idnits 2.17.1 draft-jasinska-ix-bgp-route-server-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 (March 15, 2011) is 4790 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) == Unused Reference: 'RFC1863' is defined on line 405, but no explicit reference was found in the text == Unused Reference: 'RFC4223' is defined on line 408, but no explicit reference was found in the text == Unused Reference: 'RFC4760' is defined on line 411, but no explicit reference was found in the text == Unused Reference: 'RFC5065' is defined on line 415, but no explicit reference was found in the text == Unused Reference: 'RFC5226' is defined on line 418, but no explicit reference was found in the text == Outdated reference: A later version (-08) exists of draft-ietf-grow-diverse-bgp-path-dist-03 ** Downref: Normative reference to an Informational draft: draft-ietf-grow-diverse-bgp-path-dist (ref. 'I-D.ietf-grow-diverse-bgp-path-dist') == Outdated reference: A later version (-15) exists of draft-ietf-idr-add-paths-04 -- Obsolete informational reference (is this intentional?): RFC 1863 (Obsoleted by RFC 4223) -- Obsolete informational reference (is this intentional?): RFC 5226 (Obsoleted by RFC 8126) Summary: 1 error (**), 0 flaws (~~), 8 warnings (==), 3 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 IDR Working Group E. Jasinska 3 Internet-Draft Limelight Networks 4 Intended status: Standards Track N. Hilliard 5 Expires: September 16, 2011 INEX 6 R. Raszuk 7 Cisco Systems 8 N. Bakker 9 AMS-IX B.V. 10 March 15, 2011 12 Internet Exchange Route Server 13 draft-jasinska-ix-bgp-route-server-02 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 exterior 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 September 16, 2011. 43 Copyright Notice 45 Copyright (c) 2011 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 . . . . . . . . . 3 61 1.1. Specification of Requirements . . . . . . . . . . . . . . 3 62 2. Technical Considerations for Route Server Implementations . . 4 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 70 Interconnection . . . . . . . . . . . . . . . . . . . . . 5 71 2.3.1. Path Hiding on a Route Server . . . . . . . . . . . . 6 72 2.3.2. Implementing Per-Client Policy Control . . . . . . . . 7 73 2.3.2.1. Multiple Route Server RIBs . . . . . . . . . . . . 7 74 2.3.2.2. Advertising Multiple Paths . . . . . . . . . . . . 7 75 3. Security Considerations . . . . . . . . . . . . . . . . . . . 8 76 4. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 8 77 5. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 9 78 6. References . . . . . . . . . . . . . . . . . . . . . . . . . . 9 79 6.1. Normative References . . . . . . . . . . . . . . . . . . . 9 80 6.2. Informative References . . . . . . . . . . . . . . . . . . 9 81 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 10 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 exterior BGP sessions between exchange participants 93 were previously the most common means of exchanging reachability 94 information, the overhead associated with dense interconnection has 95 caused substantial operational scaling problems for Internet exchange 96 point participants. 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 exterior BGP, and 104 the route server in turn forwards this information to every 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, which are outlined in Section 2 of this 114 document. 116 The term "route server" is often in a different context used to 117 describe a BGP node whose purpose is to accept BGP feeds from 118 multiple clients for the purpose of operational analysis and 119 troubleshooting. A system of this form may alternatively be known as 120 a "route collector" or a "route-views server". This document uses 121 the term "route server" exclusively to describe multilateral peering 122 brokerage systems. 124 1.1. Specification of Requirements 126 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 127 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 128 "OPTIONAL" in this document are to be interpreted as described in 129 [RFC2119]. 131 2. Technical Considerations for Route Server Implementations 133 2.1. Client UPDATE Messages 135 A route server MUST accept all UPDATE messages received from each of 136 its clients for inclusion in its Adj-RIB-In. These UPDATE messages 137 MAY by omitted from the route server's Loc-RIB or Loc-RIBs, due to 138 filters configured for the purposes of implementing routing policy. 