idnits 2.17.1 draft-gredler-bgp-te-00.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- == There are 1 instance of lines with private range IPv4 addresses in the document. If these are generic example addresses, they should be changed to use any of the ranges defined in RFC 6890 (or successor): 192.0.2.x, 198.51.100.x or 203.0.113.x. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == The document doesn't use any RFC 2119 keywords, yet has text resembling RFC 2119 boilerplate text. -- The document date (March 3, 2011) is 4803 days in the past. Is this intentional? -- Found something which looks like a code comment -- if you have code sections in the document, please surround them with '' and '' lines. Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) -- Possible downref: Non-RFC (?) normative reference: ref. 'IANA-ISIS' ** Downref: Normative reference to an Informational RFC: RFC 4655 ** Obsolete normative reference: RFC 4893 (Obsoleted by RFC 6793) ** Obsolete normative reference: RFC 5226 (Obsoleted by RFC 8126) == Outdated reference: A later version (-27) exists of draft-ietf-alto-protocol-06 Summary: 3 errors (**), 0 flaws (~~), 4 warnings (==), 3 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Inter-Domain Routing H. Gredler 3 Internet-Draft J. Medved 4 Intended status: Standards Track Juniper Networks, Inc. 5 Expires: September 4, 2011 March 3, 2011 7 Advertising Traffic Engineering Information in BGP 8 draft-gredler-bgp-te-00 10 Abstract 12 This document defines a new Border Gateway Protocol Network Layer 13 Reachability Information (BGP NLRI) encoding format that can be used 14 to distribute Traffic Engineering (TE) link information. Links can 15 be either physical links connecting physical nodes, or virtual paths 16 between physical or abstract nodes. The TE information is carried 17 via the BGP, thereby reusing protocol algorithms, operational 18 experience, and administrative processes, such as inter-provider 19 peering agreements. 21 The BGP protocol carrying Traffic Engineering (TE) information would 22 provide a well-defined, uniform, policy-controlled interface from the 23 network to outside servers that need to learn the network topology in 24 real-time, for example an ALTO Server or a Path Computation Server. 25 Having TE information from remote areas and/or Autonomous Systems 26 would allow path computation for inter-area and/or inter-AS source- 27 routed unicast and multicast tunnels. 29 Requirements Language 31 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 32 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 33 document are to be interpreted as described in RFC 2119 [RFC2119] 35 Status of this Memo 37 This Internet-Draft is submitted in full conformance with the 38 provisions of BCP 78 and BCP 79. 40 Internet-Drafts are working documents of the Internet Engineering 41 Task Force (IETF). Note that other groups may also distribute 42 working documents as Internet-Drafts. The list of current Internet- 43 Drafts is at http://datatracker.ietf.org/drafts/current/. 45 Internet-Drafts are draft documents valid for a maximum of six months 46 and may be updated, replaced, or obsoleted by other documents at any 47 time. It is inappropriate to use Internet-Drafts as reference 48 material or to cite them other than as "work in progress." 49 This Internet-Draft will expire on September 4, 2011. 51 Copyright Notice 53 Copyright (c) 2011 IETF Trust and the persons identified as the 54 document authors. All rights reserved. 56 This document is subject to BCP 78 and the IETF Trust's Legal 57 Provisions Relating to IETF Documents 58 (http://trustee.ietf.org/license-info) in effect on the date of 59 publication of this document. Please review these documents 60 carefully, as they describe your rights and restrictions with respect 61 to this document. Code Components extracted from this document must 62 include Simplified BSD License text as described in Section 4.e of 63 the Trust Legal Provisions and are provided without warranty as 64 described in the Simplified BSD License. 66 Table of Contents 68 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 69 2. Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 70 3. Transcoding TE Link Information Into a BGP NLRI . . . . . . . 5 71 3.1. TLV Format . . . . . . . . . . . . . . . . . . . . . . . . 6 72 3.2. Node anchors . . . . . . . . . . . . . . . . . . . . . . . 7 73 3.2.1. Router-ID Anchoring Example: ISO Pseudonode . . . . . 