MAX_RTR_SOLICITATIONS times until an RA is received. If the 1912 Client receives no RAs, or if it receives an RA with Router Lifetime 1913 set to 0, the Client SHOULD abandon this Server and try another 1914 Server. Otherwise, the Client processes the PD information found in 1915 the RA message. 1917 Next, the Client creates a symmetric neighbor cache entry with the 1918 Server's AERO address as the network-layer address and the Server's 1919 encapsulation and/or link-layer addresses as the link-layer address. 1920 The Client records the RA Router Lifetime field value in the neighbor 1921 cache entry as the time for which the Server has committed to 1922 maintaining the MNP in the routing system via this underlying 1923 interface, and caches the other RA configuration information 1924 including Cur Hop Limit, M and O flags, Reachable Time and Retrans 1925 Timer. The Client then autoconfigures AERO addresses for each of the 1926 delegated MNPs and assigns them to the AERO interface. The Client 1927 also caches any MSPs included in Route Information Options (RIOs) 1928 [RFC4191] as MSPs to associate with the AERO link, and assigns the 1929 MTU value in the MTU option to the underlying interface. 1931 The Client then registers additional underlying interfaces with the 1932 Server by sending RS messages via each additional interface. The RS 1933 messages include the same parameters as for the initial RS/RA 1934 exchange, but with destination address set to the Server's AERO 1935 address. 1937 Following autoconfiguration, the Client sub-delegates the MNPs to its 1938 attached EUNs and/or the Client's own internal virtual interfaces as 1939 described in [I-D.templin-v6ops-pdhost] to support the Client's 1940 downstream attached "Internet of Things (IoT)". The Client 1941 subsequently sends additional RS messages over each underlying 1942 interface before the Router Lifetime received for that interface 1943 expires. 1945 After the Client registers its underlying interfaces, it may wish to 1946 change one or more registrations, e.g., if an interface changes 1947 address or becomes unavailable, if QoS preferences change, etc. To 1948 do so, the Client prepares an RS message to send over any available 1949 underlying interface. The RS includes an OMNI option with prefix 1950 registration information specific to its MNP, with an ifIndex-tuple 1951 specific to the selected underlying interface with S set to '1', and 1952 with any additional ifIndex-tuples specific to other underlying 1953 interfaces. The Client includes fresh ifIndex-tuple values to update 1954 the Server's neighbor cache entry. When the Client receives the 1955 Server's RA response, it has assurance that the Server has been 1956 updated with the new information. 1958 If the Client wishes to discontinue use of a Server it issues an RS 1959 message over any underlying interface with an OMNI option with a 1960 prefix release indication. When the Server processes the message, it 1961 releases the MNP, sets the symmetric neighbor cache entry state for 1962 the Client to DEPARTED and returns an RA reply with Router Lifetime 1963 set to 0. After a short delay (e.g., 2 seconds), the Server 1964 withdraws the MNP from the routing system. 1966 3.15.3. AERO Server Behavior 1968 AERO Servers act as IP routers and support a PD service for Clients. 1969 Servers arrange to add their AERO addresses to a static map of Server 1970 addresses for the link and/or the DNS resource records for the FQDN 1971 "linkupnetworks.[domainname]" before entering service. Server 1972 addresses should be geographically and/or topologically referenced, 1973 and made available for discovery by Clients on the AERO link. 1975 When a Server receives a prospective Client's RS message on its AERO 1976 interface, it SHOULD return an immediate RA reply with Router 1977 Lifetime set to 0 if it is currently too busy or otherwise unable to 1978 service the Client. Otherwise, the Server authenticates the RS 1979 message and processes the PD parameters. The Server first determines 1980 the correct MNPs to delegate to the Client by searching the Client 1981 database. When the Server delegates the MNPs, it also creates a 1982 forwarding table entry for each MNP so that the MNPs are propagated 1983 into the routing system (see: Section 3.3). For IPv6, the Server 1984 creates an IPv6 forwarding table entry for each MNP. For IPv4, the 1985 Server creates an IPv6 forwarding table entry with the SPAN 1986 Compatibility Prefix (SCP) corresponding to the IPv4 address. 1988 The Server next creates a symmetric neighbor cache entry for the 1989 Client using the base AERO address as the network-layer address and 1990 with lifetime set to no more than the smallest PD lifetime. Next, 1991 the Server updates the neighbor cache entry by recording the 1992 information in each ifIndex-tuple in the RS OMNI option. The Server 1993 also records the actual SPAN/INET addresses in the neighbor cache 1994 entry. 1996 Next, the Server prepares an RA message using its AERO address as the 1997 network-layer source address and the network-layer source address of 1998 the RS message as the network-layer destination address. The Server 1999 sets the Router Lifetime to the time for which it will maintain both 2000 this underlying interface individually and the symmetric neighbor 2001 cache entry as a whole. The Server also sets Cur Hop Limit, M and O 2002 flags, Reachable Time and Retrans Timer to values appropriate for the 2003 AERO link. The Server includes the delegated MNPs, any other PD 2004 parameters and an OMNI option with no ifIndex-tuples. The Server 2005 then includes one or more RIOs that encode the MSPs for the AERO 2006 link, plus an MTU option (see Section 3.12). The Server finally 2007 forwards the message to the Client using SPAN/INET, INET, or NULL 2008 encapsulation as necessary. 2010 After the initial RS/RA exchange, the Server maintains a 2011 ReachableTime timer for each of the Client's underlying interfaces 2012 individually (and for the Client's symmetric neighbor cache entry 2013 collectively) set to expire after Router Lifetime seconds. If the 2014 Client (or Proxy) issues additional RS messages, the Server sends an 2015 RA response and resets ReachableTime. If the Server receives an ND 2016 message with PD release indication it sets the Client's symmetric 2017 neighbor cache entry to the DEPARTED state and withdraws the MNP from 2018 the routing system after a short delay (e.g., 2 seconds). If 2019 ReachableTime expires before a new RS is received on an individual 2020 underlying interface, the Server marks the interface as DOWN. If 2021 ReachableTime expires before any new RS is received on any individual 2022 underlying interface, the Server deletes the neighbor cache entry and 2023 withdraws the MNP without delay. 2025 The Server processes any ND/PD messages pertaining to the Client and 2026 returns an NA/RA reply in response to solicitations. The Server may 2027 also issue unsolicited RA messages, e.g., with PD reconfigure 2028 parameters to cause the Client to renegotiate its PDs, with Router 2029 Lifetime set to 0 if it can no longer service this Client, etc. 2030 Finally, If the symmetric neighbor cache entry is in the DEPARTED 2031 state, the Server deletes the entry after DepartTime expires. 2033 Note: Clients SHOULD notify former Servers of their departures, but 2034 Servers are responsible for expiring neighbor cache entries and 2035 withdrawing routes even if no departure notification is received 2036 (e.g., if the Client leaves the network unexpectedly). Servers 2037 SHOULD therefore set Router Lifetime to REACHABLE_TIME seconds in 2038 solicited RA messages to minimize persistent stale cache information 2039 in the absence of Client departure notifications. A short Router 2040 Lifetime also ensures that proactive Client/Server RS/RA messaging 2041 will keep any NAT state alive (see above). 2043 Note: All Servers on an AERO link MUST advertise consistent values in 2044 the RA Cur Hop Limit, M and O flags, Reachable Time and Retrans Timer 2045 fields the same as for any link, since unpredictable behavior could 2046 result if different Servers on the same link advertised different 2047 values. 2049 3.15.3.1. Lightweight DHCPv6 Relay Agent (LDRA) 2051 When DHCPv6 is used as the ND/PD service back end, AERO Clients and 2052 Servers are always on the same link (i.e., the AERO link) from the 2053 perspective of DHCPv6. However, in some implementations the DHCPv6 2054 server and ND function may be located in separate modules. In that 2055 case, the Server's AERO interface module can act as a Lightweight 2056 DHCPv6 Relay Agent (LDRA)[RFC6221] to relay PD messages to and from 2057 the DHCPv6 server module. 2059 When the LDRA receives an authentic RS message, it extracts the PD 2060 message parameters and uses them to construct an IPv6/UDP/DHCPv6 2061 message. It sets the IPv6 source address to the source address of 2062 the RS message, sets the IPv6 destination address to 2063 'All_DHCP_Relay_Agents_and_Servers' and sets the UDP fields to values 2064 that will be understood by the DHCPv6 server. 2066 The LDRA then wraps the message in a DHCPv6 'Relay-Forward' message 2067 header and includes an 'Interface-Id' option that includes enough 2068 information to allow the LDRA to forward the resulting Reply message 2069 back to the Client (e.g., the Client's link-layer addresses, a 2070 security association identifier, etc.). The LDRA also wraps the OMNI 2071 option and SLLAO into the Interface-Id option, then forwards the 2072 message to the DHCPv6 server. 2074 When the DHCPv6 server prepares a Reply message, it wraps the message 2075 in a 'Relay-Reply' message and echoes the Interface-Id option. The 2076 DHCPv6 server then delivers the Relay-Reply message to the LDRA, 2077 which discards the Relay-Reply wrapper and IPv6/UDP headers, then 2078 uses the DHCPv6 message to construct an RA response to the Client. 2079 The Server uses the information in the Interface-Id option to prepare 2080 the RA message and to cache the link-layer addresses taken from the 2081 OMNI option and SLLAO echoed in the Interface-Id option. 2083 3.16. The AERO Proxy 2085 Clients may connect to ANETs that deploy perimeter security services 2086 to facilitate communications to Servers in outside INETs. In that 2087 case, the ANET can employ an AERO Proxy. The Proxy is located at the 2088 ANET/INET border and listens for RS messages originating from or RA 2089 messages destined to ANET Clients. The Proxy acts on these control 2090 messages as follows: 2092 o when the Proxy receives an RS message from a new ANET Client, it 2093 first authenticates the message then examines the network-layer 2094 destination address. If the destination address is a Server's 2095 AERO address, the Proxy proceeds to the next step. Otherwise, if 2096 the destination is All-Routers multicast or Subnet-Router anycast, 2097 the Proxy selects a "nearby" Server that is likely to be a good 2098 candidate to serve the Client and replaces the destination address 2099 with the Server's AERO address. Next, the Proxy creates a proxy 2100 neighbor cache entry and caches the Client and Server link-layer 2101 addresses along with the OMNI option information and any other 2102 identifying information including Transaction IDs, Client 2103 Identifiers, Nonce values, etc. The Proxy finally encapsulates 2104 the (proxyed) RS message in a SPAN header with source set to the 2105 Proxy's SPAN address and destination set to the Server's SPAN 2106 address then forwards the message into the SPAN. 2108 o when the Server receives the RS, it authenticates the message then 2109 creates or updates a symmetric neighbor cache entry for the Client 2110 with the Proxy's SPAN address as the link-layer address. The 2111 Server then sends an RA message back to the Proxy via the SPAN. 2113 o when the Proxy receives the RA, it authenticates the message and 2114 matches it with the proxy neighbor cache entry created by the RS. 2115 The Proxy then caches the PD route information as a mapping from 2116 the Client's MNPs to the Client's ANET address, caches the 2117 Server's advertised Router Lifetime and sets the neighbor cache 2118 entry state to REACHABLE. The Proxy then sets the P bit in the RA 2119 flags field, optionally rewrites the Router Lifetime and forwards 2120 the (proxyed) message to the Client. The Proxy finally includes 2121 an MTU option (if necessary) with an MTU to use for the underlying 2122 ANET interface. 2124 After the initial RS/RA exchange, the Proxy forwards any Client data 2125 packets for which there is no matching asymmetric neighbor cache 2126 entry to a Bridge using SPAN encapsulation with its own SPAN address 2127 as the source and the SPAN address corresponding to the Client as the 2128 destination. The Proxy instead forwards any Client data destined to 2129 an asymmetric neighbor cache target directly to the target according 2130 to the SPAN/link-layer information - the process of establishing 2131 asymmetric neighbor cache entries is specified in Section 3.17. 2133 While the Client is still attached to the ANET, the Proxy sends NS, 2134 RS and/or unsolicited NA messages to update the Server's symmetric 2135 neighbor cache entries on behalf of the Client and/or to convey QoS 2136 updates. This allows for higher-frequency Proxy-initiated RS/RA 2137 messaging over well-connected INET infrastructure supplemented by 2138 lower-frequency Client-initiated RS/RA messaging over constrained 2139 ANET data links. 2141 If the Server ceases to send solicited advertisements, the Proxy 2142 sends unsolicited RAs on the ANET interface with destination set to 2143 All-Nodes multicast (ff02::1) and with Router Lifetime set to zero to 2144 inform Clients that the Server has failed. Although the Proxy 2145 engages in ND exchanges on behalf of the Client, the Client can also 2146 send ND messages on its own behalf, e.g., if it is in a better 2147 position than the Proxy to convey QoS changes, etc. For this reason, 2148 the Proxy marks any Client-originated solicitation messages (e.g. by 2149 inserting a Nonce option) so that it can return the solicited 2150 advertisement to the Client instead of processing it locally. 2152 If the Client becomes unreachable, the Proxy sets the neighbor cache 2153 entry state to DEPARTED and retains the entry for DEPART_TIME 2154 seconds. While the state is DEPARTED, the Proxy forwards any packets 2155 destined to the Client to a Bridge via SPAN encapsulation with the 2156 Client's current Server as the destination. The Bridge in turn 2157 forwards the packets to the Client's current Server. When DepartTime 2158 expires, the Proxy deletes the neighbor cache entry and discards any 2159 further packets destined to this (now forgotten) Client. 2161 In some ANETs that employ a Proxy, the Client's MNP can be injected 2162 into the ANET routing system. In that case, the Client can send data 2163 messages without encapsulation so that the ANET routing system 2164 transports the unencapsulated packets to the Proxy. This can be very 2165 beneficial, e.g., if the Client connects to the ANET via low-end data 2166 links such as some aviation wireless links. 