139 The route server SHOULD perform one or more BGP Decision Processes to 140 select routes for subsequent advertisement to its clients, taking 141 into account possible configuration to provide multiple NLRI paths to 142 a particular client as described in Section 2.3.2.2 or multiple Loc- 143 RIBs as described in Section 2.3.2.1. The route server SHOULD 144 forward UPDATE messages where appropriate from its Loc-RIB or Loc- 145 RIBs to its clients. 147 2.2. Attribute Transparency 149 As a route server primarily performs a brokering service, 150 modification of attributes could cause route server clients to alter 151 their BGP best path selection process for received prefix 152 reachability information, thereby changing the intended routing 153 policies of exchange participants. Therefore, contrary to what is 154 specified in section 5. of [RFC4271], route servers SHOULD NOT update 155 well-known BGP attributes received from route server clients before 156 redistributing them to their other route server clients. Optional 157 recognized and unrecognized BGP attributes, whether transitive or 158 non-transitive, SHOULD NOT be updated by the route server and SHOULD 159 be passed on to other route server clients. 161 2.2.1. NEXT_HOP Attribute 163 The NEXT_HOP, a well-known mandatory BGP attribute, defines the IP 164 address of the router used as the next hop to the destinations listed 165 in the Network Layer Reachability Information field of the UPDATE 166 message. As the route server does not participate in the actual 167 routing of traffic, the NEXT_HOP attribute MUST be passed unmodified 168 to the route server clients, similar to the "third party" next hop 169 feature described in section 5.1.3. of [RFC4271]. 171 2.2.2. AS_PATH Attribute 173 AS_PATH is a well-known mandatory attribute which identifies the 174 autonomous systems through which routing information carried in the 175 UPDATE message has passed. 177 As a route server does not participate in the process of forwarding 178 data between client routers, and because modification of the AS_PATH 179 attribute could affect route server client best path calculations, 180 the route server SHOULD NOT prepend its own AS number to the AS_PATH 181 segment nor modify the AS_PATH segment in any other way. 183 2.2.3. MULTI_EXIT_DISC Attribute 185 MULTI_EXIT_DISC is an optional non-transitive attribute intended to 186 be used on external (inter-AS) links to discriminate among multiple 187 exit or entry points to the same neighboring AS. If applied to an 188 NLRI UPDATE sent to a route server, the attribute (contrary to 189 section 5.1.4 of [RFC4271]) SHOULD be propagated to other route 190 server clients and the route server SHOULD NOT modify the value of 191 this attribute. 193 2.2.4. Communities Attributes 195 The BGP COMMUNITIES ([RFC1997]) and Extended Communities ([RFC4360]) 196 attributes are attributes intended for labeling information carried 197 in BGP UPDATE messages. Transitive as well as non-transitive 198 Communities attributes applied to an NLRI UPDATE sent to a route 199 server SHOULD NOT be modified, processed or removed. However, if 200 such an attribute is intended for processing by the route server 201 itself, it MAY be modified or removed. 203 2.3. Per-Client Policy Control in Multilateral Interconnection 205 While IXP participants often use route servers with the intention of 206 interconnecting with as many other route server participants as 207 possible, there are circumstances where control of path distribution 208 on a per-client basis is important for ensuring that desired 209 interconnection policies are met. 211 The control of path distribution on a per-client basis can lead to a 212 path being hidden from the route server client. We refer to this as 213 "path hiding", which is described in Section 2.3.1. 215 Route server implementations SHOULD implement one of the methods 216 described in Section 2.3.2, for the operator to be able to allow the 217 control of path distribution on a per-client basis without the 218 occurrence of "path hiding". 220 2.3.1. Path Hiding on a Route Server 222 ___ ___ 223 / \ / \ 224 ..| AS1 |..| AS2 |.. 225 : \___/ \___/ : 226 : \ / | : 227 : \ / | : 228 : IXP \/ | : 229 : /\ | : 230 : / \ | : 231 : ___/____\_|_ : 232 : / \ / \ : 233 ..| AS3 |..| AS4 |.. 234 \___/ \___/ 236 Figure 1: Per-Client Policy Controlled Interconnection at an IXP 238 Using the example in Figure 1, AS1 does not directly exchange prefix 239 information with either AS2 or AS3 at the IXP, but only interconnects 240 with AS4. 