idnits 2.17.1 draft-ietf-lisp-map-versioning-02.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 : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (July 5, 2011) is 4680 days in the past. Is this intentional? Checking references for intended status: Experimental ---------------------------------------------------------------------------- == Outdated reference: A later version (-24) exists of draft-ietf-lisp-14 == Outdated reference: A later version (-10) exists of draft-ietf-lisp-alt-07 == Outdated reference: A later version (-06) exists of draft-ietf-lisp-interworking-02 == Outdated reference: A later version (-16) exists of draft-ietf-lisp-ms-09 == Outdated reference: A later version (-15) exists of draft-ietf-lisp-threats-00 Summary: 0 errors (**), 0 flaws (~~), 6 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group L. Iannone 3 Internet-Draft TU Berlin - Deutsche Telekom 4 Intended status: Experimental Laboratories AG 5 Expires: January 6, 2012 D. Saucez 6 O. Bonaventure 7 Universite catholique de Louvain 8 July 5, 2011 10 LISP Map-Versioning 11 draft-ietf-lisp-map-versioning-02.txt 13 Abstract 15 This document describes the LISP Map-Versioning mechanism, which 16 provides in-packet information about EID-to-RLOC mappings used to 17 encapsulate LISP data packets. The proposed approach is based on 18 associating a version number to EID-to-RLOC mappings and transport 19 such a version number in the LISP specific header of LISP- 20 encapsulated packets. LISP Map-Versioning is particularly useful to 21 inform communicating xTRs about modifications of the mappings used to 22 encapsulate packets. The mechanism is transparent to legacy 23 implementations, since in the LISP-specific header and in the Map 24 Records, bits used for Map-Versioning can be safely ignored by xTRs 25 that do not support the mechanism. 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 January 6, 2012. 44 Copyright Notice 46 Copyright (c) 2011 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. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 62 2. Requirements Notation . . . . . . . . . . . . . . . . . . . . 4 63 3. Definitions of Terms . . . . . . . . . . . . . . . . . . . . . 4 64 4. EID-to-RLOC Map-Version number . . . . . . . . . . . . . . . . 4 65 4.1. The Null Map-Version . . . . . . . . . . . . . . . . . . . 5 66 5. Dealing with Map-Version numbers . . . . . . . . . . . . . . . 6 67 5.1. Handling Destination Map-Version number . . . . . . . . . 7 68 5.2. Handling Source Map-Version number . . . . . . . . . . . . 8 69 6. LISP header and Map-Version numbers . . . . . . . . . . . . . 9 70 7. Map Record and Map-Version . . . . . . . . . . . . . . . . . . 10 71 8. Benefits and case studies for Map-Versioning . . . . . . . . . 11 72 8.1. Synchronization of different xTRs . . . . . . . . . . . . 11 73 8.2. Map-Versioning and unidirectional traffic . . . . . . . . 12 74 8.3. Map-Versioning and interworking . . . . . . . . . . . . . 12 75 8.3.1. Map-Versioning and Proxy-ITRs . . . . . . . . . . . . 13 76 8.3.2. Map-Versioning and LISP-NAT . . . . . . . . . . . . . 13 77 8.3.3. Map-Versioning and Proxy-ETRs . . . . . . . . . . . . 13 78 8.4. RLOC shutdown/withdraw . . . . . . . . . . . . . . . . . . 14 79 8.5. Map-Version for lightweight LISP implementation . . . . . 14 80 9. Incremental deployment and implementation status . . . . . . . 15 81 10. Security Considerations . . . . . . . . . . . . . . . . . . . 15 82 10.1. Map-Versioning against traffic disruption . . . . . . . . 16 83 10.2. Map-Versioning against reachability information DoS . . . 16 84 11. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 17 85 12. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 17 86 13. References . . . . . . . . . . . . . . . . . . . . . . . . . . 17 87 13.1. Normative References . . . . . . . . . . . . . . . . . . . 17 88 13.2. Informative References . . . . . . . . . . . . . . . . . . 17 89 Appendix A. Estimation of time before Map-Version wrap-around . . 18 90 Appendix B. Document Change Log . . . . . . . . . . . . . . . . . 19 91 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 20 93 1. Introduction 95 This document describes the Map-Versioning mechanism used to provide 96 information on changes in the EID-to-RLOC mappings used in the LISP 97 ([I-D.ietf-lisp]) context to perform packet encapsulation. The 98 mechanism is totally transparent to xTRs not supporting such 99 functionality. It is not meant to replace any existing LISP 100 mechanism, but rather to complete them providing new functionalities. 101 The basic mechanism is to associate a Map-Version number to each LISP 102 EID-to-RLOC mapping and transport such a version number in the LISP- 103 specific header. When a mapping changes, a new version number is 104 assigned to the updated mapping. A change in an EID-to-RLOC mapping 105 can be a change in the RLOCs set, by adding or removing one or more 106 RLOCs, but it can also be a change in the priority or weight of one 107 or more RLOCs. 