idnits 2.17.1 draft-ietf-lisp-deployment-04.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 == Line 473 has weird spacing: '...ructure x ...' -- The document date (September 12, 2012) is 4244 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Outdated reference: A later version (-24) exists of draft-ietf-lisp-23 == Outdated reference: A later version (-13) exists of draft-ietf-lisp-eid-block-02 == Outdated reference: A later version (-29) exists of draft-ietf-lisp-sec-02 Summary: 0 errors (**), 0 flaws (~~), 5 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group L. Jakab 3 Internet-Draft A. Cabellos-Aparicio 4 Intended status: Informational F. Coras 5 Expires: March 16, 2013 J. Domingo-Pascual 6 Technical University of 7 Catalonia 8 D. Lewis 9 Cisco Systems 10 September 12, 2012 12 LISP Network Element Deployment Considerations 13 draft-ietf-lisp-deployment-04.txt 15 Abstract 17 This document discusses the different scenarios for the deployment of 18 the new network elements introduced by the Locator/Identifier 19 Separation Protocol (LISP). 21 Status of this Memo 23 This Internet-Draft is submitted in full conformance with the 24 provisions of BCP 78 and BCP 79. 26 Internet-Drafts are working documents of the Internet Engineering 27 Task Force (IETF). Note that other groups may also distribute 28 working documents as Internet-Drafts. The list of current Internet- 29 Drafts is at http://datatracker.ietf.org/drafts/current/. 31 Internet-Drafts are draft documents valid for a maximum of six months 32 and may be updated, replaced, or obsoleted by other documents at any 33 time. It is inappropriate to use Internet-Drafts as reference 34 material or to cite them other than as "work in progress." 36 This Internet-Draft will expire on March 16, 2013. 38 Copyright Notice 40 Copyright (c) 2012 IETF Trust and the persons identified as the 41 document authors. All rights reserved. 43 This document is subject to BCP 78 and the IETF Trust's Legal 44 Provisions Relating to IETF Documents 45 (http://trustee.ietf.org/license-info) in effect on the date of 46 publication of this document. Please review these documents 47 carefully, as they describe your rights and restrictions with respect 48 to this document. Code Components extracted from this document must 49 include Simplified BSD License text as described in Section 4.e of 50 the Trust Legal Provisions and are provided without warranty as 51 described in the Simplified BSD License. 53 Table of Contents 55 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 56 2. Tunnel Routers . . . . . . . . . . . . . . . . . . . . . . . . 3 57 2.1. Customer Edge . . . . . . . . . . . . . . . . . . . . . . 4 58 2.2. Provider Edge . . . . . . . . . . . . . . . . . . . . . . 5 59 2.3. Split ITR/ETR . . . . . . . . . . . . . . . . . . . . . . 6 60 2.4. Inter-Service Provider Traffic Engineering . . . . . . . . 8 61 2.5. Tunnel Routers Behind NAT . . . . . . . . . . . . . . . . 10 62 2.5.1. ITR . . . . . . . . . . . . . . . . . . . . . . . . . 10 63 2.5.2. ETR . . . . . . . . . . . . . . . . . . . . . . . . . 10 64 2.6. Summary and Feature Matrix . . . . . . . . . . . . . . . . 11 65 3. Map-Resolvers and Map-Servers . . . . . . . . . . . . . . . . 11 66 3.1. Map-Servers . . . . . . . . . . . . . . . . . . . . . . . 11 67 3.2. Map-Resolvers . . . . . . . . . . . . . . . . . . . . . . 12 68 4. Proxy Tunnel Routers . . . . . . . . . . . . . . . . . . . . . 13 69 4.1. P-ITR . . . . . . . . . . . . . . . . . . . . . . . . . . 13 70 4.2. P-ETR . . . . . . . . . . . . . . . . . . . . . . . . . . 14 71 5. Migration to LISP . . . . . . . . . . . . . . . . . . . . . . 15 72 5.1. LISP+BGP . . . . . . . . . . . . . . . . . . . . . . . . . 16 73 5.2. Mapping Service Provider (MSP) P-ITR Service . . . . . . . 16 74 5.3. Proxy-ITR Route Distribution (PITR-RD) . . . . . . . . . . 17 75 5.4. Migration Summary . . . . . . . . . . . . . . . . . . . . 19 76 6. Step-by-Step Example BGP to LISP Migration Procedure . . . . . 20 77 6.1. Customer Pre-Install and Pre-Turn-up Checklist . . . . . . 20 78 6.2. Customer Activating LISP Service . . . . . . . . . . . . . 21 79 6.3. Cut-Over Provider Preparation and Changes . . . . . . . . 22 80 7. Security Considerations . . . . . . . . . . . . . . . . . . . 22 81 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 23 82 9. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 23 83 10. Informative References . . . . . . . . . . . . . . . . . . . . 23 84 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 24 86 1. Introduction 88 The Locator/Identifier Separation Protocol (LISP) addresses the 89 scaling issues of the global Internet routing system by separating 90 the current addressing scheme into Endpoint IDentifiers (EIDs) and 91 Routing LOCators (RLOCs). The main protocol specification 92 [I-D.ietf-lisp] describes how the separation is achieved, which new 93 network elements are introduced, and details the packet formats for 94 the data and control planes. 96 LISP assumes that such separation is between the edge and core. 97 While the boundary between both is not strictly defined, one widely 98 accepted definition places it at the border routers of stub 99 autonomous systems, which may carry a partial or complete default- 100 free zone (DFZ) routing table. The initial design of LISP took this 101 location as a baseline for protocol development. However, the 102 applications of LISP go beyond of just decreasing the size of the DFZ 103 routing table, and include improved multihoming and ingress traffic 104 engineering (TE) support for edge networks, and even individual 105 hosts. Throughout the draft we will use the term LISP site to refer 106 to these networks/hosts behind a LISP Tunnel Router. We formally 107 define it as: 109 LISP site: A single host or a set of network elements in an edge 110 network under the administrative control of a single organization, 111 delimited from other networks by LISP Tunnel Router(s). 113 Since LISP is a protocol which can be used for different purposes, it 114 is important to identify possible deployment scenarios and the 115 additional requirements they may impose on the protocol specification 116 and other protocols. Additionally, this document is intended as a 117 guide for the operational community for LISP deployments in their 118 networks. It is expected to evolve as LISP deployment progresses, 119 and the described scenarios are better understood or new scenarios 120 are discovered. 