242 In the traditional bilateral interconnection model, per-client policy 243 control to a third party exchange participant is accomplished either 244 by not engaging in a bilateral interconnection with that participant 245 or else by implementing outbound filtering on the BGP session towards 246 that participant. However, in a multilateral interconnection 247 environment, only the route server can perform outbound filtering in 248 the direction of the route server client; route server clients depend 249 on the route server to perform their outbound filtering for them. 251 Assuming a traditional best path selection, when the same prefix is 252 advertised to a route server from multiple route server clients, the 253 route server will select a single best path for propagation to all 254 connected clients. If, however, the route server has been configured 255 to filter the calculated best path from reaching a particular route 256 server client, then that client will not receive a path for that 257 prefix, although alternate paths received by the route server might 258 have been policy compliant for that client. This phenomenon is 259 referred to as "path hiding". 261 For example, in Figure 1, if the same prefix were sent to the route 262 server via AS2 and AS4, and the route via AS2 was preferred according 263 to BGP's traditional best path selection, but AS1's policy prevents 264 AS2's path from being accepted, then AS1 would never receive a path 265 to this prefix, even though the route server had previously received 266 a valid alternative path via AS4. This happens because the best path 267 selection is performed only once on the route server for all clients. 269 Path hiding will only occur on route servers which employ per-client 270 policy control; if an IXP operator deploys a route server without the 271 possibility for policy control, then path hiding does not occur, as 272 all paths are considered equally valid from the point of view of the 273 route server. 275 2.3.2. Implementing Per-Client Policy Control 277 In this section, we describe the alternatives to provide per-client 278 policy control while preventing the occurrence of path hiding. 280 2.3.2.1. Multiple Route Server RIBs 282 The most portable means to allow for per-client policy control 283 without the occurrence of path hiding, is by using a route server BGP 284 implementation which performs the per-client best path calculation 285 for each set of paths to a prefix, which results after the route 286 server's client policies have been taken into consideration. This 287 can be implemented by using per-client Loc-RIBs, with path filtering 288 implemented between the Adj-RIB-In and the per-client Loc-RIB. 289 Implementations MAY optimize this by maintaining paths not subject to 290 filtering policies in a global Loc-RIB, with per-client Loc-RIBs 291 stored as deltas. 293 This implementation is highly portable, as it makes no assumptions 294 about the feature capabilities of the route server clients. 296 2.3.2.2. Advertising Multiple Paths 298 The path distribution model described above assumes standard BGP 299 session encoding where the route server sends a single path to its 300 client for any given prefix. This path is selected using the BGP 301 path selection decision process described in [RFC4271]. If, however, 302 it were possible for the route server to send more than a single path 303 to a route server client, then route server clients would no longer 304 depend on receiving a single best path to a particular prefix; 305 consequently, the path hiding problem described in Section 2.3.1 306 would disappear. 308 We present two methods which describe how such increased path 309 diversity could be implemented. 311 2.3.2.2.1. Diverse BGP Path Approach 313 The Diverse BGP Path proposal as defined in 314 [I-D.ietf-grow-diverse-bgp-path-dist] is a simple way to distribute 315 multiple prefix paths from a route server to a route server client by 316 using a separate BGP session from the route server to a client for 317 each different path. 319 The number of paths which may be distributed to a client is 320 constrained by the number of BGP sessions which the server and the 321 client are willing to establish with each other. The distributed 322 paths may be established from the global BGP Loc-RIB on the route 323 server in addition to any per-client Loc-RIB. As there may be more 324 potential paths to a given prefix than configured BGP sessions, this 325 method is not guaranteed to eliminate the path hiding problem in all 326 situations. Furthermore, this method may significantly increase the 327 number of BGP sessions handled by the route server, which may 328 negatively impact its performance. 