8 74 3.2.2. Router-ID Anchoring Example: OSPFv2 to IS-IS 75 Migration . . . . . . . . . . . . . . . . . . . . . . 8 76 3.3. Link Descriptors . . . . . . . . . . . . . . . . . . . . . 8 77 3.4. Link Attributes . . . . . . . . . . . . . . . . . . . . . 9 78 3.4.1. TE Default Metric TLV . . . . . . . . . . . . . . . . 10 79 3.4.2. IGP Link Metric TLV . . . . . . . . . . . . . . . . . 10 80 3.4.3. Shared Risk Link Group TLV . . . . . . . . . . . . . . 11 81 3.5. IGP Area Information . . . . . . . . . . . . . . . . . . . 11 82 3.6. Inter-AS Links . . . . . . . . . . . . . . . . . . . . . . 12 83 4. Link to Path Aggregation . . . . . . . . . . . . . . . . . . . 12 84 4.1. Example: No Link Aggregation . . . . . . . . . . . . . . . 12 85 4.2. Example: ASBR to ASBR Path Aggregation . . . . . . . . . . 13 86 4.3. Example: Multi-AS Path Aggregation . . . . . . . . . . . . 13 87 5. Originating the TED NLRI . . . . . . . . . . . . . . . . . . . 13 88 6. Receiving the TED NLRI . . . . . . . . . . . . . . . . . . . . 14 89 7. Use Cases . . . . . . . . . . . . . . . . . . . . . . . . . . 14 90 7.1. MPLS TE . . . . . . . . . . . . . . . . . . . . . . . . . 14 91 7.2. ALTO Server Network API . . . . . . . . . . . . . . . . . 15 92 7.3. Path Computation Element (PCE) TED Synchronization 93 Protocol . . . . . . . . . . . . . . . . . . . . . . . . . 16 94 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 16 95 9. Security Considerations . . . . . . . . . . . . . . . . . . . 16 96 10. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 16 97 11. References . . . . . . . . . . . . . . . . . . . . . . . . . . 17 98 11.1. Normative References . . . . . . . . . . . . . . . . . . . 17 99 11.2. Informative References . . . . . . . . . . . . . . . . . . 18 100 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 18 102 1. Introduction 104 Today, the contents of the traffic engineering database usually has 105 the scope of an IGP area. There are several use cases that could 106 benefit from knowing the topology or Traffic Engineering (TE) data in 107 a remote area or Autonomous System, but today no mechanism exists to 108 distribute this information beyond an IGP area. This draft proposes 109 to use BGP as the distribution mechanism for traffic engineering data 110 between routers in different IGP areas and/or Autonomous Systems. 111 The mechanism can also be used to exchange topology and TE data 112 between the network and external network-aware applications, such as 113 the Alto Servers. 115 The Border Gateway Protocol (BGP [RFC4271]) has grown beyond its 116 original intention of disseminating IPv4 Inter-domain routing paths. 117 A modern BGP implementation can be viewed as a ubiquitous database 118 replication mechanism, which allows replication of many different 119 state information types across arbitrary distribution graphs. Its 120 built-in loop protection mechanism (AS path, Cluster List attributes) 121 enables building of stable and redundant distribution topologies. In 122 addition to IP routing, applications that use BGP for state 123 distribution are L2VPN, VPLS, MAC-VPN, Route-target information, and 124 Flowspec for firewalling. Using BGP as a dissemination protocol for 125 Traffic Engineering data is a logical consequence. 127 A router maintains a database for storing Traffic Engineering related 128 data and link information. The Traffic Engineering Database (TED) is 129 populated by a link-state IGP routing protocol that supports TE 130 extensions: IS-IS or OSPF. The TED can be seen as a protocol-neutral 131 representation of links in the area. Link attributes stored in the 132 TED are: local/remote IP addresses, local/remote interface indices, 133 metric, link bandwidth, reservable bandwidth, per CoS class 134 reservation state, preemption and Shared Risk Link Groups (SRLG). 135 The router's BGP process can retrieve the TE data from the TED 136 database and distribute it to peer BGP Speakers using the encoding 137 specified in this draft. 139 A BGP Speaker may distribute the real physical topology from the TED, 140 or create an abstracted topology, where virtual, aggregated nodes are 141 connected by virtual paths. Aggregated nodes can be created, for 142 example, out of multiple routers in a POP. Abstracted topology can 143 also be a mix of physical and virtual nodes and physical and virtual 144 links. 146 Consumers of the TE data are peer routers in other areas either in 147 the router's own AS or in remote ASes, or entities outside the 148 network that may need network and/or TE data to optimize their 149 behavior. 