idnits 2.17.1 draft-ietf-lisp-21.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 : ---------------------------------------------------------------------------- == There are 10 instances of lines with private range IPv4 addresses in the document. If these are generic example addresses, they should be changed to use any of the ranges defined in RFC 6890 (or successor): 192.0.2.x, 198.51.100.x or 203.0.113.x. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == The document seems to use 'NOT RECOMMENDED' as an RFC 2119 keyword, but does not include the phrase in its RFC 2119 key words list. -- The document date (February 7, 2012) is 4462 days in the past. Is this intentional? Checking references for intended status: Experimental ---------------------------------------------------------------------------- == Outdated reference: A later version (-10) exists of draft-ietf-lisp-alt-09 == Outdated reference: A later version (-16) exists of draft-ietf-lisp-ms-12 ** Obsolete normative reference: RFC 2460 (Obsoleted by RFC 8200) ** Obsolete normative reference: RFC 5226 (Obsoleted by RFC 8126) == Outdated reference: A later version (-12) exists of draft-ietf-6man-udpzero-04 == Outdated reference: A later version (-09) exists of draft-ietf-lisp-map-versioning-05 == Outdated reference: A later version (-06) exists of draft-ietf-lisp-interworking-02 == Outdated reference: A later version (-10) exists of draft-ietf-karp-design-guide-06 == Outdated reference: A later version (-10) exists of draft-farinacci-lisp-lcaf-06 == Outdated reference: A later version (-12) exists of draft-ietf-lisp-deployment-02 == Outdated reference: A later version (-13) exists of draft-ietf-lisp-mib-02 == Outdated reference: A later version (-16) exists of draft-meyer-lisp-mn-06 == Outdated reference: A later version (-29) exists of draft-ietf-lisp-sec-00 -- Unexpected draft version: The latest known version of draft-meyer-loc-id-implications is -01, but you're referring to -02. == Outdated reference: A later version (-14) exists of draft-ietf-lisp-multicast-10 == Outdated reference: A later version (-09) exists of draft-lear-lisp-nerd-08 Summary: 2 errors (**), 0 flaws (~~), 16 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group D. Farinacci 3 Internet-Draft V. Fuller 4 Intended status: Experimental D. Meyer 5 Expires: August 10, 2012 D. Lewis 6 cisco Systems 7 February 7, 2012 9 Locator/ID Separation Protocol (LISP) 10 draft-ietf-lisp-21 12 Abstract 14 This draft describes a network layer based protocol that enables 15 separation of IP addresses into two new numbering spaces: Endpoint 16 Identifiers (EIDs) and Routing Locators (RLOCs). No changes are 17 required to either host protocol stacks or to the "core" of the 18 Internet infrastructure. LISP can be incrementally deployed, without 19 a "flag day", and offers traffic engineering, multi-homing, and 20 mobility benefits to early adopters, even when there are relatively 21 few LISP-capable sites. 23 Design and development of LISP was largely motivated by the problem 24 statement produced by the October 2006 IAB Routing and Addressing 25 Workshop. 27 Status of this Memo 29 This Internet-Draft is submitted in full conformance with the 30 provisions of BCP 78 and BCP 79. 32 Internet-Drafts are working documents of the Internet Engineering 33 Task Force (IETF). Note that other groups may also distribute 34 working documents as Internet-Drafts. The list of current Internet- 35 Drafts is at http://datatracker.ietf.org/drafts/current/. 37 Internet-Drafts are draft documents valid for a maximum of six months 38 and may be updated, replaced, or obsoleted by other documents at any 39 time. It is inappropriate to use Internet-Drafts as reference 40 material or to cite them other than as "work in progress." 42 This Internet-Draft will expire on August 10, 2012. 44 Copyright Notice 46 Copyright (c) 2012 IETF Trust and the persons identified as the 47 document authors. All rights reserved. 49 This document is subject to BCP 78 and the IETF Trust's Legal 50 Provisions Relating to IETF Documents 51 (http://trustee.ietf.org/license-info) in effect on the date of 52 publication of this document. Please review these documents 53 carefully, as they describe your rights and restrictions with respect 54 to this document. Code Components extracted from this document must 55 include Simplified BSD License text as described in Section 4.e of 56 the Trust Legal Provisions and are provided without warranty as 57 described in the Simplified BSD License. 59 Table of Contents 61 1. Requirements Notation . . . . . . . . . . . . . . . . . . . . 5 62 2. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 6 63 3. Definition of Terms . . . . . . . . . . . . . . . . . . . . . 8 64 4. Basic Overview . . . . . . . . . . . . . . . . . . . . . . . . 14 65 4.1. Packet Flow Sequence . . . . . . . . . . . . . . . . . . . 16 66 5. LISP Encapsulation Details . . . . . . . . . . . . . . . . . . 18 67 5.1. LISP IPv4-in-IPv4 Header Format . . . . . . . . . . . . . 19 68 5.2. LISP IPv6-in-IPv6 Header Format . . . . . . . . . . . . . 19 69 5.3. Tunnel Header Field Descriptions . . . . . . . . . . . . . 21 70 5.4. Dealing with Large Encapsulated Packets . . . . . . . . . 25 71 5.4.1. A Stateless Solution to MTU Handling . . . . . . . . . 25 72 5.4.2. A Stateful Solution to MTU Handling . . . . . . . . . 26 73 5.5. Using Virtualization and Segmentation with LISP . . . . . 26 74 6. EID-to-RLOC Mapping . . . . . . . . . . . . . . . . . . . . . 28 75 6.1. LISP IPv4 and IPv6 Control Plane Packet Formats . . . . . 28 76 6.1.1. LISP Packet Type Allocations . . . . . . . . . . . . . 30 77 6.1.2. Map-Request Message Format . . . . . . . . . . . . . . 30 78 6.1.3. EID-to-RLOC UDP Map-Request Message . . . . . . . . . 33 79 6.1.4. Map-Reply Message Format . . . . . . . . . . . . . . . 34 80 6.1.5. EID-to-RLOC UDP Map-Reply Message . . . . . . . . . . 38 81 6.1.6. Map-Register Message Format . . . . . . . . . . . . . 40 82 6.1.7. Map-Notify Message Format . . . . . . . . . . . . . . 42 83 6.1.8. Encapsulated Control Message Format . . . . . . . . . 43 84 6.2. Routing Locator Selection . . . . . . . . . . . . . . . . 45 85 6.3. Routing Locator Reachability . . . . . . . . . . . . . . . 47 86 6.3.1. Echo Nonce Algorithm . . . . . . . . . . . . . . . . . 49 87 6.3.2. RLOC Probing Algorithm . . . . . . . . . . . . . . . . 50 88 6.4. EID Reachability within a LISP Site . . . . . . . . . . . 51 89 6.5. Routing Locator Hashing . . . . . . . . . . . . . . . . . 52 90 6.6. Changing the Contents of EID-to-RLOC Mappings . . . . . . 53 91 6.6.1. Clock Sweep . . . . . . . . . . . . . . . . . . . . . 54 92 6.6.2. Solicit-Map-Request (SMR) . . . . . . . . . . . . . . 54 93 6.6.3. Database Map Versioning . . . . . . . . . . . . . . . 56 94 7. Router Performance Considerations . . . . . . . . . . . . . . 57 95 8. Deployment Scenarios . . . . . . . . . . . . . . . . . . . . . 58 96 8.1. First-hop/Last-hop Tunnel Routers . . . . . . . . . . . . 59 97 8.2. Border/Edge Tunnel Routers . . . . . . . . . . . . . . . . 59 98 8.3. ISP Provider-Edge (PE) Tunnel Routers . . . . . . . . . . 60 99 8.4. LISP Functionality with Conventional NATs . . . . . . . . 60 100 8.5. Packets Egressing a LISP Site . . . . . . . . . . . . . . 61 101 9. Traceroute Considerations . . . . . . . . . . . . . . . . . . 62 102 9.1. IPv6 Traceroute . . . . . . . . . . . . . . . . . . . . . 63 103 9.2. IPv4 Traceroute . . . . . . . . . . . . . . . . . . . . . 63 104 9.3. Traceroute using Mixed Locators . . . . . . . . . . . . . 63 105 10. Mobility Considerations . . . . . . . . . . . . . . . . . . . 65 106 10.1. Site Mobility . . . . . . . . . . . . . . . . . . . . . . 65 107 10.2. Slow Endpoint Mobility . . . . . . . . . . . . . . . . . . 65 108 10.3. Fast Endpoint Mobility . . . . . . . . . . . . . . . . . . 65 109 10.4. Fast Network Mobility . . . . . . . . . . . . . . . . . . 67 110 10.5. LISP Mobile Node Mobility . . . . . . . . . . . . . . . . 67 111 11. Multicast Considerations . . . . . . . . . . . . . . . . . . . 69 112 12. Security Considerations . . . . . . . . . . . . . . . . . . . 70 113 13. Network Management Considerations . . . . . . . . . . . . . . 72 114 14. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 73 115 14.1. LISP ACT and Flag Fields . . . . . . . . . . . . . . . . . 73 116 14.2. LISP Address Type Codes . . . . . . . . . . . . . . . . . 73 117 14.3. LISP UDP Port Numbers . . . . . . . . . . . . . . . . . . 73 118 14.4. LISP Key ID Numbers . . . . . . . . . . . . . . . . . . . 74 119 15. Known Open Issues and Areas of Future Work . . . . . . . . . . 75 120 16. References . . . . . . . . . . . . . . . . . . . . . . . . . . 77 121 16.1. Normative References . . . . . . . . . . . . . . . . . . . 77 122 16.2. Informative References . . . . . . . . . . . . . . . . . . 78 123 Appendix A. Acknowledgments . . . . . . . . . . . . . . . . . . . 82 124 Appendix B. Document Change Log . . . . . . . . . . . . . . . . . 83 125 B.1. Changes to draft-ietf-lisp-21.txt . . . . . . . . . . . . 83 126 B.2. Changes to draft-ietf-lisp-20.txt . . . . . . . . . . . . 83 127 B.3. Changes to draft-ietf-lisp-19.txt . . . . . . . . . . . . 83 128 B.4. Changes to draft-ietf-lisp-18.txt . . . . . . . . . . . . 83 129 B.5. Changes to draft-ietf-lisp-17.txt . . . . . . . . . . . . 83 130 B.6. Changes to draft-ietf-lisp-16.txt . . . . . . . . . . . . 83 131 B.7. Changes to draft-ietf-lisp-15.txt . . . . . . . . . . . . 83 132 B.8. Changes to draft-ietf-lisp-14.txt . . . . . . . . . . . . 84 133 B.9. Changes to draft-ietf-lisp-13.txt . . . . . . . . . . . . 84 134 B.10. Changes to draft-ietf-lisp-12.txt . . . . . . . . . . . . 85 135 B.11. Changes to draft-ietf-lisp-11.txt . . . . . . . . . . . . 86 136 B.12. Changes to draft-ietf-lisp-10.txt . . . . . . . . . . . . 87 137 B.13. Changes to draft-ietf-lisp-09.txt . . . . . . . . . . . . 87 138 B.14. Changes to draft-ietf-lisp-08.txt . . . . . . . . . . . . 88 139 B.15. Changes to draft-ietf-lisp-07.txt . . . . . . . . . . . . 90 140 B.16. Changes to draft-ietf-lisp-06.txt . . . . . . . . . . . . 91 141 B.17. Changes to draft-ietf-lisp-05.txt . . . . . . . . . . . . 92 142 B.18. Changes to draft-ietf-lisp-04.txt . . . . . . . . . . . . 93 143 B.19. Changes to draft-ietf-lisp-03.txt . . . . . . . . . . . . 95 144 B.20. Changes to draft-ietf-lisp-02.txt . . . . . . . . . . . . 95 145 B.21. Changes to draft-ietf-lisp-01.txt . . . . . . . . . . . . 95 146 B.22. Changes to draft-ietf-lisp-00.txt . . . . . . . . . . . . 96 147 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 97 149 1. Requirements Notation 151 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 152 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 153 document are to be interpreted as described in [RFC2119]. 155 2. Introduction 157 This document describes the Locator/Identifier Separation Protocol 158 (LISP), which provides a set of functions for routers to exchange 159 information used to map from non globally routeable Endpoint 160 Identifiers (EIDs) to routeable Routing Locators (RLOCs). It also 161 defines a mechanism for these LISP routers to encapsulate IP packets 162 addressed with EIDs for transmission across the< Internet that uses 163 RLOCs for routing and forwarding. 165 Creation of LISP was initially motivated by discussions during the 166 IAB-sponsored Routing and Addressing Workshop held in Amsterdam in 167 October, 2006 (see [RFC4984]). A key conclusion of the workshop was 168 that the Internet routing and addressing system was not scaling well 169 in the face of the explosive growth of new sites; one reason for this 170 poor scaling is the increasing number of multi-homed and other sites 171 that cannot be addressed as part of topologically- or provider-based 172 aggregated prefixes. Additional work that more completely described 173 the problem statement may be found in [RADIR]. 175 A basic observation, made many years ago in early networking research 176 such as that documented in [CHIAPPA] and [RFC4984], is that using a 177 single address field for both identifying a device and for 178 determining where it is topologically located in the network requires 179 optimization along two conflicting axes: for routing to be efficient, 180 the address must be assigned topologically; for collections of 181 devices to be easily and effectively managed, without the need for 182 renumbering in response to topological change (such as that caused by 183 adding or removing attachment points to the network or by mobility 184 events), the address must explicitly not be tied to the topology. 186 The approach that LISP takes to solving the routing scalability 187 problem is to replace IP addresses with two new types of numbers: 188 Routing Locators (RLOCs), which are topologically assigned to network 189 attachment points (and are therefore amenable to aggregation) and 190 used for routing and forwarding of packets through the network; and 191 Endpoint Identifiers (EIDs), which are assigned independently from 192 the network topology, are used for numbering devices, and are 193 aggregated along administrative boundaries. LISP then defines 194 functions for mapping between the two numbering spaces and for 195 encapsulating traffic originated by devices using non-routeable EIDs 196 for transport across a network infrastructure that routes and 197 forwards using RLOCs. Both RLOCs and EIDs are syntactically- 198 identical to IP addresses; it is the semantics of how they are used 199 that differs. 201 This document describes the protocol that implements these functions. 202 The database which stores the mappings between EIDs and RLOCs is 203 explicitly a separate "module" to facilitate experimentation with a 204 variety of approaches. One database design that is being developed 205 and prototyped as part of the LISP working group work is [ALT]. 206 Others that have been described but not implemented include [CONS], 207 [EMACS], [NERD]. Finally, [LISP-MS], documents a general-purpose 208 service interface for accessing a mapping database; this interface is 209 intended to make the mapping database modular so that different 210 approaches can be tried without the need to modify installed LISP 211 capable devices in LISP sites. 213 This experimental specification has areas that require additional 214 experience and measurement. It is NOT RECOMMENDED for deployment 215 beyond experimental situations. Results of experimentation may lead 216 to modifications and enhancements of protocol mechanisms defined in 217 this document. See Section 15 for specific, known issues that are in 218 need of further work during development, implementation, and 219 experimentation. 221 An examination of the implications of LISP on Internet traffic, 222 applications, routers, and security is for future study. This 223 analysis will explain what role LISP can play in scalable routing and 224 will also look at scalability and levels of state required for 225 encapsulation, decapsulation, liveness, and so on. 227 3. Definition of Terms 229 Provider Independent (PI) Addresses: PI addresses are an address 230 block assigned from a pool where blocks are not associated with 231 any particular location in the network (e.g. from a particular 232 service provider), and is therefore not topologically aggregatable 233 in the routing system. 235 Provider Assigned (PA) Addresses: PA addresses are an address block 236 assigned to a site by each service provider to which a site 237 connects. Typically, each block is sub-block of a service 238 provider Classless Inter-Domain Routing (CIDR) [RFC4632] block and 239 is aggregated into the larger block before being advertised into 240 the global Internet. Traditionally, IP multihoming has been 241 implemented by each multi-homed site acquiring its own, globally- 242 visible prefix. LISP uses only topologically-assigned and 243 aggregatable address blocks for RLOCs, eliminating this 244 demonstrably non-scalable practice. 246 Routing Locator (RLOC): A RLOC is an IPv4 [RFC0791] or IPv6 247 [RFC2460] address of an egress tunnel router (ETR). A RLOC is the 248 output of an EID-to-RLOC mapping lookup. An EID maps to one or 249 more RLOCs. Typically, RLOCs are numbered from topologically- 250 aggregatable blocks that are assigned to a site at each point to 251 which it attaches to the global Internet; where the topology is 252 defined by the connectivity of provider networks, RLOCs can be 253 thought of as PA addresses. Multiple RLOCs can be assigned to the 254 same ETR device or to multiple ETR devices at a site. 256 Endpoint ID (EID): An EID is a 32-bit (for IPv4) or 128-bit (for 257 IPv6) value used in the source and destination address fields of 258 the first (most inner) LISP header of a packet. The host obtains 259 a destination EID the same way it obtains an destination address 260 today, for example through a Domain Name System (DNS) [RFC1034] 261 lookup or Session Invitation Protocol (SIP) [RFC3261] exchange. 262 The source EID is obtained via existing mechanisms used to set a 263 host's "local" IP address. An EID used on the public Internet 264 must have the same properties as any other IP address used in that 265 manner; this means, among other things, that it must be globally 266 unique. An EID is allocated to a host from an EID-prefix block 267 associated with the site where the host is located. An EID can be 268 used by a host to refer to other hosts. EIDs MUST NOT be used as 269 LISP RLOCs. Note that EID blocks MAY be assigned in a 270 hierarchical manner, independent of the network topology, to 271 facilitate scaling of the mapping database. In addition, an EID 272 block assigned to a site may have site-local structure 273 (subnetting) for routing within the site; this structure is not 274 visible to the global routing system. In theory, the bit string 275 that represents an EID for one device can represent an RLOC for a 276 different device. As the architecture is realized, if a given bit 277 string is both an RLOC and an EID, it must refer to the same 278 entity in both cases. When used in discussions with other 279 Locator/ID separation proposals, a LISP EID will be called a 280 "LEID". Throughout this document, any references to "EID" refers 281 to an LEID. 283 EID-prefix: An EID-prefix is a power-of-two block of EIDs which are 284 allocated to a site by an address allocation authority. EID- 285 prefixes are associated with a set of RLOC addresses which make up 286 a "database mapping". EID-prefix allocations can be broken up 287 into smaller blocks when an RLOC set is to be associated with the 288 larger EID-prefix block. A globally routed address block (whether 289 PI or PA) is not inherently an EID-prefix. A globally routed 290 address block MAY be used by its assignee as an EID block. The 291 converse is not supported. That is, a site which receives an 292 explicitly allocated EID-prefix may not use that EID-prefix as a 293 globally routed prefix. This would require coordination and 294 cooperation with the entities managing the mapping infrastructure. 295 Once this has been done, that block could be removed from the 296 globally routed IP system, if other suitable transition and access 297 mechanisms are in place. Discussion of such transition and access 298 mechanisms can be found in [INTERWORK] and [LISP-DEPLOY]. 300 End-system: An end-system is an IPv4 or IPv6 device that originates 301 packets with a single IPv4 or IPv6 header. The end-system 302 supplies an EID value for the destination address field of the IP 303 header when communicating globally (i.e. outside of its routing 304 domain). An end-system can be a host computer, a switch or router 305 device, or any network appliance. 307 Ingress Tunnel Router (ITR): An ITR is a router that resides in a 308 LISP site. Packets sent by sources inside of the LISP site to 309 destinations outside of the site are candidates for encapsulation 310 by the ITR. The ITR treats the IP destination address as an EID 311 and performs an EID-to-RLOC mapping lookup. The router then 312 prepends an "outer" IP header with one of its globally-routable 313 RLOCs in the source address field and the result of the mapping 314 lookup in the destination address field. Note that this 315 destination RLOC MAY be an intermediate, proxy device that has 316 better knowledge of the EID-to-RLOC mapping closer to the 317 destination EID. In general, an ITR receives IP packets from site 318 end-systems on one side and sends LISP-encapsulated IP packets 319 toward the Internet on the other side. 321 Specifically, when a service provider prepends a LISP header for 322 Traffic Engineering purposes, the router that does this is also 323 regarded as an ITR. The outer RLOC the ISP ITR uses can be based 324 on the outer destination address (the originating ITR's supplied 325 RLOC) or the inner destination address (the originating hosts 326 supplied EID). 328 TE-ITR: A TE-ITR is an ITR that is deployed in a service provider 329 network that prepends an additional LISP header for Traffic 330 Engineering purposes. 332 Egress Tunnel Router (ETR): An ETR is a router that accepts an IP 333 packet where the destination address in the "outer" IP header is 334 one of its own RLOCs. The router strips the "outer" header and 335 forwards the packet based on the next IP header found. In 336 general, an ETR receives LISP-encapsulated IP packets from the 337 Internet on one side and sends decapsulated IP packets to site 338 end-systems on the other side. ETR functionality does not have to 339 be limited to a router device. A server host can be the endpoint 340 of a LISP tunnel as well. 342 TE-ETR: A TE-ETR is an ETR that is deployed in a service provider 343 network that strips an outer LISP header for Traffic Engineering 344 purposes. 346 xTR: A xTR is a reference to an ITR or ETR when direction of data 347 flow is not part of the context description. xTR refers to the 348 router that is the tunnel endpoint. Used synonymously with the 349 term "Tunnel Router". For example, "An xTR can be located at the 350 Customer Edge (CE) router", meaning both ITR and ETR functionality 351 is at the CE router. 353 LISP Router: A LISP router is a router that performs the functions 354 of any or all of ITR, ETR, PITR, or PETR. 356 EID-to-RLOC Cache: The EID-to-RLOC cache is a short-lived, on- 357 demand table in an ITR that stores, tracks, and is responsible for 358 timing-out and otherwise validating EID-to-RLOC mappings. This 359 cache is distinct from the full "database" of EID-to-RLOC 360 mappings, it is dynamic, local to the ITR(s), and relatively small 361 while the database is distributed, relatively static, and much 362 more global in scope. 364 EID-to-RLOC Database: The EID-to-RLOC database is a global 365 distributed database that contains all known EID-prefix to RLOC 366 mappings. Each potential ETR typically contains a small piece of 367 the database: the EID-to-RLOC mappings for the EID prefixes 368 "behind" the router. These map to one of the router's own, 369 globally-visible, IP addresses. The same database mapping entries 370 MUST be configured on all ETRs for a given site. In a steady 371 state the EID-prefixes for the site and the locator-set for each 372 EID-prefix MUST be the same on all ETRs. Procedures to enforce 373 and/or verify this are outside the scope of this document. Note 374 that there MAY be transient conditions when the EID-prefix for the 375 site and locator-set for each EID-prefix may not be the same on 376 all ETRs. This has no negative implications since a partial set 377 of locators can be used. 379 Recursive Tunneling: Recursive tunneling occurs when a packet has 380 more than one LISP IP header. Additional layers of tunneling MAY 381 be employed to implement traffic engineering or other re-routing 382 as needed. When this is done, an additional "outer" LISP header 383 is added and the original RLOCs are preserved in the "inner" 384 header. Any references to tunnels in this specification refers to 385 dynamic encapsulating tunnels and they are never statically 386 configured. 