idnits 2.17.1 draft-ietf-idr-ix-bgp-route-server-11.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 (June 10, 2016) is 2849 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) -- Obsolete informational reference (is this intentional?): RFC 1863 (Obsoleted by RFC 4223) Summary: 0 errors (**), 0 flaws (~~), 1 warning (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 IDR Working Group E. Jasinska 3 Internet-Draft BigWave IT 4 Intended status: Standards Track N. Hilliard 5 Expires: December 12, 2016 INEX 6 R. Raszuk 7 Bloomberg LP 8 N. Bakker 9 Akamai Technologies B.V. 10 June 10, 2016 12 Internet Exchange BGP Route Server 13 draft-ietf-idr-ix-bgp-route-server-11 15 Abstract 17 This document outlines a specification for multilateral 18 interconnections at Internet exchange points (IXPs). Multilateral 19 interconnection is a method of exchanging routing information between 20 three or more external BGP speakers using a single intermediate 21 broker system, referred to as a route server. Route servers are 22 typically used on shared access media networks, such as Internet 23 exchange points (IXPs), to facilitate simplified interconnection 24 between multiple Internet routers. 26 Status of This Memo 28 This Internet-Draft is submitted in full conformance with the 29 provisions of BCP 78 and BCP 79. 31 Internet-Drafts are working documents of the Internet Engineering 32 Task Force (IETF). Note that other groups may also distribute 33 working documents as Internet-Drafts. The list of current Internet- 34 Drafts is at http://datatracker.ietf.org/drafts/current/. 36 Internet-Drafts are draft documents valid for a maximum of six months 37 and may be updated, replaced, or obsoleted by other documents at any 38 time. It is inappropriate to use Internet-Drafts as reference 39 material or to cite them other than as "work in progress." 41 This Internet-Draft will expire on December 12, 2016. 43 Copyright Notice 45 Copyright (c) 2016 IETF Trust and the persons identified as the 46 document authors. All rights reserved. 48 This document is subject to BCP 78 and the IETF Trust's Legal 49 Provisions Relating to IETF Documents 50 (http://trustee.ietf.org/license-info) in effect on the date of 51 publication of this document. Please review these documents 52 carefully, as they describe your rights and restrictions with respect 53 to this document. Code Components extracted from this document must 54 include Simplified BSD License text as described in Section 4.e of 55 the Trust Legal Provisions and are provided without warranty as 56 described in the Simplified BSD License. 58 Table of Contents 60 1. Introduction to Multilateral Interconnection . . . . . . . . 2 61 1.1. Notational Conventions . . . . . . . . . . . . . . . . . 3 62 2. Technical Considerations for Route Server Implementations . . 3 63 2.1. Client UPDATE Messages . . . . . . . . . . . . . . . . . 4 64 2.2. Attribute Transparency . . . . . . . . . . . . . . . . . 4 65 2.2.1. NEXT_HOP Attribute . . . . . . . . . . . . . . . . . 4 66 2.2.2. AS_PATH Attribute . . . . . . . . . . . . . . . . . . 4 67 2.2.3. MULTI_EXIT_DISC Attribute . . . . . . . . . . . . . . 5 68 2.2.4. Communities Attributes . . . . . . . . . . . . . . . 5 69 2.3. Per-Client Policy Control in Multilateral Interconnection 5 70 2.3.1. Path Hiding on a Route Server . . . . . . . . . . . . 6 71 2.3.2. Mitigation of Path Hiding . . . . . . . . . . . . . . 7 72 2.3.2.1. Multiple Route Server RIBs . . . . . . . . . . . 7 73 2.3.2.2. Advertising Multiple Paths . . . . . . . . . . . 7 74 2.3.3. Implementation Suggestions . . . . . . . . . . . . . 8 75 3. Security Considerations . . . . . . . . . . . . . . . . . . . 9 76 4. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 9 77 5. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 9 78 6. References . . . . . . . . . . . . . . . . . . . . . . . . . 10 79 6.1. Normative References . . . . . . . . . . . . . . . . . . 10 80 6.2. Informative References . . . . . . . . . . . . . . . . . 10 81 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 11 83 1. Introduction to Multilateral Interconnection 85 Internet exchange points (IXPs) provide IP data interconnection 86 facilities for their participants, typically using shared Layer-2 87 networking media such as Ethernet. The Border Gateway Protocol (BGP) 88 [RFC4271], an inter-Autonomous System routing protocol, is commonly 89 used to facilitate exchange of network reachability information over 90 such media. 92 While bilateral external BGP sessions between exchange participants 93 were previously the most common means of exchanging reachability 94 information, the overhead associated with dense interconnection can 95 cause substantial operational scaling problems for partipants of 96 larger Internet exchange points. 