122 Each subsection considers an element type, discussing the impact of 123 deployment scenarios on the protocol specification. For definition 124 of terms, please refer to the appropriate documents (as cited in the 125 respective sections). 127 2. Tunnel Routers 129 LISP is a map-and-encap protocol, with the main goal of improving 130 global routing scalability. To achieve its goal, it introduces 131 several new network elements, each performing specific functions 132 necessary to separate the edge from the core. The device that is the 133 gateway between the edge and the core is called Tunnel Router (xTR), 134 performing one or both of two separate functions: 136 1. Encapsulating packets originating from an end host to be 137 transported over intermediary (transit) networks towards the 138 other end-point of the communication 140 2. Decapsulating packets entering from intermediary (transit) 141 networks, originated at a remote end host. 143 The first function is performed by an Ingress Tunnel Router (ITR), 144 the second by an Egress Tunnel Router (ETR). 146 Section 8 of the main LISP specification [I-D.ietf-lisp] has a short 147 discussion of where Tunnel Routers can be deployed and some of the 148 associated advantages and disadvantages. This section adds more 149 detail to the scenarios presented there, and provides additional 150 scenarios as well. 152 2.1. Customer Edge 154 LISP was designed with deployment at the core-edge boundary in mind, 155 which can be approximated as the set of DFZ routers belonging to non- 156 transit ASes. For the purposes of this document, we will consider 157 this boundary to be consisting of the routers connecting LISP sites 158 to their upstreams. As such, this is the most common expected 159 scenario for xTRs, and this document considers it the reference 160 location, comparing the other scenarios to this one. 162 ISP1 ISP2 163 | | 164 | | 165 +----+ +----+ 166 +--|xTR1|--|xTR2|--+ 167 | +----+ +----+ | 168 | | 169 | LISP site | 170 +------------------+ 172 Figure 1: xTRs at the customer edge 174 From the LISP site perspective the main advantage of this type of 175 deployment (compared to the one described in the next section) is 176 having direct control over its ingress traffic engineering. This 177 makes it is easy to set up and maintain active/active, active/backup, 178 or more complex TE policies, without involving third parties. 180 Being under the same administrative control, reachability information 181 of all ETRs is easier to synchronize, because the necessary control 182 traffic can be allowed between the locators of the ETRs. A correct 183 synchronous global view of the reachability status is thus available, 184 and the Loc-Status-Bits can be set correctly in the LISP data header 185 of outgoing packets. 187 By placing the tunnel router at the edge of the site, existing 188 internal network configuration does not need to be modified. 189 Firewall rules, router configurations and address assignments inside 190 the LISP site remain unchanged. This helps with incremental 191 deployment and allows a quick upgrade path to LISP. For larger sites 192 with many external connections, distributed in geographically diverse 193 PoPs, and complex internal topology, it may however make more sense 194 to both encapsulate and decapsulate as soon as possible, to benefit 195 from the information in the IGP to choose the best path (see 196 Section 2.3 for a discussion of this scenario). 198 Another thing to consider when placing tunnel routers are MTU issues. 199 Since encapsulating packets increases overhead, the MTU of the end- 200 to-end path may decrease, when encapsulated packets need to travel 201 over segments having close to minimum MTU. Some transit networks are 202 known to provide larger MTU than the typical value of 1500 bytes of 203 popular access technologies used at end hosts (e.g., IEEE 802.3 and 204 802.11). However, placing the LISP router connecting to such a 205 network at the customer edge could possibly bring up MTU issues, 206 depending on the link type to the provider as opposed to the 207 following scenario. 209 2.2. Provider Edge 211 The other location at the core-edge boundary for deploying LISP 212 routers is at the Internet service provider edge. The main incentive 213 for this case is that the customer does not have to upgrade the CE 214 router(s), or change the configuration of any equipment. 215 Encapsulation/decapsulation happens in the provider's network, which 216 may be able to serve several customers with a single device. For 217 large ISPs with many residential/business customers asking for LISP 218 this can lead to important savings, since there is no need to upgrade 219 the software (or hardware, if it's the case) at each client's 220 location. Instead, they can upgrade the software (or hardware) on a 221 few PE routers serving the customers. This scenario is depicted in 222 Figure 2. 224 +----------+ +------------------+ 225 | ISP1 | | ISP2 | 226 | | | | 227 | +----+ | | +----+ +----+ | 228 +--|xTR1|--+ +--|xTR2|--|xTR3|--+ 229 +----+ +----+ +----+ 230 | | | 231 | | | 232 +--<[LISP site]>---+-------+ 234 Figure 2: xTR at the PE 236 While this approach can make transition easy for customers and may be 237 cheaper for providers, the LISP site looses one of the main benefits 238 of LISP: ingress traffic engineering. Since the provider controls 239 the ETRs, additional complexity would be needed to allow customers to 240 modify their mapping entries. 242 The problem is aggravated when the LISP site is multihomed. Consider 243 the scenario in Figure 2: whenever a change to TE policies is 244 required, the customer contacts both ISP1 and ISP2 to make the 245 necessary changes on the routers (if they provide this possibility). 246 It is however unlikely, that both ISPs will apply changes 247 simultaneously, which may lead to inconsistent state for the mappings 248 of the LISP site. Since the different upstream ISPs are usually 249 competing business entities, the ETRs may even be configured to 250 compete, either to attract all the traffic or to get no traffic. The 251 former will happen if the customer pays per volume, the latter if the 252 connectivity has a fixed price. A solution could be to have the 253 mappings in the Map-Server(s), and have their operator give control 254 over the entries to customer, much like in the Domain Name System at 255 the time of this writing. 