388 Reencapsulating Tunnels: Reencapsulating tunneling occurs when an 389 ETR removes a LISP header, then acts as an ITR to prepend another 390 LISP header. Doing this allows a packet to be re-routed by the 391 re-encapsulating router without adding the overhead of additional 392 tunnel headers. Any references to tunnels in this specification 393 refers to dynamic encapsulating tunnels and they are never 394 statically configured. When using multiple mapping database 395 systems, care must be taken to not create reencapsulation loops. 397 LISP Header: a term used in this document to refer to the outer 398 IPv4 or IPv6 header, a UDP header, and a LISP-specific 8-octet 399 header that follows the UDP header, an ITR prepends or an ETR 400 strips. 402 Address Family Identifier (AFI): a term used to describe an address 403 encoding in a packet. An address family currently pertains to an 404 IPv4 or IPv6 address. See [AFI]/[AFI-REGISTRY] and [RFC3232] for 405 details. An AFI value of 0 used in this specification indicates 406 an unspecified encoded address where the length of the address is 407 0 octets following the 16-bit AFI value of 0. 409 Negative Mapping Entry: A negative mapping entry, also known as a 410 negative cache entry, is an EID-to-RLOC entry where an EID-prefix 411 is advertised or stored with no RLOCs. That is, the locator-set 412 for the EID-to-RLOC entry is empty or has an encoded locator count 413 of 0. This type of entry could be used to describe a prefix from 414 a non-LISP site, which is explicitly not in the mapping database. 415 There are a set of well defined actions that are encoded in a 416 Negative Map-Reply (Section 6.1.5). 418 Data Probe: A data-probe is a LISP-encapsulated data packet where 419 the inner header destination address equals the outer header 420 destination address used to trigger a Map-Reply by a decapsulating 421 ETR. In addition, the original packet is decapsulated and 422 delivered to the destination host if the destination EID is in the 423 EID-prefix range configured on the ETR. Otherwise, the packet is 424 discarded. A Data Probe is used in some of the mapping database 425 designs to "probe" or request a Map-Reply from an ETR; in other 426 cases, Map-Requests are used. See each mapping database design 427 for details. When using Data Probes, by sending Map-Requests on 428 the underlying routing system, EID-prefixes must be advertised. 429 However, this is discouraged if the core is to scale by having 430 less EID-prefixes stored in the core router's routing tables. 432 Proxy ITR (PITR): A PITR is defined and described in [INTERWORK], a 433 PITR acts like an ITR but does so on behalf of non-LISP sites 434 which send packets to destinations at LISP sites. 436 Proxy ETR (PETR): A PETR is defined and described in [INTERWORK], a 437 PETR acts like an ETR but does so on behalf of LISP sites which 438 send packets to destinations at non-LISP sites. 440 Route-returnability: is an assumption that the underlying routing 441 system will deliver packets to the destination. When combined 442 with a nonce that is provided by a sender and returned by a 443 receiver, this limits off-path data insertion. A route- 444 returnability check is verified when a message is sent with a 445 nonce, another message is returned with the same nonce, and the 446 destination of the original message appears as the source of the 447 returned message. 449 LISP site: is a set of routers in an edge network that are under a 450 single technical administration. LISP routers which reside in the 451 edge network are the demarcation points to separate the edge 452 network from the core network. 454 Client-side: a term used in this document to indicate a connection 455 initiation attempt by an EID. The ITR(s) at the LISP site are the 456 first to get involved in obtaining database map cache entries by 457 sending Map-Request messages. 459 Server-side: a term used in this document to indicate a connection 460 initiation attempt is being accepted for a destination EID. The 461 ETR(s) at the destination LISP site are the first to send Map- 462 Replies to the source site initiating the connection. The ETR(s) 463 at this destination site can obtain mappings by gleaning 464 information from Map-Requests, Data-Probes, or encapsulated 465 packets. 467 Locator Status Bits (LSBs): Locator status bits are present in the 468 LISP header. They are used by ITRs to inform ETRs about the up/ 469 down status of all ETRs at the local site. These bits are used as 470 a hint to convey up/down router status and not path reachability 471 status. The LSBs can be verified by use of one of the Locator 472 Reachability Algorithms described in Section 6.3. 474 Anycast Address: a term used in this document to refer to the same 475 IPv4 or IPv6 address configured and used on multiple systems at 476 the same time. An EID or RLOC can be an anycast address in each 477 of their own address spaces. 479 4. Basic Overview 481 One key concept of LISP is that end-systems (hosts) operate the same 482 way they do today. The IP addresses that hosts use for tracking 483 sockets, connections, and for sending and receiving packets do not 484 change. In LISP terminology, these IP addresses are called Endpoint 485 Identifiers (EIDs). 487 Routers continue to forward packets based on IP destination 488 addresses. When a packet is LISP encapsulated, these addresses are 489 referred to as Routing Locators (RLOCs). Most routers along a path 490 between two hosts will not change; they continue to perform routing/ 491 forwarding lookups on the destination addresses. For routers between 492 the source host and the ITR as well as routers from the ETR to the 493 destination host, the destination address is an EID. For the routers 494 between the ITR and the ETR, the destination address is an RLOC. 496 Another key LISP concept is the "Tunnel Router". A tunnel router 497 prepends LISP headers on host-originated packets and strips them 498 prior to final delivery to their destination. The IP addresses in 499 this "outer header" are RLOCs. During end-to-end packet exchange 500 between two Internet hosts, an ITR prepends a new LISP header to each 501 packet and an egress tunnel router strips the new header. The ITR 502 performs EID-to-RLOC lookups to determine the routing path to the 503 ETR, which has the RLOC as one of its IP addresses. 505 Some basic rules governing LISP are: 507 o End-systems (hosts) only send to addresses which are EIDs. They 508 don't know addresses are EIDs versus RLOCs but assume packets get 509 to their intended destinations. In a system where LISP is 510 deployed, LISP routers intercept EID addressed packets and assist 511 in delivering them across the network core where EIDs cannot be 512 routed. The procedure a host uses to send IP packets does not 513 change. 515 o EIDs are always IP addresses assigned to hosts. 517 o LISP routers mostly deal with Routing Locator addresses. See 518 details later in Section 4.1 to clarify what is meant by "mostly". 520 o RLOCs are always IP addresses assigned to routers; preferably, 521 topologically-oriented addresses from provider CIDR (Classless 522 Inter-Domain Routing) blocks. 524 o When a router originates packets it may use as a source address 525 either an EID or RLOC. When acting as a host (e.g. when 526 terminating a transport session such as SSH, TELNET, or SNMP), it 527 may use an EID that is explicitly assigned for that purpose. An 528 EID that identifies the router as a host MUST NOT be used as an 529 RLOC; an EID is only routable within the scope of a site. A 530 typical BGP configuration might demonstrate this "hybrid" EID/RLOC 531 usage where a router could use its "host-like" EID to terminate 532 iBGP sessions to other routers in a site while at the same time 533 using RLOCs to terminate eBGP sessions to routers outside the 534 site. 536 o Packets with EIDs in them are not expected to be delivered end-to- 537 end in the absence of an EID-to-RLOC mapping operation. They are 538 expected to be used locally for intra-site communication or to be 539 encapsulated for inter-site communication. 541 o EID prefixes are likely to be hierarchically assigned in a manner 542 which is optimized for administrative convenience and to 543 facilitate scaling of the EID-to-RLOC mapping database. The 544 hierarchy is based on a address allocation hierarchy which is 545 independent of the network topology. 547 o EIDs may also be structured (subnetted) in a manner suitable for 548 local routing within an autonomous system. 550 An additional LISP header MAY be prepended to packets by a TE-ITR 551 when re-routing of the path for a packet is desired. A potential 552 use-case for this would be an ISP router that needs to perform 553 traffic engineering for packets flowing through its network. In such 554 a situation, termed Recursive Tunneling, an ISP transit acts as an 555 additional ingress tunnel router and the RLOC it uses for the new 556 prepended header would be either a TE-ETR within the ISP (along 557 intra-ISP traffic engineered path) or a TE-ETR within another ISP (an 558 inter-ISP traffic engineered path, where an agreement to build such a 559 path exists). 561 In order to avoid excessive packet overhead as well as possible 562 encapsulation loops, this document mandates that a maximum of two 563 LISP headers can be prepended to a packet. For initial LISP 564 deployments, it is assumed two headers is sufficient, where the first 565 prepended header is used at a site for Location/Identity separation 566 and second prepended header is used inside a service provider for 567 Traffic Engineering purposes. 569 Tunnel Routers can be placed fairly flexibly in a multi-AS topology. 570 For example, the ITR for a particular end-to-end packet exchange 571 might be the first-hop or default router within a site for the source 572 host. Similarly, the egress tunnel router might be the last-hop 573 router directly-connected to the destination host. Another example, 574 perhaps for a VPN service out-sourced to an ISP by a site, the ITR 575 could be the site's border router at the service provider attachment 576 point. Mixing and matching of site-operated, ISP-operated, and other 577 tunnel routers is allowed for maximum flexibility. See Section 8 for 578 more details. 580 4.1. Packet Flow Sequence 582 This section provides an example of the unicast packet flow with the 583 following conditions: 585 o Source host "host1.abc.example.com" is sending a packet to 586 "host2.xyz.example.com", exactly what host1 would do if the site 587 was not using LISP. 589 o Each site is multi-homed, so each tunnel router has an address 590 (RLOC) assigned from the service provider address block for each 591 provider to which that particular tunnel router is attached. 593 o The ITR(s) and ETR(s) are directly connected to the source and 594 destination, respectively, but the source and destination can be 595 located anywhere in LISP site. 597 o Map-Requests can be sent on the underlying routing system 598 topology, to a mapping database system, or directly over an 599 alternative topology [ALT]. A Map-Request is sent for an external 600 destination when the destination is not found in the forwarding 601 table or matches a default route. 603 o Map-Replies are sent on the underlying routing system topology. 605 Client host1.abc.example.com wants to communicate with server 606 host2.xyz.example.com: 608 1. host1.abc.example.com wants to open a TCP connection to 609 host2.xyz.example.com. It does a DNS lookup on 610 host2.xyz.example.com. An A/AAAA record is returned. This 611 address is the destination EID. The locally-assigned address of 612 host1.abc.example.com is used as the source EID. An IPv4 or IPv6 613 packet is built and forwarded through the LISP site as a normal 614 IP packet until it reaches a LISP ITR. 616 2. The LISP ITR must be able to map the destination EID to an RLOC 617 of one of the ETRs at the destination site. The specific method 618 used to do this is not described in this example. See [ALT] or 619 [CONS] for possible solutions. 621 3. The ITR will send a LISP Map-Request. Map-Requests SHOULD be 622 rate-limited. 624 4. When an alternate mapping system is not in use, the Map-Request 625 packet is routed through the underlying routing system. 626 Otherwise, the Map-Request packet is routed on an alternate 627 logical topology, for example the [ALT] database mapping system. 628 In either case, when the Map-Request arrives at one of the ETRs 629 at the destination site, it will process the packet as a control 630 message. 632 5. The ETR looks at the destination EID of the Map-Request and 633 matches it against the prefixes in the ETR's configured EID-to- 634 RLOC mapping database. This is the list of EID-prefixes the ETR 635 is supporting for the site it resides in. If there is no match, 636 the Map-Request is dropped. Otherwise, a LISP Map-Reply is 637 returned to the ITR. 639 6. The ITR receives the Map-Reply message, parses the message (to 640 check for format validity) and stores the mapping information 641 from the packet. This information is stored in the ITR's EID-to- 642 RLOC mapping cache. Note that the map cache is an on-demand 643 cache. An ITR will manage its map cache in such a way that 644 optimizes for its resource constraints. 646 7. Subsequent packets from host1.abc.example.com to 647 host2.xyz.example.com will have a LISP header prepended by the 648 ITR using the appropriate RLOC as the LISP header destination 649 address learned from the ETR. Note the packet MAY be sent to a 650 different ETR than the one which returned the Map-Reply due to 651 the source site's hashing policy or the destination site's 652 locator-set policy. 654 8. The ETR receives these packets directly (since the destination 655 address is one of its assigned IP addresses), checks the validity 656 of the addresses, strips the LISP header, and forwards packets to 657 the attached destination host. 659 In order to defer the need for a mapping lookup in the reverse 660 direction, an ETR MAY create a cache entry that maps the source EID 661 (inner header source IP address) to the source RLOC (outer header 662 source IP address) in a received LISP packet. Such a cache entry is 663 termed a "gleaned" mapping and only contains a single RLOC for the 664 EID in question. More complete information about additional RLOCs 665 SHOULD be verified by sending a LISP Map-Request for that EID. Both 666 ITR and the ETR may also influence the decision the other makes in 667 selecting an RLOC. See Section 6 for more details. 669 5. LISP Encapsulation Details 671 Since additional tunnel headers are prepended, the packet becomes 672 larger and can exceed the MTU of any link traversed from the ITR to 673 the ETR. It is RECOMMENDED in IPv4 that packets do not get 674 fragmented as they are encapsulated by the ITR. Instead, the packet 675 is dropped and an ICMP Too Big message is returned to the source. 677 This specification RECOMMENDS that implementations provide support 678 for one of the proposed fragmentation and reassembly schemes. Two 679 existing schemes are detailed in Section 5.4. 681 Since IPv4 or IPv6 addresses can be either EIDs or RLOCs, the LISP 682 architecture supports IPv4 EIDs with IPv6 RLOCs (where the inner 683 header is in IPv4 packet format and the other header is in IPv6 684 packet format) or IPv6 EIDs with IPv4 RLOCs (where the inner header 685 is in IPv6 packet format and the other header is in IPv4 packet 686 format). The next sub-sections illustrate packet formats for the 687 homogeneous case (IPv4-in-IPv4 and IPv6-in-IPv6) but all 4 688 combinations MUST be supported. 690 5.1. LISP IPv4-in-IPv4 Header Format 692 0 1 2 3 693 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 1 694 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 695 / |Version| IHL |Type of Service| Total Length | 696 / +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 697 | | Identification |Flags| Fragment Offset | 698 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 699 OH | Time to Live | Protocol = 17 | Header Checksum | 700 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 701 | | Source Routing Locator | 702 \ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 703 \ | Destination Routing Locator | 704 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 705 / | Source Port = xxxx | Dest Port = 4341 | 706 UDP +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 707 \ | UDP Length | UDP Checksum | 708 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 709 L |N|L|E|V|I|flags| Nonce/Map-Version | 710 I \ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 711 S / | Instance ID/Locator Status Bits | 712 P +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 713 / |Version| IHL |Type of Service| Total Length | 714 / +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 715 | | Identification |Flags| Fragment Offset | 716 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 717 IH | Time to Live | Protocol | Header Checksum | 718 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 719 | | Source EID | 720 \ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 721 \ | Destination EID | 722 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 724 5.2. LISP IPv6-in-IPv6 Header Format 726 0 1 2 3 727 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 1 728 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 729 / |Version| Traffic Class | Flow Label | 730 / +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 732 | | Payload Length | Next Header=17| Hop Limit | 733 v +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 734 | | 735 O + + 736 u | | 737 t + Source Routing Locator + 738 e | | 739 r + + 740 | | 741 H +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 742 d | | 743 r + + 744 | | 745 ^ + Destination Routing Locator + 746 | | | 747 \ + + 748 \ | | 749 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 750 / | Source Port = xxxx | Dest Port = 4341 | 751 UDP +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 752 \ | UDP Length | UDP Checksum | 753 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 754 L |N|L|E|V|I|flags| Nonce/Map-Version | 755 I \ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 756 S / | Instance ID/Locator Status Bits | 757 P +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 758 / |Version| Traffic Class | Flow Label | 759 / +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 760 / | Payload Length | Next Header | Hop Limit | 761 v +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 762 | | 763 I + + 764 n | | 765 n + Source EID + 766 e | | 767 r + + 768 | | 769 H +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 770 d | | 771 r + + 772 | | 773 ^ + Destination EID + 774 \ | | 775 \ + + 776 \ | | 777 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 779 5.3. Tunnel Header Field Descriptions 781 Inner Header (IH): The inner header is the header on the datagram 782 received from the originating host. The source and destination IP 783 addresses are EIDs, [RFC0791], [RFC2460]. 785 Outer Header: (OH) The outer header is a new header prepended by an 786 ITR. The address fields contain RLOCs obtained from the ingress 787 router's EID-to-RLOC cache. The IP protocol number is "UDP (17)" 788 from [RFC0768]. The setting of the DF bit Flags field is 789 according to rules in Section 5.4.1 and Section 5.4.2. 791 UDP Header: The UDP header contains an ITR selected source port when 792 encapsulating a packet. See Section 6.5 for details on the hash 793 algorithm used to select a source port based on the 5-tuple of the 794 inner header. The destination port MUST be set to the well-known 795 IANA assigned port value 4341. 797 UDP Checksum: The UDP checksum field SHOULD be transmitted as zero 798 by an ITR for either IPv4 [RFC0768] or IPv6 encapsulation 799 [UDP-TUNNELS] [UDP-ZERO]. When a packet with a zero UDP checksum 800 is received by an ETR, the ETR MUST accept the packet for 801 decapsulation. When an ITR transmits a non-zero value for the UDP 802 checksum, it MUST send a correctly computed value in this field. 803 When an ETR receives a packet with a non-zero UDP checksum, it MAY 804 choose to verify the checksum value. If it chooses to perform 805 such verification, and the verification fails, the packet MUST be 806 silently dropped. If the ETR chooses not to perform the 807 verification, or performs the verification successfully, the 808 packet MUST be accepted for decapsulation. The handling of UDP 809 checksums for all tunneling protocols, including LISP, is under 810 active discussion within the IETF. When that discussion 811 concludes, any necessary changes will be made to align LISP with 812 the outcome of the broader discussion. 814 UDP Length: The UDP length field is set for an IPv4 encapsulated 815 packet to be the sum of the inner header IPv4 Total Length plus 816 the UDP and LISP header lengths. For an IPv6 encapsulated packet, 817 the UDP length field is the sum of the inner header IPv6 Payload 818 Length, the size of the IPv6 header (40 octets), and the size of 819 the UDP and LISP headers. 821 N: The N bit is the nonce-present bit. When this bit is set to 1, 822 the low-order 24-bits of the first 32-bits of the LISP header 823 contains a Nonce. See Section 6.3.1 for details. Both N and V 824 bits MUST NOT be set in the same packet. If they are, a 825 decapsulating ETR MUST treat the "Nonce/Map-Version" field as 826 having a Nonce value present. 828 L: The L bit is the Locator Status Bits field enabled bit. When this 829 bit is set to 1, the Locator Status Bits in the second 32-bits of 830 the LISP header are in use. 832 x 1 x x 0 x x x 833 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 834 |N|L|E|V|I|flags| Nonce/Map-Version | 835 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 836 | Locator Status Bits | 837 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 839 E: The E bit is the echo-nonce-request bit. This bit MUST be ignored 840 and has no meaning when the N bit is set to 0. When the N bit is 841 set to 1 and this bit is set to 1, means an ITR is requesting for 842 the nonce value in the Nonce field to be echoed back in LISP 843 encapsulated packets when the ITR is also an ETR. See 844 Section 6.3.1 for details. 846 V: The V bit is the Map-Version present bit. When this bit is set to 847 1, the N bit MUST be 0. Refer to Section 6.6.3 for more details. 848 This bit indicates that the LISP header is encoded in this case 849 as: 851 0 x 0 1 x x x x 852 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 853 |N|L|E|V|I|flags| Source Map-Version | Dest Map-Version | 854 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 855 | Instance ID/Locator Status Bits | 856 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 858 I: The I bit is the Instance ID bit. See Section 5.5 for more 859 details. When this bit is set to 1, the Locator Status Bits field 860 is reduced to 8-bits and the high-order 24-bits are used as an 861 Instance ID. If the L-bit is set to 0, then the low-order 8 bits 862 are transmitted as zero and ignored on receipt. The format of the 863 LISP header would look like in this case: 865 x x x x 1 x x x 866 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 867 |N|L|E|V|I|flags| Nonce/Map-Version | 868 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 869 | Instance ID | LSBs | 870 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 872 flags: The flags field is a 3-bit field is reserved for future flag 873 use. It MUST be set to 0 on transmit and MUST be ignored on 874 receipt. 876 LISP Nonce: The LISP nonce field is a 24-bit value that is randomly 877 generated by an ITR when the N-bit is set to 1. Nonce generation 878 algorithms are an implementation matter but are required to 879 generate different nonces when sending to different destinations. 