98 Multilateral interconnection is a method of interconnecting BGP 99 speaking routers using a third party brokering system, commonly 100 referred to as a route server and typically managed by the IXP 101 operator. Each of the multilateral interconnection participants 102 (usually referred to as route server clients) announces network 103 reachability information to the route server using external BGP, and 104 the route server in turn forwards this information to each other 105 route server client connected to it, according to its configuration. 106 Although a route server uses BGP to exchange reachability information 107 with each of its clients, it does not forward traffic itself and is 108 therefore not a router. 110 A route server can be viewed as similar in function to an [RFC4456] 111 route reflector, except that it operates using EBGP instead of iBGP. 112 Certain adaptions to [RFC4271] are required to enable an EBGP router 113 to operate as a route server; these are outlined in Section 2 of this 114 document. Route server functionality is not mandatory in BGP 115 implementations.`` 117 The term "route server" is often in a different context used to 118 describe a BGP node whose purpose is to accept BGP feeds from 119 multiple clients for the purpose of operational analysis and 120 troubleshooting. A system of this form may alternatively be known as 121 a "route collector" or a "route-views server". This document uses 122 the term "route server" exclusively to describe multilateral peering 123 brokerage systems. 125 1.1. Notational Conventions 127 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 128 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 129 "OPTIONAL" in this document are to be interpreted as described in 130 [RFC2119]. 132 2. Technical Considerations for Route Server Implementations 134 A route server uses the [RFC4271] Border Gateway Protocol to broker 135 network reachability information between its clients. There are some 136 differences between the behaviour of a BGP route server and a BGP 137 implementation which is strictly compliant with [RFC4271]. These 138 differences are described as follows. 140 2.1. Client UPDATE Messages 142 A route server MUST accept all UPDATE messages received from each of 143 its clients for inclusion in its Adj-RIB-In. These UPDATE messages 144 MAY be omitted from the route server's Loc-RIB or Loc-RIBs, due to 145 filters configured for the purposes of implementing routing policy. 146 The route server SHOULD perform one or more BGP Decision Processes to 147 select routes for subsequent advertisement to its clients, taking 148 into account possible configuration to provide multiple NLRI paths to 149 a particular client as described in Section 2.3.2.2 or multiple Loc- 150 RIBs as described in Section 2.3.2.1. The route server SHOULD 151 forward UPDATE messages from its Loc-RIB or Loc-RIBs to its clients 152 as determined by local policy. 154 2.2. Attribute Transparency 156 As a route server primarily performs a brokering service, 157 modification of attributes could cause route server clients to alter 158 their BGP Decision Process for received prefix reachability 159 information, thereby changing the intended routing policies of 160 exchange participants. Therefore, contrary to what is specified in 161 section 5. of [RFC4271], route servers SHOULD NOT by default (unless 162 explicitly configured) update well-known BGP attributes received from 163 route server clients before redistributing them to their other route 164 server clients. Optional recognized and unrecognized BGP attributes, 165 whether transitive or non-transitive, SHOULD NOT be updated by the 166 route server (unless enforced by local IXP operator configuration) 167 and SHOULD be passed on to other route server clients. 169 2.2.1. NEXT_HOP Attribute 171 The NEXT_HOP is a well-known mandatory BGP attribute which defines 172 the IP address of the router used as the next hop to the destinations 173 listed in the Network Layer Reachability Information field of the 174 UPDATE message. As the route server does not participate in the 175 actual routing of traffic, the NEXT_HOP attribute MUST be passed 176 unmodified to the route server clients, similar to the "third party" 177 next hop feature described in section 5.1.3. of [RFC4271]. 179 2.2.2. AS_PATH Attribute 181 AS_PATH is a well-known mandatory attribute which identifies the 182 autonomous systems through which routing information carried in the 183 UPDATE message has passed. 