109 When Map-Versioning is used, LISP-encapsulated data packets contain 110 the version number of the two mappings used to select the RLOCs in 111 the outer header (i.e., both source and destination). These version 112 numbers are encoded in the 24 low-order bits of the first longword of 113 the LISP header and indicated by a specific bit in the flags (first 8 114 high-order bits of the first longword of the LISP header). Note that 115 not all packets need to carry version numbers. 117 When an ITR encapsulates a data packet, with a LISP header containing 118 the Map-Version numbers, it puts in the LISP-specific header two 119 version numbers: 121 1. The version number assigned to the mapping (contained in the EID- 122 to-RLOC Database) used to select the source RLOC. 124 2. The version number assigned to the mapping (contained in the EID- 125 to-RLOC Cache) used to select the destination RLOC. 127 This operation is two-fold. On the one hand, it enables the ETR 128 receiving the packet to know if the ITR has the latest version number 129 that any ETR at the destination EID site has provided to the ITR in a 130 Map-Reply. If it is not the case the ETR can send to the ITR a Map- 131 Request containing the updated mapping or soliciting a Map-Request 132 from the ITR (both cases are already defined in [I-D.ietf-lisp]). In 133 this way the ITR can update its cache. On the other hand, it enables 134 an ETR receiving such a packet to know if it has in its EID-to-RLOC 135 Cache the latest mapping for the source EID (in case of bidirectional 136 traffic). If it is not the case a Map-Request can be send. 138 2. Requirements Notation 140 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 141 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 142 document are to be interpreted as described in [RFC2119]. 144 3. Definitions of Terms 146 The present document uses terms already defined in main LISP 147 specification [I-D.ietf-lisp]. Hereafter are defined only the terms 148 that are specific to the Map-Versioning mechanism. 150 Map-Version number: An unsigned 12-bits assigned to an EID-to-RLOC 151 mapping, not including the value 0 (0x000). 153 Null Map-Version: The 12-bits null value of 0 (0x000) is not used as 154 Map-Version number. It is used to signal that no Map-Version 155 number is assigned to the EID-to-RLOC mapping. 157 Source Map-Version number: Map-Version number of the EID-to-RLOC 158 mapping used to select the source address (RLOC) of the outer IP 159 header of LISP-encapsulated packets. 161 Destination Map-Version number: Map-Version number of the EID-to- 162 RLOC mapping used to select the destination address (RLOC) of the 163 outer IP header of LISP-encapsulated packets. 165 4. EID-to-RLOC Map-Version number 167 The EID-to-RLOC Map-Version number consists in an unsigned 12-bits 168 integer. The version number is assigned in a per-mapping fashion, 169 meaning that different mappings will have assigned a different 170 version number, which is also updated independently. An update in 171 the version number (i.e., a newer version) consist in incrementing by 172 one the older version number. Appendix A contains a rough estimation 173 of the wrap-around time for the Map-Version number. 175 The space of version numbers has a circular order where half of the 176 version numbers is greater than the current Map-Version number and 177 the other half is smaller than current Map-Version number. In a more 178 formal way, assuming we have two version numbers V1 and V2 and that 179 the numbers are expressed on N bits, the following three cases may 180 happen: 182 V1 = V2 : This is the exact match case. 184 V1 < V2 : True if and only if V1 < V2 < (V1 + 2**(N-1)). 186 V1 > V2 : True if and only if V1 > V2 > (V1 - 2**(N-1)). 188 Using 12 bits, as defined in this document, and assuming a Map- 189 Version value of 69, Map-Version numbers in the range [70; 69 + 2047] 190 are greater than 69, while Map-Version numbers in the range [69 + 191 2048; (69 + 4095) mod 4096] are smaller than 69. 193 The initial Map-Version number of a new EID-to-RLOC mapping SHOULD be 194 randomly generated. However, it MUST NOT be set to the Null Map- 195 Version value (0x000), because it has a special meaning (see 196 Section 4.1). 198 4.1. The Null Map-Version 200 The value 0x000 (zero) is not a valid Map-Version number indicating 201 the version of the EID-to-RLOC mapping. Such a value is used for 202 special purposes and is named the Null Map-Version number. 204 The Null Map-Version MAY appear in the LISP specific header as either 205 Source Map-Version number (cf. Section 5.2) or Destination Map- 206 Version number (cf. Section 5.1). When the Source Map-Version number 207 is set to the Null Map-version value it means that no map version 208 information is conveyed for the source site. This means that if a 209 mapping exists for the source EID in the EID-to-RLOC Cache, then the 210 ETR MUST NOT compare the received Null Map-Version with the content 211 of the EID-to-RLOC cache. When the Destination Map-version number is 212 set to the Null Map-version value it means that no map version 213 information is conveyed for the destination site. This means that 214 the ETR MUST NOT compare the value with the Map-Version number of the 215 mapping for the destination EID present in the EID-to-RLOC Database. 