2168 If the first-hop ANET access router is AERO-aware, the Client can 2169 avoid encapsulation for both its control and data messages. When the 2170 Client connects to the link, it can send an unencapsulated RS message 2171 with source address set to its AERO address and with destination 2172 address set to the AERO address of the Client's selected Server or to 2173 All-Routers multicast or Subnet-Router anycast. The Client includes 2174 an OMNI option formatted as specified in 2175 [I-D.templin-6man-omni-interface]. 2177 The Client then sends the unencapsulated RS message, which will be 2178 intercepted by the AERO-Aware access router. The access router then 2179 encapsulates the RS message in an ANET header with its own address as 2180 the source address and the address of a Proxy as the destination 2181 address. The access router further remembers the address of the 2182 Proxy so that it can encapsulate future data packets from the Client 2183 via the same Proxy. If the access router needs to change to a new 2184 Proxy, it simply sends another RS message toward the Server via the 2185 new Proxy on behalf of the Client. 2187 In some cases, the access router and Proxy may be one and the same 2188 node. In that case, the node would be located on the same physical 2189 link as the Client, but its message exchanges with the Server would 2190 need to pass through a security gateway at the ANET/INET border. The 2191 method for deploying access routers and Proxys (i.e. as a single node 2192 or multiple nodes) is an ANET-local administrative consideration. 2194 3.16.1. Detecting and Responding to Server Failures 2196 In environments where fast recovery from Server failure is required, 2197 Proxys SHOULD use proactive Neighbor Unreachability Detection (NUD) 2198 to track Server reachability in a similar fashion as for 2199 Bidirectional Forwarding Detection (BFD) [RFC5880]. Proxys can then 2200 quickly detect and react to failures so that cached information is 2201 re-established through alternate paths. The NUD control messaging is 2202 carried only over well-connected ground domain networks (i.e., and 2203 not low-end aeronautical radio links) and can therefore be tuned for 2204 rapid response. 2206 Proxys perform proactive NUD with Servers for which there are 2207 currently active ANET Clients by sending continuous NS messages in 2208 rapid succession, e.g., one message per second. The Proxy sends the 2209 NS message via the SPAN with the Proxy's AERO address as the source 2210 and the AERO address of the Server as the destination. When the 2211 Proxy is also sending RS messages to the Server on behalf of ANET 2212 Clients, the resulting RA responses can be considered as equivalent 2213 hints of forward progress. This means that the Proxy need not also 2214 send a periodic NS if it has already sent an RS within the same 2215 period. If the Server fails (i.e., if the Proxy ceases to receive 2216 advertisements), the Proxy can quickly inform Clients by sending 2217 multicast RA messages on the ANET interface. 2219 The Proxy sends RA messages on the ANET interface with source address 2220 set to the Server's address, destination address set to All-Nodes 2221 multicast, and Router Lifetime set to 0. The Proxy SHOULD send 2222 MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small delays 2223 [RFC4861]. Any Clients on the ANET that had been using the failed 2224 Server will receive the RA messages and associate with a new Server. 2226 3.16.2. Point-to-Multipoint Server Coordination 2228 In environments where Client messaging over ANETs is bandwidth- 2229 limited and/or expensive, Clients can enlist the services of the 2230 Proxy to coordinate with multiple Servers in a single RS/RA message 2231 exchange. The Client can send a single RS message to All-Routers 2232 multicast that includes the ID's of multiple Servers in MS-Register 2233 sub-options of the OMNI option. 2235 When the Proxy receives the RS and processes the OMNI option, it 2236 performs a separate RS/RA exchange with each MS-Register Server. 2237 When it has received the RA messages, it creates an "aggregate" RA 2238 message to return to the Client with an OMNI option with each 2239 responding Server's ID recorded in an MS-Register sub-option. 2241 Clients can thereafter employ efficient point-to-multipoint Server 2242 coordination under the assistance of the Proxy to dramatically reduce 2243 the number of messages sent over the ANET while enlisting the support 2244 of multiple Servers for fault tolerance. Clients can further include 2245 MS-Release suboptions in RS messages to request the Proxy to release 2246 from former Servers via the procedures discussed in Section 3.19.5. 2248 The OMNI interface specification [I-D.templin-6man-omni-interface] 2249 provides further discussion of the Client/Proxy RS/RA messaging 2250 involved in point-to-multipoint coordination. 2252 3.17. AERO Route Optimization 2254 While data packets are flowing between a source and target node, 2255 route optimization SHOULD be used. Route optimization is initiated 2256 by the first eligible Route Optimization Source (ROS) closest to the 2257 source as follows: 2259 o For Clients on VPNed and Direct interfaces, the Server is the ROS. 2261 o For Clients on Proxyed interfaces, the Proxy is the ROS. 2263 o For Clients on INET interfaces, the Client itself is the ROS. 2265 o For correspondent nodes on INET/EUN interfaces serviced by a 2266 Relay, the Relay is the ROS. 2268 The route optimization procedure is conducted between the ROS and the 2269 target Server/Relay acting as a Route Optimization Responder (ROR) in 2270 the same manner as for IPv6 ND Address Resolution and using the same 2271 NS/NA messaging. The target may either be a MNP Client serviced by a 2272 Server, or a non-MNP correspondent reachable via a Relay. 2274 The procedures are specified in the following sections. 2276 3.17.1. Route Optimization Initiation 2278 While data packets are flowing from the source node toward a target 2279 node, the ROS performs address resolution by sending an NS message 2280 for Address Resolution (NS(AR)) to receive a solicited NA message 2281 from the ROR. When the ROS sends an NS(AR), it includes: 2283 o the AERO address of the ROS as the source address. 2285 o the data packet's destination as the Target Address. 2287 o the Solicited-Node multicast address [RFC4291] formed from the 2288 lower 24 bits of the data packet's destination as the destination 2289 address, e.g., for 2001:db8:1:2::10:2000 the NS destination 2290 address is ff02:0:0:0:0:1:ff10:2000. 2292 The NS(AR) message includes an OMNI option with no ifIndex-tuples and 2293 no SLLAO, such that the target will not create a neighbor cache 2294 entry. 2296 The ROS then encapsulates the NS(AR) message in a SPAN header with 2297 source set to its own SPAN address and destination set to the SPAN 2298 address corresponding to the packet's final destination, then sends 2299 the message into the SPAN without decrementing the network-layer TTL/ 2300 Hop Limit field. 2302 3.17.2. Relaying the NS 2304 When the Bridge receives the NS(AR) message from the ROS, it discards 2305 the INET header and determines that the ROR is the next hop by 2306 consulting its standard IPv6 forwarding table for the SPAN header 2307 destination address. The Bridge then forwards the message toward the 2308 ROR via the SPAN the same as for any IPv6 router. The final-hop 2309 Bridge in the SPAN will deliver the message via a secured tunnel to 2310 the ROR. 2312 3.17.3. Processing the NS and Sending the NA 2314 When the ROR receives the NS(AR) message, it examines the Target 2315 Address to determine whether it has a neighbor cache entry and/or 2316 route that matches the target. If there is no match, the ROR drops 2317 the message. Otherwise, the ROR continues processing as follows: 2319 o if the target belongs to an MNP Client neighbor in the DEPARTED 2320 state the ROR changes the NS(AR) message SPAN destination address 2321 to the SPAN address of the Client's new Server, forwards the 2322 message into the SPAN and returns from processing. 2324 o If the target belongs to an MNP Client neighbor in the REACHABLE 2325 state, the ROR instead adds the AERO source address to the target 2326 Client's Report List with time set to ReportTime. 2328 o If the target belongs to a non-MNP route, the ROR continues 2329 processing without adding an entry to the Report List. 2331 The ROR then prepares a solicited NA message to send back to the ROS 2332 but does not create a neighbor cache entry. The ROR sets the NA 2333 source address to the AERO address corresponding to the target, sets 2334 the Target Address to the target of the solicitation, and sets the 2335 destination address to the source of the solicitation. 2337 The ROR then includes an OMNI option with prefix registration length 2338 set to the length of the MNP if the target is an MNP Client; 2339 otherwise, set to the maximum of the non-MNP prefix length and 64. 2340 (Note that a /64 limit is imposed to avoid causing the ROS to set 2341 short prefixes (e.g., "default") that would match destinations for 2342 which the routing system includes more-specific prefixes.) 2344 If the target is an MNP Client, the ROR next includes ifIndex-tuples 2345 in the OMNI option for each of the target Client's underlying 2346 interfaces with current information for each interface and with the S 2347 flag set to 0. The ROR then includes a TLLAO with ifIndex-tuples in 2348 one-to-one correspondence with the tuples that appear in the OMNI 2349 option. 2351 The ROR sets the Link Layer Address and Port Number (if necessary) to 2352 its own INET address for VPNed and Direct interfaces or to the INET 2353 address of the Proxy for Proxyed interface, then includes its own 2354 SPAN address or the SPAN address of the Proxy as the ultimate Segment 2355 Routing List entry. For INET interfaces, the ROR instead sets the 2356 Link Layer Address and Port Number (if necessary) to the Client's 2357 INET address then sets its own SPAN address in the penultimate 2358 Segment Routing List entry and sets the target's SPAN address in the 2359 ultimate Segment Routing List entry. 2361 The ROR then sets the NA message R flag to 1 (as a router), S flag to 2362 1 (as a response to a solicitation), and O flag to 0 (as a proxy). 2363 The ROR finally encapsulates the NA message in a SPAN header with 2364 source set to its own SPAN address and destination set to the source 2365 SPAN address of the NS(AR) message, then forwards the message into 2366 the SPAN without decrementing the network-layer TTL/Hop Limit field. 2368 3.17.4. Relaying the NA 2370 When the Bridge receives the NA message from the ROR, it discards the 2371 INET header and determines that the ROS is the next hop by consulting 2372 its standard IPv6 forwarding table for the SPAN header destination 2373 address. The Bridge then forwards the SPAN-encapsulated NA message 2374 toward the ROS the same as for any IPv6 router. The final-hop Bridge 2375 in the SPAN will deliver the message via a secured tunnel to the ROS. 2377 3.17.5. Processing the NA 2379 When the ROS receives the solicited NA message, it processes the 2380 message the same as for standard IPv6 Address Resolution [RFC4861]. 2381 In the process, it caches the source SPAN address then creates an 2382 asymmetric neighbor cache entry for the ROR and caches all 2383 information found in the OMNI and TLLAO options. The ROS finally 2384 sets the asymmetric neighbor cache entry lifetime to REACHABLE_TIME 2385 seconds. 2387 3.17.6. Route Optimization Maintenance 2389 Following route optimization, the ROS forwards future data packets 2390 destined to the target via the addresses found in the cached link- 2391 layer information. The route optimization is shared by all sources 2392 that send packets to the target via the ROS, i.e., and not just the 2393 source on behalf of which the route optimization was initiated. 2395 While new data packets destined to the target are flowing through the 2396 ROS, it sends additional NS(AR) messages to the ROR before 2397 ReachableTime expires to receive a fresh solicited NA message the 2398 same as described in the previous sections (route optimization 2399 refreshment strategies are an implementation matter, with a non- 2400 normative example given in Appendix B.1). The ROS uses the cached 2401 SPAN address of the ROR as the NS(AR) SPAN destination address, and 2402 sends up to MAX_MULTICAST_SOLICIT NS(AR) messages separated by 1 2403 second until an NA is received. If no NA is received, the ROS 2404 assumes that the current ROR has become unreachable and deletes the 2405 neighbor cache entry. Subsequent data packets will trigger a new 2406 route optimization per Section 3.17.1 to discover a new ROR while 2407 initial data packets travel over a suboptimal route. 2409 If an NA is received, the ROS then updates the asymmetric neighbor 2410 cache entry to refresh ReachableTime, while (for MNP destinations) 2411 the ROR adds or updates the ROS address to the target Client's Report 2412 List and with time set to ReportTime. While no data packets are 2413 flowing, the ROS instead allows ReachableTime for the asymmetric 2414 neighbor cache entry to expire. When ReachableTime expires, the ROS 2415 deletes the asymmetric neighbor cache entry. Any future data packets 2416 flowing through the ROS will again trigger a new route optimization. 2418 The ROS may also receive unsolicited NA messages from the ROR at any 2419 time (see: Section 3.19). If there is an asymmetric neighbor cache 2420 entry for the target, the ROS updates the link-layer information but 2421 does not update ReachableTime since the receipt of an unsolicited NA 2422 does not confirm that any forward paths are working. If there is no 2423 asymmetric neighbor cache entry, the ROS simply discards the 2424 unsolicited NA. 2426 In this arrangement, the ROS holds an asymmetric neighbor cache entry 2427 for the ROR, but the ROR does not hold an asymmetric neighbor cache 2428 entry for the ROS. The route optimization neighbor relationship is 2429 therefore asymmetric and unidirectional. If the target node also has 2430 packets to send back to the source node, then a separate route 2431 optimization procedure is performed in the reverse direction. But, 2432 there is no requirement that the forward and reverse paths be 2433 symmetric. 2435 3.18. Neighbor Unreachability Detection (NUD) 2437 AERO nodes perform Neighbor Unreachability Detection (NUD) per 2438 [RFC4861] either reactively in response to persistent link-layer 2439 errors (see Section 3.14) or proactively to confirm reachability. 2440 The NUD algorithm is based on periodic control message exchanges. 2441 The algorithm may further be seeded by ND hints of forward progress, 2442 but care must be taken to avoid inferring reachability based on 2443 spoofed information. For example, authentic IPv6 ND message 2444 exchanges may be considered as acceptable hints of forward progress, 2445 while spurious data packets should not be. 2447 AERO Servers, Proxys and Relays can use standard NS/NA NUD exchanges 2448 sent over the SPAN to securely test reachability without risk of DoS 2449 attacks from nodes pretending to be a neighbor; Proxys can further 2450 perform NUD to securely verify Server reachability on behalf of their 2451 proxyed Clients. However, a means for a ROS to test the unsecured 2452 forward directions of target route optimized paths is also necessary. 2454 When an ROR directs an ROS to a neighbor with one or more target 2455 link-layer addresses, the ROS can proactively test each such 2456 unsecured route optimized path by sending "loopback" NS(NUD) 2457 messages. While testing the paths, the ROS can optionally continue 2458 to send packets via the SPAN, maintain a small queue of packets until 2459 target reachability is confirmed, or (optimistically) allow packets 2460 to flow via the route optimized paths. 