151 2. Scope 153 The scope of TED NLRI are the static attributes / metrics of a path 154 between two routers. The path can be a physical link or multiple 155 links aggregated into a path. Dynamic data, such as dynamic 156 bandwidth or delay metrics, is out of scope of this draft. 158 3. Transcoding TE Link Information Into a BGP NLRI 160 The MP_REACH and MP_UNREACH attributes are BGP's containers for 161 carrying opaque information. Each TED NLRI describes a single link 162 anchored by at least a pair of router-IDs. Since there are many 163 Router-IDs formats (32 Bit IPv4 router-ID, 56 Bit ISO Node-ID and 128 164 Bit IPv6 router-ID) a link may be anchored by more than one Router-ID 165 pair. The anchoring Router-IDs are carried in the Node Anchor TLVs. 167 All TE link information shall be encoded using a TBD AFI / SAFI 1 or 168 SAFI 128 header into those attributes. SAFI 1 shall be used for 169 Internet routing (Public) and SAFI 128 shall be used for VPN routing 170 (Private) applications. 172 In order for two BGP speakers to exchange TE NLRI, they must use BGP 173 Capabilities Advertisement to ensure that they both are capable of 174 properly processing such NLRI. This is done as specified in 175 [RFC4760], by using capability code 1 (multiprotocol BGP), with an 176 AFI of TBD and an SAFI of 1 or 128. 178 0 1 2 3 179 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 180 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 181 | Total Link Length | 182 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 183 | Node Anchors (variable) | 184 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 185 | Link Descriptors (variable) | 186 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 187 | Link Attributes (variable) | 188 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 190 Figure 1: TED SAFI 1 NLRI Format 192 0 1 2 3 193 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 194 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 195 | Total Link Length | 196 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 197 | | 198 + Route Distinguisher + 199 | | 200 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 201 | Node Anchors (variable) | 202 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 203 | Link Descriptors (variable) | 204 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 205 | Link Attributes (variable) | 206 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 208 Figure 2: TED SAFI 128 NLRI Format 210 The 'Total Link Length" field contains the cumulative length of all 211 the TLVs, describing the Node Anchors, Link descriptors and Link 212 Attributes. For VPN applications it also includes the length of the 213 Route Distinguisher. 215 3.1. TLV Format 217 The Node anchor, Link descriptor and Link attribute fields are 218 described using a set of Type/Length/Value triplets. The format of 219 each TLV is shown in Figure 3 221 0 1 2 3 222 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 223 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 224 | Type | Length | 225 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 226 | | 227 | Value (variable) | 228 | | 229 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 231 Figure 3: TLV format 233 The Length field defines the length of the value portion in octets 234 (thus a TLV with no value portion would have a length of zero). The 235 TLV is not padded to four-octet alignment; Unrecognized types are 236 ignored. 238 3.2. Node anchors 240 The set of Node Anchor TLVs describes which Protocols Router-IDs will 241 be following to "anchor" the link described by the "Link attribute 242 TLVs". There must be at least one "like" router-ID pair per- 243 protocol. If a peer sends an illegal combination in this respect, 244 then this is handled as an NLRI error, described in [RFC4760]. 