217 The other use of the Null Map-Version number is in the Map Records, 218 which are part of the Map-Request, Map-Reply and Map-Register 219 messages (defined in [I-D.ietf-lisp]). Map Records that have a Null 220 Map-Version number indicate that there is no Map-Version number 221 associated with the mapping. This means that LISP encapsulated 222 packets, destined to the EID-Prefix the Map Record refers to, MUST 223 either not contain any Map-Version numbers (V bit set to 0), or if it 224 contains Map-Version numbers (V bit set to 1) then the destination 225 Map-Version number MUST be set to the Null Map-Version number. Any 226 value different from zero means that Map-Versioning is supported and 227 MAY be used. 229 The fact that the 0 value has a special meaning for the Map-Version 230 number implies that, when updating a Map-Version number because of a 231 change in the mapping, if the next value is 0 then Map-Version number 232 MUST be incremented by 2 (i.e., set to 1, which is the next valid 233 value). 235 5. Dealing with Map-Version numbers 237 The main idea of using Map-Version numbers is that whenever there is 238 a change in the mapping (e.g., adding/removing RLOCs, a change in the 239 weights due to TE policies, or a change in the priorities) or an ISP 240 realizes that one or more of its own RLOCs are not reachable anymore 241 from a local perspective (e.g., through IGP, or policy changes) the 242 ISP updates the mapping also assigning a new Map-Version number. 244 To each mapping, a version number is associated and changes each time 245 the mapping is changed. Note that map-versioning does not introduce 246 any new problem concerning the coordination of different ETRs of a 247 domain. Indeed, ETRs belonging to the same LISP site must return for 248 a specific EID-prefix the same mapping. In principle this is 249 orthogonal to whether or not map-versioning is used. The ETR's 250 synchronization problem is out of the scope of this document. 252 In order to announce in a data-driven fashion that the mapping has 253 been updated, Map-Version numbers used to create the outer IP header 254 of the LISP-encapsulated packet are embedded in the LISP-specific 255 header. This means that the header needs to contain two Map-Version 256 numbers: 258 o The Source Map-Version number of the EID-to-RLOC mapping in the 259 EID-to-RLOC Database used to select the source RLOC. 261 o The Destination Map-Version number of the EID-to-RLOC mapping in 262 the EID-to-RLOC Cache used to select the destination RLOC. 264 By embedding both Source Map-Version number and Destination Map- 265 Version number an ETR receiving a LISP packet with Map-Version 266 numbers, can perform the following checks: 268 1. The ITR that has sent the packet has an up-to-date mapping in its 269 cache for the destination EID and is performing encapsulation 270 correctly. 272 2. In case of bidirectional traffic, the mapping in the local ETR 273 EID-to-RLOC cache for the source EID is up-to-date. 275 If one or both of the above conditions do not hold, the ETR can send 276 a Map-Request either to make the ITR aware that a new mapping is 277 available (see Section 5.1) or to update the mapping in the local 278 cache (see Section 5.2). 280 5.1. Handling Destination Map-Version number 282 When an ETR receives a packet, the Destination Map-Version number 283 relates to the mapping for the destination EID for which the ETR is a 284 RLOC. This mapping is part of the ETR EID-to-RLOC Database. Since 285 the ETR is authoritative for the mapping, it has the correct and up- 286 to-date Destination Map-Version number. A check on this version 287 number can be done, where the following cases can arise: 289 1. The packets arrive with the same Destination Map-Version number 290 stored in the EID-to-RLOC Database. This is the regular case. 291 The ITR sending the packet has in its EID-to-RLOC Cache an up-to- 292 date mapping. No further actions are needed. 294 2. The packet arrives with a Destination Map-Version number greater 295 (i.e., newer) than the one stored in the EID-to-RLOC Database. 296 Since the ETR is authoritative on the mapping, this means that 297 someone is not behaving correctly w.r.t. the specifications, thus 298 the packet carries a not valid version number and SHOULD be 299 silently dropped. 301 3. The packets arrive with a Destination Map-Version number smaller 302 (i.e., older) than the one stored in the EID-to-RLOC Database. 303 This means that the ITR sending the packet has an old mapping in 304 its EID-to-RLOC Cache containing stale information. The ITR 305 sending the packet has to be informed that a newer mapping is 306 available. This is done with a Map-Request message sent back to 307 the ITR. The Map-Request will either trigger a Map-Request back 308 using the SMR bit or it will piggyback the newer mapping. These 309 are not new mechanisms; how to SMR or piggyback mappings in Map- 310 Request messages is already described in [I-D.ietf-lisp], while 311 their security is discussed in [I-D.ietf-lisp-threats]. These 312 Map-Request messages should be rate limited (rate limitation 313 policies are also described in [I-D.ietf-lisp]). The feature 314 introduced by Map-Version numbers is the possibility of blocking 315 traffic from ITRs not using the latest mapping. Indeed, after a 316 certain number of retries, if the Destination Map-Version number 317 in the packets is not updated, the ETR MAY silently drop packets 318 with a stale Map-Version number. This because either the ITR is 319 refusing to use the mapping for which the ETR is authoritative or 320 (worse) it might be some form of attack. 