2462 When the ROS sends a loopback NS(NUD) message, it uses its AERO 2463 address as both the IPv6 source and destination address, and the MNP 2464 Subnet-Router anycast address as the Target Address. The ROS 2465 includes a Nonce and Timestamp option, then encapsulates the message 2466 in SPAN/INET headers with its own SPAN address as the source and the 2467 SPAN address of the route optimization target as the destination. 2468 The ROS then forwards the message to the target (either directly to 2469 the link layer address of the target if the target is in the same 2470 SPAN segment, or via a Bridge if the target is in a different SPAN 2471 segment). 2473 When the route optimization target receives the NS(NUD) message, it 2474 notices that the IPv6 destination address is the same as the source 2475 address. It then reverses the SPAN source and destination addresses 2476 and returns the message to the ROS (either directly or via the SPAN). 2477 The route optimization target does not decrement the NS(NUD) message 2478 IPv6 Hop-Limit in the process, since the message has not exited the 2479 SPAN. 2481 When the ROS receives the NS(NUD) message, it can determine from the 2482 Nonce, Timestamp and Target Address that the message originated from 2483 itself and that it transited the forward path. The ROS need not 2484 prepare an NA response, since the destination of the response would 2485 be itself and testing the route optimization path again would be 2486 redundant. 2488 The ROS marks route optimization target paths that pass these NUD 2489 tests as "reachable", and those that do not as "unreachable". These 2490 markings inform the AERO interface forwarding algorithm specified in 2491 Section 3.13. 2493 Note that to avoid a DoS vector nodes MUST NOT return loopback 2494 NS(NUD) messages received from an unsecured link-layer source via a 2495 secured SPAN path. 2497 3.19. Mobility Management and Quality of Service (QoS) 2499 AERO is a Distributed Mobility Management (DMM) service. Each Server 2500 is responsible for only a subset of the Clients on the AERO link, as 2501 opposed to a Centralized Mobility Management (CMM) service where 2502 there is a single network mobility collective entity for all Clients. 2503 Clients coordinate with their associated Servers via RS/RA exchanges 2504 to maintain the DMM profile, and the AERO routing system tracks all 2505 current Client/Server peering relationships. 2507 Servers provide default routing and mobility/multilink services for 2508 their dependent Clients. Clients are responsible for maintaining 2509 neighbor relationships with their Servers through periodic RS/RA 2510 exchanges, which also serves to confirm neighbor reachability. When 2511 a Client's underlying interface address and/or QoS information 2512 changes, the Client is responsible for updating the Server with this 2513 new information. Note that for Proxyed interfaces, however, the 2514 Proxy can also perform some RS/RA exchanges on the Client's behalf. 2516 Mobility management considerations are specified in the following 2517 sections. 2519 3.19.1. Mobility Update Messaging 2521 Servers accommodate Client mobility/multilink and/or QoS change 2522 events by sending unsolicited NA (uNA) messages to each ROS in the 2523 target Client's Report List. When a Server sends a uNA message, it 2524 sets the IPv6 source address to the Client's AERO address, sets the 2525 destination address to All-Nodes multicast and sets the Target 2526 Address to the Client's Subnet-Router anycast address. The Server 2527 also includes an OMNI option with prefix registration information and 2528 with ifIndex-tuples for the target Client's remaining interfaces with 2529 S set to 0. The Server then includes a TLLAO with corresponding 2530 ifIndex-tuples prepared the same as for the initial route 2531 optimization event. The Server sets the NA R flag to 1, the S flag 2532 to 0 and the O flag to 0, then encapsulates the message in a SPAN 2533 header with source set to its own SPAN address and destination set to 2534 the SPAN address of the ROS and sends the message into the SPAN. 2536 As discussed in Section 7.2.6 of [RFC4861], the transmission and 2537 reception of uNA messages is unreliable but provides a useful 2538 optimization. In well-connected Internetworks with robust data links 2539 uNA messages will be delivered with high probability, but in any case 2540 the Server can optionally send up to MAX_NEIGHBOR_ADVERTISEMENT uNAs 2541 to each ROS to increase the likelihood that at least one will be 2542 received. 2544 When the ROS receives a uNA message, it ignores the message if there 2545 is no existing neighbor cache entry for the Client. Otherwise, it 2546 uses the included OMNI option and TLLAO information to update the 2547 neighbor cache entry, but does not reset ReachableTime since the 2548 receipt of an unsolicited NA message from the target Server does not 2549 provide confirmation that any forward paths to the target Client are 2550 working. 2552 If uNA messages are lost, the ROS may be left with stale address and/ 2553 or QoS information for the Client for up to REACHABLE_TIME seconds. 2554 During this time, the ROS can continue sending packets according to 2555 its stale neighbor cache information. When ReachableTime is close to 2556 expiring, the ROS will re-initiate route optimization and receive 2557 fresh link-layer address information. 2559 In addition to sending uNA messages to the current set of ROSs for 2560 the Client, the Server also sends uNAs to the former link-layer 2561 address for any ifIndex-tuple for which the link-layer address has 2562 changed. The uNA messages update Proxys that cannot easily detect 2563 (e.g., without active probing) when a formerly-active Client has 2564 departed. 2566 3.19.2. Announcing Link-Layer Address and/or QoS Preference Changes 2568 When a Client needs to change its underlying interface addresses and/ 2569 or QoS preferences (e.g., due to a mobility event), either the Client 2570 or its Proxys send RS messages to the Server via the SPAN with an 2571 OMNI option that includes an ifIndex-tuple with S set to 1 and with 2572 the new link quality and address information. 2574 Up to MAX_RTR_SOLICITATIONS RS messages MAY be sent in parallel with 2575 sending actual data packets in case one or more RAs are lost. If all 2576 RAs are lost, the Client SHOULD re-associate with a new Server. 2578 When the Server receives the Client's changes, it sends uNA messages 2579 to all nodes in the Report List the same as described in the previous 2580 section. 2582 3.19.3. Bringing New Links Into Service 2584 When a Client needs to bring new underlying interfaces into service 2585 (e.g., when it activates a new data link), it sends an RS message to 2586 the Server via the underlying interface with an OMNI option that 2587 includes an ifIndex-tuple with S set to 1 and appropriate link 2588 quality values and with link-layer address information for the new 2589 link. 2591 3.19.4. Removing Existing Links from Service 2593 When a Client needs to remove existing underlying interfaces from 2594 service (e.g., when it de-activates an existing data link), it sends 2595 an RS or uNA message to its Server with an OMNI option with 2596 appropriate link quality values. 2598 If the Client needs to send RS/uNA messages over an underlying 2599 interface other than the one being removed from service, it MUST 2600 include ifIndex-tuples with appropriate link quality values for any 2601 underlying interfaces being removed from service. 2603 3.19.5. Moving to a New Server 2605 When a Client associates with a new Server, it performs the Client 2606 procedures specified in Section 3.15.2. The Client also includes MS- 2607 Release identifiers in the RS message OMNI option per 2608 [I-D.templin-6man-omni-interface] if it wants the new Server to 2609 notify any old Servers from which the Client is departing. 2611 When the new Server receives the Client's RS message, it returns an 2612 RA as specified in Section 3.15.3 and sends up to 2613 MAX_NEIGHBOR_ADVERTIISEMENT uNA messages to any old Servers listed in 2614 OMNI option MS-Release identifiers. Each uNA message includes the 2615 Client's AERO address as the source address, the old Server's AERO 2616 address as the destination address, and an OMNI option with the 2617 Register/Release bit set to 0. The new Server wraps the uNA in a 2618 SPAN header with its own SPAN address as the source and the old 2619 Server's SPAN address as the destination, then sends the message into 2620 the SPAN. 2622 When an old Server receives the uNA, it changes the Client's neighbor 2623 cache entry state to DEPARTED, sets the link-layer address of the 2624 Client to the new Server's SPAN address, and sets DepartTime to 2625 DEPART_TIME seconds. After a short delay (e.g., 2 seconds) the old 2626 Server withdraws the Client's MNP from the routing system. After 2627 DepartTime expires, the old Server deletes the Client's neighbor 2628 cache entry. 2630 The old Server also sends unsolicited NA messages to all ROSs in the 2631 Client's Report List with an OMNI option with a single ifIndex-tuple 2632 with ifIndex set to 0 and S set to '1', and with the SPAN address of 2633 the new Server in a companion TLLAO. When the ROS receives the NA, 2634 it caches the address of the new Server in the existing asymmetric 2635 neighbor cache entry and marks the entry as STALE. Subsequent data 2636 packets will then flow according to any existing cached link-layer 2637 information and trigger a new NS(AR)/NA exchange via the new Server. 2639 Clients SHOULD NOT move rapidly between Servers in order to avoid 2640 causing excessive oscillations in the AERO routing system. Examples 2641 of when a Client might wish to change to a different Server include a 2642 Server that has gone unreachable, topological movements of 2643 significant distance, movement to a new geographic region, movement 2644 to a new SPAN segment, etc. 2646 When a Client moves to a new Server, some of the fragments of a 2647 multiple fragment packet may have already arrived at the old Server 2648 while others are en route to the new Server, however no special 2649 attention in the reassembly algorithm is necessary when re-routed 2650 fragments are simply treated as loss. 2652 3.20. Multicast 2654 The AERO Client provides an IGMP (IPv4) [RFC2236] or MLD (IPv6) 2655 [RFC3810] proxy service for its EUNs and/or hosted applications 2656 [RFC4605]. The Client forwards IGMP/MLD messages over any of its 2657 underlying interfaces for which group membership is required. The 2658 IGMP/MLD messages may be further forwarded by a first-hop ANET access 2659 router acting as an IGMP/MLD-snooping switch [RFC4541], then 2660 ultimately delivered to an AERO Proxy/Server acting as a Protocol 2661 Independent Multicast - Sparse-Mode (PIM-SM, or simply "PIM") 2662 Designated Router (DR) [RFC7761]. AERO Relays also act as PIM 2663 routers (i.e., the same as AERO Proxys/Servers) on behalf of nodes on 2664 INET/EUN networks. The behaviors identified in the following 2665 sections correspond to Source-Specific Multicast (SSM) and Any-Source 2666 Multicast (ASM) operational modes. 2668 3.20.1. Source-Specific Multicast (SSM) 2670 When an ROS (i.e., an AERO Proxy/Server/Relay) "X" acting as PIM 2671 router receives a Join/Prune message from a node on its downstream 2672 interfaces containing one or more ((S)ource, (G)roup) pairs, it 2673 updates its Multicast Routing Information Base (MRIB) accordingly. 2674 For each S belonging to a prefix reachable via X's non-AERO 2675 interfaces, X then forwards the (S, G) Join/Prune to any PIM routers 2676 on those interfaces per [RFC7761]. 2678 For each S belonging to a prefix reachable via X's AERO interface, X 2679 originates a separate copy of the Join/Prune for each (S,G) in the 2680 message using its own AERO address as the source address and ALL-PIM- 2681 ROUTERS as the destination address. X then encapsulates each message 2682 in a SPAN header with source address set to the SPAN address of X and 2683 destination address set to S then forwards the message into the SPAN. 2684 The SPAN in turn forwards the message to AERO Server/Relay "Y" that 2685 services S. At the same time, if the message was a Join, X sends a 2686 route-optimization NS message toward each S the same as discussed in 2687 Section 3.17. The resulting NAs will return the AERO address for the 2688 prefix that matches S as the network-layer source address and TLLAOs 2689 with the SPAN addresses corresponding to any ifIndex-tuples that are 2690 currently servicing S. 2692 When Y processes the Join/Prune message, if S located behind any 2693 INET, Direct, or VPNed interfaces Y acts as a PIM router and updates 2694 its MRIB to list X as the next hop in the reverse path. If S is 2695 located behind any Proxys "Z"*, Y also forwards the message to each 2696 Z* over the SPAN while continuing to use the AERO address of X as the 2697 source address. Each Z* then updates its MRIB accordingly and 2698 maintains the AERO address of X as the next hop in the reverse path. 2699 Since the Bridges in the SPAN do not examine network layer control 2700 messages, this means that the (reverse) multicast tree path is simply 2701 from each Z* (and/or Y) to X with no other multicast-aware routers in 2702 the path. If any Z* (and/or Y) is located on the same SPAN segment 2703 as X, the multicast data traffic sent to X directly using SPAN/INET 2704 encapsulation instead of via a Bridge. 2706 Following the initial Join/Prune and NS/NA messaging, X maintains an 2707 asymmetric neighbor cache entry for each S the same as if X was 2708 sending unicast data traffic to S. In particular, X performs 2709 additional NS/NA exchanges to keep the neighbor cache entry alive for 2710 up to t_periodic seconds [RFC7761]. If no new Joins are received 2711 within t_periodic seconds, X allows the neighbor cache entry to 2712 expire. Finally, if X receives any additional Join/Prune messages 2713 for (S,G) it forwards the messages to each Y and Z* in the neighbor 2714 cache entry over the SPAN. 2716 At some later time, Client C that holds an MNP for source S may 2717 depart from a first Proxy Z1 and/or connect via a new Proxy Z2. In 2718 that case, Y sends an unsolicited NA message to X the same as 2719 specified for unicast mobility in Section 3.19. When X receives the 2720 unsolicited NA message, it updates its asymmetric neighbor cache 2721 entry for the AERO address for source S and sends new Join messages 2722 to any new Proxys Z2. There is no requirement to send any Prune 2723 messages to old Proxys Z1 since source S will no longer source any 2724 multicast data traffic via Z1. Instead, the multicast state for 2725 (S,G) in Proxy Z1 will soon time out since no new Joins will arrive. 