257 Additionally, since xTR1, xTR2, and xTR3 are in different 258 administrative domains, locator reachability information is unlikely 259 to be exchanged among them, making it difficult to set Loc-Status- 260 Bits correctly on encapsulated packets. 262 Compared to the customer edge scenario, deploying LISP at the 263 provider edge might have the advantage of diminishing potential MTU 264 issues, because the tunnel router is closer to the core, where links 265 typically have higher MTUs than edge network links. 267 2.3. Split ITR/ETR 269 In a simple LISP deployment, xTRs are located at the border of the 270 LISP site (see Section 2.1). In this scenario packets are routed 271 inside the domain according to the EID. However, more complex 272 networks may want to route packets according to the destination RLOC. 273 This would enable them to choose the best egress point. 275 The LISP specification separates the ITR and ETR functionality and 276 considers that both entities can be deployed in separated network 277 equipment. ITRs can be deployed closer to the host (i.e., access 278 routers). This way packets are encapsulated as soon as possible, and 279 packets exit the network through the best egress point in terms of 280 BGP policy. In turn, ETRs can be deployed at the border routers of 281 the network, and packets are decapsulated as soon as possible. Once 282 decapsulated, packets are routed based on destination EID, according 283 to internal routing policy. 285 In the following figure we can see an example. The Source (S) 286 transmits packets using its EID and in this particular case packets 287 are encapsulated at ITR_1. The encapsulated packets are routed 288 inside the domain according to the destination RLOC, and can egress 289 the network through the best point (i.e., closer to the RLOC's AS). 290 On the other hand, inbound packets are received by ETR_1 which 291 decapsulates them. Then packets are routed towards S according to 292 the EID, again following the best path. 294 +---------------------------------------+ 295 | | 296 | +-------+ +-------+ +-------+ 297 | | ITR_1 |---------+ | ETR_1 |-RLOC_A--| ISP_A | 298 | +-------+ | +-------+ +-------+ 299 | +-+ | | | 300 | |S| | IGP | | 301 | +-+ | | | 302 | +-------+ | +-------+ +-------+ 303 | | ITR_2 |---------+ | ETR_2 |-RLOC_B--| ISP_B | 304 | +-------+ +-------+ +-------+ 305 | | 306 +---------------------------------------+ 308 Figure 3: Split ITR/ETR Scenario 310 This scenario has a set of implications: 312 o The site must carry at least partial BGP routes in order to choose 313 the best egress point, increasing the complexity of the network. 314 However, this is usually already the case for LISP sites that 315 would benefit from this scenario. 317 o If the site is multihomed to different ISPs and any of the 318 upstream ISPs is doing uRPF filtering, this scenario may become 319 impractical. ITRs need to determine the exit ETR, for setting the 320 correct source RLOC in the encapsulation header. This adds 321 complexity and reliability concerns. 323 o In LISP, ITRs set the reachability bits when encapsulating data 324 packets. Hence, ITRs need a mechanism to be aware of the liveness 325 of all ETRs serving their site. 327 o MTU within the site network must be large enough to accommodate 328 encapsulated packets. 330 o In this scenario, each ITR is serving fewer hosts than in the case 331 when it is deployed at the border of the network. It has been 332 shown that cache hit ratio grows logarithmically with the amount 333 of users [cache]. Taking this into account, when ITRs are 334 deployed closer to the host the effectiveness of the mapping cache 335 may be lower (i.e., the miss ratio is higher). Another 336 consequence of this is that the site may transmit a higher amount 337 of Map-Requests, increasing the load on the distributed mapping 338 database. 340 2.4. Inter-Service Provider Traffic Engineering 342 With LISP, two LISP sites can route packets among them and control 343 their ingress TE policies. Typically, LISP is seen as applicable to 344 stub networks, however the LISP protocol can also be applied to 345 transit networks recursively. 347 Consider the scenario depicted in Figure 4. Packets originating from 348 the LISP site Stub1, client of ISP_A, with destination Stub4, client 349 of ISP_B, are LISP encapsulated at their entry point into the ISP_A's 350 network. The external IP header now has as the source RLOC an IP 351 from ISP_A's address space and destination RLOC from ISP_B's address 352 space. One or more ASes separate ISP_A from ISP_B. With a single 353 level of LISP encapsulation, Stub4 has control over its ingress 354 traffic. However, at the time of this writing, ISP_B has only BGP 355 tools (such as prefix deaggregation) to control on which of his own 356 upstream or peering links should packets enter. This is either not 357 feasible (if fine-grained per-customer control is required, the very 358 specific prefixes may not be propagated) or increases DFZ table size. 360 _.--. 361 Stub1 ... +-------+ ,-'' `--. +-------+ ... Stub3 362 \ | R_A1|----,' `. ---|R_B1 | / 363 --| R_A2|---( Transit ) | |-- 364 Stub2 .../ | R_A3|-----. ,' ---|R_B2 | \... Stub4 365 +-------+ `--. _.-' +-------+ 366 ... ISP_A `--'' ISP_B ... 368 Figure 4: Inter-Service provider TE scenario 370 A solution for this is to apply LISP recursively. ISP_A and ISP_B 371 may reach a bilateral agreement to deploy their own private mapping 372 system. ISP_A then encapsulates packets destined for the prefixes of 373 ISP_B, which are listed in the shared mapping system. Note that in 374 this case the packet is double-encapsulated (using R_A1, R_A2 or R_A3 375 as source and R_B1 or R_B2 as destination in the example above). 376 ISP_B's ETR removes the outer, second layer of LISP encapsulation 377 from the incoming packet, and routes it towards the original RLOC, 378 the ETR of Stub4, which does the final decapsulation. 380 If ISP_A and ISP_B agree to share a private distributed mapping 381 database, both can control their ingress TE without the need of 382 deaggregating prefixes. In this scenario the private database 383 contains RLOC-to-RLOC bindings. The convergence time on the TE 384 policies updates is expected to be fast, since ISPs only have to 385 update/query a mapping to/from the database. 