322 The rule in the third case MAY be more restrictive. If the mapping 323 has been the same for a period of time as long as the TTL (defined in 324 [I-D.ietf-lisp]) of the previous version of the mapping, all packets 325 arriving with an old Map-Version SHOULD be silently dropped right 326 away without issuing any Map-Request. The reason that allows such 327 action is the fact that if the new mapping with the updated version 328 number has been unchanged for at least the same time as the TTL of 329 the older mapping, all the entries in the caches of ITRs must have 330 expired. Hence, all ITRs sending traffic should have refreshed the 331 mapping according to [I-D.ietf-lisp]. If packets with old Map- 332 Version number are still received, then either someone has not 333 respected the TTL, or it is a form of spoof/attack. In both cases 334 this is not valid behavior w.r.t. the specifications and the packet 335 SHOULD be silently dropped. 337 LISP-encapsulated packets with the V-bit set, when the original 338 mapping in the EID-to-RLOC Database has version number set to the 339 Null Map-Version value, MAY be silently dropped. As explained in 340 Section 4.1, if an EID-to-RLOC mapping has a Null Map-Version, it 341 means that ITRs, using the mapping for encapsulation, MUST NOT use 342 Map-Version number in the LISP-specific header. 344 For LISP-encapsulated packets with the V-bit set, when the original 345 mapping in the EID-to-RLOC Database has version number set to a value 346 different from the Null Map-Version value, a Destination Map-Version 347 number equal to the Null Map-Version value means that the Destination 348 Map-Version number MUST be ignored. 350 5.2. Handling Source Map-Version number 352 When an ETR receives a packet, the Source Map-Version number relates 353 to the mapping for the source EID for which the ITR that sent the 354 packet is authoritative. If the ETR has an entry in its EID-to-RLOC 355 Cache for the source EID, then a check can be performed and the 356 following cases can arise: 358 1. The packet arrives with the same Source Map-Version number stored 359 in the EID-to-RLOC Cache. This is the correct regular case. The 360 ITR has in its cache an up-to-date copy of the mapping. No 361 further actions are needed. 363 2. The packet arrives with a Source Map-Version number greater 364 (i.e., newer) than the one stored in the local EID-to-RLOC Cache. 365 This means that ETR has in its cache a mapping that is stale and 366 needs to be updated. A Map-Request SHOULD be sent to get the new 367 mapping for the source EID. This is a normal Map-Request message 368 sent through the mapping system and MUST respect the 369 specifications in [I-D.ietf-lisp], including rate limitation 370 policies. 372 3. The packet arrives with a Source Map-Version number smaller 373 (i.e., older) than the one stored in the local EID-to-RLOC Cache. 374 Such a case is not valid w.r.t. the specifications. Indeed, if 375 the mapping is already present in the EID-to-RLOC Cache, this 376 means that an explicit Map-Request has been sent and a Map-Reply 377 has been received from an authoritative source. Assuming that 378 the mapping system is not corrupted anyhow, the Map-Version in 379 the EID-to-RLOC Cache is the correct one and the packet MAY be 380 silently dropped. 382 If the ETR does not have an entry in the EID-to-RLOC Cache for the 383 source EID (e.g., in case of unidirectional traffic) then the Source 384 Map-Version number can be safely ignored. 386 For LISP-encapsulated packets with the V-bit set, if the Source Map- 387 Version number is the Null Map-Version value, it means that the 388 Source Map-Version number MUST be ignored. 390 6. LISP header and Map-Version numbers 392 In order for the versioning approach to work, the LISP specific 393 header has to carry both Source Map-Version number and Destination 394 Map-Version number. This is done by setting the V-bit in the LISP 395 specific header. When the V-bit is set the low-order 24-bits of the 396 first longword (which usually contains the nonce) are used to 397 transport both source and destination Map-Version numbers. In 398 particular the first 12 bits are used for Source Map-Version number 399 and the second 12 bits for the Destination Map-Version number. 401 Hereafter is the example of LISP header carrying version numbers in 402 the case of IPv4-in-IPv4 encapsulation. The same setting can be used 403 for any other case (IPv4-in-IPv6, IPv6-in-IPv4, and IPv6-in-IPv6). 405 0 1 2 3 406 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 1 407 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 408 / |N|L|E|V|I|flags| Source Map-Version |Destination Map-Version| 409 LISP+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 410 \ | Instance ID/Locator Status Bits | 411 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 413 Source Map-Version number (12 bits): Map-Version of the mapping used 414 by the ITR to select the RLOC present in the "Source Routing 415 Locator" field. How to set on transmission and handle on 416 reception this value is described in Section 5.2. 