2727 After some later time, C may move to a new Server Y2 and depart from 2728 old Sever Y1. In that case, Y1 sends Join messages for any of C's 2729 active (S,G) groups to Y2 while including its own AERO address as the 2730 source address. This causes Y2 to include Y1 in the multicast 2731 forwarding tree during the interim time that Y1's symmetric neighbor 2732 cache entry for C is in the DEPARTED state. At the same time, Y1 2733 sends an unsolicited NA message to X with an OMNI option and TLLAO 2734 with ifIndex-tuple set to 0 and a release indication to cause X to 2735 release its asymmetric neighbor cache entry. X then sends a new Join 2736 message to S via the SPAN and re-initiates route optimization the 2737 same as if it were receiving a fresh Join message from a node on a 2738 downstream link. 2740 3.20.2. Any-Source Multicast (ASM) 2742 When an ROS X acting as a PIM router receives a Join/Prune from a 2743 node on its downstream interfaces containing one or more (*,G) pairs, 2744 it updates its Multicast Routing Information Base (MRIB) accordingly. 2745 X then forwards a copy of the message to the Rendezvous Point (RP) R 2746 for each G over the SPAN. X uses its own AERO address as the source 2747 address and ALL-PIM-ROUTERS as the destination address, then 2748 encapsulates each message in a SPAN header with source address set to 2749 the SPAN address of X and destination address set to R, then sends 2750 the message into the SPAN. At the same time, if the message was a 2751 Join X initiates NS/NA route optimization the same as for the SSM 2752 case discussed in Section 3.20.1. 2754 For each source S that sends multicast traffic to group G via R, the 2755 Proxy/Server Z* for the Client that aggregates S encapsulates the 2756 packets in PIM Register messages and forwards them to R via the SPAN. 2757 R may then elect to send a PIM Join to Z* over the SPAN. This will 2758 result in an (S,G) tree rooted at Z* with R as the next hop so that R 2759 will begin to receive two copies of the packet; one native copy from 2760 the (S, G) tree and a second copy from the pre-existing (*, G) tree 2761 that still uses PIM Register encapsulation. R can then issue a PIM 2762 Register-stop message to suppress the Register-encapsulated stream. 2763 At some later time, if C moves to a new Proxy/Server Z*, it resumes 2764 sending packets via PIM Register encapsulation via the new Z*. 2766 At the same time, as multicast listeners discover individual S's for 2767 a given G, they can initiate an (S,G) Join for each S under the same 2768 procedures discussed in Section 3.20.1. Once the (S,G) tree is 2769 established, the listeners can send (S, G) Prune messages to R so 2770 that multicast packets for group G sourced by S will only be 2771 delivered via the (S, G) tree and not from the (*, G) tree rooted at 2772 R. All mobility considerations discussed for SSM apply. 2774 3.20.3. Bi-Directional PIM (BIDIR-PIM) 2776 Bi-Directional PIM (BIDIR-PIM) [RFC5015] provides an alternate 2777 approach to ASM that treats the Rendezvous Point (RP) as a Designated 2778 Forwarder (DF). Further considerations for BIDIR-PIM are out of 2779 scope. 2781 3.21. Operation over Multiple AERO Links (VLANs) 2783 An AERO Client can connect to multiple AERO links the same as for any 2784 data link service. In that case, the Client maintains a distinct 2785 AERO interface for each link, e.g., 'aero0' for the first link, 2786 'aero1' for the second, 'aero2' for the third, etc. Each AERO link 2787 would include its own distinct set of Bridges, Servers and Proxys, 2788 thereby providing redundancy in case of failures. 2790 The Bridges, Servers and Proxys on each AERO link can assign AERO and 2791 SPAN addresses that use the same or different numberings from those 2792 on other links. Since the links are mutually independent there is no 2793 requirement for avoiding inter-link address duplication, e.g., the 2794 same AERO address such as fe80::1000 could be used to number distinct 2795 nodes that connect to different AERO links. 2797 Each AERO link could utilize the same or different ANET connections. 2798 The links can be distinguished at the link-layer via the SSP in a 2799 similar fashion as for Virtual Local Area Network (VLAN) tagging 2800 (e.g., IEEE 802.1Q) and/or through assignment of distinct sets of 2801 MSPs on each link. This gives rise to the opportunity for supporting 2802 multiple redundant networked paths, where each VLAN is distinguished 2803 by a different SRT color (see: Section 3.5.1). In particular, the 2804 Client can tag its RS messages with the appropriate label to cause 2805 the network to select the desired VLAN. 2807 The Client's IP layer can select the outgoing AERO interface 2808 appropriate for a given traffic profile while (in the reverse 2809 direction) correspondent nodes must have some way of steering their 2810 packets destined to a target via the correct AERO link. 2812 In a first alternative, if each AERO link services different MSPs, 2813 then the Client can receive a distinct MNP from each of the links. 2814 IP routing will therefore assure that the correct Red/Green/Blue/etc. 2815 network is used for both outbound and inbound traffic. This can be 2816 accomplished using existing technologies and approaches, and without 2817 requiring any special supporting code in correspondent nodes or 2818 Bridges. 2820 In a second alternative, if each AERO link services the same MSP(s) 2821 then each link could assign a distinct "AERO Link Anycast" address 2822 that is configured by all Bridges on the link. Correspondent nodes 2823 can then perform segment routing at the SPAN layer 2824 [RFC8402][RFC8754]. Segment Routing in the correct SRT will then 2825 direct the packet over multiple hops to the target. 2827 3.22. DNS Considerations 2829 AERO Client MNs and INET correspondent nodes consult the Domain Name 2830 System (DNS) the same as for any Internetworking node. When 2831 correspondent nodes and Client MNs use different IP protocol versions 2832 (e.g., IPv4 correspondents and IPv6 MNs), the INET DNS must maintain 2833 A records for IPv4 address mappings to MNs which must then be 2834 populated in Relay NAT64 mapping caches. In that way, an IPv4 2835 correspondent node can send packets to the IPv4 address mapping of 2836 the target MN, and the Relay will translate the IPv4 header and 2837 destination address into an IPv6 header and IPv6 destination address 2838 of the MN. 2840 When an AERO Client registers with an AERO Server, the Server can 2841 return the address(es) of DNS servers in RDNSS options [RFC6106]. 2842 The DNS server provides the IP addresses of other MNs and 2843 correspondent nodes in AAAA records for IPv6 or A records for IPv4. 2845 3.23. Transition Considerations 2847 The SPAN ensures that dissimilar INET partitions can be joined into a 2848 single unified AERO link, even though the partitions themselves may 2849 have differing protocol versions and/or incompatible addressing 2850 plans. However, a commonality can be achieved by incrementally 2851 distributing globally routable (i.e., native) IP prefixes to 2852 eventually reach all nodes (both mobile and fixed) in all SPAN 2853 segments. This can be accomplished by incrementally deploying AERO 2854 Relays on each INET partition, with each Relay distributing its MNPs 2855 and/or discovering non-MNP prefixes on its INET links. 2857 This gives rise to the opportunity to eventually distribute native IP 2858 addresses to all nodes, and to present a unified AERO link view 2859 (bridged by the SPAN) even if the INET partitions remain in their 2860 current protocol and addressing plans. In that way, the AERO link 2861 can serve the dual purpose of providing a mobility/multilink service 2862 and a transition service. Or, if an INET partition is transitioned 2863 to a native IP protocol version and addressing scheme that is 2864 compatible with the AERO link MNP-based addressing scheme, the 2865 partition and AERO link can be joined by Relays. 2867 Relays that connect INETs/EUNs with dissimilar IP protocol versions 2868 may need to employ a network address and protocol translation 2869 function such as NAT64[RFC6146]. 2871 3.24. Detecting and Reacting to Server and Bridge Failures 2873 In environments where rapid failure recovery is required, Servers and 2874 Bridges SHOULD use Bidirectional Forwarding Detection (BFD) 2875 [RFC5880]. Nodes that use BFD can quickly detect and react to 2876 failures so that cached information is re-established through 2877 alternate nodes. BFD control messaging is carried only over well- 2878 connected ground domain networks (i.e., and not low-end radio links) 2879 and can therefore be tuned for rapid response. 2881 Servers and Bridges maintain BFD sessions in parallel with their BGP 2882 peerings. If a Server or Bridge fails, BGP peers will quickly re- 2883 establish routes through alternate paths the same as for common BGP 2884 deployments. Similarly, Proxys maintain BFD sessions with their 2885 associated Bridges even though they do not establish BGP peerings 2886 with them. 2888 Proxys SHOULD use proactive NUD for Servers for which there are 2889 currently active ANET Clients in a manner that parallels BFD, i.e., 2890 by sending unicast NS messages in rapid succession to receive 2891 solicited NA messages. When the Proxy is also sending RS messages on 2892 behalf of ANET Clients, the RS/RA messaging can be considered as 2893 equivalent hints of forward progress. This means that the Proxy need 2894 not also send a periodic NS if it has already sent an RS within the 2895 same period. If a Server fails, the Proxy will cease to receive 2896 advertisements and can quickly inform Clients of the outage by 2897 sending multicast RA messages on the ANET interface. 2899 The Proxy sends multicast RA messages with source address set to the 2900 Server's address, destination address set to All-Nodes multicast, and 2901 Router Lifetime set to 0. The Proxy SHOULD send 2902 MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small delays 2903 [RFC4861]. Any Clients on the ANET interface that have been using 2904 the (now defunct) Server will receive the RA messages and associate 2905 with a new Server. 2907 3.25. AERO Clients on the Open Internet 2909 AERO Clients that connect to the open Internet via INET interfaces 2910 can establish a VPN or direct link to securely connect to a Server in 2911 a "tethered" arrangement with all of the Client's traffic transiting 2912 the Server. Alternatively, the Client can associate with an INET 2913 Server using UDP/IP encapsulation and asymmetric securing services as 2914 discussed in the following sections. 2916 When a Client's AERO interface enables INET underlying interfaces, it 2917 sends a UDP/IP-encapsulated RS message with IPv6 source address set 2918 to its AERO address, with IPv6 destination set to All-Routers 2919 multicast, with an OMNI option and with a Teredo Authentication 2920 option to provide message authentication. The Client also includes 2921 an SLLAO with Link Layer Address set according to the address of the 2922 underlying interface used for INET encapsulation. If the underlying 2923 interface address is IPv6, the Client sets the FMT according to 2924 whether the Port Number also must be included as discussed in 2925 Section 3.6. If the underlying address is IPv4, the Client includes 2926 a Teredo address [RFC4380] using the prefix fe80::/32 with the 2927 Server's IPv4 address, and with the IP address and Port Number used 2928 for INET encapsulation written in obfuscated form and sets FMT to 2929 NAT. The Client then sets the UDP/IP source to its INET address and 2930 UDP port, and sets the destination to the Server's INET address and 2931 the AERO service port number (8060), then sends the message to the 2932 Server. 2934 When the Server receives the RS message, it authenticates the message 2935 and registers the Client's MNP and INET interface information 2936 according to the OMNI option parameters. The Server then returns an 2937 RA message with IPv6 source and destination set corresponding to the 2938 addresses in the RS, and with a Teredo Authentication option. For 2939 IPv4 INET interfaces, the Server also includes a Teredo Origin option 2940 with the mapped and obfuscated Client observed IP address and port 2941 number for the Client, and with a Teredo Authentication option. The 2942 Server then sends the message to the Client and records the Client's 2943 IPv6 address (for IPv6 INET interfaces) or fe80:: Teredo address (for 2944 IPv4 INET interfaces) as the link layer address in the neighbor 2945 cache. For IPv4, if the Client's IPv4 address and port from the 2946 SLLAO match the UDP/IPv4 header information (and if the IPv4 address 2947 is global unicast) the Server notes the FMT for this Client as Teredo 2948 on the open INET. 2950 When the Client receives the RA message, for IPv6 INET interfaces the 2951 Client proceeds under the assumption that any NATs on the path to the 2952 Server will behave according to the Cone NAT principle. For IPv4 2953 INET interfaces, the Client instead compares the mapped IP address 2954 and port from the Teredo Origin option with its own address. If the 2955 addresses are the same, the Client assumes the Cone NAT principle; if 2956 the addresses are different, the Client instead assumes that further 2957 Server qualification procedures are necessary to detect the type of 2958 NAT and proceeds according to standard Teredo procedures. 2960 After the Client has registered its INET interfaces in such RS/RA 2961 exchanges it sends periodic RS messages to receive fresh RA messages 2962 before the Router Lifetime received on each INET interface expires. 2963 The Client also maintains default routes via its Servers, i.e., the 2964 same as described in earlier sections. 2966 When the Client sends messages to target IP addresses, it also 2967 invokes route optimization per Section 3.17 using IPv6 ND address 2968 resolution messaging. The Client sends the NS(AR) message to the 2969 Server wrapped in a UDP/IP header with a Teredo Authentication option 2970 with the NS source address set to the Client's AERO address and 2971 destination address set to the target solicited node multicast 2972 address. The Server authenticates the message and sends a 2973 corresponding NS(AR) message over the SPAN the same as if it were the 2974 ROS, but with the SPAN source address set to the Server's SPAN 2975 address and destination set to the SPAN address of the target. When 2976 the ROR receives the NS(AR), it adds the Server's SPAN address and 2977 Client's AERO address to the target's Report List, and returns an NA 2978 with OMNI and TLLAO information for the target. The Server then 2979 returns a UDP/IP encapsulated NA message with a Teredo Authentication 2980 option to the Client. 2982 Following route optimization, for targets in the same SPAN segment if 2983 the target's Link Layer Address is native IPv6 or a Teredo address on 2984 the open INET, the Client forwards data packets directly to the 2985 target INET address. If the Link Layer Address is a Teredo address 2986 for a peer behind a NAT, the Client first establishes NAT state for 2987 the Link Layer Address using the "bubble" mechanisms specified in 2988 [RFC6081][RFC4380]. The Client continues to send data packets via 2989 its Server until NAT state is populated, then begins forwarding 2990 packets via the direct path through the NAT to the target. For 2991 targets in different SPAN segments, the Client inserts a Segment 2992 Routing header and forwards data packets to the Bridge that returned 2993 the NA message. 2995 The ROR may return uNAs via the Server if the target moves, and the 2996 Server will send corresponding Teredo Authentication-protected uNAs 2997 to the Client. The Client can also send "loopback" NS(NUD) messages 2998 to test forward path reachability even though there is no security 2999 association between the Client and the target. 