387 This deployment scenario includes two important caveats. First, it 388 is intended to be deployed between only two ISPs (ISP_A and ISP_B in 389 Figure 4). If more than two ISPs use this approach, then the xTRs 390 deployed at the participating ISPs must either query multiple mapping 391 systems, or the ISPs must agree on a common shared mapping system. 392 Second, the scenario is only recommended for ISPs providing 393 connectivity to LISP sites, such that source RLOCs of packets to be 394 reencapsulated belong to said ISP. Otherwise the participating ISPs 395 must register prefixes they do not own in the above mentioned private 396 mapping system. Failure to follow these recommendations may lead to 397 operational and security issues when deploying this scenario. 399 Besides these recommendations, the main disadvantages of this 400 deployment case are: 402 o Extra LISP header is needed. This increases the packet size and 403 requires that the MTU between both ISPs accommodates double- 404 encapsulated packets. 406 o The ISP ITR must encapsulate packets and therefore must know the 407 RLOC-to-RLOC binding. These bindings are stored in a mapping 408 database and may be cached in the ITR's mapping cache. Cache 409 misses lead to an additional lookup latency, unless a push based 410 mapping system is used for the private mapping system. 412 o The operational overhead of maintaining the shared mapping 413 database. 415 o If an IPv6 address block is reserved for EID use, as specified in 416 [I-D.ietf-lisp-eid-block], the EID-to-RLOC encapsulation (first 417 level) can avoid LISP processing altogether for non-LISP 418 destinations. The ISP tunnel routers however will not be able to 419 take advantage of this optimization, all RLOC-to-RLOC mappings 420 need a lookup in the private database (or map-cache, once results 421 are cached). 423 2.5. Tunnel Routers Behind NAT 425 NAT in this section refers to IPv4 network address and port 426 translation. 428 2.5.1. ITR 430 Packets encapsulated by an ITR are just UDP packets from a NAT 431 device's point of view, and they are handled like any UDP packet, 432 there are no additional requirements for LISP data packets. 434 Map-Requests sent by an ITR, which create the state in the NAT table, 435 have a different 5-tuple in the IP header than the Map-Reply 436 generated by the authoritative ETR. Since the source address of this 437 packet is different from the destination address of the request 438 packet, no state will be matched in the NAT table and the packet will 439 be dropped. To avoid this, the NAT device has to do the following: 441 o Send all UDP packets with source port 4342, regardless of the 442 destination port, to the RLOC of the ITR. The most simple way to 443 achieve this is configuring 1:1 NAT mode from the external RLOC of 444 the NAT device to the ITR's RLOC (Called "DMZ" mode in consumer 445 broadband routers). 447 o Rewrite the ITR-AFI and "Originating ITR RLOC Address" fields in 448 the payload. 450 This setup supports a single ITR behind the NAT device. 452 2.5.2. ETR 454 An ETR placed behind NAT is reachable from the outside by the 455 Internet-facing locator of the NAT device. It needs to know this 456 locator (and configure a loopback interface with it), so that it can 457 use it in Map-Reply and Map-Register messages. Thus support for 458 dynamic locators for the mapping database is needed in LISP 459 equipment. 461 Again, only one ETR behind the NAT device is supported. 463 An implication of the issues described above is that LISP sites with 464 xTRs can not be behind carrier based NATs, since two different sites 465 would collide on the port forwarding. 467 2.6. Summary and Feature Matrix 469 Feature CE PE Split Rec. 470 -------------------------------------------------------- 471 Control of ingress TE x - x x 472 No modifications to existing 473 int. network infrastructure x x - - 474 Loc-Status-Bits sync x - x x 475 MTU/PMTUD issues minimized - x - x 477 3. Map-Resolvers and Map-Servers 479 3.1. Map-Servers 481 The Map-Server learns EID-to-RLOC mapping entries from an 482 authoritative source and publishes them in the distributed mapping 483 database. These entries are learned through authenticated Map- 484 Register messages sent by authoritative ETRs. Also, upon reception 485 of a Map-Request, the Map-Server verifies that the destination EID 486 matches an EID-prefix for which it is authoritative for, and then re- 487 encapsulates and forwards it to a matching ETR. Map-Server 488 functionality is described in detail in [I-D.ietf-lisp-ms]. 490 The Map-Server is provided by a Mapping Service Provider (MSP). A 491 MSP can be any of the following: 493 o EID registrar. Since the IPv4 address space is nearing 494 exhaustion, IPv4 EIDs will come from already allocated Provider 495 Independent (PI) space. The registrars in this case remain the 496 current five Regional Internet Registries (RIRs). In the case of 497 IPv6, the possibility of reserving a /16 block as EID space is 498 currently under consideration [I-D.ietf-lisp-eid-block]. If 499 granted by IANA, the community will have to determine the body 500 responsible for allocations from this block, and the associated 501 policies. Existing allocation policies apply to EIDs outside this 502 block. 504 o Third parties. Participating in the LISP mapping system is 505 similar to participating in global routing or DNS: as long as 506 there is at least another already participating entity willing to 507 forward the newcomer's traffic, there is no barrier to entry. 508 Still, just like routing and DNS, LISP mappings have the issue of 509 trust, with efforts underway to make the published information 510 verifiable. When these mechanisms will be deployed in the LISP 511 mapping system, the burden of providing and verifying trust should 512 be kept away from MSPs, which will simply host the secured 513 mappings. This will keep the low barrier of entry to become an 514 MSP for third parties. 516 In all cases, the MSP configures its Map-Server(s) to publish the 517 prefixes of its clients in the distributed mapping database and start 518 encapsulating and forwarding Map-Requests to the ETRs of the AS. 519 These ETRs register their prefix(es) with the Map-Server(s) through 520 periodic authenticated Map-Register messages. In this context, for 521 some LISP end sites, there is a need for mechanisms to: 523 o Automatically distribute EID prefix(es) shared keys between the 524 ETRs and the EID-registrar Map-Server. 