185 As a route server does not participate in the process of forwarding 186 data between client routers, and because modification of the AS_PATH 187 attribute could affect route server client BGP Decision Process, the 188 route server SHOULD NOT prepend its own AS number to the AS_PATH 189 segment nor modify the AS_PATH segment in any other way. This 190 differs from the behaviour specified in section 5.1.2 of [RFC4271], 191 which requires that the BGP speaker prepends its own AS number as the 192 last element of the AS_PATH segment. 194 In contrast to what is recommended in section 6.3 of [RFC4271], route 195 server clients need to be able to accept UPDATE messages where the 196 leftmost AS in the AS_PATH attribute is not equal to the AS number of 197 the route server that sent the UPDATE message. If the route server 198 client BGP system has implemented a check for this, the BGP 199 implementation MUST allow this check to be disabled and SHOULD allow 200 the check to be disabled on a per-peer basis. 202 2.2.3. MULTI_EXIT_DISC Attribute 204 MULTI_EXIT_DISC is an optional non-transitive attribute intended to 205 be used on external (inter-AS) links to discriminate among multiple 206 exit or entry points to the same neighboring AS. Contrary to section 207 5.1.4 of [RFC4271], if applied to an NLRI UPDATE sent to a route 208 server, this attribute SHOULD be propagated to other route server 209 clients and the route server SHOULD NOT modify its value. 211 2.2.4. Communities Attributes 213 The BGP COMMUNITIES ([RFC1997]) and Extended Communities ([RFC4360]) 214 attributes are attributes intended for labeling information carried 215 in BGP UPDATE messages. Transitive as well as non-transitive 216 Communities attributes applied to an NLRI UPDATE sent to a route 217 server SHOULD NOT be modified, processed or removed, except as 218 defined by local policy. If a Communities Attribute is intended for 219 processing by the route server itself, as determined by local policy, 220 it MAY be modified or removed. 222 2.3. Per-Client Policy Control in Multilateral Interconnection 224 While IXP participants often use route servers with the intention of 225 interconnecting with as many other route server participants as 226 possible, there are circumstances where control of path distribution 227 on a per-client basis is important to ensure that desired 228 interconnection policies are met. 230 The control of path distribution on a per-client basis can lead to a 231 path being hidden from the route server client. We refer to this as 232 "path hiding". 234 Neither Section 2.3 nor its subsections form part of the normative 235 specification of this document, but are included for information 236 purposes only. 238 2.3.1. Path Hiding on a Route Server 240 ___ ___ 241 / \ / \ 242 ..| AS1 |..| AS2 |.. 243 : \___/ \___/ : 244 : \ / | : 245 : \ / | : 246 : IXP \/ | : 247 : /\ | : 248 : / \ | : 249 : ___/____\_|_ : 250 : / \ / \ : 251 ..| AS3 |..| AS4 |.. 252 \___/ \___/ 254 Figure 1: Per-Client Policy Controlled Interconnection at an IXP 256 Using the example in Figure 1, AS1 does not directly exchange prefix 257 information with either AS2 or AS3 at the IXP, but only interconnects 258 with AS4. 260 In the traditional bilateral interconnection model, per-client policy 261 control to a third party exchange participant is accomplished either 262 by not engaging in a bilateral interconnection with that participant 263 or else by implementing outbound filtering on the BGP session towards 264 that participant. However, in a multilateral interconnection 265 environment, only the route server can perform outbound filtering in 266 the direction of the route server client; route server clients depend 267 on the route server to perform their outbound filtering for them. 269 Assuming the [RFC4271] BGP Decision Process is used, when the same 270 prefix is advertised to a route server from multiple route server 271 clients, the route server will select a single path for propagation 272 to all connected clients. If, however, the route server has been 273 configured to filter the calculated best path from reaching a 274 particular route server client, then that client will not receive a 275 path for that prefix, although alternate paths received by the route 276 server might have been policy compliant for that client. This 277 phenomenon is referred to as "path hiding". 279 For example, in Figure 1, if the same prefix were sent to the route 280 server via AS2 and AS4, and the route via AS2 was preferred according 281 to the BGP Decision Process on the route server, but AS2's policy 282 prevented the route server from sending the path to AS1, then AS1 283 would never receive a path to this prefix, even though the route 284 server had previously received a valid alternative path via AS4. 