880 However, the same nonce can be used for a period of time to the 881 same destination. The nonce is also used when the E-bit is set to 882 request the nonce value to be echoed by the other side when 883 packets are returned. When the E-bit is clear but the N-bit is 884 set, a remote ITR is either echoing a previously requested echo- 885 nonce or providing a random nonce. See Section 6.3.1 for more 886 details. 888 LISP Locator Status Bits (LSBs): When the L-bit is also set, the 889 locator status bits field in the LISP header is set by an ITR to 890 indicate to an ETR the up/down status of the Locators in the 891 source site. Each RLOC in a Map-Reply is assigned an ordinal 892 value from 0 to n-1 (when there are n RLOCs in a mapping entry). 893 The Locator Status Bits are numbered from 0 to n-1 from the least 894 significant bit of field. The field is 32-bits when the I-bit is 895 set to 0 and is 8 bits when the I-bit is set to 1. When a Locator 896 Status Bit is set to 1, the ITR is indicating to the ETR the RLOC 897 associated with the bit ordinal has up status. See Section 6.3 898 for details on how an ITR can determine the status of the ETRs at 899 the same site. When a site has multiple EID-prefixes which result 900 in multiple mappings (where each could have a different locator- 901 set), the Locator Status Bits setting in an encapsulated packet 902 MUST reflect the mapping for the EID-prefix that the inner-header 903 source EID address matches. If the LSB for an anycast locator is 904 set to 1, then there is at least one RLOC with that address the 905 ETR is considered 'up'. 907 When doing ITR/PITR encapsulation: 909 o The outer header Time to Live field (or Hop Limit field, in case 910 of IPv6) SHOULD be copied from the inner header Time to Live 911 field. 913 o The outer header Type of Service field (or the Traffic Class 914 field, in the case of IPv6) SHOULD be copied from the inner header 915 Type of Service field (with one exception, see below). 917 When doing ETR/PETR decapsulation: 919 o The inner header Time to Live field (or Hop Limit field, in case 920 of IPv6) SHOULD be copied from the outer header Time to Live 921 field, when the Time to Live field of the outer header is less 922 than the Time to Live of the inner header. Failing to perform 923 this check can cause the Time to Live of the inner header to 924 increment across encapsulation/decapsulation cycle. This check is 925 also performed when doing initial encapsulation when a packet 926 comes to an ITR or PITR destined for a LISP site. 928 o The inner header Type of Service field (or the Traffic Class 929 field, in the case of IPv6) SHOULD be copied from the outer header 930 Type of Service field (with one exception, see below). 932 Note if an ETR/PETR is also an ITR/PITR and choose to reencapsulate 933 after decapsulating, the net effect of this is that the new outer 934 header will carry the same Time to Live as the old outer header minus 935 1. 937 Copying the TTL serves two purposes: first, it preserves the distance 938 the host intended the packet to travel; second, and more importantly, 939 it provides for suppression of looping packets in the event there is 940 a loop of concatenated tunnels due to misconfiguration. See 941 Section 9.3 for TTL exception handling for traceroute packets. 943 The ECN field occupies bits 6 and 7 of both the IPv4 Type of Service 944 field and the IPv6 Traffic Class field [RFC3168]. The ECN field 945 requires special treatment in order to avoid discarding indications 946 of congestion [RFC3168]. ITR encapsulation MUST copy the 2-bit ECN 947 field from the inner header to the outer header. Re-encapsulation 948 MUST copy the 2-bit ECN field from the stripped outer header to the 949 new outer header. If the ECN field contains a congestion indication 950 codepoint (the value is '11', the Congestion Experienced (CE) 951 codepoint), then ETR decapsulation MUST copy the 2-bit ECN field from 952 the stripped outer header to the surviving inner header that is used 953 to forward the packet beyond the ETR. These requirements preserve 954 Congestion Experienced (CE) indications when a packet that uses ECN 955 traverses a LISP tunnel and becomes marked with a CE indication due 956 to congestion between the tunnel endpoints. 958 5.4. Dealing with Large Encapsulated Packets 960 This section proposes two mechanisms to deal with packets that exceed 961 the path MTU between the ITR and ETR. 963 It is left to the implementor to decide if the stateless or stateful 964 mechanism should be implemented. Both or neither can be used since 965 it is a local decision in the ITR regarding how to deal with MTU 966 issues, and sites can interoperate with differing mechanisms. 968 Both stateless and stateful mechanisms also apply to Reencapsulating 969 and Recursive Tunneling. So any actions below referring to an ITR 970 also apply to an TE-ITR. 972 5.4.1. A Stateless Solution to MTU Handling 974 An ITR stateless solution to handle MTU issues is described as 975 follows: 977 1. Define H to be the size, in octets, of the outer header an ITR 978 prepends to a packet. This includes the UDP and LISP header 979 lengths. 981 2. Define L to be the size, in octets, of the maximum sized packet 982 an ITR can send to an ETR without the need for the ITR or any 983 intermediate routers to fragment the packet. 985 3. Define an architectural constant S for the maximum size of a 986 packet, in octets, an ITR must receive so the effective MTU can 987 be met. That is, S = L - H. 989 When an ITR receives a packet from a site-facing interface and adds H 990 octets worth of encapsulation to yield a packet size greater than L 991 octets, it resolves the MTU issue by first splitting the original 992 packet into 2 equal-sized fragments. A LISP header is then prepended 993 to each fragment. The size of the encapsulated fragments is then 994 (S/2 + H), which is less than the ITR's estimate of the path MTU 995 between the ITR and its correspondent ETR. 997 When an ETR receives encapsulated fragments, it treats them as two 998 individually encapsulated packets. It strips the LISP headers then 999 forwards each fragment to the destination host of the destination 1000 site. The two fragments are reassembled at the destination host into 1001 the single IP datagram that was originated by the source host. Note 1002 that reassembly can happen at the ETR if the encapsulated packet was 1003 fragmented at or after the ITR. 1005 This behavior is performed by the ITR when the source host originates 1006 a packet with the DF field of the IP header is set to 0. When the DF 1007 field of the IP header is set to 1, or the packet is an IPv6 packet 1008 originated by the source host, the ITR will drop the packet when the 1009 size is greater than L, and sends an ICMP Too Big message to the 1010 source with a value of S, where S is (L - H). 1012 When the outer header encapsulation uses an IPv4 header, an 1013 implementation SHOULD set the DF bit to 1 so ETR fragment reassembly 1014 can be avoided. An implementation MAY set the DF bit in such headers 1015 to 0 if it has good reason to believe there are unresolvable path MTU 1016 issues between the sending ITR and the receiving ETR. 1018 This specification RECOMMENDS that L be defined as 1500. 1020 5.4.2. A Stateful Solution to MTU Handling 1022 An ITR stateful solution to handle MTU issues is described as follows 1023 and was first introduced in [OPENLISP]: 1025 1. The ITR will keep state of the effective MTU for each locator per 1026 mapping cache entry. The effective MTU is what the core network 1027 can deliver along the path between ITR and ETR. 1029 2. When an IPv6 encapsulated packet or an IPv4 encapsulated packet 1030 with DF bit set to 1, exceeds what the core network can deliver, 1031 one of the intermediate routers on the path will send an ICMP Too 1032 Big message to the ITR. The ITR will parse the ICMP message to 1033 determine which locator is affected by the effective MTU change 1034 and then record the new effective MTU value in the mapping cache 1035 entry. 1037 3. When a packet is received by the ITR from a source inside of the 1038 site and the size of the packet is greater than the effective MTU 1039 stored with the mapping cache entry associated with the 1040 destination EID the packet is for, the ITR will send an ICMP Too 1041 Big message back to the source. The packet size advertised by 1042 the ITR in the ICMP Too Big message is the effective MTU minus 1043 the LISP encapsulation length. 1045 Even though this mechanism is stateful, it has advantages over the 1046 stateless IP fragmentation mechanism, by not involving the 1047 destination host with reassembly of ITR fragmented packets. 1049 5.5. Using Virtualization and Segmentation with LISP 1051 When multiple organizations inside of a LISP site are using private 1052 addresses [RFC1918] as EID-prefixes, their address spaces MUST remain 1053 segregated due to possible address duplication. An Instance ID in 1054 the address encoding can aid in making the entire AFI based address 1055 unique. See IANA Considerations Section 14.2 for details for 1056 possible address encodings. 1058 An Instance ID can be carried in a LISP encapsulated packet. An ITR 1059 that prepends a LISP header, will copy a 24-bit value, used by the 1060 LISP router to uniquely identify the address space. The value is 1061 copied to the Instance ID field of the LISP header and the I-bit is 1062 set to 1. 1064 When an ETR decapsulates a packet, the Instance ID from the LISP 1065 header is used as a table identifier to locate the forwarding table 1066 to use for the inner destination EID lookup. 1068 For example, a 802.1Q VLAN tag or VPN identifier could be used as a 1069 24-bit Instance ID. 1071 6. EID-to-RLOC Mapping 1073 6.1. LISP IPv4 and IPv6 Control Plane Packet Formats 1075 The following UDP packet formats are used by the LISP control-plane. 1077 0 1 2 3 1078 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 1079 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1080 |Version| IHL |Type of Service| Total Length | 1081 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1082 | Identification |Flags| Fragment Offset | 1083 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1084 | Time to Live | Protocol = 17 | Header Checksum | 1085 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1086 | Source Routing Locator | 1087 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1088 | Destination Routing Locator | 1089 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1090 / | Source Port | Dest Port | 1091 UDP +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1092 \ | UDP Length | UDP Checksum | 1093 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1094 | | 1095 | LISP Message | 1096 | | 1097 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1099 0 1 2 3 1100 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 1101 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1102 |Version| Traffic Class | Flow Label | 1103 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1104 | Payload Length | Next Header=17| Hop Limit | 1105 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1106 | | 1107 + + 1108 | | 1109 + Source Routing Locator + 1110 | | 1111 + + 1112 | | 1113 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1114 | | 1115 + + 1116 | | 1117 + Destination Routing Locator + 1118 | | 1119 + + 1120 | | 1121 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1122 / | Source Port | Dest Port | 1123 UDP +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1124 \ | UDP Length | UDP Checksum | 1125 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1126 | | 1127 | LISP Message | 1128 | | 1129 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1131 The LISP UDP-based messages are the Map-Request and Map-Reply 1132 messages. When a UDP Map-Request is sent, the UDP source port is 1133 chosen by the sender and the destination UDP port number is set to 1134 4342. When a UDP Map-Reply is sent, the source UDP port number is 1135 set to 4342 and the destination UDP port number is copied from the 1136 source port of either the Map-Request or the invoking data packet. 1137 Implementations MUST be prepared to accept packets when either the 1138 source port or destination UDP port is set to 4342 due to NATs 1139 changing port number values. 1141 The UDP Length field will reflect the length of the UDP header and 1142 the LISP Message payload. 1144 The UDP Checksum is computed and set to non-zero for Map-Request, 1145 Map-Reply, Map-Register and ECM control messages. It MUST be checked 1146 on receipt and if the checksum fails, the packet MUST be dropped. 1148 The format of control messages includes the UDP header so the 1149 checksum and length fields can be used to protect and delimit message 1150 boundaries. 1152 6.1.1. LISP Packet Type Allocations 1154 This section will be the authoritative source for allocating LISP 1155 Type values and for defining LISP control message formats. Current 1156 allocations are: 1158 Reserved: 0 b'0000' 1159 LISP Map-Request: 1 b'0001' 1160 LISP Map-Reply: 2 b'0010' 1161 LISP Map-Register: 3 b'0011' 1162 LISP Map-Notify: 4 b'0100' 1163 LISP Encapsulated Control Message: 8 b'1000' 1165 6.1.2. Map-Request Message Format 1167 0 1 2 3 1168 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 1169 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1170 |Type=1 |A|M|P|S|p|s| Reserved | IRC | Record Count | 1171 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1172 | Nonce . . . | 1173 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1174 | . . . Nonce | 1175 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1176 | Source-EID-AFI | Source EID Address ... | 1177 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1178 | ITR-RLOC-AFI 1 | ITR-RLOC Address 1 ... | 1179 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1180 | ... | 1181 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1182 | ITR-RLOC-AFI n | ITR-RLOC Address n ... | 1183 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1184 / | Reserved | EID mask-len | EID-prefix-AFI | 1185 Rec +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1186 \ | EID-prefix ... | 1187 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1188 | Map-Reply Record ... | 1189 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1191 Packet field descriptions: 1193 Type: 1 (Map-Request) 1195 A: This is an authoritative bit, which is set to 0 for UDP-based Map- 1196 Requests sent by an ITR. Set to 1 when an ITR wants the 1197 destination site to return the Map-Reply rather than the mapping 1198 database system. 1200 M: This is the map-data-present bit, when set, it indicates a Map- 1201 Reply Record segment is included in the Map-Request. 1203 P: This is the probe-bit which indicates that a Map-Request SHOULD be 1204 treated as a locator reachability probe. The receiver SHOULD 1205 respond with a Map-Reply with the probe-bit set, indicating the 1206 Map-Reply is a locator reachability probe reply, with the nonce 1207 copied from the Map-Request. See Section 6.3.2 for more details. 1209 S: This is the Solicit-Map-Request (SMR) bit. See Section 6.6.2 for 1210 details. 1212 p: This is the PITR bit. This bit is set to 1 when a PITR sends a 1213 Map-Request. 1215 s: This is the SMR-invoked bit. This bit is set to 1 when an xTR is 1216 sending a Map-Request in response to a received SMR-based Map- 1217 Request. 1219 Reserved: It MUST be set to 0 on transmit and MUST be ignored on 1220 receipt. 1222 IRC: This 5-bit field is the ITR-RLOC Count which encodes the 1223 additional number of (ITR-RLOC-AFI, ITR-RLOC Address) fields 1224 present in this message. At least one (ITR-RLOC-AFI, ITR-RLOC- 1225 Address) pair MUST be encoded. Multiple ITR-RLOC Address fields 1226 are used so a Map-Replier can select which destination address to 1227 use for a Map-Reply. The IRC value ranges from 0 to 31. For a 1228 value of 0, there is 1 ITR-RLOC address encoded, and for a value 1229 of 1, there are 2 ITR-RLOC addresses encoded and so on up to 31 1230 which encodes a total of 32 ITR-RLOC addresses. 1232 Record Count: The number of records in this Map-Request message. A 1233 record is comprised of the portion of the packet that is labeled 1234 'Rec' above and occurs the number of times equal to Record Count. 1235 For this version of the protocol, a receiver MUST accept and 1236 process Map-Requests that contain one or more records, but a 1237 sender MUST only send Map-Requests containing one record. Support 1238 for requesting multiple EIDs in a single Map-Request message will 1239 be specified in a future version of the protocol. 1241 Nonce: An 8-octet random value created by the sender of the Map- 1242 Request. This nonce will be returned in the Map-Reply. The 1243 security of the LISP mapping protocol depends critically on the 1244 strength of the nonce in the Map-Request message. The nonce 1245 SHOULD be generated by a properly seeded pseudo-random (or strong 1246 random) source. See [RFC4086] for advice on generating security- 1247 sensitive random data. 1249 Source-EID-AFI: Address family of the "Source EID Address" field. 1251 Source EID Address: This is the EID of the source host which 1252 originated the packet which is caused the Map-Request. When Map- 1253 Requests are used for refreshing a map-cache entry or for RLOC- 1254 probing, an AFI value 0 is used and this field is of zero length. 1256 ITR-RLOC-AFI: Address family of the "ITR-RLOC Address" field that 1257 follows this field. 1259 ITR-RLOC Address: Used to give the ETR the option of selecting the 1260 destination address from any address family for the Map-Reply 1261 message. This address MUST be a routable RLOC address of the 1262 sender of the Map-Request message. 1264 EID mask-len: Mask length for EID prefix. 1266 EID-prefix-AFI: Address family of EID-prefix according to [AFI] 1268 EID-prefix: 4 octets if an IPv4 address-family, 16 octets if an IPv6 1269 address-family. When a Map-Request is sent by an ITR because a 1270 data packet is received for a destination where there is no 1271 mapping entry, the EID-prefix is set to the destination IP address 1272 of the data packet. And the 'EID mask-len' is set to 32 or 128 1273 for IPv4 or IPv6, respectively. When an xTR wants to query a site 1274 about the status of a mapping it already has cached, the EID- 1275 prefix used in the Map-Request has the same mask-length as the 1276 EID-prefix returned from the site when it sent a Map-Reply 1277 message. 1279 Map-Reply Record: When the M bit is set, this field is the size of a 1280 single "Record" in the Map-Reply format. This Map-Reply record 1281 contains the EID-to-RLOC mapping entry associated with the Source 1282 EID. This allows the ETR which will receive this Map-Request to 1283 cache the data if it chooses to do so. 1285 6.1.3. EID-to-RLOC UDP Map-Request Message 1287 A Map-Request is sent from an ITR when it needs a mapping for an EID, 1288 wants to test an RLOC for reachability, or wants to refresh a mapping 1289 before TTL expiration. For the initial case, the destination IP 1290 address used for the Map-Request is the data packet's destination 1291 address (i.e. the destination-EID) which had a mapping cache lookup 1292 failure. For the latter two cases, the destination IP address used 1293 for the Map-Request is one of the RLOC addresses from the locator-set 1294 of the map cache entry. The source address is either an IPv4 or IPv6 1295 RLOC address depending if the Map-Request is using an IPv4 versus 1296 IPv6 header, respectively. In all cases, the UDP source port number 1297 for the Map-Request message is an ITR/PITR selected 16-bit value and 1298 the UDP destination port number is set to the well-known destination 1299 port number 4342. A successful Map-Reply, which is one that has a 1300 nonce that matches an outstanding Map-Request nonce, will update the 1301 cached set of RLOCs associated with the EID prefix range. 1303 One or more Map-Request (ITR-RLOC-AFI, ITR-RLOC-Address) fields MUST 1304 be filled in by the ITR. The number of fields (minus 1) encoded MUST 1305 be placed in the IRC field. The ITR MAY include all locally 1306 configured locators in this list or just provide one locator address 1307 from each address family it supports. If the ITR erroneously 1308 provides no ITR-RLOC addresses, the Map-Replier MUST drop the Map- 1309 Request. 1311 Map-Requests can also be LISP encapsulated using UDP destination port 1312 4342 with a LISP type value set to "Encapsulated Control Message", 1313 when sent from an ITR to a Map-Resolver. Likewise, Map-Requests are 1314 LISP encapsulated the same way from a Map-Server to an ETR. Details 1315 on encapsulated Map-Requests and Map-Resolvers can be found in 1316 [LISP-MS]. 1318 Map-Requests MUST be rate-limited. It is RECOMMENDED that a Map- 1319 Request for the same EID-prefix be sent no more than once per second. 1321 An ITR that is configured with mapping database information (i.e. it 1322 is also an ETR) MAY optionally include those mappings in a Map- 1323 Request. When an ETR configured to accept and verify such 1324 "piggybacked" mapping data receives such a Map-Request and it does 1325 not have this mapping in the map-cache, it MAY originate a "verifying 1326 Map-Request", addressed to the map-requesting ITR and the ETR MAY add 1327 a map-cache entry. If the ETR has a map-cache entry that matches the 1328 "piggybacked" EID and the RLOC is in the locator-set for the entry, 1329 then it may send the "verifying Map-Request" directly to the 1330 originating Map-Request source. If the RLOC is not in the locator- 1331 set, then the ETR MUST send the "verifying Map-Request" to the 1332 "piggybacked" EID. Doing this forces the "verifying Map-Request" to 1333 go through the mapping database system to reach the authoritative 1334 source of information about that EID, guarding against RLOC-spoofing 1335 in in the "piggybacked" mapping data. 1337 6.1.4. Map-Reply Message Format 1339 0 1 2 3 1340 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 1341 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1342 |Type=2 |P|E|S| Reserved | Record Count | 1343 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1344 | Nonce . . . | 1345 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1346 | . . . Nonce | 1347 +-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1348 | | Record TTL | 1349 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1350 R | Locator Count | EID mask-len | ACT |A| Reserved | 1351 e +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1352 c | Rsvd | Map-Version Number | EID-prefix-AFI | 1353 o +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1354 r | EID-prefix | 1355 d +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1356 | /| Priority | Weight | M Priority | M Weight | 1357 | L +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1358 | o | Unused Flags |L|p|R| Loc-AFI | 1359 | c +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1360 | \| Locator | 1361 +-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1363 Packet field descriptions: 1365 Type: 2 (Map-Reply) 1367 P: This is the probe-bit which indicates that the Map-Reply is in 1368 response to a locator reachability probe Map-Request. The nonce 1369 field MUST contain a copy of the nonce value from the original 1370 Map-Request. See Section 6.3.2 for more details. 1372 E: Indicates that the ETR which sends this Map-Reply message is 1373 advertising that the site is enabled for the Echo-Nonce locator 1374 reachability algorithm. See Section 6.3.1 for more details. 1376 S: This is the Security bit. When set to 1 the following 1377 authentication information will be appended to the end of the Map- 1378 Reply. The detailed format of the Authentication Data Content is 1379 for further study. 1381 0 1 2 3 1382 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 1383 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1384 | AD Type | Authentication Data Content . . . | 1385 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1387 Reserved: It MUST be set to 0 on transmit and MUST be ignored on 1388 receipt. 1390 Record Count: The number of records in this reply message. A record 1391 is comprised of that portion of the packet labeled 'Record' above 1392 and occurs the number of times equal to Record count. 