418 Destination Map-Version number (12 bits): Map-Version of the mapping 419 used by the ITR to select the RLOC present in the "Destination 420 Routing Locator" field. How to set on transmission and handle on 421 reception this value is described in Section 5.1. 423 The present document just specifies how to use the low-order 24-bits 424 of the first longword of the LISP-specific header when the V-bit is 425 set to 1. All other cases, including the bit fields of the rest of 426 the LISP-specific header and the whole LISP packet format are 427 specified in [I-D.ietf-lisp]. Not all of the LISP encapsulated 428 packets need to carry version numbers. When Map-Version numbers are 429 carried the V-bit MUST be set to 1. All legal combinations of the 430 flags, when the V-bit is set to 1, are described in [I-D.ietf-lisp]. 432 7. Map Record and Map-Version 434 To accommodate the proposed mechanism, the Map Records that are 435 transported on Map-Request/Map-Reply/Map-Register messages need to 436 carry the Map-Version number as well. For this purpose the 12-bits 437 before the EID-AFI field in the Record that describe a mapping is 438 used. This is defined in [I-D.ietf-lisp] and reported here as 439 example. 441 0 1 2 3 442 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 1 443 +-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 444 | | Record TTL | 445 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 446 R | Locator Count | EID mask-len | ACT |A| Reserved | 447 e +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 448 c | Rsvd | Map-Version Number | EID-AFI | 449 o +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 450 r | EID-prefix | 451 d +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 452 | /| Priority | Weight | M Priority | M Weight | 453 | L +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 454 | o | Unused Flags |L|p|R| Loc-AFI | 455 | c +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 456 | \| Locator | 457 +-> +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 459 Map-Version Number: Map-Version of the mapping contained in the 460 Record. As explained in Section 4.1 this field can be zero (0), 461 meaning that no Map-Version is associated to the mapping, hence 462 packets that are LISP-encapsulated using this mapping MUST NOT 463 contain Map-Version numbers in the LISP specific header and the 464 V-bit MUST be set to 0. 466 This packet format works perfectly with xTRs that do not support Map- 467 Versioning, since they can simply ignore those bits. 469 8. Benefits and case studies for Map-Versioning 471 In the following sections we provide more discussion on various 472 aspects and use of the Map-Versioning. Security observations are 473 instead grouped in Section 10. 475 8.1. Synchronization of different xTRs 477 Map-Versioning does not require additional synchronization mechanism 478 compared to the normal functioning of LISP without Map-Versioning. 479 Clearly all the ETRs have to reply with the same Map-Version number, 480 otherwise there can be an inconsistency that creates additional 481 control traffic, instabilities, traffic disruptions. It is the same 482 without Map-Versioning, with ETRs that have to reply with the same 483 mapping, otherwise the same problems can arise. 485 As an example, let's consider the topology of Figure 1 where ITR A.1 486 of domain A is sending unidirectional traffic to the domain B, while 487 A.2 of domain A exchange bidirectional traffic with domain B. In 488 particular, ITR A.2 send traffic to ETR B and ETR A.2 receives 489 traffic from ITR B. 491 +-----------------+ +-----------------+ 492 | Domain A | | Domain B | 493 | +---------+ | | 494 | | ITR A.1 |--- | | 495 | +---------+ \ +---------+ | 496 | | ------->| ETR B | | 497 | | ------->| | | 498 | +---------+ / | | | 499 | | ITR A.2 |--- -----| ITR B | | 500 | | | / +---------+ | 501 | | ETR A.2 |<----- | | 502 | +---------+ | | 503 | | | | 504 +-----------------+ +-----------------+ 506 Figure 1 508 Obviously in the case of Map-Versioning both ITR A.1 and ITR A.2 of 509 domain A must use the same value otherwise the ETR of domain B will 510 start to send Map-Requests. 512 The same problem can, however, arise without Map-Versioning. For 513 instance, if the two ITRs of domain A send different Loc Status Bits. 514 In this case either the traffic is disrupted, if the ETR B trusts the 515 Locator Status Bits, or if ETR B does not trusts the Locator Status 516 Bits it will start sending Map-Requests to confirm the each change in 517 the reachability. 519 So far, LISP does not provide any specific synchronization mechanism, 520 but assumes that synchronization is provided by configuring the 521 different xTRs consistently. The same applies for Map-Versioning. 522 If in the future any synchronization mechanism is provided, Map- 523 Versioning will take advantage of it automatically since it is 524 included in the Record format, as described in Section 7. 