3001 The Client sends Teredo UDP/IP encapsulated IPv6 packets no larger 3002 than 1280 bytes in one piece. In order to accommodate larger IPv6 3003 packets (up to the AERO interface 9180 MTU), the Client inserts a 3004 SPAN header with source set to its own SPAN address and destination 3005 set to the SPAN address of the target and uses IPv6 fragmentation 3006 according to Section 3.12. The Client then encapsulates each 3007 fragment in a UDP/IP header and sends the fragments to the next hop. 3009 3.25.1. Use of SEND and CGA 3011 In some environments, use of the Teredo Authentication option alone 3012 may be sufficient for assuring IPv6 ND message authentication between 3013 Clients and Servers. When additional protection is necessary, nodes 3014 should employ SEcure Neighbor Discovery (SEND) [RFC3971] with 3015 Cryptographically-Generated Addresses (CGA) [RFC3972]. 3017 When SEND/CGA are used, the Client prepares RS messages with its 3018 link-local CGA as the IPv6 source and All-Routers as the IPv6 3019 Destination, includes any SEND options and wraps the message in a 3020 SPAN header. The Client sets the SPAN source address to its own SPAN 3021 address and sets the SPAN destination address to the "All-Routers" 3022 SPAN address. The Client then wraps the RS message in UDP/IP headers 3023 according to the Teredo format and sends the message to the Server. 3025 When the Server receives the message, it first verifies the Teredo 3026 Authentication option (if present) then uses the SPAN source address 3027 to determine the MNP of the Client. The Server then processes the 3028 SEND options to authenticate the RS message and prepares an RA 3029 message response. The Server prepares the RA with its own link-local 3030 CGA and the CGA of the Client as the IPv6 source and destination, 3031 includes any SEND options and wraps the message in a SPAN header. 3032 The Server sets the SPAN source address to its own SPAN address and 3033 sets the SPAN destination address to the Client's SPAN address. The 3034 Server then wraps the RA message in UDP/IP headers according to the 3035 Teredo format and sends the message to the Client. Thereafter, the 3036 Client/Server send additional RS/RA messages to maintain their 3037 association and any NAT state. 3039 The Client and Server also may exchange NS/NA messages using their 3040 own CGA as the source and with SPAN encapsulation as above. When a 3041 Client sends an NS(AR), it sets the IPv6 source to its CGA and sets 3042 the IPv6 destination to the Solicited-Node Multicast address of the 3043 target. The Client then wraps the message in a SPAN header with its 3044 own SPAN address as the source and the SPAN address of the target as 3045 the destination and sends it to the Server. The Server authenticates 3046 the message, then changes the IPv6 source address to the Client's 3047 AERO address, removes the SEND options, and sends a corresponding 3048 NS(AR) into the SPAN. When the Server receives the corresponding 3049 SPAN-encapsulated NA, it changes the IPv6 destination address to the 3050 Client's CGA, inserts SEND options, then wraps the message in UDP/IP 3051 headers and sends it to the Client. 3053 When a Client sends a uNA, it sets the IPv6 source address to its own 3054 CGA and sets the IPv6 destination address to All-Nodes multicast, 3055 includes SEND options, wraps the message in SPAN and UDP/IP headers 3056 and sends the message to the Server. The Server authenticates the 3057 message, then changes the IPv6 address to the Client's AERO address, 3058 removes the SEND options and forwards the message the same as 3059 discussed in Section 3.19.1. In the reverse direction, when the 3060 Server forwards a uNA to the Client, it changes the IPv6 address to 3061 its own CGA and inserts SEND options then forwards the message to the 3062 Client. 3064 When a Client sends an NS(NUD), it sets both the IPv6 source and 3065 destination address to its own AERO address, wraps the message in a 3066 SPAN header and UDP/IP headers, then sends the message directly to 3067 the peer which will loop the message back. In this case alone, the 3068 Client does not use the Server as a trust broker for forwarding the 3069 ND message. 3071 3.26. Time-Varying MNPs 3073 In some use cases, it is desirable, beneficial and efficient for the 3074 Client to receive a constant MNP that travels with the Client 3075 wherever it moves. For example, this would allow air traffic 3076 controllers to easily track aircraft, etc. In other cases, however 3077 (e.g., intelligent transportation systems), the MN may be willing to 3078 sacrifice a modicum of efficiency in order to have time-varying MNPs 3079 that can be changed every so often to defeat adversarial tracking. 3081 The DHCPv6-PD service offers a way for Clients that desire time- 3082 varying MNPs to obtain short-lived prefixes (e.g., on the order of a 3083 small number of minutes). In that case, the identity of the Client 3084 would not be bound to the MNP but rather the Client's identity would 3085 be bound to the DHCPv6 Device Unique Identifier (DUID) and used as 3086 the seed for Prefix Delegation. The Client would then be obligated 3087 to renumber its internal networks whenever its MNP (and therefore 3088 also its AERO address) changes. This should not present a challenge 3089 for Clients with automated network renumbering services, however 3090 presents limits for the durations of ongoing sessions that would 3091 prefer to use a constant address. 3093 4. Implementation Status 3095 An AERO implementation based on OpenVPN (https://openvpn.net/) was 3096 announced on the v6ops mailing list on January 10, 2018 and an 3097 initial public release of the AERO proof-of-concept source code was 3098 announced on the intarea mailing list on August 21, 2015. 3100 As of 4/1/2020, more recent updated implementations are under 3101 internal development and testing with plans to release in the near 3102 future. 3104 5. IANA Considerations 3106 The IANA has assigned a 4-octet Private Enterprise Number "45282" for 3107 AERO in the "enterprise-numbers" registry. 3109 The IANA has assigned the UDP port number "8060" for an earlier 3110 experimental version of AERO [RFC6706]. This document obsoletes 3111 [RFC6706] and claims the UDP port number "8060" for all future use. 3113 No further IANA actions are required. 3115 6. Security Considerations 3117 AERO Bridges configure secured tunnels with AERO Servers and Proxys 3118 within their local SPAN segments. Applicable secured tunnel 3119 alternatives include IPsec [RFC4301], TLS/SSL [RFC8446], DTLS 3120 [RFC6347], WireGuard, etc. The AERO Bridges of all SPAN segments in 3121 turn configure secured tunnels for their neighboring AERO Bridges 3122 across the SPAN. Therefore, control messages that traverse the SPAN 3123 between any pair of AERO link neighbors are already secured. 3125 AERO Servers, Relays and Proxys targeted by a route optimization may 3126 also receive packets directly from the INET partitions instead of via 3127 the SPAN. For INET partitions that apply effective ingress filtering 3128 to defeat source address spoofing, the simple data origin 3129 authentication procedures in Section 3.11 can be applied. 3131 For INET partitions that cannot apply effective ingress filtering, 3132 the two options for securing communications include 1) disable route 3133 optimization so that all traffic is conveyed over secured tunnels via 3134 the SPAN, or 2) enable on-demand secure tunnel creation between INET 3135 partition neighbors. Option 1) would result in longer routes than 3136 necessary and traffic concentration on critical infrastructure 3137 elements. Option 2) could be coordinated by establishing a secured 3138 tunnel on-demand instead of performing an NS/NA exchange in the route 3139 optimization procedures. Procedures for establishing on-demand 3140 secured tunnels are out of scope. 3142 AERO Clients that connect to secured ANETs need not apply security to 3143 their ND messages, since the messages will be intercepted by a 3144 perimeter Proxy that applies security on its INET-facing interface. 3145 AERO Clients connected to the open INET can use symmetric network 3146 and/or transport layer security services such as VPNs or can by some 3147 other means establish a direct link. When a VPN or direct link may 3148 be impractical, however, an asymmetric security service such as 3149 SEcure Neighbor Discovery (SEND) [RFC3971] with Cryptographically 3150 Generated Addresses (CGAs) [RFC3972] and/or the Teredo Authentication 3151 option [RFC4380] may be necessary. 3153 Application endpoints SHOULD use application-layer security services 3154 such as TLS/SSL, DTLS or SSH [RFC4251] to assure the same level of 3155 protection as for critical secured Internet services. AERO Clients 3156 that require host-based VPN services SHOULD use symmetric network 3157 and/or transport layer security services such as IPsec, TLS/SSL, 3158 DTLS, etc. AERO Proxys and Servers can also provide a network-based 3159 VPN service on behalf of the Client, e.g., if the Client is located 3160 within a secured enclave and cannot establish a VPN on its own 3161 behalf. 3163 AERO Servers and Bridges present targets for traffic amplification 3164 Denial of Service (DoS) attacks. This concern is no different than 3165 for widely-deployed VPN security gateways in the Internet, where 3166 attackers could send spoofed packets to the gateways at high data 3167 rates. This can be mitigated by connecting Servers and Bridges over 3168 dedicated links with no connections to the Internet and/or when 3169 connections to the Internet are only permitted through well-managed 3170 firewalls. Traffic amplification DoS attacks can also target an AERO 3171 Client's low data rate links. This is a concern not only for Clients 3172 located on the open Internet but also for Clients in secured 3173 enclaves. AERO Servers and Proxys can institute rate limits that 3174 protect Clients from receiving packet floods that could DoS low data 3175 rate links. 3177 AERO Relays must implement ingress filtering to avoid a spoofing 3178 attack in which spurious SPAN messages are injected into an AERO link 3179 from an outside attacker. AERO Clients MUST ensure that their 3180 connectivity is not used by unauthorized nodes on their EUNs to gain 3181 access to a protected network, i.e., AERO Clients that act as routers 3182 MUST NOT provide routing services for unauthorized nodes. (This 3183 concern is no different than for ordinary hosts that receive an IP 3184 address delegation but then "share" the address with other nodes via 3185 some form of Internet connection sharing such as tethering.) 3187 The MAP list MUST be well-managed and secured from unauthorized 3188 tampering, even though the list contains only public information. 3189 The MAP list can be conveyed to the Client in a similar fashion as in 3190 [RFC5214] (e.g., through layer 2 data link login messaging, secure 3191 upload of a static file, DNS lookups, etc.). 3193 Although public domain and commercial SEND implementations exist, 3194 concerns regarding the strength of the cryptographic hash algorithm 3195 have been documented [RFC6273] [RFC4982]. 3197 Segment routing provides authentication facilities that can be used 3198 to authenticate the information in the SRH [RFC8754]. 3200 Security considerations for accepting link-layer ICMP messages and 3201 reflected packets are discussed throughout the document. 3203 7. Acknowledgements 3205 Discussions in the IETF, aviation standards communities and private 3206 exchanges helped shape some of the concepts in this work. 3207 Individuals who contributed insights include Mikael Abrahamsson, Mark 3208 Andrews, Fred Baker, Bob Braden, Stewart Bryant, Brian Carpenter, 3209 Wojciech Dec, Pavel Drasil, Ralph Droms, Adrian Farrel, Nick Green, 3210 Sri Gundavelli, Brian Haberman, Bernhard Haindl, Joel Halpern, Tom 3211 Herbert, Sascha Hlusiak, Lee Howard, Zdenek Jaron, Andre Kostur, 3212 Hubert Kuenig, Ted Lemon, Andy Malis, Satoru Matsushima, Tomek 3213 Mrugalski, Madhu Niraula, Alexandru Petrescu, Behcet Saikaya, Michal 3214 Skorepa, Joe Touch, Bernie Volz, Ryuji Wakikawa, Tony Whyman, Lloyd 3215 Wood and James Woodyatt. Members of the IESG also provided valuable 3216 input during their review process that greatly improved the document. 3217 Special thanks go to Stewart Bryant, Joel Halpern and Brian Haberman 3218 for their shepherding guidance during the publication of the AERO 3219 first edition. 3221 This work has further been encouraged and supported by Boeing 3222 colleagues including Kyle Bae, M. Wayne Benson, Dave Bernhardt, Cam 3223 Brodie, John Bush, Balaguruna Chidambaram, Irene Chin, Bruce Cornish, 3224 Claudiu Danilov, Don Dillenburg, Joe Dudkowski, Wen Fang, Samad 3225 Farooqui, Anthony Gregory, Jeff Holland, Seth Jahne, Brian Jaury, 3226 Greg Kimberly, Ed King, Madhuri Madhava Badgandi, Laurel Matthew, 3227 Gene MacLean III, Rob Muszkiewicz, Sean O'Sullivan, Vijay 3228 Rajagopalan, Greg Saccone, Rod Santiago, Kent Shuey, Brian Skeen, 3229 Mike Slane, Carrie Spiker, Katie Tran, Brendan Williams, Amelia 3230 Wilson, Julie Wulff, Yueli Yang, Eric Yeh and other members of the 3231 Boeing mobility, networking and autonomy teams. Kyle Bae, Wayne 3232 Benson, Katie Tran and Eric Yeh are especially acknowledged for 3233 implementing the AERO functions as extensions to the public domain 3234 OpenVPN distribution. 3236 Earlier works on NBMA tunneling approaches are found in 3237 [RFC2529][RFC5214][RFC5569]. 3239 Many of the constructs presented in this second edition of AERO are 3240 based on the author's earlier works, including: 3242 o The Internet Routing Overlay Network (IRON) 3243 [RFC6179][I-D.templin-ironbis] 3245 o Virtual Enterprise Traversal (VET) 3246 [RFC5558][I-D.templin-intarea-vet] 3248 o The Subnetwork Encapsulation and Adaptation Layer (SEAL) 3249 [RFC5320][I-D.templin-intarea-seal] 3251 o AERO, First Edition [RFC6706] 3253 Note that these works cite numerous earlier efforts that are not also 3254 cited here due to space limitations. The authors of those earlier 3255 works are acknowledged for their insights. 3257 This work is aligned with the NASA Safe Autonomous Systems Operation 3258 (SASO) program under NASA contract number NNA16BD84C. 3260 This work is aligned with the FAA as per the SE2025 contract number 3261 DTFAWA-15-D-00030. 3263 This work is aligned with the Boeing Commercial Airplanes (BCA) 3264 Internet of Things (IoT) and autonomy programs. 3266 This work is aligned with the Boeing Information Technology (BIT) 3267 MobileNet program. 3269 8. References 3270 8.1. Normative References 3272 [I-D.templin-6man-omni-interface] 3273 Templin, F. and T. Whyman, "Transmission of IPv6 Packets 3274 over Overlay Multilink Network (OMNI) Interfaces", draft- 3275 templin-6man-omni-interface-20 (work in progress), May 3276 2020. 3278 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 3279 DOI 10.17487/RFC0791, September 1981, 3280 . 3282 [RFC0792] Postel, J., "Internet Control Message Protocol", STD 5, 3283 RFC 792, DOI 10.17487/RFC0792, September 1981, 3284 . 3286 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 3287 Requirement Levels", BCP 14, RFC 2119, 3288 DOI 10.17487/RFC2119, March 1997, 3289 . 3291 [RFC2473] Conta, A. and S. Deering, "Generic Packet Tunneling in 3292 IPv6 Specification", RFC 2473, DOI 10.17487/RFC2473, 3293 December 1998, . 