1394 Nonce: A 24-bit value set in a Data-Probe packet or a 64-bit value 1395 from the Map-Request is echoed in this Nonce field of the Map- 1396 Reply. When a 24-bit value is supplied, it resides in the low- 1397 order 64 bits of the nonce field. 1399 Record TTL: The time in minutes the recipient of the Map-Reply will 1400 store the mapping. If the TTL is 0, the entry SHOULD be removed 1401 from the cache immediately. If the value is 0xffffffff, the 1402 recipient can decide locally how long to store the mapping. 1404 Locator Count: The number of Locator entries. A locator entry 1405 comprises what is labeled above as 'Loc'. The locator count can 1406 be 0 indicating there are no locators for the EID-prefix. 1408 EID mask-len: Mask length for EID prefix. 1410 ACT: This 3-bit field describes negative Map-Reply actions. In any 1411 other message type, these bits are set to 0 and ignored on 1412 receipt. These bits are used only when the 'Locator Count' field 1413 is set to 0. The action bits are encoded only in Map-Reply 1414 messages. The actions defined are used by an ITR or PITR when a 1415 destination EID matches a negative mapping cache entry. 1416 Unassigned values should cause a map-cache entry to be created 1417 and, when packets match this negative cache entry, they will be 1418 dropped. The current assigned values are: 1420 (0) No-Action: The map-cache is kept alive and no packet 1421 encapsulation occurs. 1423 (1) Natively-Forward: The packet is not encapsulated or dropped 1424 but natively forwarded. 1426 (2) Send-Map-Request: The packet invokes sending a Map-Request. 1428 (3) Drop: A packet that matches this map-cache entry is dropped. 1429 An ICMP Unreachable message SHOULD be sent. 1431 A: The Authoritative bit, when sent is always set to 1 by an ETR. 1432 When a Map-Server is proxy Map-Replying [LISP-MS] for a LISP site, 1433 the Authoritative bit is set to 0. This indicates to requesting 1434 ITRs that the Map-Reply was not originated by a LISP node managed 1435 at the site that owns the EID-prefix. 1437 Map-Version Number: When this 12-bit value is non-zero the Map-Reply 1438 sender is informing the ITR what the version number is for the 1439 EID-record contained in the Map-Reply. The ETR can allocate this 1440 number internally but MUST coordinate this value with other ETRs 1441 for the site. When this value is 0, there is no versioning 1442 information conveyed. The Map-Version Number can be included in 1443 Map-Request and Map-Register messages. See Section 6.6.3 for more 1444 details. 1446 EID-prefix-AFI: Address family of EID-prefix according to [AFI]. 1448 EID-prefix: 4 octets if an IPv4 address-family, 16 octets if an IPv6 1449 address-family. 1451 Priority: each RLOC is assigned a unicast priority. Lower values 1452 are more preferable. When multiple RLOCs have the same priority, 1453 they MAY be used in a load-split fashion. A value of 255 means 1454 the RLOC MUST NOT be used for unicast forwarding. 1456 Weight: when priorities are the same for multiple RLOCs, the weight 1457 indicates how to balance unicast traffic between them. Weight is 1458 encoded as a relative weight of total unicast packets that match 1459 the mapping entry. For example if there are 4 locators in a 1460 locator set, where the weights assigned are 30, 20, 20, and 10, 1461 the first locator will get 37.5% of the traffic, the 2nd and 3rd 1462 locators will get 25% of traffic and the 4th locator will get 1463 12.5% of the traffic. If all weights for a locator-set are equal, 1464 receiver of the Map-Reply will decide how to load-split traffic. 1465 See Section 6.5 for a suggested hash algorithm to distribute load 1466 across locators with same priority and equal weight values. 1468 M Priority: each RLOC is assigned a multicast priority used by an 1469 ETR in a receiver multicast site to select an ITR in a source 1470 multicast site for building multicast distribution trees. A value 1471 of 255 means the RLOC MUST NOT be used for joining a multicast 1472 distribution tree. For more details, see [MLISP]. 1474 M Weight: when priorities are the same for multiple RLOCs, the 1475 weight indicates how to balance building multicast distribution 1476 trees across multiple ITRs. The weight is encoded as a relative 1477 weight (similar to the unicast Weights) of total number of trees 1478 built to the source site identified by the EID-prefix. If all 1479 weights for a locator-set are equal, the receiver of the Map-Reply 1480 will decide how to distribute multicast state across ITRs. For 1481 more details, see [MLISP]. 1483 Unused Flags: set to 0 when sending and ignored on receipt. 1485 L: when this bit is set, the locator is flagged as a local locator to 1486 the ETR that is sending the Map-Reply. When a Map-Server is doing 1487 proxy Map-Replying [LISP-MS] for a LISP site, the L bit is set to 1488 0 for all locators in this locator-set. 1490 p: when this bit is set, an ETR informs the RLOC-probing ITR that the 1491 locator address, for which this bit is set, is the one being RLOC- 1492 probed and MAY be different from the source address of the Map- 1493 Reply. An ITR that RLOC-probes a particular locator, MUST use 1494 this locator for retrieving the data structure used to store the 1495 fact that the locator is reachable. The "p" bit is set for a 1496 single locator in the same locator set. If an implementation sets 1497 more than one "p" bit erroneously, the receiver of the Map-Reply 1498 MUST select the first locator. The "p" bit MUST NOT be set for 1499 locator-set records sent in Map-Request and Map-Register messages. 1501 R: set when the sender of a Map-Reply has a route to the locator in 1502 the locator data record. This receiver may find this useful to 1503 know if the locator is up but not necessarily reachable from the 1504 receiver's point of view. See also Section 6.4 for another way 1505 the R-bit may be used. 1507 Locator: an IPv4 or IPv6 address (as encoded by the 'Loc-AFI' field) 1508 assigned to an ETR. Note that the destination RLOC address MAY be 1509 an anycast address. A source RLOC can be an anycast address as 1510 well. The source or destination RLOC MUST NOT be the broadcast 1511 address (255.255.255.255 or any subnet broadcast address known to 1512 the router), and MUST NOT be a link-local multicast address. The 1513 source RLOC MUST NOT be a multicast address. The destination RLOC 1514 SHOULD be a multicast address if it is being mapped from a 1515 multicast destination EID. 1517 6.1.5. EID-to-RLOC UDP Map-Reply Message 1519 A Map-Reply returns an EID-prefix with a prefix length that is less 1520 than or equal to the EID being requested. The EID being requested is 1521 either from the destination field of an IP header of a Data-Probe or 1522 the EID record of a Map-Request. The RLOCs in the Map-Reply are 1523 globally-routable IP addresses of all ETRs for the LISP site. Each 1524 RLOC conveys status reachability but does not convey path 1525 reachability from a requesters perspective. Separate testing of path 1526 reachability is required, See Section 6.3 for details. 1528 Note that a Map-Reply may contain different EID-prefix granularity 1529 (prefix + length) than the Map-Request which triggers it. This might 1530 occur if a Map-Request were for a prefix that had been returned by an 1531 earlier Map-Reply. In such a case, the requester updates its cache 1532 with the new prefix information and granularity. For example, a 1533 requester with two cached EID-prefixes that are covered by a Map- 1534 Reply containing one, less-specific prefix, replaces the entry with 1535 the less-specific EID-prefix. Note that the reverse, replacement of 1536 one less-specific prefix with multiple more-specific prefixes, can 1537 also occur but not by removing the less-specific prefix rather by 1538 adding the more-specific prefixes which during a lookup will override 1539 the less-specific prefix. 1541 When an ETR is configured with overlapping EID-prefixes, a Map- 1542 Request with an EID that longest matches any EID-prefix MUST be 1543 returned in a single Map-Reply message. For instance, if an ETR had 1544 database mapping entries for EID-prefixes: 1546 10.0.0.0/8 1547 10.1.0.0/16 1548 10.1.1.0/24 1549 10.1.2.0/24 1551 A Map-Request for EID 10.1.1.1 would cause a Map-Reply with a record 1552 count of 1 to be returned with a mapping record EID-prefix of 1553 10.1.1.0/24. 1555 A Map-Request for EID 10.1.5.5, would cause a Map-Reply with a record 1556 count of 3 to be returned with mapping records for EID-prefixes 1557 10.1.0.0/16, 10.1.1.0/24, and 10.1.2.0/24. 1559 Note that not all overlapping EID-prefixes need to be returned, only 1560 the more specifics (note in the second example above 10.0.0.0/8 was 1561 not returned for requesting EID 10.1.5.5) entries for the matching 1562 EID-prefix of the requesting EID. When more than one EID-prefix is 1563 returned, all SHOULD use the same Time-to-Live value so they can all 1564 time out at the same time. When a more specific EID-prefix is 1565 received later, its Time-to-Live value in the Map-Reply record can be 1566 stored even when other less specifics exist. When a less specific 1567 EID-prefix is received later, its map-cache expiration time SHOULD be 1568 set to the minimum expiration time of any more specific EID-prefix in 1569 the map-cache. This is done so the integrity of the EID-prefix set 1570 is wholly maintained so no more-specific entries are removed from the 1571 map-cache while keeping less-specific entries. 1573 Map-Replies SHOULD be sent for an EID-prefix no more often than once 1574 per second to the same requesting router. For scalability, it is 1575 expected that aggregation of EID addresses into EID-prefixes will 1576 allow one Map-Reply to satisfy a mapping for the EID addresses in the 1577 prefix range thereby reducing the number of Map-Request messages. 1579 Map-Reply records can have an empty locator-set. A negative Map- 1580 Reply is a Map-Reply with an empty locator-set. Negative Map-Replies 1581 convey special actions by the sender to the ITR or PITR which have 1582 solicited the Map-Reply. There are two primary applications for 1583 Negative Map-Replies. The first is for a Map-Resolver to instruct an 1584 ITR or PITR when a destination is for a LISP site versus a non-LISP 1585 site. And the other is to source quench Map-Requests which are sent 1586 for non-allocated EIDs. 1588 For each Map-Reply record, the list of locators in a locator-set MUST 1589 appear in the same order for each ETR that originates a Map-Reply 1590 message. The locator-set MUST be sorted in order of ascending IP 1591 address where an IPv4 locator address is considered numerically 'less 1592 than' an IPv6 locator address. 1594 When sending a Map-Reply message, the destination address is copied 1595 from the one of the ITR-RLOC fields from the Map-Request. The ETR 1596 can choose a locator address from one of the address families it 1597 supports. For Data-Probes, the destination address of the Map-Reply 1598 is copied from the source address of the Data-Probe message which is 1599 invoking the reply. The source address of the Map-Reply is one of 1600 the local IP addresses chosen to allow uRPF checks to succeed in the 1601 upstream service provider. The destination port of a Map-Reply 1602 message is copied from the source port of the Map-Request or Data- 1603 Probe and the source port of the Map-Reply message is set to the 1604 well-known UDP port 4342. 1606 6.1.5.1. Traffic Redirection with Coarse EID-Prefixes 1608 When an ETR is misconfigured or compromised, it could return coarse 1609 EID-prefixes in Map-Reply messages it sends. The EID-prefix could 1610 cover EID-prefixes which are allocated to other sites redirecting 1611 their traffic to the locators of the compromised site. 1613 To solve this problem, there are two basic solutions that could be 1614 used. The first is to have Map-Servers proxy-map-reply on behalf of 1615 ETRs so their registered EID-prefixes are the ones returned in Map- 1616 Replies. Since the interaction between an ETR and Map-Server is 1617 secured with shared-keys, it is easier for an ETR to detect 1618 misbehavior. The second solution is to have ITRs and PITRs cache 1619 EID-prefixes with mask-lengths that are greater than or equal to a 1620 configured prefix length. This limits the damage to a specific width 1621 of any EID-prefix advertised, but needs to be coordinated with the 1622 allocation of site prefixes. These solutions can be used 1623 independently or at the same time. 1625 At the time of this writing, other approaches are being considered 1626 and researched. 1628 6.1.6. Map-Register Message Format 1630 The usage details of the Map-Register message can be found in 1631 specification [LISP-MS]. This section solely defines the message 1632 format. 1634 The message is sent in UDP with a destination UDP port of 4342 and a 1635 randomly selected UDP source port number. 1637 The Map-Register message format is: 1639 0 1 2 3 1640 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 1641 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1642 |Type=3 |P| Reserved |M| Record Count | 1643 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1644 | Nonce . . . | 1645 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1646 | . . . Nonce | 1647 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1648 | Key ID | Authentication Data Length | 1649 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1650 ~ Authentication Data ~ 1651 +-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1652 | | Record TTL | 1653 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1654 R | Locator Count | EID mask-len | ACT |A| Reserved | 1655 e +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1656 c | Rsvd | Map-Version Number | EID-prefix-AFI | 1657 o +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1658 r | EID-prefix | 1659 d +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1660 | /| Priority | Weight | M Priority | M Weight | 1661 | L +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1662 | o | Unused Flags |L|p|R| Loc-AFI | 1663 | c +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1664 | \| Locator | 1665 +-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1667 Packet field descriptions: 1669 Type: 3 (Map-Register) 1671 P: This is the proxy-map-reply bit, when set to 1 an ETR sends a Map- 1672 Register message requesting for the Map-Server to proxy Map-Reply. 1673 The Map-Server will send non-authoritative Map-Replies on behalf 1674 of the ETR. Details on this usage can be found in [LISP-MS]. 1676 Reserved: It MUST be set to 0 on transmit and MUST be ignored on 1677 receipt. 1679 M: This is the want-map-notify bit, when set to 1 an ETR is 1680 requesting for a Map-Notify message to be returned in response to 1681 sending a Map-Register message. The Map-Notify message sent by a 1682 Map-Server is used to an acknowledge receipt of a Map-Register 1683 message. 1685 Record Count: The number of records in this Map-Register message. A 1686 record is comprised of that portion of the packet labeled 'Record' 1687 above and occurs the number of times equal to Record count. 1689 Nonce: This 8-octet Nonce field is set to 0 in Map-Register 1690 messages. Since the Map-Register message is authenticated, the 1691 nonce field is not currently used for any security function but 1692 may be in the future as part of an anti-replay solution. 1694 Key ID: A configured ID to find the configured Message 1695 Authentication Code (MAC) algorithm and key value used for the 1696 authentication function. See Section 14.4 for codepoint 1697 assignments. 1699 Authentication Data Length: The length in octets of the 1700 Authentication Data field that follows this field. The length of 1701 the Authentication Data field is dependent on the Message 1702 Authentication Code (MAC) algorithm used. The length field allows 1703 a device that doesn't know the MAC algorithm to correctly parse 1704 the packet. 1706 Authentication Data: The message digest used from the output of the 1707 Message Authentication Code (MAC) algorithm. The entire Map- 1708 Register payload is authenticated with this field preset to 0. 1709 After the MAC is computed, it is placed in this field. 1710 Implementations of this specification MUST include support for 1711 HMAC-SHA-1-96 [RFC2404] and support for HMAC-SHA-256-128 [RFC6234] 1712 is RECOMMENDED. 1714 The definition of the rest of the Map-Register can be found in the 1715 Map-Reply section. 1717 6.1.7. Map-Notify Message Format 1719 The usage details of the Map-Notify message can be found in 1720 specification [LISP-MS]. This section solely defines the message 1721 format. 1723 The message is sent inside a UDP packet with source and destination 1724 UDP ports equal to 4342. 1726 The Map-Notify message format is: 1728 0 1 2 3 1729 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 1730 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1731 |Type=4 | Reserved | Record Count | 1732 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1733 | Nonce . . . | 1734 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1735 | . . . Nonce | 1736 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1737 | Key ID | Authentication Data Length | 1738 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1739 ~ Authentication Data ~ 1740 +-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1741 | | Record TTL | 1742 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1743 R | Locator Count | EID mask-len | ACT |A| Reserved | 1744 e +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1745 c | Rsvd | Map-Version Number | EID-prefix-AFI | 1746 o +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1747 r | EID-prefix | 1748 d +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1749 | /| Priority | Weight | M Priority | M Weight | 1750 | L +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1751 | o | Unused Flags |L|p|R| Loc-AFI | 1752 | c +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1753 | \| Locator | 1754 +-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1756 Packet field descriptions: 1758 Type: 4 (Map-Notify) 1760 The Map-Notify message has the same contents as a Map-Register 1761 message. See Map-Register section for field descriptions. 1763 6.1.8. Encapsulated Control Message Format 1765 An Encapsulated Control Message (ECM) is used to encapsulate control 1766 packets sent between xTRs and the mapping database system described 1767 in [LISP-MS]. 1769 0 1 2 3 1770 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 1771 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1772 / | IPv4 or IPv6 Header | 1773 OH | (uses RLOC addresses) | 1774 \ | | 1775 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1776 / | Source Port = xxxx | Dest Port = 4342 | 1777 UDP +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1778 \ | UDP Length | UDP Checksum | 1779 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1780 LH |Type=8 |S| Reserved | 1781 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1782 / | IPv4 or IPv6 Header | 1783 IH | (uses RLOC or EID addresses) | 1784 \ | | 1785 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1786 / | Source Port = xxxx | Dest Port = yyyy | 1787 UDP +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1788 \ | UDP Length | UDP Checksum | 1789 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1790 LCM | LISP Control Message | 1791 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1793 Packet header descriptions: 1795 OH: The outer IPv4 or IPv6 header which uses RLOC addresses in the 1796 source and destination header address fields. 1798 UDP: The outer UDP header with destination port 4342. The source 1799 port is randomly allocated. The checksum field MUST be non-zero. 1801 LH: Type 8 is defined to be a "LISP Encapsulated Control Message" 1802 and what follows is either an IPv4 or IPv6 header as encoded by 1803 the first 4 bits after the reserved field. 1805 S: This is the Security bit. When set to 1 the field following the 1806 Reserved field will have the following format. The detailed 1807 format of the Authentication Data Content is for further study. 1809 0 1 2 3 1810 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 1811 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1812 | AD Type | Authentication Data Content . . . | 1813 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1815 IH: The inner IPv4 or IPv6 header which can use either RLOC or EID 1816 addresses in the header address fields. When a Map-Request is 1817 encapsulated in this packet format the destination address in this 1818 header is an EID. 1820 UDP: The inner UDP header where the port assignments depends on the 1821 control packet being encapsulated. When the control packet is a 1822 Map-Request or Map-Register, the source port is ITR/PITR selected 1823 and the destination port is 4342. When the control packet is a 1824 Map-Reply, the source port is 4342 and the destination port is 1825 assigned from the source port of the invoking Map-Request. Port 1826 number 4341 MUST NOT be assigned to either port. The checksum 1827 field MUST be non-zero. 1829 LCM: The format is one of the control message formats described in 1830 this section. At this time, only Map-Request messages are allowed 1831 to be encapsulated. And in the future, PIM Join-Prune messages 1832 [MLISP] might be allowed. Encapsulating other types of LISP 1833 control messages are for further study. When Map-Requests are 1834 sent for RLOC-probing purposes (i.e the probe-bit is set), they 1835 MUST NOT be sent inside Encapsulated Control Messages. 1837 6.2. Routing Locator Selection 1839 Both client-side and server-side may need control over the selection 1840 of RLOCs for conversations between them. This control is achieved by 1841 manipulating the Priority and Weight fields in EID-to-RLOC Map-Reply 1842 messages. Alternatively, RLOC information MAY be gleaned from 1843 received tunneled packets or EID-to-RLOC Map-Request messages. 1845 The following enumerates different scenarios for choosing RLOCs and 1846 the controls that are available: 1848 o Server-side returns one RLOC. Client-side can only use one RLOC. 1849 Server-side has complete control of the selection. 1851 o Server-side returns a list of RLOC where a subset of the list has 1852 the same best priority. Client can only use the subset list 1853 according to the weighting assigned by the server-side. In this 1854 case, the server-side controls both the subset list and load- 1855 splitting across its members. The client-side can use RLOCs 1856 outside of the subset list if it determines that the subset list 1857 is unreachable (unless RLOCs are set to a Priority of 255). Some 1858 sharing of control exists: the server-side determines the 1859 destination RLOC list and load distribution while the client-side 1860 has the option of using alternatives to this list if RLOCs in the 1861 list are unreachable. 1863 o Server-side sets weight of 0 for the RLOC subset list. In this 1864 case, the client-side can choose how the traffic load is spread 1865 across the subset list. Control is shared by the server-side 1866 determining the list and the client determining load distribution. 1867 Again, the client can use alternative RLOCs if the server-provided 1868 list of RLOCs are unreachable. 1870 o Either side (more likely on the server-side ETR) decides not to 1871 send a Map-Request. For example, if the server-side ETR does not 1872 send Map-Requests, it gleans RLOCs from the client-side ITR, 1873 giving the client-side ITR responsibility for bidirectional RLOC 1874 reachability and preferability. Server-side ETR gleaning of the 1875 client-side ITR RLOC is done by caching the inner header source 1876 EID and the outer header source RLOC of received packets. The 1877 client-side ITR controls how traffic is returned and can alternate 1878 using an outer header source RLOC, which then can be added to the 1879 list the server-side ETR uses to return traffic. Since no 1880 Priority or Weights are provided using this method, the server- 1881 side ETR MUST assume each client-side ITR RLOC uses the same best 1882 Priority with a Weight of zero. In addition, since EID-prefix 1883 encoding cannot be conveyed in data packets, the EID-to-RLOC cache 1884 on tunnel routers can grow to be very large. 