330 2.3.2.2.2. BGP ADD-PATH Approach 332 The [I-D.ietf-idr-add-paths] Internet draft proposes a different 333 approach to multiple path propagation, by allowing a BGP speaker to 334 forward multiple paths for the same prefix on a single BGP session. 335 As [RFC4271] specifies that a BGP listener must implement an implicit 336 withdraw when it receives an UPDATE message for a prefix which 337 already exists in its Adj-RIB-In, this approach requires explicit 338 support for the feature both on the route server and on its clients. 340 If the ADD-PATH capability is negotiated bidirectionally between the 341 route server and a route server client, and the route server client 342 propagates multiple paths for the same prefix to the route server, 343 then this could potentially cause the propagation of inactive, 344 invalid or suboptimal paths to the route server, thereby causing loss 345 of reachability to other route server clients. For this reason, ADD- 346 PATH implementations on a route server SHOULD enforce send-only mode 347 with the route server clients, which would result in negotiating 348 receive-only mode from the client to the route server. 350 3. Security Considerations 352 The path hiding problem outlined in section Section 2.3.1 can be used 353 in certain circumstances to proactively block third party path 354 announcements from other route server clients. 356 4. IANA Considerations 358 The new set of mechanism for route servers does not require any new 359 allocations from IANA. 361 5. Acknowledgments 363 The authors would like to thank Ryan Bickhart, Steven Bakker, Chris 364 Hall, Bruno Decraene and Pierre Francois for their valuable input. 366 In addition, the authors would like to acknowledge the developers of 367 BIRD, OpenBGPD and Quagga, whose open source BGP implementations 368 include route server capabilities which are compliant with this 369 document. 371 6. References 373 6.1. Normative References 375 [I-D.ietf-grow-diverse-bgp-path-dist] 376 Raszuk, R., Fernando, R., Patel, K., McPherson, D., and K. 377 Kumaki, "Distribution of diverse BGP paths.", 378 draft-ietf-grow-diverse-bgp-path-dist-03 (work in 379 progress), January 2011. 381 [I-D.ietf-idr-add-paths] 382 Walton, D., Retana, A., Chen, E., and J. Scudder, 383 "Advertisement of Multiple Paths in BGP", 384 draft-ietf-idr-add-paths-04 (work in progress), 385 August 2010. 387 [RFC1997] Chandrasekeran, R., Traina, P., and T. Li, "BGP 388 Communities Attribute", RFC 1997, August 1996. 390 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 391 Requirement Levels", BCP 14, RFC 2119, March 1997. 393 [RFC4271] Rekhter, Y., Li, T., and S. Hares, "A Border Gateway 394 Protocol 4 (BGP-4)", RFC 4271, January 2006. 396 [RFC4360] Sangli, S., Tappan, D., and Y. Rekhter, "BGP Extended 397 Communities Attribute", RFC 4360, February 2006. 399 [RFC4456] Bates, T., Chen, E., and R. Chandra, "BGP Route 400 Reflection: An Alternative to Full Mesh Internal BGP 401 (IBGP)", RFC 4456, April 2006. 403 6.2. Informative References 405 [RFC1863] Haskin, D., "A BGP/IDRP Route Server alternative to a full 406 mesh routing", RFC 1863, October 1995. 408 [RFC4223] Savola, P., "Reclassification of RFC 1863 to Historic", 409 RFC 4223, October 2005. 411 [RFC4760] Bates, T., Chandra, R., Katz, D., and Y. Rekhter, 412 "Multiprotocol Extensions for BGP-4", RFC 4760, 413 January 2007. 415 [RFC5065] Traina, P., McPherson, D., and J. Scudder, "Autonomous 416 System Confederations for BGP", RFC 5065, August 2007. 418 [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an 419 IANA Considerations Section in RFCs", BCP 26, RFC 5226, 420 May 2008. 422 Authors' Addresses 424 Elisa Jasinska 425 Limelight Networks 426 2220 W 14th St 427 Tempe, AZ 85281 428 US 430 Email: elisa@llnw.com 432 Nick Hilliard 433 INEX 434 4027 Kingswood Road 435 Dublin 24 436 IE 438 Email: nick@inex.ie 440 Robert Raszuk 441 Cisco Systems 442 170 West Tasman Drive 443 San Jose, CA 95134 444 US 446 Email: raszuk@cisco.com 447 Niels Bakker 448 AMS-IX B.V. 449 Westeinde 12 450 Amsterdam, NH 1017 ZN 451 NL 453 Email: niels.bakker@ams-ix.net