3295 [RFC2474] Nichols, K., Blake, S., Baker, F., and D. Black, 3296 "Definition of the Differentiated Services Field (DS 3297 Field) in the IPv4 and IPv6 Headers", RFC 2474, 3298 DOI 10.17487/RFC2474, December 1998, 3299 . 3301 [RFC3971] Arkko, J., Ed., Kempf, J., Zill, B., and P. Nikander, 3302 "SEcure Neighbor Discovery (SEND)", RFC 3971, 3303 DOI 10.17487/RFC3971, March 2005, 3304 . 3306 [RFC3972] Aura, T., "Cryptographically Generated Addresses (CGA)", 3307 RFC 3972, DOI 10.17487/RFC3972, March 2005, 3308 . 3310 [RFC4191] Draves, R. and D. Thaler, "Default Router Preferences and 3311 More-Specific Routes", RFC 4191, DOI 10.17487/RFC4191, 3312 November 2005, . 3314 [RFC4193] Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast 3315 Addresses", RFC 4193, DOI 10.17487/RFC4193, October 2005, 3316 . 3318 [RFC4380] Huitema, C., "Teredo: Tunneling IPv6 over UDP through 3319 Network Address Translations (NATs)", RFC 4380, 3320 DOI 10.17487/RFC4380, February 2006, 3321 . 3323 [RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman, 3324 "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861, 3325 DOI 10.17487/RFC4861, September 2007, 3326 . 3328 [RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless 3329 Address Autoconfiguration", RFC 4862, 3330 DOI 10.17487/RFC4862, September 2007, 3331 . 3333 [RFC5175] Haberman, B., Ed. and R. Hinden, "IPv6 Router 3334 Advertisement Flags Option", RFC 5175, 3335 DOI 10.17487/RFC5175, March 2008, 3336 . 3338 [RFC6081] Thaler, D., "Teredo Extensions", RFC 6081, 3339 DOI 10.17487/RFC6081, January 2011, 3340 . 3342 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 3343 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 3344 May 2017, . 3346 [RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6 3347 (IPv6) Specification", STD 86, RFC 8200, 3348 DOI 10.17487/RFC8200, July 2017, 3349 . 3351 [RFC8415] Mrugalski, T., Siodelski, M., Volz, B., Yourtchenko, A., 3352 Richardson, M., Jiang, S., Lemon, T., and T. Winters, 3353 "Dynamic Host Configuration Protocol for IPv6 (DHCPv6)", 3354 RFC 8415, DOI 10.17487/RFC8415, November 2018, 3355 . 3357 8.2. Informative References 3359 [BGP] Huston, G., "BGP in 2015, http://potaroo.net", January 3360 2016. 3362 [I-D.ietf-6man-segment-routing-header] 3363 Filsfils, C., Dukes, D., Previdi, S., Leddy, J., 3364 Matsushima, S., and D. Voyer, "IPv6 Segment Routing Header 3365 (SRH)", draft-ietf-6man-segment-routing-header-26 (work in 3366 progress), October 2019. 3368 [I-D.ietf-dmm-distributed-mobility-anchoring] 3369 Chan, A., Wei, X., Lee, J., Jeon, S., and C. Bernardos, 3370 "Distributed Mobility Anchoring", draft-ietf-dmm- 3371 distributed-mobility-anchoring-15 (work in progress), 3372 March 2020. 3374 [I-D.ietf-intarea-gue] 3375 Herbert, T., Yong, L., and O. Zia, "Generic UDP 3376 Encapsulation", draft-ietf-intarea-gue-09 (work in 3377 progress), October 2019. 3379 [I-D.ietf-intarea-gue-extensions] 3380 Herbert, T., Yong, L., and F. Templin, "Extensions for 3381 Generic UDP Encapsulation", draft-ietf-intarea-gue- 3382 extensions-06 (work in progress), March 2019. 3384 [I-D.ietf-intarea-tunnels] 3385 Touch, J. and M. Townsley, "IP Tunnels in the Internet 3386 Architecture", draft-ietf-intarea-tunnels-10 (work in 3387 progress), September 2019. 3389 [I-D.ietf-rtgwg-atn-bgp] 3390 Templin, F., Saccone, G., Dawra, G., Lindem, A., and V. 3391 Moreno, "A Simple BGP-based Mobile Routing System for the 3392 Aeronautical Telecommunications Network", draft-ietf- 3393 rtgwg-atn-bgp-05 (work in progress), January 2020. 3395 [I-D.templin-6man-dhcpv6-ndopt] 3396 Templin, F., "A Unified Stateful/Stateless Configuration 3397 Service for IPv6", draft-templin-6man-dhcpv6-ndopt-09 3398 (work in progress), January 2020. 3400 [I-D.templin-intarea-grefrag] 3401 Templin, F., "GRE Tunnel Level Fragmentation", draft- 3402 templin-intarea-grefrag-04 (work in progress), July 2016. 3404 [I-D.templin-intarea-seal] 3405 Templin, F., "The Subnetwork Encapsulation and Adaptation 3406 Layer (SEAL)", draft-templin-intarea-seal-68 (work in 3407 progress), January 2014. 3409 [I-D.templin-intarea-vet] 3410 Templin, F., "Virtual Enterprise Traversal (VET)", draft- 3411 templin-intarea-vet-40 (work in progress), May 2013. 3413 [I-D.templin-ironbis] 3414 Templin, F., "The Interior Routing Overlay Network 3415 (IRON)", draft-templin-ironbis-16 (work in progress), 3416 March 2014. 3418 [I-D.templin-v6ops-pdhost] 3419 Templin, F., "IPv6 Prefix Delegation and Multi-Addressing 3420 Models", draft-templin-v6ops-pdhost-25 (work in progress), 3421 January 2020. 3423 [OVPN] OpenVPN, O., "http://openvpn.net", October 2016. 3425 [RFC1035] Mockapetris, P., "Domain names - implementation and 3426 specification", STD 13, RFC 1035, DOI 10.17487/RFC1035, 3427 November 1987, . 3429 [RFC1122] Braden, R., Ed., "Requirements for Internet Hosts - 3430 Communication Layers", STD 3, RFC 1122, 3431 DOI 10.17487/RFC1122, October 1989, 3432 . 3434 [RFC1191] Mogul, J. and S. Deering, "Path MTU discovery", RFC 1191, 3435 DOI 10.17487/RFC1191, November 1990, 3436 . 3438 [RFC1812] Baker, F., Ed., "Requirements for IP Version 4 Routers", 3439 RFC 1812, DOI 10.17487/RFC1812, June 1995, 3440 . 3442 [RFC2003] Perkins, C., "IP Encapsulation within IP", RFC 2003, 3443 DOI 10.17487/RFC2003, October 1996, 3444 . 3446 [RFC2236] Fenner, W., "Internet Group Management Protocol, Version 3447 2", RFC 2236, DOI 10.17487/RFC2236, November 1997, 3448 . 3450 [RFC2492] Armitage, G., Schulter, P., and M. Jork, "IPv6 over ATM 3451 Networks", RFC 2492, DOI 10.17487/RFC2492, January 1999, 3452 . 3454 [RFC2529] Carpenter, B. and C. Jung, "Transmission of IPv6 over IPv4 3455 Domains without Explicit Tunnels", RFC 2529, 3456 DOI 10.17487/RFC2529, March 1999, 3457 . 3459 [RFC2764] Gleeson, B., Lin, A., Heinanen, J., Armitage, G., and A. 3460 Malis, "A Framework for IP Based Virtual Private 3461 Networks", RFC 2764, DOI 10.17487/RFC2764, February 2000, 3462 . 3464 [RFC2784] Farinacci, D., Li, T., Hanks, S., Meyer, D., and P. 3465 Traina, "Generic Routing Encapsulation (GRE)", RFC 2784, 3466 DOI 10.17487/RFC2784, March 2000, 3467 . 3469 [RFC2890] Dommety, G., "Key and Sequence Number Extensions to GRE", 3470 RFC 2890, DOI 10.17487/RFC2890, September 2000, 3471 . 3473 [RFC2923] Lahey, K., "TCP Problems with Path MTU Discovery", 3474 RFC 2923, DOI 10.17487/RFC2923, September 2000, 3475 . 3477 [RFC2983] Black, D., "Differentiated Services and Tunnels", 3478 RFC 2983, DOI 10.17487/RFC2983, October 2000, 3479 . 3481 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 3482 of Explicit Congestion Notification (ECN) to IP", 3483 RFC 3168, DOI 10.17487/RFC3168, September 2001, 3484 . 3486 [RFC3810] Vida, R., Ed. and L. Costa, Ed., "Multicast Listener 3487 Discovery Version 2 (MLDv2) for IPv6", RFC 3810, 3488 DOI 10.17487/RFC3810, June 2004, 3489 . 3491 [RFC3819] Karn, P., Ed., Bormann, C., Fairhurst, G., Grossman, D., 3492 Ludwig, R., Mahdavi, J., Montenegro, G., Touch, J., and L. 3493 Wood, "Advice for Internet Subnetwork Designers", BCP 89, 3494 RFC 3819, DOI 10.17487/RFC3819, July 2004, 3495 . 3497 [RFC4213] Nordmark, E. and R. Gilligan, "Basic Transition Mechanisms 3498 for IPv6 Hosts and Routers", RFC 4213, 3499 DOI 10.17487/RFC4213, October 2005, 3500 . 3502 [RFC4251] Ylonen, T. and C. Lonvick, Ed., "The Secure Shell (SSH) 3503 Protocol Architecture", RFC 4251, DOI 10.17487/RFC4251, 3504 January 2006, . 3506 [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A 3507 Border Gateway Protocol 4 (BGP-4)", RFC 4271, 3508 DOI 10.17487/RFC4271, January 2006, 3509 . 3511 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 3512 Architecture", RFC 4291, DOI 10.17487/RFC4291, February 3513 2006, . 3515 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 3516 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 3517 December 2005, . 3519 [RFC4389] Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery 3520 Proxies (ND Proxy)", RFC 4389, DOI 10.17487/RFC4389, April 3521 2006, . 3523 [RFC4443] Conta, A., Deering, S., and M. Gupta, Ed., "Internet 3524 Control Message Protocol (ICMPv6) for the Internet 3525 Protocol Version 6 (IPv6) Specification", STD 89, 3526 RFC 4443, DOI 10.17487/RFC4443, March 2006, 3527 . 3529 [RFC4511] Sermersheim, J., Ed., "Lightweight Directory Access 3530 Protocol (LDAP): The Protocol", RFC 4511, 3531 DOI 10.17487/RFC4511, June 2006, 3532 . 3534 [RFC4541] Christensen, M., Kimball, K., and F. Solensky, 3535 "Considerations for Internet Group Management Protocol 3536 (IGMP) and Multicast Listener Discovery (MLD) Snooping 3537 Switches", RFC 4541, DOI 10.17487/RFC4541, May 2006, 3538 . 3540 [RFC4605] Fenner, B., He, H., Haberman, B., and H. Sandick, 3541 "Internet Group Management Protocol (IGMP) / Multicast 3542 Listener Discovery (MLD)-Based Multicast Forwarding 3543 ("IGMP/MLD Proxying")", RFC 4605, DOI 10.17487/RFC4605, 3544 August 2006, . 3546 [RFC4607] Holbrook, H. and B. Cain, "Source-Specific Multicast for 3547 IP", RFC 4607, DOI 10.17487/RFC4607, August 2006, 3548 . 3550 [RFC4963] Heffner, J., Mathis, M., and B. Chandler, "IPv4 Reassembly 3551 Errors at High Data Rates", RFC 4963, 3552 DOI 10.17487/RFC4963, July 2007, 3553 . 3555 [RFC4982] Bagnulo, M. and J. Arkko, "Support for Multiple Hash 3556 Algorithms in Cryptographically Generated Addresses 3557 (CGAs)", RFC 4982, DOI 10.17487/RFC4982, July 2007, 3558 . 3560 [RFC5015] Handley, M., Kouvelas, I., Speakman, T., and L. Vicisano, 3561 "Bidirectional Protocol Independent Multicast (BIDIR- 3562 PIM)", RFC 5015, DOI 10.17487/RFC5015, October 2007, 3563 . 3565 [RFC5214] Templin, F., Gleeson, T., and D. Thaler, "Intra-Site 3566 Automatic Tunnel Addressing Protocol (ISATAP)", RFC 5214, 3567 DOI 10.17487/RFC5214, March 2008, 3568 . 3570 [RFC5320] Templin, F., Ed., "The Subnetwork Encapsulation and 3571 Adaptation Layer (SEAL)", RFC 5320, DOI 10.17487/RFC5320, 3572 February 2010, . 3574 [RFC5522] Eddy, W., Ivancic, W., and T. Davis, "Network Mobility 3575 Route Optimization Requirements for Operational Use in 3576 Aeronautics and Space Exploration Mobile Networks", 3577 RFC 5522, DOI 10.17487/RFC5522, October 2009, 3578 . 3580 [RFC5558] Templin, F., Ed., "Virtual Enterprise Traversal (VET)", 3581 RFC 5558, DOI 10.17487/RFC5558, February 2010, 3582 . 3584 [RFC5569] Despres, R., "IPv6 Rapid Deployment on IPv4 3585 Infrastructures (6rd)", RFC 5569, DOI 10.17487/RFC5569, 3586 January 2010, . 3588 [RFC5880] Katz, D. and D. Ward, "Bidirectional Forwarding Detection 3589 (BFD)", RFC 5880, DOI 10.17487/RFC5880, June 2010, 3590 . 3592 [RFC5991] Thaler, D., Krishnan, S., and J. Hoagland, "Teredo 3593 Security Updates", RFC 5991, DOI 10.17487/RFC5991, 3594 September 2010, . 3596 [RFC6106] Jeong, J., Park, S., Beloeil, L., and S. Madanapalli, 3597 "IPv6 Router Advertisement Options for DNS Configuration", 3598 RFC 6106, DOI 10.17487/RFC6106, November 2010, 3599 . 3601 [RFC6146] Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful 3602 NAT64: Network Address and Protocol Translation from IPv6 3603 Clients to IPv4 Servers", RFC 6146, DOI 10.17487/RFC6146, 3604 April 2011, . 3606 [RFC6179] Templin, F., Ed., "The Internet Routing Overlay Network 3607 (IRON)", RFC 6179, DOI 10.17487/RFC6179, March 2011, 3608 . 3610 [RFC6221] Miles, D., Ed., Ooghe, S., Dec, W., Krishnan, S., and A. 3611 Kavanagh, "Lightweight DHCPv6 Relay Agent", RFC 6221, 3612 DOI 10.17487/RFC6221, May 2011, 3613 . 3615 [RFC6273] Kukec, A., Krishnan, S., and S. Jiang, "The Secure 3616 Neighbor Discovery (SEND) Hash Threat Analysis", RFC 6273, 3617 DOI 10.17487/RFC6273, June 2011, 3618 . 3620 [RFC6347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer 3621 Security Version 1.2", RFC 6347, DOI 10.17487/RFC6347, 3622 January 2012, . 3624 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 3625 for Equal Cost Multipath Routing and Link Aggregation in 3626 Tunnels", RFC 6438, DOI 10.17487/RFC6438, November 2011, 3627 . 3629 [RFC6706] Templin, F., Ed., "Asymmetric Extended Route Optimization 3630 (AERO)", RFC 6706, DOI 10.17487/RFC6706, August 2012, 3631 . 3633 [RFC6864] Touch, J., "Updated Specification of the IPv4 ID Field", 3634 RFC 6864, DOI 10.17487/RFC6864, February 2013, 3635 . 3637 [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and 3638 UDP Checksums for Tunneled Packets", RFC 6935, 3639 DOI 10.17487/RFC6935, April 2013, 3640 . 3642 [RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement 3643 for the Use of IPv6 UDP Datagrams with Zero Checksums", 3644 RFC 6936, DOI 10.17487/RFC6936, April 2013, 3645 . 3647 [RFC7269] Chen, G., Cao, Z., Xie, C., and D. Binet, "NAT64 3648 Deployment Options and Experience", RFC 7269, 3649 DOI 10.17487/RFC7269, June 2014, 3650 . 3652 [RFC7333] Chan, H., Ed., Liu, D., Seite, P., Yokota, H., and J. 3653 Korhonen, "Requirements for Distributed Mobility 3654 Management", RFC 7333, DOI 10.17487/RFC7333, August 2014, 3655 . 3657 [RFC7421] Carpenter, B., Ed., Chown, T., Gont, F., Jiang, S., 3658 Petrescu, A., and A. Yourtchenko, "Analysis of the 64-bit 3659 Boundary in IPv6 Addressing", RFC 7421, 3660 DOI 10.17487/RFC7421, January 2015, 3661 . 3663 [RFC7761] Fenner, B., Handley, M., Holbrook, H., Kouvelas, I., 3664 Parekh, R., Zhang, Z., and L. Zheng, "Protocol Independent 3665 Multicast - Sparse Mode (PIM-SM): Protocol Specification 3666 (Revised)", STD 83, RFC 7761, DOI 10.17487/RFC7761, March 3667 2016, . 3669 [RFC8086] Yong, L., Ed., Crabbe, E., Xu, X., and T. Herbert, "GRE- 3670 in-UDP Encapsulation", RFC 8086, DOI 10.17487/RFC8086, 3671 March 2017, . 3673 [RFC8201] McCann, J., Deering, S., Mogul, J., and R. Hinden, Ed., 3674 "Path MTU Discovery for IP version 6", STD 87, RFC 8201, 3675 DOI 10.17487/RFC8201, July 2017, 3676 . 3678 [RFC8402] Filsfils, C., Ed., Previdi, S., Ed., Ginsberg, L., 3679 Decraene, B., Litkowski, S., and R. Shakir, "Segment 3680 Routing Architecture", RFC 8402, DOI 10.17487/RFC8402, 3681 July 2018, . 3683 [RFC8446] Rescorla, E., "The Transport Layer Security (TLS) Protocol 3684 Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018, 3685 . 3687 [RFC8754] Filsfils, C., Ed., Dukes, D., Ed., Previdi, S., Leddy, J., 3688 Matsushima, S., and D. Voyer, "IPv6 Segment Routing Header 3689 (SRH)", RFC 8754, DOI 10.17487/RFC8754, March 2020, 3690 . 3692 Appendix A. AERO Alternate Encapsulations 3694 When GUE encapsulation is not needed, AERO can use common 3695 encapsulations such as IP-in-IP [RFC2003][RFC2473][RFC4213], Generic 3696 Routing Encapsulation (GRE) [RFC2784][RFC2890] and others. The 3697 encapsulation is therefore only differentiated from non-AERO tunnels 3698 through the application of AERO control messaging and not through, 3699 e.g., a well-known UDP port number. 3701 As for GUE encapsulation, alternate AERO encapsulation formats may 3702 require encapsulation layer fragmentation. For simple IP-in-IP 3703 encapsulation, an IPv6 fragment header is inserted directly between 3704 the inner and outer IP headers when needed, i.e., even if the outer 3705 header is IPv4. The IPv6 Fragment Header is identified to the outer 3706 IP layer by its IP protocol number, and the Next Header field in the 3707 IPv6 Fragment Header identifies the inner IP header version. For GRE 3708 encapsulation, a GRE fragment header is inserted within the GRE 3709 header [I-D.templin-intarea-grefrag]. 