1886 o A "gleaned" map-cache entry, one learned from the source RLOC of a 1887 received encapsulated packet, is only stored and used for a few 1888 seconds, pending verification. Verification is performed by 1889 sending a Map-Request to the source EID (the inner header IP 1890 source address) of the received encapsulated packet. A reply to 1891 this "verifying Map-Request" is used to fully populate the map- 1892 cache entry for the "gleaned" EID and is stored and used for the 1893 time indicated from the TTL field of a received Map-Reply. When a 1894 verified map-cache entry is stored, data gleaning no longer occurs 1895 for subsequent packets which have a source EID that matches the 1896 EID-prefix of the verified entry. 1898 RLOCs that appear in EID-to-RLOC Map-Reply messages are assumed to be 1899 reachable when the R-bit for the locator record is set to 1. When 1900 the R-bit is set to 0, an ITR or PITR MUST NOT encapsulate to the 1901 RLOC. Neither the information contained in a Map-Reply or that 1902 stored in the mapping database system provides reachability 1903 information for RLOCs. Note that reachability is not part of the 1904 mapping system and is determined using one or more of the Routing 1905 Locator Reachability Algorithms described in the next section. 1907 6.3. Routing Locator Reachability 1909 Several mechanisms for determining RLOC reachability are currently 1910 defined: 1912 1. An ETR may examine the Locator Status Bits in the LISP header of 1913 an encapsulated data packet received from an ITR. If the ETR is 1914 also acting as an ITR and has traffic to return to the original 1915 ITR site, it can use this status information to help select an 1916 RLOC. 1918 2. An ITR may receive an ICMP Network or ICMP Host Unreachable 1919 message for an RLOC it is using. This indicates that the RLOC is 1920 likely down. Note, trusting ICMP messages may not be desirable 1921 but neither is ignoring them completely. Implementations are 1922 encouraged to follow current best practices in treating these 1923 conditions. 1925 3. An ITR which participates in the global routing system can 1926 determine that an RLOC is down if no BGP RIB route exists that 1927 matches the RLOC IP address. 1929 4. An ITR may receive an ICMP Port Unreachable message from a 1930 destination host. This occurs if an ITR attempts to use 1931 interworking [INTERWORK] and LISP-encapsulated data is sent to a 1932 non-LISP-capable site. 1934 5. An ITR may receive a Map-Reply from an ETR in response to a 1935 previously sent Map-Request. The RLOC source of the Map-Reply is 1936 likely up since the ETR was able to send the Map-Reply to the 1937 ITR. 1939 6. When an ETR receives an encapsulated packet from an ITR, the 1940 source RLOC from the outer header of the packet is likely up. 1942 7. An ITR/ETR pair can use the Locator Reachability Algorithms 1943 described in this section, namely Echo-Noncing or RLOC-Probing. 1945 When determining Locator up/down reachability by examining the 1946 Locator Status Bits from the LISP encapsulated data packet, an ETR 1947 will receive up to date status from an encapsulating ITR about 1948 reachability for all ETRs at the site. CE-based ITRs at the source 1949 site can determine reachability relative to each other using the site 1950 IGP as follows: 1952 o Under normal circumstances, each ITR will advertise a default 1953 route into the site IGP. 1955 o If an ITR fails or if the upstream link to its PE fails, its 1956 default route will either time-out or be withdrawn. 1958 Each ITR can thus observe the presence or lack of a default route 1959 originated by the others to determine the Locator Status Bits it sets 1960 for them. 1962 RLOCs listed in a Map-Reply are numbered with ordinals 0 to n-1. The 1963 Locator Status Bits in a LISP encapsulated packet are numbered from 0 1964 to n-1 starting with the least significant bit. For example, if an 1965 RLOC listed in the 3rd position of the Map-Reply goes down (ordinal 1966 value 2), then all ITRs at the site will clear the 3rd least 1967 significant bit (xxxx x0xx) of the Locator Status Bits field for the 1968 packets they encapsulate. 1970 When an ETR decapsulates a packet, it will check for any change in 1971 the Locator Status Bits field. When a bit goes from 1 to 0, the ETR 1972 if acting also as an ITR, will refrain from encapsulating packets to 1973 an RLOC that is indicated as down. It will only resume using that 1974 RLOC if the corresponding Locator Status Bit returns to a value of 1. 1975 Locator Status Bits are associated with a locator-set per EID-prefix. 1976 Therefore, when a locator becomes unreachable, the Locator Status Bit 1977 that corresponds to that locator's position in the list returned by 1978 the last Map-Reply will be set to zero for that particular EID- 1979 prefix. 1981 When ITRs at the site are not deployed in CE routers, the IGP can 1982 still be used to determine the reachability of Locators provided they 1983 are injected into the IGP. This is typically done when a /32 address 1984 is configured on a loopback interface. 1986 When ITRs receive ICMP Network or Host Unreachable messages as a 1987 method to determine unreachability, they will refrain from using 1988 Locators which are described in Locator lists of Map-Replies. 1989 However, using this approach is unreliable because many network 1990 operators turn off generation of ICMP Unreachable messages. 1992 If an ITR does receive an ICMP Network or Host Unreachable message, 1993 it MAY originate its own ICMP Unreachable message destined for the 1994 host that originated the data packet the ITR encapsulated. 1996 Also, BGP-enabled ITRs can unilaterally examine the RIB to see if a 1997 locator address from a locator-set in a mapping entry matches a 1998 prefix. If it does not find one and BGP is running in the Default 1999 Free Zone (DFZ), it can decide to not use the locator even though the 2000 Locator Status Bits indicate the locator is up. In this case, the 2001 path from the ITR to the ETR that is assigned the locator is not 2002 available. More details are in [LOC-ID-ARCH]. 2004 Optionally, an ITR can send a Map-Request to a Locator and if a Map- 2005 Reply is returned, reachability of the Locator has been determined. 2006 Obviously, sending such probes increases the number of control 2007 messages originated by tunnel routers for active flows, so Locators 2008 are assumed to be reachable when they are advertised. 2010 This assumption does create a dependency: Locator unreachability is 2011 detected by the receipt of ICMP Host Unreachable messages. When an 2012 Locator has been determined to be unreachable, it is not used for 2013 active traffic; this is the same as if it were listed in a Map-Reply 2014 with priority 255. 2016 The ITR can test the reachability of the unreachable Locator by 2017 sending periodic Requests. Both Requests and Replies MUST be rate- 2018 limited. Locator reachability testing is never done with data 2019 packets since that increases the risk of packet loss for end-to-end 2020 sessions. 2022 When an ETR decapsulates a packet, it knows that it is reachable from 2023 the encapsulating ITR because that is how the packet arrived. In 2024 most cases, the ETR can also reach the ITR but cannot assume this to 2025 be true due to the possibility of path asymmetry. In the presence of 2026 unidirectional traffic flow from an ITR to an ETR, the ITR SHOULD NOT 2027 use the lack of return traffic as an indication that the ETR is 2028 unreachable. Instead, it MUST use an alternate mechanisms to 2029 determine reachability. 2031 6.3.1. Echo Nonce Algorithm 2033 When data flows bidirectionally between locators from different 2034 sites, a data-plane mechanism called "nonce echoing" can be used to 2035 determine reachability between an ITR and ETR. When an ITR wants to 2036 solicit a nonce echo, it sets the N and E bits and places a 24-bit 2037 nonce [RFC4086] in the LISP header of the next encapsulated data 2038 packet. 2040 When this packet is received by the ETR, the encapsulated packet is 2041 forwarded as normal. When the ETR next sends a data packet to the 2042 ITR, it includes the nonce received earlier with the N bit set and E 2043 bit cleared. The ITR sees this "echoed nonce" and knows the path to 2044 and from the ETR is up. 2046 The ITR will set the E-bit and N-bit for every packet it sends while 2047 in echo-nonce-request state. The time the ITR waits to process the 2048 echoed nonce before it determines the path is unreachable is variable 2049 and a choice left for the implementation. 2051 If the ITR is receiving packets from the ETR but does not see the 2052 nonce echoed while being in echo-nonce-request state, then the path 2053 to the ETR is unreachable. This decision may be overridden by other 2054 locator reachability algorithms. Once the ITR determines the path to 2055 the ETR is down it can switch to another locator for that EID-prefix. 2057 Note that "ITR" and "ETR" are relative terms here. Both devices MUST 2058 be implementing both ITR and ETR functionality for the echo nonce 2059 mechanism to operate. 2061 The ITR and ETR may both go into echo-nonce-request state at the same 2062 time. The number of packets sent or the time during which echo nonce 2063 requests are sent is an implementation specific setting. However, 2064 when an ITR is in echo-nonce-request state, it can echo the ETR's 2065 nonce in the next set of packets that it encapsulates and then 2066 subsequently, continue sending echo-nonce-request packets. 2068 This mechanism does not completely solve the forward path 2069 reachability problem as traffic may be unidirectional. That is, the 2070 ETR receiving traffic at a site may not be the same device as an ITR 2071 which transmits traffic from that site or the site to site traffic is 2072 unidirectional so there is no ITR returning traffic. 2074 The echo-nonce algorithm is bilateral. That is, if one side sets the 2075 E-bit and the other side is not enabled for echo-noncing, then the 2076 echoing of the nonce does not occur and the requesting side may 2077 regard the locator unreachable erroneously. An ITR SHOULD only set 2078 the E-bit in a encapsulated data packet when it knows the ETR is 2079 enabled for echo-noncing. This is conveyed by the E-bit in the Map- 2080 Reply message. 2082 Note that other locator reachability mechanisms are being researched 2083 and can be used to compliment or even override the Echo Nonce 2084 Algorithm. See next section for an example of control-plane probing. 2086 6.3.2. RLOC Probing Algorithm 2088 RLOC Probing is a method that an ITR or PITR can use to determine the 2089 reachability status of one or more locators that it has cached in a 2090 map-cache entry. The probe-bit of the Map-Request and Map-Reply 2091 messages are used for RLOC Probing. 2093 RLOC probing is done in the control-plane on a timer basis where an 2094 ITR or PITR will originate a Map-Request destined to a locator 2095 address from one of its own locator addresses. A Map-Request used as 2096 an RLOC-probe is NOT encapsulated and NOT sent to a Map-Server or on 2097 the ALT like one would when soliciting mapping data. The EID record 2098 encoded in the Map-Request is the EID-prefix of the map-cache entry 2099 cached by the ITR or PITR. The ITR may include a mapping data record 2100 for its own database mapping information which contains the local 2101 EID-prefixes and RLOCs for its site. RLOC-probes are sent 2102 periodically using a jittered timer interval. 2104 When an ETR receives a Map-Request message with the probe-bit set, it 2105 returns a Map-Reply with the probe-bit set. The source address of 2106 the Map-Reply is set according to the procedure described in 2107 Section 6.1.5. The Map-Reply SHOULD contain mapping data for the 2108 EID-prefix contained in the Map-Request. This provides the 2109 opportunity for the ITR or PITR, which sent the RLOC-probe to get 2110 mapping updates if there were changes to the ETR's database mapping 2111 entries. 2113 There are advantages and disadvantages of RLOC Probing. The greatest 2114 benefit of RLOC Probing is that it can handle many failure scenarios 2115 allowing the ITR to determine when the path to a specific locator is 2116 reachable or has become unreachable, thus providing a robust 2117 mechanism for switching to using another locator from the cached 2118 locator. RLOC Probing can also provide rough RTT estimates between a 2119 pair of locators which can be useful for network management purposes 2120 as well as for selecting low delay paths. The major disadvantage of 2121 RLOC Probing is in the number of control messages required and the 2122 amount of bandwidth used to obtain those benefits, especially if the 2123 requirement for failure detection times are very small. 2125 Continued research and testing will attempt to characterize the 2126 tradeoffs of failure detection times versus message overhead. 2128 6.4. EID Reachability within a LISP Site 2130 A site may be multihomed using two or more ETRs. The hosts and 2131 infrastructure within a site will be addressed using one or more EID 2132 prefixes that are mapped to the RLOCs of the relevant ETRs in the 2133 mapping system. One possible failure mode is for an ETR to lose 2134 reachability to one or more of the EID prefixes within its own site. 2135 When this occurs when the ETR sends Map-Replies, it can clear the 2136 R-bit associated with its own locator. And when the ETR is also an 2137 ITR, it can clear its locator-status-bit in the encapsulation data 2138 header. 2140 It is recognized there are no simple solutions to the site 2141 partitioning problem because it is hard to know which part of the 2142 EID-prefix range is partitioned. And which locators can reach any 2143 sub-ranges of the EID-prefixes. This problem is under investigation 2144 with the expectation that experiments will tell us more. Note, this 2145 is not a new problem introduced by the LISP architecture. The 2146 problem exists today when a multi-homed site uses BGP to advertise 2147 its reachability upstream. 2149 6.5. Routing Locator Hashing 2151 When an ETR provides an EID-to-RLOC mapping in a Map-Reply message to 2152 a requesting ITR, the locator-set for the EID-prefix may contain 2153 different priority values for each locator address. When more than 2154 one best priority locator exists, the ITR can decide how to load 2155 share traffic against the corresponding locators. 2157 The following hash algorithm may be used by an ITR to select a 2158 locator for a packet destined to an EID for the EID-to-RLOC mapping: 2160 1. Either a source and destination address hash can be used or the 2161 traditional 5-tuple hash which includes the source and 2162 destination addresses, source and destination TCP, UDP, or SCTP 2163 port numbers and the IP protocol number field or IPv6 next- 2164 protocol fields of a packet a host originates from within a LISP 2165 site. When a packet is not a TCP, UDP, or SCTP packet, the 2166 source and destination addresses only from the header are used to 2167 compute the hash. 2169 2. Take the hash value and divide it by the number of locators 2170 stored in the locator-set for the EID-to-RLOC mapping. 2172 3. The remainder will yield a value of 0 to "number of locators 2173 minus 1". Use the remainder to select the locator in the 2174 locator-set. 2176 Note that when a packet is LISP encapsulated, the source port number 2177 in the outer UDP header needs to be set. Selecting a hashed value 2178 allows core routers which are attached to Link Aggregation Groups 2179 (LAGs) to load-split the encapsulated packets across member links of 2180 such LAGs. Otherwise, core routers would see a single flow, since 2181 packets have a source address of the ITR, for packets which are 2182 originated by different EIDs at the source site. A suggested setting 2183 for the source port number computed by an ITR is a 5-tuple hash 2184 function on the inner header, as described above. 2186 Many core router implementations use a 5-tuple hash to decide how to 2187 balance packet load across members of a LAG. The 5-tuple hash 2188 includes the source and destination addresses of the packet and the 2189 source and destination ports when the protocol number in the packet 2190 is TCP or UDP. For this reason, UDP encoding is used for LISP 2191 encapsulation. 2193 6.6. Changing the Contents of EID-to-RLOC Mappings 2195 Since the LISP architecture uses a caching scheme to retrieve and 2196 store EID-to-RLOC mappings, the only way an ITR can get a more up-to- 2197 date mapping is to re-request the mapping. However, the ITRs do not 2198 know when the mappings change and the ETRs do not keep track of which 2199 ITRs requested its mappings. For scalability reasons, we want to 2200 maintain this approach but need to provide a way for ETRs change 2201 their mappings and inform the sites that are currently communicating 2202 with the ETR site using such mappings. 2204 When adding a new locator record in lexicographic order to the end of 2205 a locator-set, it is easy to update mappings. We assume new mappings 2206 will maintain the same locator ordering as the old mapping but just 2207 have new locators appended to the end of the list. So some ITRs can 2208 have a new mapping while other ITRs have only an old mapping that is 2209 used until they time out. When an ITR has only an old mapping but 2210 detects bits set in the loc-status-bits that correspond to locators 2211 beyond the list it has cached, it simply ignores them. However, this 2212 can only happen for locator addresses that are lexicographically 2213 greater than the locator addresses in the existing locator-set. 2215 When a locator record is inserted in the middle of a locator-set, to 2216 maintain lexicographic order, the SMR procedure in Section 6.6.2 is 2217 used to inform ITRs and PITRs of the new locator-status-bit mappings. 2219 When a locator record is removed from a locator-set, ITRs that have 2220 the mapping cached will not use the removed locator because the xTRs 2221 will set the loc-status-bit to 0. So even if the locator is in the 2222 list, it will not be used. For new mapping requests, the xTRs can 2223 set the locator AFI to 0 (indicating an unspecified address), as well 2224 as setting the corresponding loc-status-bit to 0. This forces ITRs 2225 with old or new mappings to avoid using the removed locator. 2227 If many changes occur to a mapping over a long period of time, one 2228 will find empty record slots in the middle of the locator-set and new 2229 records appended to the locator-set. At some point, it would be 2230 useful to compact the locator-set so the loc-status-bit settings can 2231 be efficiently packed. 2233 We propose here three approaches for locator-set compaction, one 2234 operational and two protocol mechanisms. The operational approach 2235 uses a clock sweep method. The protocol approaches use the concept 2236 of Solicit-Map-Requests and Map-Versioning. 2238 6.6.1. Clock Sweep 2240 The clock sweep approach uses planning in advance and the use of 2241 count-down TTLs to time out mappings that have already been cached. 2242 The default setting for an EID-to-RLOC mapping TTL is 24 hours. So 2243 there is a 24 hour window to time out old mappings. The following 2244 clock sweep procedure is used: 2246 1. 24 hours before a mapping change is to take effect, a network 2247 administrator configures the ETRs at a site to start the clock 2248 sweep window. 2250 2. During the clock sweep window, ETRs continue to send Map-Reply 2251 messages with the current (unchanged) mapping records. The TTL 2252 for these mappings is set to 1 hour. 2254 3. 24 hours later, all previous cache entries will have timed out, 2255 and any active cache entries will time out within 1 hour. During 2256 this 1 hour window the ETRs continue to send Map-Reply messages 2257 with the current (unchanged) mapping records with the TTL set to 2258 1 minute. 2260 4. At the end of the 1 hour window, the ETRs will send Map-Reply 2261 messages with the new (changed) mapping records. So any active 2262 caches can get the new mapping contents right away if not cached, 2263 or in 1 minute if they had the mapping cached. The new mappings 2264 are cached with a time to live equal to the TTL in the Map-Reply. 2266 6.6.2. Solicit-Map-Request (SMR) 2268 Soliciting a Map-Request is a selective way for ETRs, at the site 2269 where mappings change, to control the rate they receive requests for 2270 Map-Reply messages. SMRs are also used to tell remote ITRs to update 2271 the mappings they have cached. 2273 Since the ETRs don't keep track of remote ITRs that have cached their 2274 mappings, they do not know which ITRs need to have their mappings 2275 updated. As a result, an ETR will solicit Map-Requests (called an 2276 SMR message) from those sites to which it has been sending 2277 encapsulated data to for the last minute. In particular, an ETR will 2278 send an SMR an ITR to which it has recently sent encapsulated data. 2280 An SMR message is simply a bit set in a Map-Request message. An ITR 2281 or PITR will send a Map-Request when they receive an SMR message. 2282 Both the SMR sender and the Map-Request responder MUST rate-limited 2283 these messages. Rate-limiting can be implemented as a global rate- 2284 limiter or one rate-limiter per SMR destination. 2286 The following procedure shows how a SMR exchange occurs when a site 2287 is doing locator-set compaction for an EID-to-RLOC mapping: 2289 1. When the database mappings in an ETR change, the ETRs at the site 2290 begin to send Map-Requests with the SMR bit set for each locator 2291 in each map-cache entry the ETR caches. 2293 2. A remote ITR which receives the SMR message will schedule sending 2294 a Map-Request message to the source locator address of the SMR 2295 message or to the mapping database system. A newly allocated 2296 random nonce is selected and the EID-prefix used is the one 2297 copied from the SMR message. If the source locator is the only 2298 locator in the cached locator-set, the remote ITR SHOULD send a 2299 Map-Request to the database mapping system just in case the 2300 single locator has changed and may no longer be reachable to 2301 accept the Map-Request. 2303 3. The remote ITR MUST rate-limit the Map-Request until it gets a 2304 Map-Reply while continuing to use the cached mapping. When Map 2305 Versioning is used, described in Section 6.6.3, an SMR sender can 2306 detect if an ITR is using the most up to date database mapping. 2308 4. The ETRs at the site with the changed mapping will reply to the 2309 Map-Request with a Map-Reply message that has a nonce from the 2310 SMR-invoked Map-Request. The Map-Reply messages SHOULD be rate 2311 limited. This is important to avoid Map-Reply implosion. 2313 5. The ETRs, at the site with the changed mapping, record the fact 2314 that the site that sent the Map-Request has received the new 2315 mapping data in the mapping cache entry for the remote site so 2316 the loc-status-bits are reflective of the new mapping for packets 2317 going to the remote site. The ETR then stops sending SMR 2318 messages. 2320 Experimentation is in progress to determine the appropriate rate- 2321 limit parameters. 2323 For security reasons an ITR MUST NOT process unsolicited Map-Replies. 