285 This happens because the BGP Decision Process is performed only once 286 on the route server for all clients. 288 Path hiding will only occur on route servers which employ per-client 289 policy control; if an IXP operator deploys a route server without 290 implementing a per-client routing policy control system, then path 291 hiding does not occur as all paths are considered equally valid from 292 the point of view of the route server. 294 2.3.2. Mitigation of Path Hiding 296 There are several approaches which can be taken to mitigate against 297 path hiding. 299 2.3.2.1. Multiple Route Server RIBs 301 The most portable method to allow for per-client policy control 302 without the occurrence of path hiding, is by using a route server BGP 303 implementation which performs the per-client best path calculation 304 for each set of paths to a prefix, which results after the route 305 server's client policies have been taken into consideration. This 306 can be implemented by using per-client Loc-RIBs, with path filtering 307 implemented between the Adj-RIB-In and the per-client Loc-RIB. 308 Implementations can optimize this by maintaining paths not subject to 309 filtering policies in a global Loc-RIB, with per-client Loc-RIBs 310 stored as deltas. 312 This implementation is highly portable, as it makes no assumptions 313 about the feature capabilities of the route server clients. 315 2.3.2.2. Advertising Multiple Paths 317 The path distribution model described above assumes standard BGP 318 session encoding where the route server sends a single path to its 319 client for any given prefix. This path is selected using the BGP 320 path selection decision process described in [RFC4271]. If, however, 321 it were possible for the route server to send more than a single path 322 to a route server client, then route server clients would no longer 323 depend on receiving a single path to a particular prefix; 324 consequently, the path hiding problem described in Section 2.3.1 325 would disappear. 327 We present two methods which describe how such increased path 328 diversity could be implemented. 330 2.3.2.2.1. Diverse BGP Path Approach 332 The Diverse BGP Path proposal as defined in [RFC6774] is a simple way 333 to distribute multiple prefix paths from a route server to a route 334 server client by using a separate BGP session from the route server 335 to a client for each different path. 337 The number of paths which may be distributed to a client is 338 constrained by the number of BGP sessions which the server and the 339 client are willing to establish with each other. The distributed 340 paths may be established from the global BGP Loc-RIB on the route 341 server in addition to any per-client Loc-RIB. As there may be more 342 potential paths to a given prefix than configured BGP sessions, this 343 method is not guaranteed to eliminate the path hiding problem in all 344 situations. Furthermore, this method may significantly increase the 345 number of BGP sessions handled by the route server, which may 346 negatively impact its performance. 348 2.3.2.2.2. BGP ADD-PATH Approach 350 The [I-D.ietf-idr-add-paths] Internet draft proposes a different 351 approach to multiple path propagation, by allowing a BGP speaker to 352 forward multiple paths for the same prefix on a single BGP session. 353 As [RFC4271] specifies that a BGP listener must implement an implicit 354 withdraw when it receives an UPDATE message for a prefix which 355 already exists in its Adj-RIB-In, this approach requires explicit 356 support for the feature both on the route server and on its clients. 358 If the ADD-PATH capability is negotiated bidirectionally between the 359 route server and a route server client, and the route server client 360 propagates multiple paths for the same prefix to the route server, 361 then this could potentially cause the propagation of inactive, 362 invalid or suboptimal paths to the route server, thereby causing loss 363 of reachability to other route server clients. For this reason, ADD- 364 PATH implementations on a route server should enforce send-only mode 365 with the route server clients, which would result in negotiating 366 receive-only mode from the client to the route server. 368 2.3.3. Implementation Suggestions 370 Route server implementation authors may wish to consider one of the 371 methods described in Section 2.3.2 to allow per-client route server 372 policy control without "path hiding". 374 Recommendations for route server operations are described separately 375 in [I-D.ietf-grow-ix-bgp-route-server-operations]. 