526 o Dynamically obtain the address of the Map-Server in the ETR of the 527 AS. 529 The Map-Server plays a key role in the reachability of the EID- 530 prefixes it is serving. On the one hand it is publishing these 531 prefixes into the distributed mapping database and on the other hand 532 it is encapsulating and forwarding Map-Requests to the authoritative 533 ETRs of these prefixes. ITRs encapsulating towards EIDs under the 534 responsibility of a failed Map-Server will be unable to look up any 535 of their covering prefixes. The only exception are the ITRs that 536 already contain the mappings in their local cache. In this case ITRs 537 can reach ETRs until the entry expires (typically 24 hours). For 538 this reason, redundant Map-Server deployments are desirable. A set 539 of Map-Servers providing high-availability service to the same set of 540 prefixes is called a redundancy group. ETRs are configured to send 541 Map-Register messages to all Map-Servers in the redundancy group. To 542 achieve fail-over (or load-balancing, if desired), known mapping 543 system specific best practices should be used. 545 Additionally, if a Map-Server has no reachability for any ETR serving 546 a given EID block, it should not originate that block into the 547 mapping system. 549 3.2. Map-Resolvers 551 A Map-Resolver a is a network infrastructure component which accepts 552 LISP encapsulated Map-Requests, typically from an ITR, and finds the 553 appropriate EID-to-RLOC mapping by either consulting its local cache 554 or by consulting the distributed mapping database. Map-Resolver 555 functionality is described in detail in [I-D.ietf-lisp-ms]. 557 Anyone with access to the distributed mapping database can set up a 558 Map-Resolver and provide EID-to-RLOC mapping lookup service. 559 Database access setup is mapping system specific. 561 For performance reasons, it is recommended that LISP sites use Map- 562 Resolvers that are topologically close to their ITRs. ISPs 563 supporting LISP will provide this service to their customers, 564 possibly restricting access to their user base. LISP sites not in 565 this position can use open access Map-Resolvers, if available. 566 However, regardless of the availability of open access resolvers, the 567 MSP providing the Map-Server(s) for a LISP site should also make 568 available Map-Resolver(s) for the use of that site. 570 In medium to large-size ASes, ITRs must be configured with the RLOC 571 of a Map-Resolver, operation which can be done manually. However, in 572 Small Office Home Office (SOHO) scenarios a mechanism for 573 autoconfiguration should be provided. 575 One solution to avoid manual configuration in LISP sites of any size 576 is the use of anycast RLOCs for Map-Resolvers similar to the DNS root 577 server infrastructure. Since LISP uses UDP encapsulation, the use of 578 anycast would not affect reliability. LISP routers are then shipped 579 with a preconfigured list of well know Map-Resolver RLOCs, which can 580 be edited by the network administrator, if needed. 582 The use of anycast also helps improving mapping lookup performance. 583 Large MSPs can increase the number and geographical diversity of 584 their Map-Resolver infrastructure, using a single anycasted RLOC. 585 Once LISP deployment is advanced enough, very large content providers 586 may also be interested running this kind of setup, to ensure minimal 587 connection setup latency for those connecting to their network from 588 LISP sites. 590 While Map-Servers and Map-Resolvers implement different 591 functionalities within the LISP mapping system, they can coexist on 592 the same device. For example, MSPs offering both services, can 593 deploy a single Map-Resolver/Map-Server in each PoP where they have a 594 presence. 596 4. Proxy Tunnel Routers 598 4.1. P-ITR 600 Proxy Ingress Tunnel Routers (P-ITRs) are part of the non-LISP/LISP 601 transition mechanism, allowing non-LISP sites to reach LISP sites. 602 They announce via BGP certain EID prefixes (aggregated, whenever 603 possible) to attract traffic from non-LISP sites towards EIDs in the 604 covered range. They do the mapping system lookup, and encapsulate 605 received packets towards the appropriate ETR. Note that for the 606 reverse path LISP sites can reach non-LISP sites simply by not 607 encapsulating traffic. See [I-D.ietf-lisp-interworking] for a 608 detailed description of P-ITR functionality. 610 The success of new protocols depends greatly on their ability to 611 maintain backwards compatibility and inter-operate with the 612 protocol(s) they intend to enhance or replace, and on the incentives 613 to deploy the necessary new software or equipment. A LISP site needs 614 an interworking mechanism to be reachable from non-LISP sites. A 615 P-ITR can fulfill this role, enabling early adopters to see the 616 benefits of LISP, similar to tunnel brokers helping the transition 617 from IPv4 to IPv6. A site benefits from new LISP functionality 618 (proportionally with existing global LISP deployment) when going 619 LISP, so it has the incentives to deploy the necessary tunnel 620 routers. In order to be reachable from non-LISP sites it has two 621 options: keep announcing its prefix(es) with BGP, or have a P-ITR 622 announce prefix(es) covering them. 624 If the goal of reducing the DFZ routing table size is to be reached, 625 the second option is preferred. Moreover, the second option allows 626 LISP-based ingress traffic engineering from all sites. However, the 627 placement of P-ITRs significantly influences performance and 628 deployment incentives. Section 5 is dedicated to the migration to a 629 LISP-enabled Internet, and includes deployment scenarios for P-ITRs. 631 4.2. P-ETR 633 In contrast to P-ITRs, P-ETRs are not required for the correct 634 functioning of all LISP sites. There are two cases, where they can 635 be of great help: 637 o LISP sites with unicast reverse path forwarding (uRPF) 638 restrictions, and 640 o Communication between sites using different address family RLOCs. 