2324 To avoid map-cache entry corruption by a third-party, a sender of an 2325 SMR-based Map-Request MUST be verified. If an ITR receives an SMR- 2326 based Map-Request and the source is not in the locator-set for the 2327 stored map-cache entry, then the responding Map-Request MUST be sent 2328 with an EID destination to the mapping database system. Since the 2329 mapping database system is more secure to reach an authoritative ETR, 2330 it will deliver the Map-Request to the authoritative source of the 2331 mapping data. 2333 When an ITR receives an SMR-based Map-Request for which it does not 2334 have a cached mapping for the EID in the SMR message, it MAY not send 2335 a SMR-invoked Map-Request. This scenario can occur when an ETR sends 2336 SMR messages to all locators in the locator-set it has stored in its 2337 map-cache but the remote ITRs that receive the SMR may not be sending 2338 packets to the site. There is no point in updating the ITRs until 2339 they need to send, in which case, they will send Map-Requests to 2340 obtain a map-cache entry. 2342 6.6.3. Database Map Versioning 2344 When there is unidirectional packet flow between an ITR and ETR, and 2345 the EID-to-RLOC mappings change on the ETR, it needs to inform the 2346 ITR so encapsulation can stop to a removed locator and start to a new 2347 locator in the locator-set. 2349 An ETR, when it sends Map-Reply messages, conveys its own Map-Version 2350 number. This is known as the Destination Map-Version Number. ITRs 2351 include the Destination Map-Version Number in packets they 2352 encapsulate to the site. When an ETR decapsulates a packet and 2353 detects the Destination Map-Version Number is less than the current 2354 version for its mapping, the SMR procedure described in Section 6.6.2 2355 occurs. 2357 An ITR, when it encapsulates packets to ETRs, can convey its own Map- 2358 Version number. This is known as the Source Map-Version Number. 2359 When an ETR decapsulates a packet and detects the Source Map-Version 2360 Number is greater than the last Map-Version Number sent in a Map- 2361 Reply from the ITR's site, the ETR will send a Map-Request to one of 2362 the ETRs for the source site. 2364 A Map-Version Number is used as a sequence number per EID-prefix. So 2365 values that are greater, are considered to be more recent. A value 2366 of 0 for the Source Map-Version Number or the Destination Map-Version 2367 Number conveys no versioning information and an ITR does no 2368 comparison with previously received Map-Version Numbers. 2370 A Map-Version Number can be included in Map-Register messages as 2371 well. This is a good way for the Map-Server can assure that all ETRs 2372 for a site registering to it will be Map-Version number synchronized. 2374 See [VERSIONING] for a more detailed analysis and description of 2375 Database Map Versioning. 2377 7. Router Performance Considerations 2379 LISP is designed to be very hardware-based forwarding friendly. A 2380 few implementation techniques can be used to incrementally implement 2381 LISP: 2383 o When a tunnel encapsulated packet is received by an ETR, the outer 2384 destination address may not be the address of the router. This 2385 makes it challenging for the control plane to get packets from the 2386 hardware. This may be mitigated by creating special FIB entries 2387 for the EID-prefixes of EIDs served by the ETR (those for which 2388 the router provides an RLOC translation). These FIB entries are 2389 marked with a flag indicating that control plane processing should 2390 be performed. The forwarding logic of testing for particular IP 2391 protocol number value is not necessary. There are a few proven 2392 cases where no changes to existing deployed hardware were needed 2393 to support the LISP data-plane. 2395 o On an ITR, prepending a new IP header consists of adding more 2396 octets to a MAC rewrite string and prepending the string as part 2397 of the outgoing encapsulation procedure. Routers that support GRE 2398 tunneling [RFC2784] or 6to4 tunneling [RFC3056] may already 2399 support this action. 2401 o A packet's source address or interface the packet was received on 2402 can be used to select a VRF (Virtual Routing/Forwarding). The 2403 VRF's routing table can be used to find EID-to-RLOC mappings. 2405 For performance issues related to map-cache management, see section 2406 Section 12. 2408 8. Deployment Scenarios 2410 This section will explore how and where ITRs and ETRs can be deployed 2411 and will discuss the pros and cons of each deployment scenario. For 2412 a more detailed deployment recommendation, refer to [LISP-DEPLOY]. 2414 There are two basic deployment trade-offs to consider: centralized 2415 versus distributed caches and flat, recursive, or re-encapsulating 2416 tunneling. When deciding on centralized versus distributed caching, 2417 the following issues should be considered: 2419 o Are the tunnel routers spread out so that the caches are spread 2420 across all the memories of each router? A centralized cache is 2421 when an ITR keeps a cache for all the EIDs it is encapsulating to. 2422 The packet takes a direct path to the destination locator. A 2423 distributed cache is when an ITR needs help from other re- 2424 encapsulating routers because it does not store all the cache 2425 entries for the EIDs is it encapsulating to. So the packet takes 2426 a path through re-encapsulating routers that have a different set 2427 of cache entries. 2429 o Should management "touch points" be minimized by choosing few 2430 tunnel routers, just enough for redundancy? 2432 o In general, using more ITRs doesn't increase management load, 2433 since caches are built and stored dynamically. On the other hand, 2434 more ETRs does require more management since EID-prefix-to-RLOC 2435 mappings need to be explicitly configured. 2437 When deciding on flat, recursive, or re-encapsulation tunneling, the 2438 following issues should be considered: 2440 o Flat tunneling implements a single tunnel between source site and 2441 destination site. This generally offers better paths between 2442 sources and destinations with a single tunnel path. 2444 o Recursive tunneling is when tunneled traffic is again further 2445 encapsulated in another tunnel, either to implement VPNs or to 2446 perform Traffic Engineering. When doing VPN-based tunneling, the 2447 site has some control since the site is prepending a new tunnel 2448 header. In the case of TE-based tunneling, the site may have 2449 control if it is prepending a new tunnel header, but if the site's 2450 ISP is doing the TE, then the site has no control. Recursive 2451 tunneling generally will result in suboptimal paths but at the 2452 benefit of steering traffic to resource available parts of the 2453 network. 2455 o The technique of re-encapsulation ensures that packets only 2456 require one tunnel header. So if a packet needs to be rerouted, 2457 it is first decapsulated by the ETR and then re-encapsulated with 2458 a new tunnel header using a new RLOC. 2460 The next sub-sections will survey where tunnel routers can reside in 2461 the network. 2463 8.1. First-hop/Last-hop Tunnel Routers 2465 By locating tunnel routers close to hosts, the EID-prefix set is at 2466 the granularity of an IP subnet. So at the expense of more EID- 2467 prefix-to-RLOC sets for the site, the caches in each tunnel router 2468 can remain relatively small. But caches always depend on the number 2469 of non-aggregated EID destination flows active through these tunnel 2470 routers. 2472 With more tunnel routers doing encapsulation, the increase in control 2473 traffic grows as well: since the EID-granularity is greater, more 2474 Map-Requests and Map-Replies are traveling between more routers. 2476 The advantage of placing the caches and databases at these stub 2477 routers is that the products deployed in this part of the network 2478 have better price-memory ratios then their core router counterparts. 2479 Memory is typically less expensive in these devices and fewer routes 2480 are stored (only IGP routes). These devices tend to have excess 2481 capacity, both for forwarding and routing state. 2483 LISP functionality can also be deployed in edge switches. These 2484 devices generally have layer-2 ports facing hosts and layer-3 ports 2485 facing the Internet. Spare capacity is also often available in these 2486 devices as well. 2488 8.2. Border/Edge Tunnel Routers 2490 Using customer-edge (CE) routers for tunnel endpoints allows the EID 2491 space associated with a site to be reachable via a small set of RLOCs 2492 assigned to the CE routers for that site. This is the default 2493 behavior envisioned in the rest of this specification. 2495 This offers the opposite benefit of the first-hop/last-hop tunnel 2496 router scenario: the number of mapping entries and network management 2497 touch points are reduced, allowing better scaling. 2499 One disadvantage is that less of the network's resources are used to 2500 reach host endpoints thereby centralizing the point-of-failure domain 2501 and creating network choke points at the CE router. 2503 Note that more than one CE router at a site can be configured with 2504 the same IP address. In this case an RLOC is an anycast address. 2505 This allows resilience between the CE routers. That is, if a CE 2506 router fails, traffic is automatically routed to the other routers 2507 using the same anycast address. However, this comes with the 2508 disadvantage where the site cannot control the entrance point when 2509 the anycast route is advertised out from all border routers. Another 2510 disadvantage of using anycast locators is the limited advertisement 2511 scope of /32 (or /128 for IPv6) routes. 2513 8.3. ISP Provider-Edge (PE) Tunnel Routers 2515 Use of ISP PE routers as tunnel endpoint routers is not the typical 2516 deployment scenario envisioned in the specification. This section 2517 attempts to capture some of reasoning behind this preference of 2518 implementing LISP on CE routers. 2520 Use of ISP PE routers as tunnel endpoint routers gives an ISP, rather 2521 than a site, control over the location of the egress tunnel 2522 endpoints. That is, the ISP can decide if the tunnel endpoints are 2523 in the destination site (in either CE routers or last-hop routers 2524 within a site) or at other PE edges. The advantage of this case is 2525 that two tunnel headers can be avoided. By having the PE be the 2526 first router on the path to encapsulate, it can choose a TE path 2527 first, and the ETR can decapsulate and re-encapsulate for a tunnel to 2528 the destination end site. 2530 An obvious disadvantage is that the end site has no control over 2531 where its packets flow or the RLOCs used. Other disadvantages 2532 include the difficulty in synchronizing path liveness updates between 2533 CE and PE routers. 2535 As mentioned in earlier sections a combination of these scenarios is 2536 possible at the expense of extra packet header overhead, if both site 2537 and provider want control, then recursive or re-encapsulating tunnels 2538 are used. 2540 8.4. LISP Functionality with Conventional NATs 2542 LISP routers can be deployed behind Network Address Translator (NAT) 2543 devices to provide the same set of packet services hosts have today 2544 when they are addressed out of private address space. 2546 It is important to note that a locator address in any LISP control 2547 message MUST be a globally routable address and therefore SHOULD NOT 2548 contain [RFC1918] addresses. If a LISP router is configured with 2549 private addresses, they MUST be used only in the outer IP header so 2550 the NAT device can translate properly. Otherwise, EID addresses MUST 2551 be translated before encapsulation is performed. Both NAT 2552 translation and LISP encapsulation functions could be co-located in 2553 the same device. 2555 More details on LISP address translation can be found in [INTERWORK]. 2557 8.5. Packets Egressing a LISP Site 2559 When a LISP site is using two ITRs for redundancy, the failure of one 2560 ITR will likely shift outbound traffic to the second. This second 2561 ITR's cache may not not be populated with the same EID-to-RLOC 2562 mapping entries as the first. If this second ITR does not have these 2563 mappings, traffic will be dropped while the mappings are retrieved 2564 from the mapping system. The retrieval of these messages may 2565 increase the load of requests being sent into the mapping system. 2566 Deployment and experimentation will determine whether this issue 2567 requires more attention. 2569 9. Traceroute Considerations 2571 When a source host in a LISP site initiates a traceroute to a 2572 destination host in another LISP site, it is highly desirable for it 2573 to see the entire path. Since packets are encapsulated from ITR to 2574 ETR, the hop across the tunnel could be viewed as a single hop. 2575 However, LISP traceroute will provide the entire path so the user can 2576 see 3 distinct segments of the path from a source LISP host to a 2577 destination LISP host: 2579 Segment 1 (in source LISP site based on EIDs): 2581 source-host ---> first-hop ... next-hop ---> ITR 2583 Segment 2 (in the core network based on RLOCs): 2585 ITR ---> next-hop ... next-hop ---> ETR 2587 Segment 3 (in the destination LISP site based on EIDs): 2589 ETR ---> next-hop ... last-hop ---> destination-host 2591 For segment 1 of the path, ICMP Time Exceeded messages are returned 2592 in the normal manner as they are today. The ITR performs a TTL 2593 decrement and test for 0 before encapsulating. So the ITR hop is 2594 seen by the traceroute source has an EID address (the address of 2595 site-facing interface). 2597 For segment 2 of the path, ICMP Time Exceeded messages are returned 2598 to the ITR because the TTL decrement to 0 is done on the outer 2599 header, so the destination of the ICMP messages are to the ITR RLOC 2600 address, the source RLOC address of the encapsulated traceroute 2601 packet. The ITR looks inside of the ICMP payload to inspect the 2602 traceroute source so it can return the ICMP message to the address of 2603 the traceroute client as well as retaining the core router IP address 2604 in the ICMP message. This is so the traceroute client can display 2605 the core router address (the RLOC address) in the traceroute output. 2606 The ETR returns its RLOC address and responds to the TTL decrement to 2607 0 like the previous core routers did. 2609 For segment 3, the next-hop router downstream from the ETR will be 2610 decrementing the TTL for the packet that was encapsulated, sent into 2611 the core, decapsulated by the ETR, and forwarded because it isn't the 2612 final destination. If the TTL is decremented to 0, any router on the 2613 path to the destination of the traceroute, including the next-hop 2614 router or destination, will send an ICMP Time Exceeded message to the 2615 source EID of the traceroute client. The ICMP message will be 2616 encapsulated by the local ITR and sent back to the ETR in the 2617 originated traceroute source site, where the packet will be delivered 2618 to the host. 2620 9.1. IPv6 Traceroute 2622 IPv6 traceroute follows the procedure described above since the 2623 entire traceroute data packet is included in ICMP Time Exceeded 2624 message payload. Therefore, only the ITR needs to pay special 2625 attention for forwarding ICMP messages back to the traceroute source. 2627 9.2. IPv4 Traceroute 2629 For IPv4 traceroute, we cannot follow the above procedure since IPv4 2630 ICMP Time Exceeded messages only include the invoking IP header and 8 2631 octets that follow the IP header. Therefore, when a core router 2632 sends an IPv4 Time Exceeded message to an ITR, all the ITR has in the 2633 ICMP payload is the encapsulated header it prepended followed by a 2634 UDP header. The original invoking IP header, and therefore the 2635 identity of the traceroute source is lost. 2637 The solution we propose to solve this problem is to cache traceroute 2638 IPv4 headers in the ITR and to match them up with corresponding IPv4 2639 Time Exceeded messages received from core routers and the ETR. The 2640 ITR will use a circular buffer for caching the IPv4 and UDP headers 2641 of traceroute packets. It will select a 16-bit number as a key to 2642 find them later when the IPv4 Time Exceeded messages are received. 2643 When an ITR encapsulates an IPv4 traceroute packet, it will use the 2644 16-bit number as the UDP source port in the encapsulating header. 2645 When the ICMP Time Exceeded message is returned to the ITR, the UDP 2646 header of the encapsulating header is present in the ICMP payload 2647 thereby allowing the ITR to find the cached headers for the 2648 traceroute source. The ITR puts the cached headers in the payload 2649 and sends the ICMP Time Exceeded message to the traceroute source 2650 retaining the source address of the original ICMP Time Exceeded 2651 message (a core router or the ETR of the site of the traceroute 2652 destination). 2654 The signature of a traceroute packet comes in two forms. The first 2655 form is encoded as a UDP message where the destination port is 2656 inspected for a range of values. The second form is encoded as an 2657 ICMP message where the IP identification field is inspected for a 2658 well-known value. 2660 9.3. Traceroute using Mixed Locators 2662 When either an IPv4 traceroute or IPv6 traceroute is originated and 2663 the ITR encapsulates it in the other address family header, you 2664 cannot get all 3 segments of the traceroute. Segment 2 of the 2665 traceroute can not be conveyed to the traceroute source since it is 2666 expecting addresses from intermediate hops in the same address format 2667 for the type of traceroute it originated. Therefore, in this case, 2668 segment 2 will make the tunnel look like one hop. All the ITR has to 2669 do to make this work is to not copy the inner TTL to the outer, 2670 encapsulating header's TTL when a traceroute packet is encapsulated 2671 using an RLOC from a different address family. This will cause no 2672 TTL decrement to 0 to occur in core routers between the ITR and ETR. 2674 10. Mobility Considerations 2676 There are several kinds of mobility of which only some might be of 2677 concern to LISP. Essentially they are as follows. 2679 10.1. Site Mobility 2681 A site wishes to change its attachment points to the Internet, and 2682 its LISP Tunnel Routers will have new RLOCs when it changes upstream 2683 providers. Changes in EID-RLOC mappings for sites are expected to be 2684 handled by configuration, outside of the LISP protocol. 2686 10.2. Slow Endpoint Mobility 2688 An individual endpoint wishes to move, but is not concerned about 2689 maintaining session continuity. Renumbering is involved. LISP can 2690 help with the issues surrounding renumbering [RFC4192] [LISA96] by 2691 decoupling the address space used by a site from the address spaces 2692 used by its ISPs. [RFC4984] 2694 10.3. Fast Endpoint Mobility 2696 Fast endpoint mobility occurs when an endpoint moves relatively 2697 rapidly, changing its IP layer network attachment point. Maintenance 2698 of session continuity is a goal. This is where the Mobile IPv4 2699 [RFC5944] and Mobile IPv6 [RFC6275] [RFC4866] mechanisms are used, 2700 and primarily where interactions with LISP need to be explored. 2702 The problem is that as an endpoint moves, it may require changes to 2703 the mapping between its EID and a set of RLOCs for its new network 2704 location. When this is added to the overhead of mobile IP binding 2705 updates, some packets might be delayed or dropped. 2707 In IPv4 mobility, when an endpoint is away from home, packets to it 2708 are encapsulated and forwarded via a home agent which resides in the 2709 home area the endpoint's address belongs to. The home agent will 2710 encapsulate and forward packets either directly to the endpoint or to 2711 a foreign agent which resides where the endpoint has moved to. 2712 Packets from the endpoint may be sent directly to the correspondent 2713 node, may be sent via the foreign agent, or may be reverse-tunneled 2714 back to the home agent for delivery to the mobile node. As the 2715 mobile node's EID or available RLOC changes, LISP EID-to-RLOC 2716 mappings are required for communication between the mobile node and 2717 the home agent, whether via foreign agent or not. As a mobile 2718 endpoint changes networks, up to three LISP mapping changes may be 2719 required: 2721 o The mobile node moves from an old location to a new visited 2722 network location and notifies its home agent that it has done so. 2723 The Mobile IPv4 control packets the mobile node sends pass through 2724 one of the new visited network's ITRs, which needs an EID-RLOC 2725 mapping for the home agent. 2727 o The home agent might not have the EID-RLOC mappings for the mobile 2728 node's "care-of" address or its foreign agent in the new visited 2729 network, in which case it will need to acquire them. 2731 o When packets are sent directly to the correspondent node, it may 2732 be that no traffic has been sent from the new visited network to 2733 the correspondent node's network, and the new visited network's 2734 ITR will need to obtain an EID-RLOC mapping for the correspondent 2735 node's site. 2737 In addition, if the IPv4 endpoint is sending packets from the new 2738 visited network using its original EID, then LISP will need to 2739 perform a route-returnability check on the new EID-RLOC mapping for 2740 that EID. 2742 In IPv6 mobility, packets can flow directly between the mobile node 2743 and the correspondent node in either direction. The mobile node uses 2744 its "care-of" address (EID). In this case, the route-returnability 2745 check would not be needed but one more LISP mapping lookup may be 2746 required instead: 2748 o As above, three mapping changes may be needed for the mobile node 2749 to communicate with its home agent and to send packets to the 2750 correspondent node. 2752 o In addition, another mapping will be needed in the correspondent 2753 node's ITR, in order for the correspondent node to send packets to 2754 the mobile node's "care-of" address (EID) at the new network 2755 location. 2757 When both endpoints are mobile the number of potential mapping 2758 lookups increases accordingly. 2760 As a mobile node moves there are not only mobility state changes in 2761 the mobile node, correspondent node, and home agent, but also state 2762 changes in the ITRs and ETRs for at least some EID-prefixes. 2764 The goal is to support rapid adaptation, with little delay or packet 2765 loss for the entire system. Also IP mobility can be modified to 2766 require fewer mapping changes. In order to increase overall system 2767 performance, there may be a need to reduce the optimization of one 2768 area in order to place fewer demands on another. 2770 In LISP, one possibility is to "glean" information. When a packet 2771 arrives, the ETR could examine the EID-RLOC mapping and use that 2772 mapping for all outgoing traffic to that EID. It can do this after 2773 performing a route-returnability check, to ensure that the new 2774 network location does have a internal route to that endpoint. 2775 However, this does not cover the case where an ITR (the node assigned 2776 the RLOC) at the mobile-node location has been compromised. 2778 Mobile IP packet exchange is designed for an environment in which all 2779 routing information is disseminated before packets can be forwarded. 2780 In order to allow the Internet to grow to support expected future 2781 use, we are moving to an environment where some information may have 2782 to be obtained after packets are in flight. Modifications to IP 2783 mobility should be considered in order to optimize the behavior of 2784 the overall system. Anything which decreases the number of new EID- 2785 RLOC mappings needed when a node moves, or maintains the validity of 2786 an EID-RLOC mapping for a longer time, is useful. 2788 10.4. Fast Network Mobility 2790 In addition to endpoints, a network can be mobile, possibly changing 2791 xTRs. A "network" can be as small as a single router and as large as 2792 a whole site. This is different from site mobility in that it is 2793 fast and possibly short-lived, but different from endpoint mobility 2794 in that a whole prefix is changing RLOCs. However, the mechanisms 2795 are the same and there is no new overhead in LISP. A map request for 2796 any endpoint will return a binding for the entire mobile prefix. 