526 8.2. Map-Versioning and unidirectional traffic 528 When using Map-Versioning the LISP specific header carries two Map- 529 Version numbers, for both source and destination mappings. This can 530 raise the question on what will happen in the case of unidirectional 531 flows, like for instance in the case presented in Figure 2, since 532 LISP specification do not mandate for ETR to have a mapping for the 533 source EID. 535 +-----------------+ +-----------------+ 536 | Domain A | | Domain B | 537 | +---------+ +---------+ | 538 | | ITR A |----------->| ETR B | | 539 | +---------+ +---------+ | 540 | | | | 541 +-----------------+ +-----------------+ 543 Figure 2 545 For what concerns the ITR, it is able to put both source and 546 destination version number in the LISP header since the Source Map- 547 Version number is in ITR's database, while the Destination Map- 548 Version number is in ITR's cache. 550 For what concerns the ETR, it simply checks only the Destination Map- 551 Version number in the same way as described in Section 5, ignoring 552 the Source Map-Version number. 554 8.3. Map-Versioning and interworking 556 Map-Versioning is compatible with the LISP interworking between LISP 557 and non-LISP sites as defined in [I-D.ietf-lisp-interworking]. LISP 558 interworking defines three techniques to make LISP sites and non-LISP 559 sites, namely Proxy-ITR, LISP-NAT, and Proxy-ETR. Hereafter it is 560 described how Map-Versioning relates to these three mechanisms. 562 8.3.1. Map-Versioning and Proxy-ITRs 564 The purpose of the Proxy-ITR (PITR) is to encapsulate traffic 565 originating in a non-LISP site in order to deliver the packet to one 566 of the ETRs of the LISP site (cf. Figure 3). This case is very 567 similar to the unidirectional traffic case described in Section 8.2, 568 hence similar rules apply. 570 +----------+ +-------------+ 571 | LISP | | non-LISP | 572 | Domain A | | Domain B | 573 | +-------+ +-----------+ | | 574 | | ETR A |<-------| Proxy ITR |<-------| | 575 | +-------+ +-----------+ | | 576 | | | | 577 +----------+ +-------------+ 579 Figure 3 581 The main difference is that a Proxy-ITR does not have any mapping, 582 since it just encapsulate packets arriving from non-LISP site, thus 583 cannot provide a Source Map-Version. In this case, the proxy-ITR 584 will just put the Null Map-Version value as Source Map-Version 585 number, while the receiving ETR will ignore the field. 587 With this setup the LISP Domain A is able to check whether or not the 588 PITR is using the latest mapping. If this is not the case the 589 mapping for LISP Domain A on the PITR can be updated using one of the 590 mechanisms defined in [I-D.ietf-lisp] and 591 [I-D.ietf-lisp-interworking]. 593 8.3.2. Map-Versioning and LISP-NAT 595 The LISP-NAT mechanism is based on address translation from non- 596 routable EIDs to routable EIDs and does not involve any form of 597 encapsulation. As such Map-Versioning does not apply in this case. 599 8.3.3. Map-Versioning and Proxy-ETRs 601 The purpose of the Proxy-ETR (PETR) is to decapsulate traffic 602 originating in a LISP site in order to deliver the packet to the non- 603 LISP site (cf. Figure 4). One of the main reasons of deploy PETRs is 604 to bypass uRPF (Unicast Reverse Path Forwarding) checks on the 605 provider edge. 607 +----------+ +-------------+ 608 | LISP | | non-LISP | 609 | Domain A | | Domain B | 610 | +-------+ +-----------+ | | 611 | | ITR A |------->| Proxy ETR |------->| | 612 | +-------+ +-----------+ | | 613 | | | | 614 +----------+ +-------------+ 616 Figure 4 618 A Proxy-ETR does not have any mapping, since it just decapsulate 619 packets arriving from LISP site. In this case, the ITR will just put 620 the Null Map-Version value as Destination Map-Version number, while 621 the receiving Proxy-ETR will ignore the field. 623 With this setup the Proxy-ETR is able to check whether or not the 624 mapping has changed. If this is the case the mapping for LISP Domain 625 A on the PETR can be updated using one of the mechanisms defined in 626 [I-D.ietf-lisp] and [I-D.ietf-lisp-interworking]. 628 8.4. RLOC shutdown/withdraw 630 Map-Versioning can be even used to perform a graceful shutdown or 631 withdraw of a specific RLOC. This is achieved by simply issuing a 632 new mapping, with an updated Map-Version number, where the specific 633 RLOC to be shut down is withdrawn or announced as unreachable (R bit 634 in the Map Record, see [I-D.ietf-lisp]), but without actually turning 635 it off. 637 Once no more traffic is received by the RLOC, because all sites have 638 updated the mapping, it can be shut down safely. 640 It should be pointed out that for frequent up/down changes such a 641 mechanism should not be used since this can generate excessive load 642 on the Mapping System. 644 8.5. Map-Version for lightweight LISP implementation 646 The use of Map-Versioning can help in developing a lightweight 647 implementation of LISP. This comes with the price of not supporting 648 Loc-Status-Bit, which are useful in some contexts. 650 In the current LISP specifications the set of RLOCs must always be 651 maintained ordered and consistent with the content of the Loc Status 652 Bits (see section 6.5 of [I-D.ietf-lisp]). With Map-Versioning such 653 type of mechanisms can be avoided. When a new RLOC is added to a 654 mapping, it is not necessary to "append" new locators to the existing 655 ones as explained in Section 6.5 of [I-D.ietf-lisp]. A new mapping 656 with a new Map-Version number will be issued, and since the old 657 locators are still valid the transition will be with no disruptions. 