246 +------+--------------------------+--------+ 247 | Type | Description | Length | 248 +------+--------------------------+--------+ 249 | 256 | Local Autonomous System | 4 | 250 | 257 | Local IPv4 Router-ID | 4 | 251 | 258 | Local IPv6 Router-ID | 16 | 252 | 259 | Local ISO Node-ID | 7 | 253 | 260 | Remote Autonomous System | 4 | 254 | 261 | Remote IPv4 Router-ID | 4 | 255 | 262 | Remote IPv6 Router-ID | 16 | 256 | 263 | Remote ISO Node-ID | 7 | 257 +------+--------------------------+--------+ 259 Table 1: Node Anchor TLVs 261 Local IPv4 Router ID: opaque value (can be an IPv4 address or an 32 262 Bit router ID) 264 Remote IPv4 Router ID: opaque value (can be an IPv4 address or 32 265 Bit router ID) 267 Local IPv6 Router ID: opaque value (can be an IPv6 address or 128 268 Bit router ID) 270 Remote IPv6 Router ID: opaque value (can be an IPv6 address or 128 271 Bit router ID) 273 Local ISO Node ID: ISO node-ID (6 octets ISO system-ID plus PSN 274 octet) 276 Remote ISO Node ID: ISO node-ID (6 octets ISO system-ID plus PSN 277 octet) 279 It is desirable that the Router-ID assignments inside the Node anchor 280 are globally unique. However there may be router-ID spaces (e.g. 281 ISO) where not even a global registry exists, or worse, Router-IDs 282 have been allocated following private-IP RFC 1918 [RFC1918] 283 allocation. In order to disambiguate the Router-IDs the local and 284 remote Autonomous System number TLVs of the anchor nodes may be 285 included in the NLRI. The Local and Remote Autonomous System TLVs 286 are 4 octets wide as described in [RFC4893]. 2-octet AS Numbers shall 287 be expanded to 4-octet AS Numbers by zeroing the two MSB octets. 289 3.2.1. Router-ID Anchoring Example: ISO Pseudonode 291 IS-IS Pseudonodes are a good example for the variable Router-ID 292 anchoring. Consider Figure 4. This represents a Broadcast LAN 293 between a pair of routers. The "real" (=non pseudonode) routers have 294 both an IPv4 Router-ID and IS-IS Node-ID. The pseudonode does not 295 have an IPv4 Router-ID. Two unidirectional links (Node1, Pseudonode 296 1) and (Pseudonode 1, Node 2) are being generated. 298 The NRLI for (Node1, Pseudonode1) encodes local IPv4 router-ID, local 299 ISO node-ID and remote ISO node-id) 301 The NLRI for (Pseudonode1, Node2) encodes a local ISO node-ID, remote 302 IPv4 router-ID and remote ISO node-id. 304 +-----------------+ +-----------------+ +-----------------+ 305 | Node1 | | Pseudonode 1 | | Node2 | 306 |1921.6800.1001.00|--->|1921.6800.1001.02|--->|1921.6800.1002.00| 307 | 192.168.1.1 | | | | 192.168.1.2 | 308 +-----------------+ +-----------------+ +-----------------+ 310 Figure 4: IS-IS Pseudonodes 312 3.2.2. Router-ID Anchoring Example: OSPFv2 to IS-IS Migration 314 Migrating gracefully from one IGP to another requires congruent 315 operation of both routing protocols during the migration period. The 316 target protocol (IS-IS) does support more router-ID spaces than the 317 source (OSPFv2) protocol. When advertising a point-to-point link 318 between an OSPFv2-only router and an OSPFv2 and IS-IS enabled router 319 the following link information may be generated. Note that the IS-IS 320 router also supports the IPv6 traffic engineering extensions RFC 6119 321 [RFC6119] for IS-IS. 323 The NRLI does encode local IPv4 router-id, remote IPv4 router-id, 324 remote ISO node-id and remote IPv6 node-id. 326 3.3. Link Descriptors 328 The 'Link Descriptor' field is a set of Type/Length/Value (TLV) 329 triplets. The format of each TLV is shown in Figure 3. The 'Link 330 descriptor' TLVs uniquely identify a link between a pair of anchor 331 Routers. 333 The encoding of 'Link Descriptor' TLVs, i.e. the Codepoints in 334 'Type', and the 'Length' and 'Value' fields are the same as defined 335 in [RFC5305], [RFC5307], and [RFC6119] for sub-TLVs in the Extended 336 IS reachability TLV. The Codepoints are in the IANA Protocol 337 Registry for IS-IS, sub-TLV Codepoints for TLV 22, [IANA-ISIS]. 338 Although the encodings for 'Link Descriptor' TLVs were originally 339 defined for IS-IS, the TLVs can carry data sourced either by IS-IS or 340 OSPF. 342 The following link descriptor TLVs are valid in the TED NLRI: 344 +------+-------------------------------+------------------------+ 345 | Type | Description | Defined in: | 346 +------+-------------------------------+------------------------+ 347 | 4 | Link Local/Remote Identifiers | [RFC5307], Section 1.1 | 348 | 6 | IPv4 interface address | [RFC5305], Section 3.2 | 349 | 8 | IPv4 neighbor address | [RFC5305], Section 3.3 | 350 | 12 | IPv6 interface address | [RFC6119], Section 4.2 | 351 | 13 | IPv6 neighbor address | [RFC6119], Section 4.3 | 352 +------+-------------------------------+------------------------+ 354 Table 2: Link Descriptor TLVs 356 3.4. Link Attributes 358 The 'Link Attributes' field is a set of Type/Length/Value (TLV) 359 triplets. The format of each TLV is shown in Figure 3. 