642 In the first case, uRPF filtering is applied at their upstream PE 643 router. When forwarding traffic to non-LISP sites, an ITR does not 644 encapsulate packets, leaving the original IP headers intact. As a 645 result, packets will have EIDs in their source address. Since we are 646 discussing the transition period, we can assume that a prefix 647 covering the EIDs belonging to the LISP site is advertised to the 648 global routing tables by a P-ITR, and the PE router has a route 649 towards it. However, the next hop will not be on the interface 650 towards the CE router, so non-encapsulated packets will fail uRPF 651 checks. 653 To avoid this filtering, the affected ITR encapsulates packets 654 towards the locator of the P-ETR for non-LISP destinations. Now the 655 source address of the packets, as seen by the PE router is the ITR's 656 locator, which will not fail the uRPF check. The P-ETR then 657 decapsulates and forwards the packets. 659 The second use case is IPv4-to-IPv6 transition. Service providers 660 using older access network hardware, which only supports IPv4 can 661 still offer IPv6 to their clients, by providing a CPE device running 662 LISP, and P-ETR(s) for accessing IPv6-only non-LISP sites and LISP 663 sites, with IPv6-only locators. Packets originating from the client 664 LISP site for these destinations would be encapsulated towards the 665 P-ETR's IPv4 locator. The P-ETR is in a native IPv6 network, 666 decapsulating and forwarding packets. For non-LISP destination, the 667 packet travels natively from the P-ETR. For LISP destinations with 668 IPv6-only locators, the packet will go through a P-ITR, in order to 669 reach its destination. 671 For more details on P-ETRs see the [I-D.ietf-lisp-interworking] 672 draft. 674 P-ETRs can be deployed by ISPs wishing to offer value-added services 675 to their customers. As is the case with P-ITRs, P-ETRs too may 676 introduce path stretch. Because of this the ISP needs to consider 677 the tradeoff of using several devices, close to the customers, to 678 minimize it, or few devices, farther away from the customers, 679 minimizing cost instead. 681 Since the deployment incentives for P-ITRs and P-ETRs are different, 682 it is likely they will be deployed in separate devices, except for 683 the CDN case, which may deploy both in a single device. 685 In all cases, the existence of a P-ETR involves another step in the 686 configuration of a LISP router. CPE routers, which are typically 687 configured by DHCP, stand to benefit most from P-ETRs. 688 Autoconfiguration of the P-ETR locator could be achieved by a DHCP 689 option, or adding a P-ETR field to either Map-Notifys or Map-Replies. 691 As a security measure, access to P-ETRs should be limited to 692 legitimate users by enforcing ACLs. 694 5. Migration to LISP 696 This section discusses a deployment architecture to support the 697 migration to a LISP-enabled Internet. The loosely defined terms of 698 "early transition phase", "late transition phase", and "LISP Internet 699 phase" refer to time periods when LISP sites are a minority, a 700 majority, or represent all edge networks respectively. 702 5.1. LISP+BGP 704 For sites wishing to go LISP with their PI prefix the least 705 disruptive way is to upgrade their border routers to support LISP, 706 register the prefix into the LISP mapping system, but keep announcing 707 it with BGP as well. This way LISP sites will reach them over LISP, 708 while legacy sites will be unaffected by the change. The main 709 disadvantage of this approach is that no decrease in the DFZ routing 710 table size is achieved. Still, just increasing the number of LISP 711 sites is an important gain, as an increasing LISP/non-LISP site ratio 712 will slowly decrease the need for BGP-based traffic engineering that 713 leads to prefix deaggregation. That, in turn, may lead to a decrease 714 in the DFZ size in the late transition phase. 716 This scenario is not limited to sites that already have their 717 prefixes announced with BGP. Newly allocated EID blocks could follow 718 this strategy as well during the early LISP deployment phase, 719 depending on the cost/benefit analysis of the individual networks. 720 Since this leads to an increase in the DFZ size, the following 721 architecture should be preferred for new allocations. 723 5.2. Mapping Service Provider (MSP) P-ITR Service 725 In addition to publishing their clients' registered prefixes in the 726 mapping system, MSPs with enough transit capacity can offer them 727 P-ITR service as a separate service. This service is especially 728 useful for new PI allocations, to sites without existing BGP 729 infrastructure, that wish to avoid BGP altogether. The MSP announces 730 the prefix into the DFZ, and the client benefits from ingress traffic 731 engineering without prefix deaggregation. The downside of this 732 scenario is adding path stretch. 734 Routing all non-LISP ingress traffic through a third party which is 735 not one of its ISPs is only feasible for sites with modest amounts of 736 traffic (like those using the IPv6 tunnel broker services today), 737 especially in the first stage of the transition to LISP, with a 738 significant number of legacy sites. When the LISP/non-LISP site 739 ratio becomes high enough, this approach can prove increasingly 740 attractive. 742 Compared to LISP+BGP, this approach avoids DFZ bloat caused by prefix 743 deaggregation for traffic engineering purposes, resulting in slower 744 routing table increase in the case of new allocations and potential 745 decrease for existing ones. Moreover, MSPs serving different clients 746 with adjacent aggregable prefixes may lead to additional decrease, 747 but quantifying this decrease is subject to future research study. 749 5.3. Proxy-ITR Route Distribution (PITR-RD) 751 Instead of a LISP site, or the MSP, announcing their EIDs with BGP to 752 the DFZ, this function can be outsourced to a third party, a P-ITR 753 Service Provider (PSP). This will result in a decrease of the 754 operational complexity both at the site and at the MSP. 756 The PSP manages a set of distributed P-ITR(s) that will advertise the 757 corresponding EID prefixes through BGP to the DFZ. These P-ITR(s) 758 will then encapsulate the traffic they receive for those EIDs towards 759 the RLOCs of the LISP site, ensuring their reachability from non-LISP 760 sites. 762 While it is possible for a PSP to manually configure each client's 763 EID routes to be announced, this approach offers little flexibility 764 and is not scalable. This section presents a scalable architecture 765 that offers automatic distribution of EID routes to LISP sites and 766 service providers. 