2798 If mobile networks become a more common occurrence, it may be useful 2799 to revisit the design of the mapping service and allow for dynamic 2800 updates of the database. 2802 The issue of interactions between mobility and LISP needs to be 2803 explored further. Specific improvements to the entire system will 2804 depend on the details of mapping mechanisms. Mapping mechanisms 2805 should be evaluated on how well they support session continuity for 2806 mobile nodes. 2808 10.5. LISP Mobile Node Mobility 2810 A mobile device can use the LISP infrastructure to achieve mobility 2811 by implementing the LISP encapsulation and decapsulation functions 2812 and acting as a simple ITR/ETR. By doing this, such a "LISP mobile 2813 node" can use topologically-independent EID IP addresses that are not 2814 advertised into and do not impose a cost on the global routing 2815 system. These EIDs are maintained at the edges of the mapping system 2816 (in LISP Map-Servers and Map-Resolvers) and are provided on demand to 2817 only the correspondents of the LISP mobile node. 2819 Refer to the LISP Mobility Architecture specification [LISP-MN] for 2820 more details. 2822 11. Multicast Considerations 2824 A multicast group address, as defined in the original Internet 2825 architecture is an identifier of a grouping of topologically 2826 independent receiver host locations. The address encoding itself 2827 does not determine the location of the receiver(s). The multicast 2828 routing protocol, and the network-based state the protocol creates, 2829 determines where the receivers are located. 2831 In the context of LISP, a multicast group address is both an EID and 2832 a Routing Locator. Therefore, no specific semantic or action needs 2833 to be taken for a destination address, as it would appear in an IP 2834 header. Therefore, a group address that appears in an inner IP 2835 header built by a source host will be used as the destination EID. 2836 The outer IP header (the destination Routing Locator address), 2837 prepended by a LISP router, will use the same group address as the 2838 destination Routing Locator. 2840 Having said that, only the source EID and source Routing Locator 2841 needs to be dealt with. Therefore, an ITR merely needs to put its 2842 own IP address in the source Routing Locator field when prepending 2843 the outer IP header. This source Routing Locator address, like any 2844 other Routing Locator address MUST be globally routable. 2846 Therefore, an EID-to-RLOC mapping does not need to be performed by an 2847 ITR when a received data packet is a multicast data packet or when 2848 processing a source-specific Join (either by IGMPv3 or PIM). But the 2849 source Routing Locator is decided by the multicast routing protocol 2850 in a receiver site. That is, an EID to Routing Locator translation 2851 is done at control-time. 2853 Another approach is to have the ITR not encapsulate a multicast 2854 packet and allow the host built packet to flow into the core even if 2855 the source address is allocated out of the EID namespace. If the 2856 RPF-Vector TLV [RFC5496] is used by PIM in the core, then core 2857 routers can RPF to the ITR (the Locator address which is injected 2858 into core routing) rather than the host source address (the EID 2859 address which is not injected into core routing). 2861 To avoid any EID-based multicast state in the network core, the first 2862 approach is chosen for LISP-Multicast. Details for LISP-Multicast 2863 and Interworking with non-LISP sites is described in specification 2864 [MLISP]. 2866 12. Security Considerations 2868 It is believed that most of the security mechanisms will be part of 2869 the mapping database service when using control plane procedures for 2870 obtaining EID-to-RLOC mappings. For data plane triggered mappings, 2871 as described in this specification, protection is provided against 2872 ETR spoofing by using Return-Routability (see Section 3) mechanisms 2873 evidenced by the use of a 24-bit Nonce field in the LISP 2874 encapsulation header and a 64-bit Nonce field in the LISP control 2875 message. 2877 The nonce, coupled with the ITR accepting only solicited Map-Replies 2878 provides a basic level of security, in many ways similar to the 2879 security experienced in the current Internet routing system. It is 2880 hard for off-path attackers to launch attacks against these LISP 2881 mechanisms, as they do not have the nonce values. Sending a large 2882 number of packets to accidentally find the right nonce value is 2883 possible, but would already by itself be a denial-of-service attack. 2884 On-path attackers can perform far more serious attacks, but on-path 2885 attackers can launch serious attacks in the current Internet as well, 2886 including eavesdropping, blocking or redirecting traffic. See more 2887 discussion on this topic in Section 6.1.5.1. 2889 LISP does not rely on a PKI or a more heavy weight authentication 2890 system. These systems challenge the scalability of LISP which was a 2891 primary design goal. 2893 DoS attack prevention will depend on implementations rate-limiting 2894 Map-Requests and Map-Replies to the control plane as well as rate- 2895 limiting the number of data-triggered Map-Replies. 2897 An incorrectly implemented or malicious ITR might choose to ignore 2898 the priority and weights provided by the ETR in its Map-Reply. This 2899 traffic steering would be limited to the traffic that is sent by this 2900 ITR's site, and no more severe than if the site initiated a bandwidth 2901 DoS attack on (one of) the ETR's ingress links. The ITR's site would 2902 typically gain no benefit from not respecting the weights, and would 2903 likely to receive better service by abiding by them. 2905 To deal with map-cache exhaustion attempts in an ITR/PITR, the 2906 implementation should consider putting a maximum cap on the number of 2907 entries stored with a reserve list for special or frequently accessed 2908 sites. This should be a configuration policy control set by the 2909 network administrator who manages ITRs and PITRs. When overlapping 2910 EID-prefixes occur across multiple map-cache entries, the integrity 2911 of the set must be wholly maintained. So if a more-specific entry 2912 cannot be added due to reaching the maximum cap, then none of the 2913 less specifics should be stored in the map-cache. 2915 Given that the ITR/PITR maintains a cache of EID-to-RLOC mappings, 2916 cache sizing and maintenance is an issue to be kept in mind during 2917 implementation. It is a good idea to have instrumentation in place 2918 to detect thrashing of the cache. Implementation experimentation 2919 will be used to determine which cache management strategies work 2920 best. In general, it is difficult to defend against cache trashing 2921 attacks. It should be noted that an undersized cache in an ITR/PITR 2922 not only causes adverse affect on the site or region they support, 2923 but may also cause increased Map-Request load on the mapping system. 2925 "Piggybacked" mapping data discussed in Section 6.1.3 specifies how 2926 to handle such mappings and includes the possibility for an ETR to 2927 temporarily accept such a mapping before verification when running in 2928 "trusted" environments. In such cases, there is a potential threat 2929 that a fake mapping could be inserted (even if only for a short 2930 period) into a map-cache. As noted in Section 6.1.3, an ETR MUST be 2931 specifically configured to run in such a mode and might usefully only 2932 consider some specific ITRs as also running in that same trusted 2933 environment. 2935 There is a security risk implicit in the fact that ETRs generate the 2936 EID prefix to which they are responding. An ETR can claim a shorter 2937 prefix than it is actually responsible for. Various mechanisms to 2938 ameliorate or resolve this issue will be examined in the future, 2939 [LISP-SEC]. 2941 Spoofing of inner header addresses of LISP encapsulated packets is 2942 possible like with any tunneling mechanism. ITRs MUST verify the 2943 source address of a packet to be an EID that belongs to the site's 2944 EID-prefix range prior to encapsulation. An ETR must only 2945 decapsulate and forward datagrams with an inner header destination 2946 that matches one of its EID-prefix ranges. If, upon receipt and 2947 decapsulation, the destination EID of a datagram does not match one 2948 of the ETR's configured EID-prefixes, the ETR MUST drop the datagram. 2949 If a LISP encapsulated packet arrives at an ETR, it SHOULD compare 2950 the inner header source EID address and the outer header source RLOC 2951 address with the mapping that exists in the mapping database. Then 2952 when spoofing attacks occur, the outer header source RLOC address can 2953 be used to trace back the attack to the source site, using existing 2954 operational tools. 2956 This experimental specification does not address automated key 2957 management (AKM). BCP 107 provides guidance in this area. In 2958 addition, at the time of this writing, substantial work is being 2959 undertaken to improve security of the routing system [KARP], [RPKI], 2960 [BGP-SEC], [LISP-SEC]. Future work on LISP should address BCP-107 as 2961 well as other open security considerations, which may require changes 2962 to this specification. 2964 13. Network Management Considerations 2966 Considerations for Network Management tools exist so the LISP 2967 protocol suite can be operationally managed. The mechanisms can be 2968 found in [LISP-MIB] and [LISP-LIG]. 2970 14. IANA Considerations 2972 This section provides guidance to the Internet Assigned Numbers 2973 Authority (IANA) regarding registration of values related to the LISP 2974 specification, in accordance with BCP 26 and RFC 5226 [RFC5226]. 2976 There are four name spaces in LISP that require registration: 2978 o LISP IANA registry allocations should not be made for purposes 2979 unrelated to LISP routing or transport protocols. 2981 o The following policies are used here with the meanings defined in 2982 BCP 26: "Specification Required", "IETF Review", "Experimental 2983 Use", "First Come First Served". 2985 14.1. LISP ACT and Flag Fields 2987 New ACT values (Section 6.1.4) can be allocated through IETF review 2988 or IESG approval. Four values have already been allocated by this 2989 specification (Section 6.1.4). 2991 In addition, the LISP protocol has a number of flag and reserved 2992 fields, such as the LISP header flags field (Section 5.3). New bits 2993 for flags can be taken into use from these fields through IETF review 2994 or IESG approval, but these need not be managed by IANA. 2996 14.2. LISP Address Type Codes 2998 LISP Address [LCAF] type codes have a range from 0 to 255. New type 2999 codes MUST be allocated consecutively starting at 0. Type Codes 0 - 3000 127 are to be assigned by IETF review or IESG approval. 3002 Type Codes 128 - 255 are available on a First Come First Served 3003 policy. 3005 This registry, initially empty, is constructed for future-use 3006 experimental work of LCAF values. See [LCAF] for details for other 3007 possible unapproved address encodings. The unapproved LCAF encodings 3008 are an area for further study and experimentation. 3010 14.3. LISP UDP Port Numbers 3012 The IANA registry has allocated UDP port numbers 4341 and 4342 for 3013 LISP data-plane and control-plane operation, respectively. 3015 14.4. LISP Key ID Numbers 3017 The following Key ID values are defined by this specification as used 3018 in any packet type that references a Key ID field: 3020 Name Number Defined in 3021 ----------------------------------------------- 3022 None 0 n/a 3023 HMAC-SHA-1-96 1 [RFC2404] 3024 HMAC-SHA-256-128 2 [RFC6234] 3026 15. Known Open Issues and Areas of Future Work 3028 As an experimental specification, this work is, by definition, 3029 incomplete. Specific areas where additional experience and work are 3030 needed include: 3032 o At present, only [ALT] is defined for implementing a database of 3033 EID-to-RLOC mapping information. Additional research on other 3034 mapping database systems is strongly encouraged. 3036 o Failure and recovery of LISP site partitioning (see Section 6.4), 3037 in the presence of redundant configuration (see Section 8.5) needs 3038 further research and experimentation. 3040 o The characteristics of map-cache management under exceptional 3041 conditions, such as denial-of-service attacks are not fully 3042 understood. Further experience is needed to determine whether 3043 current caching methods are practical or in need of further 3044 development. In particular, the performance, scaling and security 3045 characteristics of the map-cache will be discovered as part of 3046 this experiment. Performance metrics to be observed are packet 3047 reordering associated with the LISP data probe and loss of the 3048 first packet in a flow associated with map-caching. The impact of 3049 these upon TCP will be observed. See Section 12 for additional 3050 thoughts and considerations. 3052 o Preliminary work has been done to ensure that sites employing LISP 3053 can interconnect with the rest of the Internet. This work is 3054 documented in [INTERWORK], but further experimentation and 3055 experience is needed. 3057 o At present, no mechanism for automated key management for message 3058 authentication is defined. Addressing automated key management is 3059 necessary before this specification could be developed into a 3060 standards track RFC. See Section 12 for further details regarding 3061 security considerations. 3063 o In order to maintain security and stability, Internet Protocols 3064 typically isolate the control and data planes. Therefore, user 3065 activity cannot cause control plane state to be created or 3066 destroyed. LISP does not maintain this separation. The degree to 3067 which the loss of separation impacts security and stability is a 3068 topic for experimental observation. 3070 o LISP allows for different mapping database systems to be used. 3071 While only one [ALT] is currently well-defined, each mapping 3072 database will likely have some impact on the security of the EID- 3073 to-RLOC mappings. How each mapping database system's security 3074 properties impact on LISP overall is for further study. 3076 o An examination of the implications of LISP on Internet traffic, 3077 applications, routers, and security is needed. This will help to 3078 understand the consequences for network stability, routing 3079 protocol function, routing scalability, migration and backward 3080 compatibility, and implementation scalability (as influenced by 3081 additional protocol components, additional state, and additional 3082 processing for encapsulation, decapsulation, liveness). 3084 o Experiments need to verify that LISP produces no significant 3085 change in the behavior of protocols run between end-systems over a 3086 LISP infrastructure versus being run directly between those same 3087 end-systems. 3089 o Experiments need to verify that the issues raised in the Critique 3090 section of [RFC6115] are either insignificant or have been 3091 addressed by updates to the LISP protocol. 3093 Other LISP documents may also include open issues and areas for 3094 future work. 3096 16. References 3098 16.1. Normative References 3100 [ALT] Farinacci, D., Fuller, V., Meyer, D., and D. Lewis, "LISP 3101 Alternative Topology (LISP-ALT)", 3102 draft-ietf-lisp-alt-09.txt (work in progress). 3104 [LISP-MS] Farinacci, D. and V. Fuller, "LISP Map Server", 3105 draft-ietf-lisp-ms-12.txt (work in progress). 3107 [RFC0768] Postel, J., "User Datagram Protocol", STD 6, RFC 768, 3108 August 1980. 3110 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 3111 September 1981. 3113 [RFC1918] Rekhter, Y., Moskowitz, R., Karrenberg, D., Groot, G., and 3114 E. Lear, "Address Allocation for Private Internets", 3115 BCP 5, RFC 1918, February 1996. 3117 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 3118 Requirement Levels", BCP 14, RFC 2119, March 1997. 3120 [RFC2404] Madson, C. and R. Glenn, "The Use of HMAC-SHA-1-96 within 3121 ESP and AH", RFC 2404, November 1998. 3123 [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 3124 (IPv6) Specification", RFC 2460, December 1998. 3126 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 3127 of Explicit Congestion Notification (ECN) to IP", 3128 RFC 3168, September 2001. 3130 [RFC3232] Reynolds, J., "Assigned Numbers: RFC 1700 is Replaced by 3131 an On-line Database", RFC 3232, January 2002. 3133 [RFC4086] Eastlake, D., Schiller, J., and S. Crocker, "Randomness 3134 Requirements for Security", BCP 106, RFC 4086, June 2005. 3136 [RFC4632] Fuller, V. and T. Li, "Classless Inter-domain Routing 3137 (CIDR): The Internet Address Assignment and Aggregation 3138 Plan", BCP 122, RFC 4632, August 2006. 3140 [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an 3141 IANA Considerations Section in RFCs", BCP 26, RFC 5226, 3142 May 2008. 3144 [RFC5496] Wijnands, IJ., Boers, A., and E. Rosen, "The Reverse Path 3145 Forwarding (RPF) Vector TLV", RFC 5496, March 2009. 3147 [RFC5944] Perkins, C., "IP Mobility Support for IPv4, Revised", 3148 RFC 5944, November 2010. 3150 [RFC6115] Li, T., "Recommendation for a Routing Architecture", 3151 RFC 6115, February 2011. 3153 [RFC6234] Eastlake, D. and T. Hansen, "US Secure Hash Algorithms 3154 (SHA and SHA-based HMAC and HKDF)", RFC 6234, May 2011. 3156 [RFC6275] Perkins, C., Johnson, D., and J. Arkko, "Mobility Support 3157 in IPv6", RFC 6275, July 2011. 3159 [UDP-TUNNELS] 3160 Eubanks, M. and P. Chimento, "UDP Checksums for Tunneled 3161 Packets", draft-eubanks-chimento-6man-01.txt (work in 3162 progress), October 2010. 3164 [UDP-ZERO] 3165 Fairhurst, G. and M. Westerland, "IPv6 UDP Checksum 3166 Considerations", draft-ietf-6man-udpzero-04.txt (work in 3167 progress), October 2011. 3169 [VERSIONING] 3170 Iannone, L., Saucez, D., and O. Bonaventure, "LISP Mapping 3171 Versioning", draft-ietf-lisp-map-versioning-05.txt (work 3172 in progress). 3174 16.2. Informative References 3176 [AFI] IANA, "Address Family Indicators (AFIs)", ADDRESS FAMILY 3177 NUMBERS 3178 http://www.iana.org/assignments/address-family-numbers. 3180 [AFI-REGISTRY] 3181 IANA, "Address Family Indicators (AFIs)", ADDRESS FAMILY 3182 NUMBER registry http://www.iana.org/assignments/ 3183 address-family-numbers/ 3184 address-family-numbers.xml#address-family-numbers-1. 3186 [BGP-SEC] Lepinski, M., "An Overview of BGPSEC", 3187 draft-lepinski-bgpsec-overview-00.txt (work in progress), 3188 March 2011. 3190 [CHIAPPA] Chiappa, J., "Endpoints and Endpoint names: A Proposed 3191 Enhancement to the Internet Architecture", Internet- 3192 Draft http://www.chiappa.net/~jnc/tech/endpoints.txt. 3194 [CONS] Farinacci, D., Fuller, V., and D. Meyer, "LISP-CONS: A 3195 Content distribution Overlay Network Service for LISP", 3196 draft-meyer-lisp-cons-04.txt (work in progress). 3198 [EMACS] Brim, S., Farinacci, D., Meyer, D., and J. Curran, "EID 3199 Mappings Multicast Across Cooperating Systems for LISP", 3200 draft-curran-lisp-emacs-00.txt (work in progress). 3202 [INTERWORK] 3203 Lewis, D., Meyer, D., Farinacci, D., and V. Fuller, 3204 "Interworking LISP with IPv4 and IPv6", 3205 draft-ietf-lisp-interworking-02.txt (work in progress). 3207 [KARP] Lebovitz, G. and M. Bhatia, "Keying and Authentication for 3208 Routing Protocols (KARP)Design Guidelines", 3209 draft-ietf-karp-design-guide-06.txt (work in progress), 3210 October 2011. 3212 [LCAF] Farinacci, D., Meyer, D., and J. Snijders, "LISP Canonical 3213 Address Format", draft-farinacci-lisp-lcaf-06.txt (work in 3214 progress). 3216 [LISA96] Lear, E., Katinsky, J., Coffin, J., and D. Tharp, 3217 "Renumbering: Threat or Menace?", Usenix . 3219 [LISP-DEPLOY] 3220 Jakab, L., Coras, F., Domingo-Pascual, J., and D. Lewis, 3221 "LISP Network Element Deployment Considerations", 3222 draft-ietf-lisp-deployment-02.txt (work in progress). 3224 [LISP-LIG] 3225 Farinacci, D. and D. Meyer, "LISP Internet Groper (LIG)", 3226 draft-ietf-lisp-lig-06.txt (work in progress). 3228 [LISP-MAIN] 3229 Farinacci, D., Fuller, V., Meyer, D., and D. Lewis, 3230 "Locator/ID Separation Protocol (LISP)", 3231 draft-farinacci-lisp-12.txt (work in progress). 3233 [LISP-MIB] 3234 Schudel, G., Jain, A., and V. Moreno, "LISP MIB", 3235 draft-ietf-lisp-mib-02.txt (work in progress). 3237 [LISP-MN] Farinacci, D., Fuller, V., Lewis, D., and D. Meyer, "LISP 3238 Mobility Architecture", draft-meyer-lisp-mn-06.txt (work 3239 in progress). 3241 [LISP-SEC] 3242 Maino, F., Ermagon, V., Cabellos, A., Sausez, D., and O. 3243 Bonaventure, "LISP-Security (LISP-SEC)", 3244 draft-ietf-lisp-sec-00.txt (work in progress). 3246 [LOC-ID-ARCH] 3247 Meyer, D. and D. Lewis, "Architectural Implications of 3248 Locator/ID Separation", 3249 draft-meyer-loc-id-implications-02.txt (work in progress). 3251 [MLISP] Farinacci, D., Meyer, D., Zwiebel, J., and S. Venaas, 3252 "LISP for Multicast Environments", 3253 draft-ietf-lisp-multicast-10.txt (work in progress). 3255 [NERD] Lear, E., "NERD: A Not-so-novel EID to RLOC Database", 3256 draft-lear-lisp-nerd-08.txt (work in progress). 3258 [OPENLISP] 3259 Iannone, L. and O. Bonaventure, "OpenLISP Implementation 3260 Report", draft-iannone-openlisp-implementation-01.txt 3261 (work in progress). 3263 [RADIR] Narten, T., "Routing and Addressing Problem Statement", 3264 draft-narten-radir-problem-statement-05.txt (work in 3265 progress). 3267 [RFC1034] Mockapetris, P., "Domain names - concepts and facilities", 3268 STD 13, RFC 1034, November 1987. 3270 [RFC2784] Farinacci, D., Li, T., Hanks, S., Meyer, D., and P. 3271 Traina, "Generic Routing Encapsulation (GRE)", RFC 2784, 3272 March 2000. 3274 [RFC3056] Carpenter, B. and K. Moore, "Connection of IPv6 Domains 3275 via IPv4 Clouds", RFC 3056, February 2001. 3277 [RFC3261] Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston, 3278 A., Peterson, J., Sparks, R., Handley, M., and E. 3279 Schooler, "SIP: Session Initiation Protocol", RFC 3261, 3280 June 2002. 3282 [RFC4192] Baker, F., Lear, E., and R. Droms, "Procedures for 3283 Renumbering an IPv6 Network without a Flag Day", RFC 4192, 3284 September 2005. 3286 [RFC4866] Arkko, J., Vogt, C., and W. Haddad, "Enhanced Route 3287 Optimization for Mobile IPv6", RFC 4866, May 2007. 3289 [RFC4984] Meyer, D., Zhang, L., and K. Fall, "Report from the IAB 3290 Workshop on Routing and Addressing", RFC 4984, 3291 September 2007. 3293 [RPKI] Lepinski, M., "An Infrastructure to Support Secure 3294 Internet Routing", draft-ietf-sidr-arch-13.txt (work in 3295 progress), February 2011. 3297 Appendix A. Acknowledgments 3299 An initial thank you goes to Dave Oran for planting the seeds for the 3300 initial ideas for LISP. His consultation continues to provide value 3301 to the LISP authors. 3303 A special and appreciative thank you goes to Noel Chiappa for 3304 providing architectural impetus over the past decades on separation 3305 of location and identity, as well as detailed review of the LISP 3306 architecture and documents, coupled with enthusiasm for making LISP a 3307 practical and incremental transition for the Internet. 3309 The authors would like to gratefully acknowledge many people who have 3310 contributed discussion and ideas to the making of this proposal. 3311 They include Scott Brim, Andrew Partan, John Zwiebel, Jason Schiller, 3312 Lixia Zhang, Dorian Kim, Peter Schoenmaker, Vijay Gill, Geoff Huston, 3313 David Conrad, Mark Handley, Ron Bonica, Ted Seely, Mark Townsley, 3314 Chris Morrow, Brian Weis, Dave McGrew, Peter Lothberg, Dave Thaler, 3315 Eliot Lear, Shane Amante, Ved Kafle, Olivier Bonaventure, Luigi 3316 Iannone, Robin Whittle, Brian Carpenter, Joel Halpern, Terry 3317 Manderson, Roger Jorgensen, Ran Atkinson, Stig Venaas, Iljitsch van 3318 Beijnum, Roland Bless, Dana Blair, Bill Lynch, Marc Woolward, Damien 3319 Saucez, Damian Lezama, Attilla De Groot, Parantap Lahiri, David 3320 Black, Roque Gagliano, Isidor Kouvelas, Jesper Skriver, Fred Templin, 3321 Margaret Wasserman, Sam Hartman, Michael Hofling, Pedro Marques, Jari 3322 Arkko, Gregg Schudel, Srinivas Subramanian, Amit Jain, Xu Xiaohu, 3323 Dhirendra Trivedi, Yakov Rekhter, John Scudder, John Drake, Dimitri 3324 Papadimitriou, Ross Callon, Selina Heimlich, Job Snijders, Vina 3325 Ermagan, Albert Cabellos, Fabio Maino, Victor Moreno, Chris White, 3326 Clarence Filsfils, and Alia Atlas. 