377 3. Security Considerations 379 The path hiding problem outlined in section Section 2.3.1 can be used 380 in certain circumstances to proactively block third party path 381 announcements from other route server clients. Route server 382 operators should be aware that security issues may arise unless steps 383 are taken to mitigate against path hiding. 385 The AS_PATH check described in Section 2.2.2 is normally enabled in 386 order to check for malformed AS paths. If this check is disabled, 387 the route server client loses the ability to check incoming UPDATE 388 messages for certain categories of problems. This could potentially 389 cause corrupted BGP UPDATE messages to be propagated where they might 390 not be propagated if the check were enabled. Regardless of any 391 problems relating to malformed UPDATE messages, this check is also 392 used to detect BGP loops, so removing the check could potentially 393 cause routing loops to be formed. Consequently, this check SHOULD 394 NOT be disabled by IXP participants unless it is needed to establish 395 BGP sessions with a route server, and if possible should only be 396 disabled for peers which are route servers. 398 Route server operators should carefully consider the security 399 practices discussed in [RFC7454], "BGP Operations and Security". 401 4. IANA Considerations 403 The new set of mechanisms for route servers does not require any new 404 allocations from IANA. 406 5. Acknowledgments 408 The authors would like to thank Ryan Bickhart, Steven Bakker, Martin 409 Pels, Chris Hall, Aleksi Suhonen, Bruno Decraene, Pierre Francois and 410 Eduardo Ascenco Reis for their valuable input. 412 In addition, the authors would like to acknowledge the developers of 413 BIRD, OpenBGPD, Quagga and IOS whose BGP implementations include 414 route server capabilities which are compliant with this document. 416 Route server functionality was described in 1995 in [RFC1863] and 417 modern route server implementations are based on concepts developed 418 in the 1990s by the Routing Arbiter Project and the Route Server Next 419 Generation Project, managed by ISI and Merit. Although the original 420 RSNG code is no longer in use at any IXPs, the IXP community owes a 421 debt of gratitude to the many people who were involved in route 422 server development in the 1990s. Note that [RFC1863] was made 423 historical by [RFC4223]. 425 6. References 427 6.1. Normative References 429 [RFC1997] Chandra, R., Traina, P., and T. Li, "BGP Communities 430 Attribute", RFC 1997, DOI 10.17487/RFC1997, August 1996, 431 . 433 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 434 Requirement Levels", BCP 14, RFC 2119, 435 DOI 10.17487/RFC2119, March 1997, 436 . 438 [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A 439 Border Gateway Protocol 4 (BGP-4)", RFC 4271, 440 DOI 10.17487/RFC4271, January 2006, 441 . 443 [RFC4360] Sangli, S., Tappan, D., and Y. Rekhter, "BGP Extended 444 Communities Attribute", RFC 4360, DOI 10.17487/RFC4360, 445 February 2006, . 447 6.2. Informative References 449 [I-D.ietf-grow-ix-bgp-route-server-operations] 450 Hilliard, N., Jasinska, E., Raszuk, R., and N. Bakker, 451 "Internet Exchange BGP Route Server Operations", draft- 452 ietf-grow-ix-bgp-route-server-operations-05 (work in 453 progress), June 2015. 455 [I-D.ietf-idr-add-paths] 456 Walton, D., Retana, A., Chen, E., and J. Scudder, 457 "Advertisement of Multiple Paths in BGP", draft-ietf-idr- 458 add-paths-15 (work in progress), May 2016. 460 [RFC1863] Haskin, D., "A BGP/IDRP Route Server alternative to a full 461 mesh routing", RFC 1863, DOI 10.17487/RFC1863, October 462 1995, . 464 [RFC4223] Savola, P., "Reclassification of RFC 1863 to Historic", 465 RFC 4223, DOI 10.17487/RFC4223, October 2005, 466 . 468 [RFC4456] Bates, T., Chen, E., and R. Chandra, "BGP Route 469 Reflection: An Alternative to Full Mesh Internal BGP 470 (IBGP)", RFC 4456, DOI 10.17487/RFC4456, April 2006, 471 . 473 [RFC6774] Raszuk, R., Ed., Fernando, R., Patel, K., McPherson, D., 474 and K. Kumaki, "Distribution of Diverse BGP Paths", 475 RFC 6774, DOI 10.17487/RFC6774, November 2012, 476 . 478 [RFC7454] Durand, J., Pepelnjak, I., and G. Doering, "BGP Operations 479 and Security", BCP 194, RFC 7454, DOI 10.17487/RFC7454, 480 February 2015, . 482 Authors' Addresses 484 Elisa Jasinska 485 BigWave IT 486 ul. Skawinska 27/7 487 Krakow, MP 31-066 488 Poland 490 Email: elisa@bigwaveit.org 492 Nick Hilliard 493 INEX 494 4027 Kingswood Road 495 Dublin 24 496 IE 498 Email: nick@inex.ie 500 Robert Raszuk 501 Bloomberg LP 502 731 Lexington Ave 503 New York City, NY 10022 504 USA 506 Email: robert@raszuk.net 508 Niels Bakker 509 Akamai Technologies B.V. 510 Kingsfordweg 151 511 Amsterdam 1043 GR 512 NL 514 Email: nbakker@akamai.com