768 The architecture requires no modification to existing LISP network 769 elements, but it introduces a new (conceptual) network element, the 770 EID Route Server, defined as a router that either propagates routes 771 learned from other EID Route Servers, or it originates EID Routes. 772 The EID-Routes that it originates are those that it is authoritative 773 for. It propagates these routes to Proxy-ITRs within the AS of the 774 EID Route Server. It is worth to note that a BGP capable router can 775 be also considered as an EID Route Server. 777 Further, an EID-Route is defined as a prefix originated via the Route 778 Server of the mapping service provider, which should be aggregated if 779 the MSP has multiple customers inside a single netblock. This prefix 780 is propagated to other P-ITRs both within the MSP and to other P-ITR 781 operators it peers with. EID Route Servers are operated either by 782 the LISP site, MSPs or PSPs, and they may be collocated with a Map- 783 Server or P-ITR, but are a functionally discrete entity. They 784 distribute EID-Routes, using BGP, to other domains, according to 785 policies set by participants. 787 MSP (AS64500) 788 RS ---> P-ITR 789 | / 790 | _.--./ 791 ,-'' /`--. 792 LISP site ---,' | v `. 793 ( | DFZ )----- Mapping system 794 non-LISP site ----. | ^ ,' 795 `--. / _.-' 796 | `--'' 797 v / 798 P-ITR 799 PSP (AS64501) 801 Figure 5: The P-ITR Route Distribution architecture 803 The architecture described above decouples EID origination from route 804 propagation, with the following benefits: 806 o Can accurately represent business relationships between P-ITR 807 operators 809 o More mapping system agnostic 811 o Minor changes to P-ITR implementation, no changes to other 812 components 814 In the example in the figure we have a MSP providing services to the 815 LISP site. The LISP site does not run BGP, and gets an EID 816 allocation directly from a RIR, or from the MSP, who may be a LIR. 817 Existing PI allocations can be migrated as well. The MSP ensures the 818 presence of the prefix in the mapping system, and runs an EID Route 819 Server to distribute it to P-ITR service providers. Since the LISP 820 site does not run BGP, the prefix will be originated with the AS 821 number of the MSP. 823 In the simple case depicted in Figure 5 the EID-Route of LISP Site 824 will be originated by the Route Server, and announced to the DFZ by 825 the PSP's P-ITRs with AS path 64501 64500. From that point on, the 826 usual BGP dynamics apply. This way, routes announced by P-ITR are 827 still originated by the authoritative Route Server. Note that the 828 peering relationships between MSP/PSPs and those in the underlying 829 forwarding plane may not be congruent, making the AS path to a P-ITR 830 shorter than it is in reality. 832 The non-LISP site will select the best path towards the EID-prefix, 833 according to its local BGP policies. Since AS-path length is usually 834 an important metric for selecting paths, a careful placement of P-ITR 835 could significantly reduce path-stretch between LISP and non-LISP 836 sites. 838 The architecture allows for flexible policies between MSP/PSPs. 839 Consider the EID Route Server networks as control plane overlays, 840 facilitating the implementation of policies necessary to reflect the 841 business relationships between participants. The results are then 842 injected to the common underlying forwarding plane. For example, 843 some MSP/PSPs may agree to exchange EID-Prefixes and only announce 844 them to each of their forwarding plane customers. Global 845 reachability of an EID-prefix depends on the MSP the LISP site buys 846 service from, and is also subject to agreement between the mentioned 847 parties. 849 In terms of impact on the DFZ, this architecture results in a slower 850 routing table increase for new allocations, since traffic engineering 851 will be done at the LISP level. For existing allocations migrating 852 to LISP, the DFZ may decrease since MSPs may be able to aggregate the 853 prefixes announced. 855 Compared to LISP+BGP, this approach avoids DFZ bloat caused by prefix 856 deaggregation for traffic engineering purposes, resulting in slower 857 routing table increase in the case of new allocations and potential 858 decrease for existing ones. Moreover, MSPs serving different clients 859 with adjacent aggregable prefixes may lead to additional decrease, 860 but quantifying this decrease is subject to future research study. 862 The flexibility and scalability of this architecture does not come 863 without a cost however: A PSP operator has to establish either 864 transit or peering relationships to improve their connectivity. 866 5.4. Migration Summary 868 The following table presents the expected effects of the different 869 transition scenarios during a certain phase on the DFZ routing table 870 size: 872 Phase | LISP+BGP | MSP P-ITR | PITR-RD 873 -----------------+--------------+-----------------+---------------- 874 Early transition | no change | slower increase | slower increase 875 Late transition | may decrease | slower increase | slower increase 876 LISP Internet | considerable decrease 878 It is expected that PITR-RD will co-exist with LISP+BGP during the 879 migration, with the latter being more popular in the early transition 880 phase. As the transition progresses and the MSP P-ITR and PITR-RD 881 ecosystem gets more ubiquitous, LISP+BGP should become less 882 attractive, slowing down the increase of the number of routes in the 883 DFZ. 885 6. Step-by-Step Example BGP to LISP Migration Procedure 887 6.1. Customer Pre-Install and Pre-Turn-up Checklist 889 1. Determine how many current physical service provider connections 890 the customer has and their existing bandwidth and traffic 891 engineering requirements. 893 This information will determine the number of routing locators, 894 and the priorities and weights that should be configured on the 895 xTRs. 897 2. Make sure customer router has LISP capabilities. 899 * Obtain output of 'show version' from the CE router. 901 This information can be used to determine if the platform is 902 appropriate to support LISP, in order to determine if a 903 software and/or hardware upgrade is required. 905 * Have customer upgrade (if necessary, software and/or hardware) 906 to be LISP capable. 