3711 Figure 6 shows the AERO IP-in-IP encapsulation format before any 3712 fragmentation is applied: 3714 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 3715 | Outer IPv4 Header | | Outer IPv6 Header | 3716 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 3717 |IPv6 Frag Header (optional)| |IPv6 Frag Header (optional)| 3718 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 3719 | Inner IP Header | | Inner IP Header | 3720 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 3721 | | | | 3722 ~ ~ ~ ~ 3723 ~ Inner Packet Body ~ ~ Inner Packet Body ~ 3724 ~ ~ ~ ~ 3725 | | | | 3726 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 3728 Minimal Encapsulation in IPv4 Minimal Encapsulation in IPv6 3730 Figure 6: Minimal Encapsulation Format using IP-in-IP 3732 Figure 7 shows the AERO GRE encapsulation format before any 3733 fragmentation is applied: 3735 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 3736 | Outer IP Header | 3737 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 3738 | GRE Header | 3739 | (with checksum, key, etc..) | 3740 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 3741 | GRE Fragment Header (optional)| 3742 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 3743 | Inner IP Header | 3744 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 3745 | | 3746 ~ ~ 3747 ~ Inner Packet Body ~ 3748 ~ ~ 3749 | | 3750 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 3752 Figure 7: Minimal Encapsulation Using GRE 3754 Alternate encapsulation may be preferred in environments where GUE 3755 encapsulation would add unnecessary overhead. For example, certain 3756 low-bandwidth wireless data links may benefit from a reduced 3757 encapsulation overhead. 3759 GUE encapsulation can traverse network paths that are inaccessible to 3760 non-UDP encapsulations, e.g., for crossing Network Address 3761 Translators (NATs). More and more, network middleboxes are also 3762 being configured to discard packets that include anything other than 3763 a well-known IP protocol such as UDP and TCP. It may therefore be 3764 necessary to determine the potential for middlebox filtering before 3765 enabling alternate encapsulation in a given environment. 3767 In addition to IP-in-IP, GRE and GUE, AERO can also use security 3768 encapsulations such as IPsec, TLS/SSL, DTLS, etc. In that case, AERO 3769 control messaging and route determination occur before security 3770 encapsulation is applied for outgoing packets and after security 3771 decapsulation is applied for incoming packets. 3773 AERO is especially well suited for use with VPN system encapsulations 3774 such as OpenVPN [OVPN]. 3776 Appendix B. Non-Normative Considerations 3778 AERO can be applied to a multitude of Internetworking scenarios, with 3779 each having its own adaptations. The following considerations are 3780 provided as non-normative guidance: 3782 B.1. Implementation Strategies for Route Optimization 3784 Route optimization as discussed in Section 3.17 results in the route 3785 optimization source (ROS) creating an asymmetric neighbor cache entry 3786 for the target neighbor. The neighbor cache entry is maintained for 3787 at most REACHABLE_TIME seconds and then deleted unless updated. In 3788 order to refresh the neighbor cache entry lifetime before the 3789 ReachableTime timer expires, the specification requires 3790 implementations to issue a new NS/NA exchange to reset ReachableTime 3791 to REACHABLE_TIME seconds while data packets are still flowing. 3792 However, the decision of when to initiate a new NS/NA exchange and to 3793 perpetuate the process is left as an implementation detail. 3795 One possible strategy may be to monitor the neighbor cache entry 3796 watching for data packets for (REACHABLE_TIME - 5) seconds. If any 3797 data packets have been sent to the neighbor within this timeframe, 3798 then send an NS to receive a new NA. If no data packets have been 3799 sent, wait for 5 additional seconds and send an immediate NS if any 3800 data packets are sent within this "expiration pending" 5 second 3801 window. If no additional data packets are sent within the 5 second 3802 window, delete the neighbor cache entry. 3804 The monitoring of the neighbor data packet traffic therefore becomes 3805 an asymmetric ongoing process during the neighbor cache entry 3806 lifetime. If the neighbor cache entry expires, future data packets 3807 will trigger a new NS/NA exchange while the packets themselves are 3808 delivered over a longer path until route optimization state is re- 3809 established. 3811 B.2. Implicit Mobility Management 3813 AERO interface neighbors MAY provide a configuration option that 3814 allows them to perform implicit mobility management in which no ND 3815 messaging is used. In that case, the Client only transmits packets 3816 over a single interface at a time, and the neighbor always observes 3817 packets arriving from the Client from the same link-layer source 3818 address. 3820 If the Client's underlying interface address changes (either due to a 3821 readdressing of the original interface or switching to a new 3822 interface) the neighbor immediately updates the neighbor cache entry 3823 for the Client and begins accepting and sending packets according to 3824 the Client's new address. This implicit mobility method applies to 3825 use cases such as cellphones with both WiFi and Cellular interfaces 3826 where only one of the interfaces is active at a given time, and the 3827 Client automatically switches over to the backup interface if the 3828 primary interface fails. 3830 B.3. Direct Underlying Interfaces 3832 When a Client's AERO interface is configured over a Direct interface, 3833 the neighbor at the other end of the Direct link can receive packets 3834 without any encapsulation. In that case, the Client sends packets 3835 over the Direct link according to QoS preferences. If the Direct 3836 interface has the highest QoS preference, then the Client's IP 3837 packets are transmitted directly to the peer without going through an 3838 ANET/INET. If other interfaces have higher QoS preferences, then the 3839 Client's IP packets are transmitted via a different interface, which 3840 may result in the inclusion of Proxys, Servers and Bridges in the 3841 communications path. Direct interfaces must be tested periodically 3842 for reachability, e.g., via NUD. 3844 B.4. Operation on AERO Links with /64 ASPs 3846 IPv6 AERO links typically have MSPs that aggregate many candidate 3847 MNPs of length /64 or shorter. However, in some cases it may be 3848 desirable to use AERO over links that have only a /64 MSP. This can 3849 be accommodated by treating all Clients on the AERO link as simple 3850 hosts that receive /128 prefix delegations. 3852 In that case, the Client sends an RS message to the Server the same 3853 as for ordinary AERO links. The Server responds with an RA message 3854 that includes one or more /128 prefixes (i.e., singleton addresses) 3855 that include the /64 MSP prefix along with an interface identifier 3856 portion to be assigned to the Client. The Client and Server then 3857 configure their AERO addresses based on the interface identifier 3858 portions of the /128s (i.e., the lower 64 bits) and not based on the 3859 /64 prefix (i.e., the upper 64 bits). 3861 For example, if the MSP for the host-only IPv6 AERO link is 3862 2001:db8:1000:2000::/64, each Client will receive one or more /128 3863 IPv6 prefix delegations such as 2001:db8:1000:2000::1/128, 3864 2001:db8:1000:2000::2/128, etc. When the Client receives the prefix 3865 delegations, it assigns the AERO addresses fe80::1, fe80::2, etc. to 3866 the AERO interface, and assigns the global IPv6 addresses (i.e., the 3867 /128s) to either the AERO interface or an internal virtual interface 3868 such as a loopback. In this arrangement, the Client conducts route 3869 optimization in the same sense as discussed in Section 3.17. 3871 This specification has applicability for nodes that act as a Client 3872 on an "upstream" AERO link, but also act as a Server on "downstream" 3873 AERO links. More specifically, if the node acts as a Client to 3874 receive a /64 prefix from the upstream AERO link it can then act as a 3875 Server to provision /128s to Clients on downstream AERO links. 3877 B.5. AERO Critical Infrastructure Considerations 3879 AERO Bridges can be either Commercial off-the Shelf (COTS) standard 3880 IP routers or virtual machines in the cloud. Bridges must be 3881 provisioned, supported and managed by the INET administrative 3882 authority, and connected to the Bridges of other INETs via inter- 3883 domain peerings. Cost for purchasing, configuring and managing 3884 Bridges is nominal even for very large AERO links. 3886 AERO Servers can be standard dedicated server platforms, but most 3887 often will be deployed as virtual machines in the cloud. The only 3888 requirements for Servers are that they can run the AERO user-level 3889 code and have at least one network interface connection to the INET. 3890 As with Bridges, Servers must be provisioned, supported and managed 3891 by the INET administrative authority. Cost for purchasing, 3892 configuring and managing Servers is nominal especially for virtual 3893 Servers hosted in the cloud. 3895 AERO Proxys are most often standard dedicated server platforms with 3896 one network interface connected to the ANET and a second interface 3897 connected to an INET. As with Servers, the only requirements are 3898 that they can run the AERO user-level code and have at least one 3899 interface connection to the INET. Proxys must be provisioned, 3900 supported and managed by the ANET administrative authority. Cost for 3901 purchasing, configuring and managing Proxys is nominal, and borne by 3902 the ANET administrative authority. 3904 AERO Relays can be any dedicated server or COTS router platform 3905 connected to INETs and/or EUNs. The Relay joins the SPAN and engages 3906 in eBGP peering with one or more Bridges as a stub AS. The Relay 3907 then injects its MNPs and/or non-MNP prefixes into the BGP routing 3908 system, and provisions the prefixes to its downstream-attached 3909 networks. The Relay can perform ROS/ROR services the same as for any 3910 Server, and can route between the MNP and non-MNP address spaces. 3912 B.6. AERO Server Failure Implications 3914 AERO Servers may appear as a single point of failure in the 3915 architecture, but such is not the case since all Servers on the link 3916 provide identical services and loss of a Server does not imply 3917 immediate and/or comprehensive communication failures. Although 3918 Clients typically associate with a single Server at a time, Server 3919 failure is quickly detected and conveyed by Bidirectional Forward 3920 Detection (BFD) and/or proactive NUD allowing Clients to migrate to 3921 new Servers. 3923 If a Server fails, ongoing packet forwarding to Clients will continue 3924 by virtue of the asymmetric neighbor cache entries that have already 3925 been established in route optimization sources (ROSs). If a Client 3926 also experiences mobility events at roughly the same time the Server 3927 fails, unsolicited NA messages may be lost but proxy neighbor cache 3928 entries in the DEPARTED state will ensure that packet forwarding to 3929 the Client's new locations will continue for up to DEPART_TIME 3930 seconds. 3932 If a Client is left without a Server for an extended timeframe (e.g., 3933 greater than REACHABLETIIME seconds) then existing asymmetric 3934 neighbor cache entries will eventually expire and both ongoing and 3935 new communications will fail. The original source will continue to 3936 retransmit until the Client has established a new Server 3937 relationship, after which time continuous communications will resume. 3939 Therefore, providing many Servers on the link with high availability 3940 profiles provides resilience against loss of individual Servers and 3941 assurance that Clients can establish new Server relationships quickly 3942 in event of a Server failure. 3944 B.7. AERO Client / Server Architecture 3946 The AERO architectural model is client / server in the control plane, 3947 with route optimization in the data plane. The same as for common 3948 Internet services, the AERO Client discovers the addresses of AERO 3949 Servers and selects one Server to connect to. The AERO service is 3950 analogous to common Internet services such as google.com, yahoo.com, 3951 cnn.com, etc. However, there is only one AERO service for the link 3952 and all Servers provide identical services. 3954 Common Internet services provide differing strategies for advertising 3955 server addresses to clients. The strategy is conveyed through the 3956 DNS resource records returned in response to name resolution queries. 3957 As of January 2020 Internet-based 'nslookup' services were used to 3958 determine the following: 3960 o When a client resolves the domainname "google.com", the DNS always 3961 returns one A record (i.e., an IPv4 address) and one AAAA record 3962 (i.e., an IPv6 address). The client receives the same addresses 3963 each time it resolves the domainname via the same DNS resolver, 3964 but may receive different addresses when it resolves the 3965 domainname via different DNS resolvers. But, in each case, 3966 exactly one A and one AAAA record are returned. 3968 o When a client resolves the domainname "ietf.org", the DNS always 3969 returns one A record and one AAAA record with the same addresses 3970 regardless of which DNS resolver is used. 3972 o When a client resolves the domainname "yahoo.com", the DNS always 3973 returns a list of 4 A records and 4 AAAA records. Each time the 3974 client resolves the domainname via the same DNS resolver, the same 3975 list of addresses are returned but in randomized order (i.e., 3976 consistent with a DNS round-robin strategy). But, interestingly, 3977 the same addresses are returned (albeit in randomized order) when 3978 the domainname is resolved via different DNS resolvers. 3980 o When a client resolves the domainname "amazon.com", the DNS always 3981 returns a list of 3 A records and no AAAA records. As with 3982 "yahoo.com", the same three A records are returned from any 3983 worldwide Internet connection point in randomized order. 3985 The above example strategies show differing approaches to Internet 3986 resilience and service distribution offered by major Internet 3987 services. The Google approach exposes only a single IPv4 and a 3988 single IPv6 address to clients. Clients can then select whichever IP 3989 protocol version offers the best response, but will always use the 3990 same IP address according to the current Internet connection point. 3991 This means that the IP address offered by the network must lead to a 3992 highly-available server and/or service distribution point. In other 3993 words, resilience is predicated on high availability within the 3994 network and with no client-initiated failovers expected (i.e., it is 3995 all-or-nothing from the client's perspective). However, Google does 3996 provide for worldwide distributed service distribution by virtue of 3997 the fact that each Internet connection point responds with a 3998 different IPv6 and IPv4 address. The IETF approach is like google 3999 (all-or-nothing from the client's perspective), but provides only a 4000 single IPv4 or IPv6 address on a worldwide basis. This means that 4001 the addresses must be made highly-available at the network level with 4002 no client failover possibility, and if there is any worldwide service 4003 distribution it would need to be conducted by a network element that 4004 is reached via the IP address acting as a service distribution point. 