3328 This work originated in the Routing Research Group (RRG) of the IRTF. 3329 The individual submission [LISP-MAIN] was converted into this IETF 3330 LISP working group draft. 3332 The LISP working group would like to give a special thanks to Jari 3333 Arkko, the Internet Area AD at the time the set of LISP documents 3334 were being prepared for IESG last call, for his meticulous review and 3335 detail commentary on the 7 working group last call drafts progressing 3336 toward experimental RFCs. 3338 Appendix B. Document Change Log 3340 B.1. Changes to draft-ietf-lisp-21.txt 3342 o Posted February 2012 to reflect DISCUSS comments from Adrian 3343 Farrel, Stewart Bryant, and Wesley Eddy. 3345 B.2. Changes to draft-ietf-lisp-20.txt 3347 o Posted January 2012 for resolution to Adrian Farrel's security 3348 comments as well as additions to the end of section 2, Elwyn 3349 Davies Gen-Art comments, and Ralph Droms' IANA and EID definition 3350 comments. 3352 B.3. Changes to draft-ietf-lisp-19.txt 3354 o Posted January 2012 for Stephen Farrell's comment resolution. 3356 B.4. Changes to draft-ietf-lisp-18.txt 3358 o Posted December 2011 after reflecting comments from IANA. 3360 o Create reference to sections 5.4.1 and 5.4.2 about DF bit setting 3361 from section 5.3. 3363 o Inserted two references for Route-Returnability and on-path 3364 attacks in Security Considerations section. 3366 B.5. Changes to draft-ietf-lisp-17.txt 3368 o Posted December 2011 after IETF last call comments. 3370 o Make Map-Notify port assignment be 4342 in both source and 3371 destination ports. This change was agreed on and put in [LISP-MS] 3372 but was not updated in this spec. 3374 B.6. Changes to draft-ietf-lisp-16.txt 3376 o Posted October 2011 after AD review by Jari. 3378 B.7. Changes to draft-ietf-lisp-15.txt 3380 o Posted July 2011. Fixing IDnits errors. 3382 o Change description on how to select a source address for RLOC- 3383 probe Map-Replies to refer to the "EID-to-RLOC Map-Reply Message" 3384 section. 3386 B.8. Changes to draft-ietf-lisp-14.txt 3388 o Post working group last call and pre-IESG last call review. 3390 o Indicate that an ICMP Unreachable message should be sent when a 3391 packet matches a drop-based negative map-cache entry. 3393 o Indicate how a map-cache set of overlapping EID-prefixes must 3394 maintain integrity when the map-cache maximum cap is reached. 3396 o Add Joel's description for the definition of an EID, that the bit 3397 string value can be an RLOC for another device in abstract but the 3398 architecture allows it to be an EID of one device and the same 3399 value as an RLOC for another device. 3401 o In the "Tunnel Encapsulation Details" section, indicate that 4 3402 combinations of encapsulation are supported. 3404 o Add what ETR should do for a Data-Probe when received for a 3405 destination EID outside of its EID-prefix range. This was added 3406 in the Data Probe definition section. 3408 o Added text indicating that more-specific EID-prefixes must not be 3409 removed when less-specific entries stay in the map-cache. This is 3410 to preserve the integrity of the EID-prefix set. 3412 o Add clarifying text in the Security Considerations section about 3413 how an ETR must not decapsulate and forward a packet that is not 3414 for its configured EID-prefix range. 3416 B.9. Changes to draft-ietf-lisp-13.txt 3418 o Posted June 2011 to complete working group last call. 3420 o Tracker item 87. Put Yakov suggested wording in the EID-prefix 3421 definition section to reference [INTERWORK] and [LISP-DEPLOY] 3422 about discussion on transition and access mechanisms. 3424 o Change "ITRs" to "ETRs" in the Locator Status Bit definition 3425 section and data packet description section per Damien's comment. 3427 o Remove the normative reference to [LISP-SEC] when describing the 3428 S-bit in the ECM and Map-Reply headers. 3430 o Tracker item 54. Added text from John Scudder in the "Packets 3431 Egressing a LISP Site" section. 3433 o Add sentence to the "Reencapsulating Tunnel" definition about how 3434 reencapsulation loops can occur when not coordinating among 3435 multiple mapping database systems. 3437 o Remove "In theory" from a sentence in the Security Considerations 3438 section. 3440 o Remove Security Area Statement title and reword section with 3441 Eliot's provided text. The text was agreed upon by LISP-WG chairs 3442 and Security ADs. 3444 o Remove word "potential" from the over-claiming paragraph of the 3445 Security Considerations section per Stephen's request. 3447 o Wordsmithing and other editorial comments from Alia. 3449 B.10. Changes to draft-ietf-lisp-12.txt 3451 o Posted April 2011. 3453 o Tracker item 87. Provided rewording how an EID-prefix can be 3454 reused in the definition section of "EID-prefix". 3456 o Tracker item 95. Change "eliminate" to "defer" in section 4.1. 3458 o Tracker item 110. Added that the Mapping Protocol Data field in 3459 the Map-Reply message is only used when needed by the particular 3460 Mapping Database System. 3462 o Tracker item 111. Indicate that if an LSB that is associated with 3463 an anycast address, that there is at least one RLOC that is up. 3465 o Tracker item 108. Make clear the R-bit does not define RLOC path 3466 reachability. 3468 o Tracker item 107. Indicate that weights are relative to each 3469 other versus requiring an addition of up to 100%. 3471 o Tracker item 46. Add a sentence how LISP products should be sized 3472 for the appropriate demand so cache thrashing is avoided. 3474 o Change some references of RFC 5226 to [AFI] per Luigi. 3476 o Per Luigi, make reference to "EID-AFI" consistent to "EID-prefix- 3477 AFI". 3479 o Tracker item 66. Indicate that appending locators to a locator- 3480 set is done when the added locators are lexicographically greater 3481 than the previous ones in the set. 3483 o Tracker item 87. Once again reword the definition of the EID- 3484 prefix to reflect recent comments. 3486 o Tracker item 70. Added text to security section on what the 3487 implications could be if an ITR does not obey priority and weights 3488 from a Map-Reply message. 3490 o Tracker item 54. Added text to the new section titled "Packets 3491 Egressing a LISP Site" to describe the implications when two or 3492 more ITRs exist at a site where only one ITR is used for egress 3493 traffic and when there is a shift of traffic to the others, how 3494 the map-cache will need to be populated in those new egress ITRs. 3496 o Tracker item 33. Make more clear in the Routing Locator Selection 3497 section what an ITR should do when it sees an R-bit of 0 in a 3498 locator-record of a Map-Reply. 3500 o Tracker item 33. Add paragraph to the EID Reachability section 3501 indicating that site partitioning is under investigation. 3503 o Tracker item 58. Added last paragraph of Security Considerations 3504 section about how to protect inner header EID address spoofing 3505 attacks. 3507 o Add suggested Sam text to indicate that all security concerns need 3508 not be addressed for moving document to Experimental RFC status. 3509 Put this in a subsection of the Security Considerations section. 3511 B.11. Changes to draft-ietf-lisp-11.txt 3513 o Posted March 30, 2011. 3515 o Change IANA URL. The URL we had pointed to a general protocol 3516 numbers page. 3518 o Added the "s" bit to the Map-Request to allow SMR-invoked Map- 3519 Requests to be sent to a MN ETR via the map-server. 3521 o Generalize text for the definition of Reencapsuatling tunnels. 3523 o Add paragraph suggested by Joel to explain how implementation 3524 experimentation will be used to determine the proper cache 3525 management techniques. 3527 o Add Yakov provided text for the definition of "EID-to-RLOC 3528 "Database". 3530 o Add reference in Section 8, Deployment Scenarios, to the 3531 draft-jakab-lisp-deploy-02.txt draft. 3533 o Clarify sentence about no hardware changes needed to support LISP 3534 encapsulation. 3536 o Add paragraph about what is the procedure when a locator is 3537 inserted in the middle of a locator-set. 3539 o Add a definition for Locator Status Bits so we can emphasize they 3540 are used as a hint for router up/down status and not path 3541 reachability. 3543 o Change "BGP RIB" to "RIB" per Clarence's comment. 3545 o Fixed complaints by IDnits. 3547 o Add subsection to Security Considerations section indicating how 3548 EID-prefix overclaiming in Map-Replies is for further study and 3549 add a reference to LISP-SEC. 3551 B.12. Changes to draft-ietf-lisp-10.txt 3553 o Posted March 2011. 3555 o Add p-bit to Map-Request so there is documentary reasons to know 3556 when a PITR has sent a Map-Request to an ETR. 3558 o Add Map-Notify message which is used to acknowledge a Map-Register 3559 message sent to a Map-Server. 3561 o Add M-bit to the Map-Register message so an ETR that wants an 3562 acknowledgment for the Map-Register can request one. 3564 o Add S-bit to the ECM and Map-Reply messages to describe security 3565 data that can be present in each message. Then refer to 3566 [LISP-SEC] for expansive details. 3568 o Add Network Management Considerations section and point to the MIB 3569 and LIG drafts. 3571 o Remove the word "simple" per Yakov's comments. 3573 B.13. Changes to draft-ietf-lisp-09.txt 3575 o Posted October 2010. 3577 o Add to IANA Consideration section about the use of LCAF Type 3578 values that accepted and maintained by the IANA registry and not 3579 the LCAF specification. 3581 o Indicate that implementations should be able to receive LISP 3582 control messages when either UDP port is 4342, so they can be 3583 robust in the face of intervening NAT boxes. 3585 o Add paragraph to SMR section to indicate that an ITR does not need 3586 to respond to an SMR-based Map-Request when it has no map-cache 3587 entry for the SMR source's EID-prefix. 3589 B.14. Changes to draft-ietf-lisp-08.txt 3591 o Posted August 2010. 3593 o In section 6.1.6, remove statement about setting TTL to 0 in Map- 3594 Register messages. 3596 o Clarify language in section 6.1.5 about Map-Replying to Data- 3597 Probes or Map-Requests. 3599 o Indicate that outer TTL should only be copied to inner TTL when it 3600 is less than inner TTL. 3602 o Indicate a source-EID for RLOC-probes are encoded with an AFI 3603 value of 0. 3605 o Indicate that SMRs can have a global or per SMR destination rate- 3606 limiter. 3608 o Add clarifications to the SMR procedures. 3610 o Add definitions for "client-side" and 'server-side" terms used in 3611 this specification. 3613 o Clear up language in section 6.4, last paragraph. 3615 o Change ACT of value 0 to "no-action". This is so we can RLOC- 3616 probe a PETR and have it return a Map-Reply with a locator-set of 3617 size 0. The way it is spec'ed the map-cache entry has action 3618 "dropped". Drop-action is set to 3. 3620 o Add statement about normalizing locator weights. 3622 o Clarify R-bit definition in the Map-Reply locator record. 3624 o Add section on EID Reachability within a LISP site. 3626 o Clarify another disadvantage of using anycast locators. 3628 o Reworded Abstract. 3630 o Change section 2.0 Introduction to remove obsolete information 3631 such as the LISP variant definitions. 3633 o Change section 5 title from "Tunneling Details" to "LISP 3634 Encapsulation Details". 3636 o Changes to section 5 to include results of network deployment 3637 experience with MTU. Recommend that implementations use either 3638 the stateful or stateless handling. 3640 o Make clarification wordsmithing to Section 7 and 8. 3642 o Identify that if there is one locator in the locator-set of a map- 3643 cache entry, that an SMR from that locator should be responded to 3644 by sending the the SMR-invoked Map-Request to the database mapping 3645 system rather than to the RLOC itself (which may be unreachable). 3647 o When describing Unicast and Multicast Weights indicate the the 3648 values are relative weights rather than percentages. So it 3649 doesn't imply the sum of all locator weights in the locator-set 3650 need to be 100. 3652 o Do some wordsmithing on copying TTL and TOS fields. 3654 o Numerous wordsmithing changes from Dave Meyer. He fine toothed 3655 combed the spec. 3657 o Removed Section 14 "Prototype Plans and Status". We felt this 3658 type of section is no longer appropriate for a protocol 3659 specification. 3661 o Add clarification text for the IRC description per Damien's 3662 commentary. 3664 o Remove text on copying nonce from SMR to SMR-invoked Map- Request 3665 per Vina's comment about a possible DoS vector. 3667 o Clarify (S/2 + H) in the stateless MTU section. 3669 o Add text to reflect Damien's comment about the description of the 3670 "ITR-RLOC Address" field in the Map-Request. that the list of RLOC 3671 addresses are local addresses of the Map-Requester. 3673 B.15. Changes to draft-ietf-lisp-07.txt 3675 o Posted April 2010. 3677 o Added I-bit to data header so LSB field can also be used as an 3678 Instance ID field. When this occurs, the LSB field is reduced to 3679 8-bits (from 32-bits). 3681 o Added V-bit to the data header so the 24-bit nonce field can also 3682 be used for source and destination version numbers. 3684 o Added Map-Version 12-bit value to the EID-record to be used in all 3685 of Map-Request, Map-Reply, and Map-Register messages. 3687 o Added multiple ITR-RLOC fields to the Map-Request packet so an ETR 3688 can decide what address to select for the destination of a Map- 3689 Reply. 3691 o Added L-bit (Local RLOC bit) and p-bit (Probe-Reply RLOC bit) to 3692 the Locator-Set record of an EID-record for a Map-Reply message. 3693 The L-bit indicates which RLOCs in the locator-set are local to 3694 the sender of the message. The P-bit indicates which RLOC is the 3695 source of a RLOC-probe Reply (Map-Reply) message. 3697 o Add reference to the LISP Canonical Address Format [LCAF] draft. 3699 o Made editorial and clarification changes based on comments from 3700 Dhirendra Trivedi. 3702 o Added wordsmithing comments from Joel Halpern on DF=1 setting. 3704 o Add John Zwiebel clarification to Echo Nonce Algorithm section 3705 6.3.1. 3707 o Add John Zwiebel comment about expanding on proxy-map-reply bit 3708 for Map-Register messages. 3710 o Add NAT section per Ron Bonica comments. 3712 o Fix IDnits issues per Ron Bonica. 3714 o Added section on Virtualization and Segmentation to explain the 3715 use if the Instance ID field in the data header. 3717 o There are too many P-bits, keep their scope to the packet format 3718 description and refer to them by name every where else in the 3719 spec. 3721 o Scanned all occurrences of "should", "should not", "must" and 3722 "must not" and uppercased them. 3724 o John Zwiebel offered text for section 4.1 to modernize the 3725 example. Thanks Z! 3727 o Make it more clear in the definition of "EID-to-RLOC Database" 3728 that all ETRs need to have the same database mapping. This 3729 reflects a comment from John Scudder. 3731 o Add a definition "Route-returnability" to the Definition of Terms 3732 section. 3734 o In section 9.2, add text to describe what the signature of 3735 traceroute packets can look like. 3737 o Removed references to Data Probe for introductory example. Data- 3738 probes are still part of the LISP design but not encouraged. 3740 o Added the definition for "LISP site" to the Definition of Terms" 3741 section. 3743 B.16. Changes to draft-ietf-lisp-06.txt 3745 Editorial based changes: 3747 o Posted December 2009. 3749 o Fix typo for flags in LISP data header. Changed from "4" to "5". 3751 o Add text to indicate that Map-Register messages must contain a 3752 computed UDP checksum. 3754 o Add definitions for PITR and PETR. 3756 o Indicate an AFI value of 0 is an unspecified address. 3758 o Indicate that the TTL field of a Map-Register is not used and set 3759 to 0 by the sender. This change makes this spec consistent with 3760 [LISP-MS]. 3762 o Change "... yield a packet size of L octets" to "... yield a 3763 packet size greater than L octets". 3765 o Clarify section 6.1.5 on what addresses and ports are used in Map- 3766 Reply messages. 3768 o Clarify that LSBs that go beyond the number of locators do not to 3769 be SMRed when the locator addresses are greater lexicographically 3770 than the locator in the existing locator-set. 3772 o Add Gregg, Srini, and Amit to acknowledgment section. 3774 o Clarify in the definition of a LISP header what is following the 3775 UDP header. 3777 o Clarify "verifying Map-Request" text in section 6.1.3. 3779 o Add Xu Xiaohu to the acknowledgment section for introducing the 3780 problem of overlapping EID-prefixes among multiple sites in an RRG 3781 email message. 3783 Design based changes: 3785 o Use stronger language to have the outer IPv4 header set DF=1 so we 3786 can avoid fragment reassembly in an ETR or PETR. This will also 3787 make IPv4 and IPv6 encapsulation have consistent behavior. 3789 o Map-Requests should not be sent in ECM with the Probe bit is set. 3790 These type of Map-Requests are used as RLOC-probes and are sent 3791 directly to locator addresses in the underlying network. 3793 o Add text in section 6.1.5 about returning all EID-prefixes in a 3794 Map-Reply sent by an ETR when there are overlapping EID-prefixes 3795 configure. 3797 o Add text in a new subsection of section 6.1.5 about dealing with 3798 Map-Replies with coarse EID-prefixes. 3800 B.17. Changes to draft-ietf-lisp-05.txt 3802 o Posted September 2009. 3804 o Added this Document Change Log appendix. 3806 o Added section indicating that encapsulated Map-Requests must use 3807 destination UDP port 4342. 3809 o Don't use AH in Map-Registers. Put key-id, auth-length, and auth- 3810 data in Map-Register payload. 3812 o Added Jari to acknowledgment section. 3814 o State the source-EID is set to 0 when using Map-Requests to 3815 refresh or RLOC-probe. 3817 o Make more clear what source-RLOC should be for a Map-Request. 3819 o The LISP-CONS authors thought that the Type definitions for CONS 3820 should be removed from this specification. 3822 o Removed nonce from Map-Register message, it wasn't used so no need 3823 for it. 3825 o Clarify what to do for unspecified Action bits for negative Map- 3826 Replies. Since No Action is a drop, make value 0 Drop. 3828 B.18. Changes to draft-ietf-lisp-04.txt 3830 o Posted September 2009. 3832 o How do deal with record count greater than 1 for a Map-Request. 3833 Damien and Joel comment. Joel suggests: 1) Specify that senders 3834 compliant with the current document will always set the count to 3835 1, and note that the count is included for future extensibility. 3836 2) Specify what a receiver compliant with the draft should do if 3837 it receives a request with a count greater than 1. Presumably, it 3838 should send some error back? 3840 o Add Fred Templin in acknowledgment section. 3842 o Add Margaret and Sam to the acknowledgment section for their great 3843 comments. 3845 o Say more about LAGs in the UDP section per Sam Hartman's comment. 3847 o Sam wants to use MAY instead of SHOULD for ignoring checksums on 3848 ETR. From the mailing list: "You'd need to word it as an ITR MAY 3849 send a zero checksum, an ETR MUST accept a 0 checksum and MAY 3850 ignore the checksum completely. And of course we'd need to 3851 confirm that can actually be implemented. In particular, hardware 3852 that verifies UDP checksums on receive needs to be checked to make 3853 sure it permits 0 checksums." 3855 o Margaret wants a reference to 3856 http://www.ietf.org/id/draft-eubanks-chimento-6man-00.txt. 3858 o Fix description in Map-Request section. Where we describe Map- 3859 Reply Record, change "R-bit" to "M-bit". 3861 o Add the mobility bit to Map-Replies. So PITRs don't probe so 3862 often for MNs but often enough to get mapping updates. 3864 o Indicate SHA1 can be used as well for Map-Registers. 3866 o More Fred comments on MTU handling. 3868 o Isidor comment about spec'ing better periodic Map-Registers. Will 3869 be fixed in draft-ietf-lisp-ms-02.txt. 3871 o Margaret's comment on gleaning: "The current specification does 3872 not make it clear how long gleaned map entries should be retained 3873 in the cache, nor does it make it clear how/ when they will be 3874 validated. The LISP spec should, at the very least, include a 3875 (short) default lifetime for gleaned entries, require that they be 3876 validated within a short period of time, and state that a new 3877 gleaned entry should never overwrite an entry that was obtained 3878 from the mapping system. The security implications of storing 3879 "gleaned" entries should also be explored in detail." 3881 o Add section on RLOC-probing per working group feedback. 3883 o Change "loc-reach-bits" to "loc-status-bits" per comment from 3884 Noel. 3886 o Remove SMR-bit from data-plane. Dino prefers to have it in the 3887 control plane only. 3889 o Change LISP header to allow a "Research Bit" so the Nonce and LSB 3890 fields can be turned off and used for another future purpose. For 3891 Luigi et al versioning convergence. 3893 o Add a N-bit to the data header suggested by Noel. Then the nonce 3894 field could be used when N is not 1. 3896 o Clarify that when E-bit is 0, the nonce field can be an echoed 3897 nonce or a random nonce. Comment from Jesper. 3899 o Indicate when doing data-gleaning that a verifying Map-Request is 3900 sent to the source-EID of the gleaned data packet so we can avoid 3901 map-cache corruption by a 3rd party. Comment from Pedro. 3903 o Indicate that a verifying Map-Request, for accepting mapping data, 3904 should be sent over the ALT (or to the EID). 3906 o Reference IPsec RFC 4302. Comment from Sam and Brian Weis. 3908 o Put E-bit in Map-Reply to tell ITRs that the ETR supports echo- 3909 noncing. Comment by Pedro and Dino. 3911 o Jesper made a comment to loosen the language about requiring the 3912 copy of inner TTL to outer TTL since the text to get mixed-AF 3913 traceroute to work would violate the "MUST" clause. Changed from 3914 MUST to SHOULD in section 5.3. 3916 B.19. Changes to draft-ietf-lisp-03.txt 3918 o Posted July 2009. 3920 o Removed loc-reach-bits longword from control packets per Damien 3921 comment. 3923 o Clarifications in MTU text from Roque. 3925 o Added text to indicate that the locator-set be sorted by locator 3926 address from Isidor. 3928 o Clarification text from John Zwiebel in Echo-Nonce section. 3930 B.20. Changes to draft-ietf-lisp-02.txt 3932 o Posted July 2009. 3934 o Encapsulation packet format change to add E-bit and make loc- 3935 reach-bits 32-bits in length. 3937 o Added Echo-Nonce Algorithm section. 3939 o Clarification how ECN bits are copied. 3941 o Moved S-bit in Map-Request. 3943 o Added P-bit in Map-Request and Map-Reply messages to anticipate 3944 RLOC-Probe Algorithm. 3946 o Added to Mobility section to reference [LISP-MN]. 3948 B.21. Changes to draft-ietf-lisp-01.txt 3950 o Posted 2 days after draft-ietf-lisp-00.txt in May 2009. 3952 o Defined LEID to be a "LISP EID". 3954 o Indicate encapsulation use IPv4 DF=0. 3956 o Added negative Map-Reply messages with drop, native-forward, and 3957 send-map-request actions. 3959 o Added Proxy-Map-Reply bit to Map-Register. 3961 B.22. Changes to draft-ietf-lisp-00.txt 3963 o Posted May 2009. 3965 o Rename of draft-farinacci-lisp-12.txt. 3967 o Acknowledgment to RRG. 3969 Authors' Addresses 3971 Dino Farinacci 3972 cisco Systems 3973 Tasman Drive 3974 San Jose, CA 95134 3975 USA 3977 Email: dino@cisco.com 3979 Vince Fuller 3980 cisco Systems 3981 Tasman Drive 3982 San Jose, CA 95134 3983 USA 3985 Email: vaf@cisco.com 3987 Dave Meyer 3988 cisco Systems 3989 170 Tasman Drive 3990 San Jose, CA 3991 USA 3993 Email: dmm@cisco.com 3995 Darrel Lewis 3996 cisco Systems 3997 170 Tasman Drive 3998 San Jose, CA 3999 USA 4001 Email: darlewis@cisco.com