908 3. Obtain current running configuration of CE router. A suggested 909 LISP router configuration example can be customized to the 910 customer's existing environment. 912 4. Verify MTU Handling 914 * Request increase in MTU to (1556) on service provider 915 connections. Prior to MTU change verify that 1500 byte packet 916 from P-xTR to RLOC with do not fragment (DF-bit) bit set. 918 * Ensure they are not filtering ICMP unreachable or time- 919 exceeded on their firewall or router. 921 LISP, like any tunneling protocol, will increase the size of 922 packets when the LISP header is appended. If increasing the MTU 923 of the access links is not possible, care must be taken that ICMP 924 is not being filtered in order to allow for Path MTU Discovery to 925 take place. 927 5. Validate member prefix allocation. 929 This step is to check if the prefix used by the customer is a 930 direct (Provider Independent), or if it is a prefix assigned by a 931 physical service provider (Provider Allocated). If the prefixes 932 are assigned by other service provivers then a Letter of 933 Agreement is required to announce prefixes through the Proxy 934 Service Provider. 936 6. Verify the member RLOCs and their reachability. 938 This step ensures that the RLOCs configured on the CE router are 939 in fact reachable and working. 941 7. Prepare for cut-over. 943 * If possible, have a host outside of all security and filtering 944 policies connected to the console port of the edge router or 945 switch. 947 * Make sure customer has access to the router in order to 948 configure it. 950 6.2. Customer Activating LISP Service 952 1. Customer configures LISP on CE router(s) from service provider 953 recommended configuration. 955 The LISP configuration consists of the EID prefix, the locators, 956 and the weights and priorities of the mapping between the two 957 values. In addition, the xTR must be configured with Map- 958 Resolver(s), Map-Server(s) and the shared key for registering to 959 Map-Server(s). If required, Proxy-ETR(s) may be configured as 960 well. 962 In addition to the LISP configuration, the following: 964 * Ensure default route(s) to next-hop external neighbors are 965 included and RLOCs are present in configuration. 967 * If two or more routers are used, ensure all RLOCs are included 968 in the LISP configuration on all routers. 970 * It will be necessary to redistribute default route via IGP 971 between the external routers. 973 2. When transition is ready perform a soft shutdown on existing eBGP 974 peer session(s) 976 * From CE router, use LIG to ensure registration is successful. 978 * To verify LISP connectivity, ping LISP connected sites. See 979 http://www.lisp4.net/ and/or http://www.lisp6.net/ for 980 potential candidates. 982 * To verify connectivity to non-LISP sites, try accessing major 983 Internet sites via a web browser. 985 6.3. Cut-Over Provider Preparation and Changes 987 1. Verify site configuration and then active registration on Map- 988 Server(s) 990 * Authentication key 992 * EID prefix 994 2. Add EID space to map-cache on proxies 996 3. Add networks to BGP advertisement on proxies 998 * Modify route-maps/policies on P-xTRs 1000 * Modify route policies on core routers (if non-connected 1001 member) 1003 * Modify ingress policers on core routers 1005 * Ensure route announcement in looking glass servers, RouteViews 1007 4. Perform traffic verification test 1009 * Ensure MTU handling is as expected (PMTUD working) 1011 * Ensure proxy-ITR map-cache population 1013 * Ensure access from traceroute/ping servers around Internet 1015 * Use a looking glass, to check for external visibility of 1016 registration via several Map-Resolvers (e.g., 1017 http://lispmon.net/). 1019 7. Security Considerations 1021 Security implications of LISP deployments are to be discussed in 1022 separate documents. [I-D.saucez-lisp-security] gives an overview of 1023 LISP threat models, while securing mapping lookups is discussed in 1024 [I-D.ietf-lisp-sec]. 1026 8. IANA Considerations 1028 This memo includes no request to IANA. 1030 9. Acknowledgements 1032 Many thanks to Margaret Wasserman for her contribution to the IETF76 1033 presentation that kickstarted this work. The authors would also like 1034 to thank Damien Saucez, Luigi Iannone, Joel Halpern, Vince Fuller, 1035 Dino Farinacci, Terry Manderson, Noel Chiappa, Hannu Flinck, and 1036 everyone else who provided input. 1038 10. Informative References 1040 [I-D.ietf-lisp] 1041 Farinacci, D., Fuller, V., Meyer, D., and D. Lewis, 1042 "Locator/ID Separation Protocol (LISP)", 1043 draft-ietf-lisp-23 (work in progress), May 2012. 1045 [I-D.ietf-lisp-eid-block] 1046 Iannone, L., Lewis, D., Meyer, D., and V. Fuller, "LISP 1047 EID Block", draft-ietf-lisp-eid-block-02 (work in 1048 progress), April 2012. 1050 [I-D.ietf-lisp-interworking] 1051 Lewis, D., Meyer, D., Farinacci, D., and V. Fuller, 1052 "Interworking LISP with IPv4 and IPv6", 1053 draft-ietf-lisp-interworking-06 (work in progress), 1054 March 2012. 1056 [I-D.ietf-lisp-ms] 1057 Fuller, V. and D. Farinacci, "LISP Map Server Interface", 1058 draft-ietf-lisp-ms-16 (work in progress), March 2012. 1060 [I-D.ietf-lisp-sec] 1061 Maino, F., Ermagan, V., Cabellos-Aparicio, A., Saucez, D., 1062 and O. Bonaventure, "LISP-Security (LISP-SEC)", 1063 draft-ietf-lisp-sec-02 (work in progress), March 2012. 1065 [I-D.saucez-lisp-security] 1066 Saucez, D., Iannone, L., and O. Bonaventure, "LISP 1067 Security Threats", draft-saucez-lisp-security-03 (work in 1068 progress), March 2011. 1070 [cache] Jung, J., Sit, E., Balakrishnan, H., and R. Morris, "DNS 1071 performance and the effectiveness of caching", 2002. 1073 Authors' Addresses 1075 Lorand Jakab 1076 Technical University of Catalonia 1077 C/Jordi Girona, s/n 1078 BARCELONA 08034 1079 Spain 1081 Email: ljakab@ac.upc.edu 1083 Albert Cabellos-Aparicio 1084 Technical University of Catalonia 1085 C/Jordi Girona, s/n 1086 BARCELONA 08034 1087 Spain 1089 Email: acabello@ac.upc.edu 1091 Florin Coras 1092 Technical University of Catalonia 1093 C/Jordi Girona, s/n 1094 BARCELONA 08034 1095 Spain 1097 Email: fcoras@ac.upc.edu 1099 Jordi Domingo-Pascual 1100 Technical University of Catalonia 1101 C/Jordi Girona, s/n 1102 BARCELONA 08034 1103 Spain 1105 Email: jordi.domingo@ac.upc.edu 1107 Darrel Lewis 1108 Cisco Systems 1109 170 Tasman Drive 1110 San Jose, CA 95134 1111 USA 1113 Email: darlewis@cisco.com