4006 In contrast to the Google and IETF philosophies, Yahoo and Amazon 4007 both provide clients with a (short) list of IP addresses with Yahoo 4008 providing both IP protocol versions and Amazon as IPv4-only. The 4009 order of the list is randomized with each name service query 4010 response, with the effect of round-robin load balancing for service 4011 distribution. With a short list of addresses, there is still 4012 expectation that the network will implement high availability for 4013 each address but in case any single address fails the client can 4014 switch over to using a different address. The balance then becomes 4015 one of function in the network vs function in the end system. 4017 The same implications observed for common highly-available services 4018 in the Internet apply also to the AERO client/server architecture. 4019 When an AERO Client connects to one or more ANETs, it discovers one 4020 or more AERO Server addresses through the mechanisms discussed in 4021 earlier sections. Each Server address presumably leads to a fault- 4022 tolerant clustering arrangement such as supported by Linux-HA, 4023 Extended Virtual Synchrony or Paxos. Such an arrangement has 4024 precedence in common Internet service deployments in lightweight 4025 virtual machines without requiring expensive hardware deployment. 4026 Similarly, common Internet service deployments set service IP 4027 addresses on service distribution points that may relay requests to 4028 many different servers. 4030 For AERO, the expectation is that a combination of the Google/IETF 4031 and Yahoo/Amazon philosophies would be employed. The AERO Client 4032 connects to different ANET access points and can receive 1-2 Server 4033 AERO addresses at each point. It then selects one AERO Server 4034 address, and engages in RS/RA exchanges with the same Server from all 4035 ANET connections. The Client remains with this Server unless or 4036 until the Server fails, in which case it can switch over to an 4037 alternate Server. The Client can likewise switch over to a different 4038 Server at any time if there is some reason for it to do so. So, the 4039 AERO expectation is for a balance of function in the network and end 4040 system, with fault tolerance and resilience at both levels. 4042 Appendix C. Change Log 4044 << RFC Editor - remove prior to publication >> 4046 Changes from draft-templin-intarea-6706bis-48 to draft-templin- 4047 intrea-6706bis-49: 4049 o SPAN Anycast address and SBM/PBM concepts introduced. 4051 Changes from draft-templin-intarea-6706bis-47 to draft-templin- 4052 intrea-6706bis-48: 4054 o SEND/CGA. 4056 Changes from draft-templin-intarea-6706bis-46 to draft-templin- 4057 intrea-6706bis-47: 4059 o Major changes to align with Teredo, including changed AERO "Relay" 4060 to "Bridge", and changed AERO "Gateway" to "Relay". The term 4061 "[Rr]elay" now refers to exactly the same thing in both AERO and 4062 Teredo. 4064 o Changed to use Teredo message authentication instead of SEND. 4066 Changes from draft-templin-intarea-6706bis-42 to draft-templin- 4067 intrea-6706bis-43: 4069 o Segment Routing. 4071 Changes from draft-templin-intarea-6706bis-39 to draft-templin- 4072 intrea-6706bis-40: 4074 o Teredo. 4076 Changes from draft-templin-intarea-6706bis-38 to draft-templin- 4077 intrea-6706bis-39: 4079 o Major clarifications and simplifications of SPAN fragmentation/ 4080 reassembly. 4082 o Revised AERO address format to support prefix lengths up to 112. 4084 o New method for forming SPAN Client Prefixes and population in the 4085 routing system. 4087 o Updates RFC4443 to set a new value in the ICMP PTB Code field. 4089 Changes from draft-templin-intarea-6706bis-35 to draft-templin- 4090 intrea-6706bis-36: 4092 o Clients in the open Internet secured using SEND/CGA. 4094 Changes from draft-templin-intarea-6706bis-32 to draft-templin- 4095 intrea-6706bis-33: 4097 o Updated Proxy discussion with "point-to-multipoint" server 4098 coordination 4100 o Significant updates to Address Resolution and NUD to include 4101 correct addresses in messages 4103 o Differentiate between NS(AR) and NS(NUD) as their addresses and 4104 use cases differ. 4106 Changes from draft-templin-intarea-6706bis-30 to draft-templin- 4107 intrea-6706bis-31: 4109 o Added "advisory PTB messages" under FAA SE2025 contract number 4110 DTFAWA-15-D-00030. 4112 Changes from draft-templin-intarea-6706bis-29 to draft-templin- 4113 intrea-6706bis-30: 4115 o Deprecate "primary" concept. Now, RS/RA keepalives are maintained 4116 over *all* underlying interfaces (i.e., and not just one primary). 4118 Changes from draft-templin-intarea-6706bis-28 to draft-templin- 4119 intrea-6706bis-29: 4121 o Changed OMNI interface citation to "draft-templin-6man-omni- 4122 interface" 4124 o Changed SPAN Service Prefix to fd80::/10. 4126 o Changed S/TLLAO format to include 'S' bit for ifIndex 4127 corresponding to the underlying interface that is Source of ND 4128 message. 4130 o Updated Path MTU 4132 Changes from draft-templin-intarea-6706bis-27 to draft-templin- 4133 intrea-6706bis-28: 4135 o MTU and fragmentation. 4137 Changes from draft-templin-intarea-6706bis-26 to draft-templin- 4138 intrea-6706bis-27: 4140 o MTU and fragmentation. 4142 o SPAN Service Prefix set to fd00::/10 4144 o Client SPAN addresses defined. 4146 Changes from draft-templin-intarea-6706bis-25 to draft-templin- 4147 intrea-6706bis-26: 4149 o MTU and RA configuration information updated. 4151 Changes from draft-templin-intarea-6706bis-24 to draft-templin- 4152 intrea-6706bis-25: 4154 o Added concept of "primary" to allow for proxyed RS/RA over only 4155 selected underlying interfaces. 4157 o General Cleanup. 4159 Changes from draft-templin-intarea-6706bis-23 to draft-templin- 4160 intrea-6706bis-24: 4162 o OMNI interface spec now a normative reference. 4164 o Use REACHABLE_TIME as the nominal Router Lifetime to return in 4165 RAs. 4167 o General cleanup. 4169 Changes from draft-templin-intarea-6706bis-22 to draft-templin- 4170 intrea-6706bis-23: 4172 o Choice of using either RS/RA or unsolicited NA for old Server 4173 notification. 4175 o General cleanup. 4177 Changes from draft-templin-intarea-6706bis-21 to draft-templin- 4178 intrea-6706bis-22: 4180 o Tightened up text on Proxy. 4182 o Removed unnecessarily restrictive texts. 4184 o General cleanup. 4186 Changes from draft-templin-intarea-6706bis-20 to draft-templin- 4187 intrea-6706bis-21: 4189 o Clarified relationship between OMNI and S/TLLAO ifIndex-tuples. 4191 o Important text in Section 13.15.3 on Servers timing out Clients 4192 that have gone silent without sending a departure notification. 4194 o New text on RS/RA as "hints of forward progress" for proactive 4195 NUD. 4197 Changes from draft-templin-intarea-6706bis-19 to draft-templin- 4198 intrea-6706bis-20: 4200 o Included new route optimization source and destination addressing 4201 strategy. Now, route optimization maintenance uses the address of 4202 the existing Server instead of the data packet destination address 4203 so that less pressure is placed on the BGP routing system 4204 convergence time and Server constancy is supported. 4206 o Included new method for releasing from old MSE without requiring 4207 Client messaging. 4209 o Included references to new OMNI interface spec (including the OMNI 4210 option). 4212 o New appendix on AERO Client/Server architecture. 4214 Changes from draft-templin-intarea-6706bis-18 to draft-templin- 4215 intrea-6706bis-19: 4217 o Changed Proxy/Server keepalives to use "proactive NUD" in a manner 4218 that parallels BFD 4220 Changes from draft-templin-intarea-6706bis-17 to draft-templin- 4221 intrea-6706bis-18: 4223 o Discuss how AERO option is used in relation to S/TLLAOs 4225 o New text on Bidirectional Forwarding Detection (BFD) 4227 o Cleaned up usage (and non-usage) of unsolicited NAs 4229 o New appendix on Server failures 4231 Changes from draft-templin-intarea-6706bis-15 to draft-templin- 4232 intrea-6706bis-17: 4234 o S/TLLAO now includes multiple link-layer addresses within a single 4235 option instead of requiring multiple options 4237 o New unsolicited NA message to inform the old link that a Client 4238 has moved to a new link 4240 Changes from draft-templin-intarea-6706bis-14 to draft-templin- 4241 intrea-6706bis-15: 4243 o MTU and fragmentation 4245 o New details in movement to new Server 4247 Changes from draft-templin-intarea-6706bis-13 to draft-templin- 4248 intrea-6706bis-14: 4250 o Security based on secured tunnels, ingress filtering, MAP list and 4251 ROS list 4253 Changes from draft-templin-intarea-6706bis-12 to draft-templin- 4254 intrea-6706bis-13: 4256 o New paragraph in Section 3.6 on AERO interface layering over 4257 secured tunnels 4259 o Removed extraneous text in Section 3.7 4261 o Added new detail to the forwarding algorithm in Section 3.9 4262 o Clarified use of fragmentation 4264 o Route optimization now supported for both MNP and non-MNP-based 4265 prefixes 4267 o Relays are now seen as link-layer elements in the architecture. 4269 o Built out multicast section in detail. 4271 o New Appendix on implementation considerations for route 4272 optimization. 4274 Changes from draft-templin-intarea-6706bis-11 to draft-templin- 4275 intrea-6706bis-12: 4277 o Introduced Gateways as a new AERO element for connecting 4278 Correspondent Nodes on INET links 4280 o Introduced terms "Access Network (ANET)" and "Internetwork (INET)" 4282 o Changed "ASP" to "MSP", and "ACP" to "MNP" 4284 o New figure on the relation of Segments to the SPAN and AERO link 4286 o New "S" bit in S/TLLAO to indicate the "Source" S/TLLAO as opposed 4287 to additional S/TLLAOs 4289 o Changed Interface ID for Servers from 255 to 0xffff 4291 o Significant updates to Route Optimization, NUD, and Mobility 4292 Management 4294 o New Section on Multicast 4296 o New Section on AERO Clients in the open Internetwork 4298 o New Section on Operation over multiple AERO links (VLANs over the 4299 SPAN) 4301 o New Sections on DNS considerations and Transition considerations 4303 o 4305 Changes from draft-templin-intarea-6706bis-10 to draft-templin- 4306 intrea-6706bis-11: 4308 o Added The SPAN 4309 Changes from draft-templin-intarea-6706bis-09 to draft-templin- 4310 intrea-6706bis-10: 4312 o Orphaned packets in flight (e.g., when a neighbor cache entry is 4313 in the DEPARTED state) are now forwarded at the link layer instead 4314 of at the network layer. Forwarding at the network layer can 4315 result in routing loops and/or excessive delays of forwarded 4316 packets while the routing system is still reconverging. 4318 o Update route optimization to clarify the unsecured nature of the 4319 first NS used for route discovery 4321 o Many cleanups and clarifications on ND messaging parameters 4323 Changes from draft-templin-intarea-6706bis-08 to draft-templin- 4324 intrea-6706bis-09: 4326 o Changed PRL to "MAP list" 4328 o For neighbor cache entries, changed "static" to "symmetric", and 4329 "dynamic" to "asymmetric" 4331 o Specified Proxy RS/RA exchanges with Servers on behalf of Clients 4333 o Added discussion of unsolicited NAs in Section 3.16, and included 4334 forward reference to Section 3.18 4336 o Added discussion of AERO Clients used as critical infrastructure 4337 elements to connect fixed networks. 4339 o Added network-based VPN under security considerations 4341 Changes from draft-templin-intarea-6706bis-07 to draft-templin- 4342 intrea-6706bis-08: 4344 o New section on AERO-Aware Access Router 4346 Changes from draft-templin-intarea-6706bis-06 to draft-templin- 4347 intrea-6706bis-07: 4349 o Added "R" bit for release of PDs. Now have a full RS/RA service 4350 that can do PD without requiring DHCPv6 messaging over-the-air 4352 o Clarifications on solicited vs unsolicited NAs 4354 o Clarified use of MAX_NEIGHBOR_ADVERTISEMENT for the purpose of 4355 increase reliability 4357 Changes from draft-templin-intarea-6706bis-05 to draft-templin- 4358 intrea-6706bis-06: 4360 o Major re-work and simplification of Route Optimization function 4362 o Added Distributed Mobility Management (DMM) and Mobility Anchor 4363 Point (MAP) terminology 4365 o New section on "AERO Critical Infrastructure Element 4366 Considerations" demonstrating low overall cost for the service 4368 o minor text revisions and deletions 4370 o removed extraneous appendices 4372 Changes from draft-templin-intarea-6706bis-04 to draft-templin- 4373 intrea-6706bis-05: 4375 o New Appendix E on S/TLLAO Extensions for special-purpose links. 4376 Discussed ATN/IPS as example. 4378 o New sentence in introduction to declare appendices as non- 4379 normative. 4381 Changes from draft-templin-intarea-6706bis-03 to draft-templin- 4382 intrea-6706bis-04: 4384 o Added definitions for Potential Router List (PRL) and secure 4385 enclave 4387 o Included text on mapping transport layer port numbers to network 4388 layer DSCP values 4390 o Added reference to DTLS and DMM Distributed Mobility Anchoring 4391 working group document 4393 o Reworked Security Considerations 4395 o Updated references. 4397 Changes from draft-templin-intarea-6706bis-02 to draft-templin- 4398 intrea-6706bis-03: 4400 o Added new section on SEND. 4402 o Clarifications on "AERO Address" section. 4404 o Updated references and added new reference for RFC8086. 4406 o Security considerations updates. 4408 o General text clarifications and cleanup. 4410 Changes from draft-templin-intarea-6706bis-01 to draft-templin- 4411 intrea-6706bis-02: 4413 o Note on encapsulation avoidance in Section 4. 4415 Changes from draft-templin-intarea-6706bis-00 to draft-templin- 4416 intrea-6706bis-01: 4418 o Remove DHCPv6 Server Release procedures that leveraged the old way 4419 Relays used to "route" between Server link-local addresses 4421 o Remove all text relating to Relays needing to do any AERO-specific 4422 operations 4424 o Proxy sends RS and receives RA from Server using SEND. Use CGAs 4425 as source addresses, and destination address of RA reply is to the 4426 AERO address corresponding to the Client's ACP. 4428 o Proxy uses SEND to protect RS and authenticate RA (Client does not 4429 use SEND, but rather relies on subnetwork security. When the 4430 Proxy receives an RS from the Client, it creates a new RS using 4431 its own addresses as the source and uses SEND with CGAs to send a 4432 new RS to the Server. 4434 o Emphasize distributed mobility management 4436 o AERO address-based RS injection of ACP into underlying routing 4437 system. 4439 Changes from draft-templin-aerolink-82 to draft-templin-intarea- 4440 6706bis-00: 4442 o Document use of NUD (NS/NA) for reliable link-layer address 4443 updates as an alternative to unreliable unsolicited NA. 4444 Consistent with Section 7.2.6 of RFC4861. 4446 o Server adds additional layer of encapsulation between outer and 4447 inner headers of NS/NA messages for transmission through Relays 4448 that act as vanilla IPv6 routers. The messages include the AERO 4449 Server Subnet Router Anycast address as the source and the Subnet 4450 Router Anycast address corresponding to the Client's ACP as the 4451 destination. 4453 o Clients use Subnet Router Anycast address as the encapsulation 4454 source address when the access network does not provide a 4455 topologically-fixed address. 4457 Author's Address 4459 Fred L. Templin (editor) 4460 Boeing Research & Technology 4461 P.O. Box 3707 4462 Seattle, WA 98124 4463 USA 4465 Email: fltemplin@acm.org