658 The same applies for the case a RLOC is withdrawn. There is no need 659 to maintain holes in the list of locators, as is the case when using 660 Locator Status Bits, for sites that are not using the RLOC that has 661 been withdrawn the transition will be with no disruptions. 663 All of these operations, as already stated, do not need to maintain 664 any consistency among Locator Status Bits, and the way RLOC are 665 stored in the cache. 667 Further, Map-Version can be used to substitute the "clock sweep" 668 operation described in Section 6.5.1 of [I-D.ietf-lisp]. Indeed, 669 every LISP site communicating to a specific LISP site that has 670 updated the mapping will be informed of the available new mapping in 671 a data-driven manner. 673 Note that what is proposed in the present section is just an example 674 and MUST NOT be considered as specifications for a lightweight LISP 675 implementation. In case the IETF decides to undertake such a work, 676 it will be documented elsewhere. 678 9. Incremental deployment and implementation status 680 Map-Versioning can be incrementally deployed without any negative 681 impact on existing LISP elements (e.g., xTRs, Map-Servers, Proxy- 682 ITRs, etc). Any LISP element that does not support Map-Versioning 683 can safely ignore them. Further, there is no need of any specific 684 mechanism to discover if an xTR supports or not Map-Versioning. This 685 information is already included in the Map Record. 687 Map-Versioning is currently implemented in OpenLISP 688 [I-D.iannone-openlisp-implementation]. 690 Note that the reference document for LISP implementation and 691 interoperability tests remains [I-D.ietf-lisp]. 693 10. Security Considerations 695 Map-Versioning does not introduce any new security issue concerning 696 both the data-plane and the control-plane. On the contrary, as 697 described in the following, if Map-Versioning may be used also to 698 update mappings in case of change in the reachability information 699 (i.e., instead of the Locator Status Bits) it is possible to reduce 700 the effects of some DoS or spoofing attacks that can happen in an 701 untrusted environment. 703 A thorough security analysis of LISP is documented in 704 [I-D.ietf-lisp-threats]. 706 10.1. Map-Versioning against traffic disruption 708 An attacker can try to disrupt ongoing communications by creating 709 LISP encapsulated packets with wrong Locator Status Bits. If the xTR 710 blindly trusts the Locator Status Bits it will change the 711 encapsulation accordingly, which can result in traffic disruption. 713 This does not happen in the case of Map-Versioning. As described in 714 Section 5, upon a version number change the xTR first issues a Map- 715 Request. The assumption is that the mapping distribution system is 716 sufficiently secure that Map-Request and Map-Reply messages and their 717 content can be trusted. Security issues concerning specific mapping 718 distribution system are out of the scope of this document. In the 719 case of Map-Versioning the attacker should "guess" a valid version 720 number that triggers a Map-Request, as described in Section 5, 721 otherwise the packet is simply dropped. Nevertheless, guessing a 722 version number that generates a Map-Request is easy, hence it is 723 important to follow the rate limitations policies described in 724 [I-D.ietf-lisp] in order to avoid DoS attacks. 726 Note that a similar level of security can be obtained with Loc Status 727 Bits, by simply making mandatory to verify any change through a Map- 728 Request. However, in this case Locator Status Bits loose their 729 meaning, because, it does not matter anymore which specific bits has 730 changed, the xTR will query the mapping system and trust the content 731 of the received Map-Reply. Furthermore there is no way to perform 732 filtering as in the Map-Versioning in order to drop packets that do 733 not carry a valid Map-Version number. In the case of Locator Status 734 Bits, any random change can trigger a Map-Request (unless rate 735 limitation is enabled which raise another type of attack discussed in 736 Section 10.2). 738 10.2. Map-Versioning against reachability information DoS 740 Attackers can try to trigger a large amount of Map-Request by simply 741 forging packets with random Map-Version or random Locator Status 742 Bits. In both cases the Map-Requests are rate limited as described 743 in [I-D.ietf-lisp]. However, differently from Locator Status Bit 744 where there is no filtering possible, in the case of Map-Versioning 745 is possible to filter not valid version numbers before triggering a 746 Map-Request, thus helping in reducing the effects of DoS attacks. In 747 other words the use of Map-Versioning enables a fine control on when 748 to update a mapping or when to notify that a mapping has been 749 updated. 751 It is clear, that Map-Versioning does not protect against DoS and 752 DDoS attacks, where an xTR looses processing power doing checks on 753 the LISP header of packets sent by attackers. This is independent 754 from Map-Versioning and is the same for Loc Status Bits. 756 11. IANA Considerations 758 This document has no actions for IANA. 760 12. Acknowledgements 762 The authors would like to thank Alia Atlas, Jesper Skriver, Pierre 763 Francois, Noel Chiappa, Dino Farinacci for their comments and review. 765 This work has been partially supported by the INFSO-ICT-216372 766 TRILOGY Project (www.trilogy-project.org). 768 13. References 770 13.1. Normative References 772 [I-D.ietf-lisp] 773 Farinacci, D., Fuller, V., Meyer, D., and D. Lewis, 774 "Locator/ID Separation Protocol (LISP)", 775 draft-ietf-lisp-14 (work in progress), June 2011. 777 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 778 Requirement Levels", BCP 14, RFC 2119, March 1997. 780 13.2. Informative References 782 [I-D.iannone-openlisp-implementation] 783 Iannone, L., Saucez, D., and O. Bonaventure, "OpenLISP 784 Implementation Report", 785 draft-iannone-openlisp-implementation-01 (work in 786 progress), July 2008. 788 [I-D.ietf-lisp-alt] 789 Fuller, V., Farinacci, D., Meyer, D., and D. Lewis, "LISP 790 Alternative Topology (LISP+ALT)", draft-ietf-lisp-alt-07 791 (work in progress), June 2011. 793 [I-D.ietf-lisp-interworking] 794 Lewis, D., Meyer, D., Farinacci, D., and V. Fuller, 795 "Interworking LISP with IPv4 and IPv6", 796 draft-ietf-lisp-interworking-02 (work in progress), 797 June 2011. 799 [I-D.ietf-lisp-ms] 800 Fuller, V. and D. Farinacci, "LISP Map Server", 801 draft-ietf-lisp-ms-09 (work in progress), June 2011. 803 [I-D.ietf-lisp-threats] 804 Saucez, D., Iannone, L., and O. Bonaventure, "LISP Threats 805 Analysis", draft-ietf-lisp-threats-00 (work in progress), 806 July 2011. 808 Appendix A. Estimation of time before Map-Version wrap-around 810 The present section proposes an estimation of the wrap-around time 811 for the proposed 12 bits size for the Map-Version number. Using a 812 granularity of seconds and assuming as worst-case that a new version 813 is issued each second, it takes slightly more than 1 hour before the 814 version wraps around. Note that the granularity of seconds is in 815 line with the rate limitation policy for Map-Request messages, as 816 proposed in the LISP main specifications ([I-D.ietf-lisp]). 817 Alternatively a granularity of minutes can also be used, as for the 818 TTL of the Map-Reply ([I-D.ietf-lisp]). In this case the worst 819 scenario is when a new version is issued every minute, leading to a 820 much longer time before wrap-around. In particular, when using 12 821 bits, the wrap-around time is almost 3 days. 823 For general information, hereafter there is a table with a rough 824 estimation of the time before wrap-around in the worst-case scenario, 825 considering different sizes (bits length) of the Map-Version number 826 and different time granularity. 828 +---------------+--------------------------------------------+ 829 |Version Number | Time before wrap around | 830 | Size (bits) +---------------------+----------------------+ 831 | |Granularity: Minutes | Granularity: Seconds | 832 | | (mapping changes | (mapping changes | 833 | | every 1 minute) | every 1 second) | 834 +-------------------------------------+----------------------+ 835 | 32 | 8171 Years | 136 Years | 836 | 30 | 2042 Years | 34 Years | 837 | 24 | 31 Years | 194 Days | 838 | 16 | 45 Days | 18 Hours | 839 | 15 | 22 Days | 9 Hours | 840 | 14 | 11 Days | 4 Hours | 841 | 13 | 5.6 Days | 2.2 Hours | 842 | 12 | 2.8 Days | 1.1 Hours | 843 +---------------+---------------------+----------------------+ 845 Figure 5: Estimation of time before wrap-around 847 Appendix B. Document Change Log 849 o Version 02 Posted July 2011. 851 * Added text in Section 5 about ETR synchronization, as suggested 852 by Alia Atlas. 854 * Modified text in Section 8.5 concerning lightweight LISP 855 implementation, as suggested by Alia Atlas. 857 * Deleted text concerning old versions of [I-D.ietf-lisp-ms] and 858 [I-D.ietf-lisp-alt] in Section 7, as pointed out by Alia Atlas. 860 * Fixed section 4.1 to be less restrictive, as suggested by 861 Jesper Skriver. 863 o Version 01 Posted March 2011. 865 * Changed the wording from "Map-Version number 0" to "Null Map- 866 Version. 868 * Clarification of the use of the Null Map-Version value as 869 Source Map-Version Number and Destination Map-Version Number. 871 * Extended the section describing Map-Versioning and LISP 872 Interworking co-existence. 874 * Reduce packet format description to avoid double definitions 875 with the main specs. 877 o Version 00 Posted September 2010. 879 * Added Section "Definitions of Terms". 881 * Editorial polishing of all sections. 883 * Added clarifications in section "Dealing with Map-Version 884 numbers" for the case of the special Map-Version number 0. 886 * Rename of draft-iannone-mapping-versioning-02.txt. 888 Authors' Addresses 890 Luigi Iannone 891 TU Berlin - Deutsche Telekom Laboratories AG 892 Ernst-Reuter Platz 7 893 Berlin 894 Germany 896 Email: luigi@net.t-labs.tu-berlin.de 898 Damien Saucez 899 Universite catholique de Louvain 900 Place St. Barbe 2 901 Louvain-la-Neuve 902 Belgium 904 Email: damien.saucez@uclouvain.be 906 Olivier Bonaventure 907 Universite catholique de Louvain 908 Place St. Barbe 2 909 Louvain-la-Neuve 910 Belgium 912 Email: olivier.bonaventure@uclouvain.be