361 For Codepoints < 255, the encoding of 'Link Attributes' TLVs, i.e. 362 the Codepoints in 'Type', and the 'Length' and 'Value' fields are the 363 same as defined in [RFC5305], [RFC5307], and [RFC6119] for sub-TLVs 364 in the Extended IS reachability TLV. The Codepoints are in the IANA 365 Protocol Registry for IS-IS, sub-TLV Codepoints for TLV 22, 366 [IANA-ISIS]. Although the encodings for 'Link Attributes' TLVs were 367 originally defined for IS-IS, the TLVs can carry data sourced either 368 by IS-IS or OSPF. 370 For Codepoints > 255, the encoding of 'Link Attributes' TLVs is 371 described in subsequent sections. 373 The following link attribute TLVs are valid in the TED NLRI: 375 +-------+--------------------------------+------------------------+ 376 | Type | Description | Defined in: | 377 +-------+--------------------------------+------------------------+ 378 | 3 | Administrative group (color) | [RFC5305], Section 3.1 | 379 | 9 | Maximum link bandwidth | [RFC5305], Section 3.3 | 380 | 10 | Max. reservable link bandwidth | [RFC5305], Section 3.5 | 381 | 11 | Unreserved bandwidth | [RFC5305], Section 3.6 | 382 | 20 | Link Protection Type | [RFC5307], Section 1.2 | 383 | 64512 | TE Default Metric | Section 3.4.1 | 384 | 64513 | IGP Link Metric | Section 3.4.2 | 385 | 64514 | Shared Risk Link Group | Section 3.4.3 | 386 +-------+--------------------------------+------------------------+ 388 Table 3: Link Attribute TLVs 390 3.4.1. TE Default Metric TLV 392 The TE Default Metric TLV (Type 64512) carries the TE Default metric 393 for this link. This TLV corresponds to the IS-IS TE Default metric 394 sub-TLV (Type 18), defined in RFC5305, Section 3.7 [RFC5305], and the 395 OSPF TE Metric sub-TLV (Type 5), defined in RFC3630, Section 2.5.5 396 [RFC3630]. If the value in the TE Default metric TLV is derived from 397 IS-IS TE Default Metric, then the upper 8 bits of this TLV are set to 398 0. 400 0 1 2 3 401 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 402 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 403 | Type | Length | 404 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 405 | TE Default Metric | 406 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 408 Figure 5: TE Default metric TLV format 410 3.4.2. IGP Link Metric TLV 412 The IGP Metric TLV (Type 64513) carries the IGP metric for this link. 413 This attribute is only present if the IGP link metric is different 414 from the TE Default Metric (Type 18). The length of this TLV is 3. 415 If the length of the IGP link metric from which the IGP Metric value 416 is derived is less than 3 (e.g. for OSPF link metrics or non-wide 417 IS-IS metric), then the upper bits of the TLV are set to 0. 419 0 1 2 3 420 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 421 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 422 | Type | Length | 423 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 424 | IGP Link Metric | 425 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 427 Figure 6: IGP Link Metric TLV format 429 3.4.3. Shared Risk Link Group TLV 431 The Shared Risk Link Group (SRLG) TLV (Type 64514) carries the Shared 432 Risk Link Group information (see Section 2.3, "Shared Risk Link Group 433 Information", of [RFC4202]). It contains a data structure consisting 434 of a (variable) list of SRLG values, where each element in the list 435 has 4 octets, as shown in Figure 7. The length of this TLV is 4 * 436 (number of SRLG values). 438 0 1 2 3 439 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 440 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 441 | Type | Length | 442 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 443 | Shared Risk Link Group Value | 444 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 445 | ............ | 446 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 447 | Shared Risk Link Group Value | 448 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 450 Figure 7: Shared Risk Link Group TLV format 452 Note that there is no SRLG TLV in OSPF-TE. In IS-IS the SRLG 453 information is carried in two different TLVs: the IPv4 (SRLG) TLV 454 (Type 138) defined in [RFC5307], and the IPv6 SRLG TLV (Type 139) 455 defined in [RFC6119]. Since the BGP TED NLRI uses variable Router-ID 456 anchoring, both IPv4 and IPv6 SRLG information can be carried in a 457 single TLV. 459 3.5. IGP Area Information 461 IGP Area information can be carried in BGP communities. An 462 implementation should support configuration that maps IGP areas to 463 BGP communities. 465 3.6. Inter-AS Links 467 The main source of TE information is the IGP, which is not active on 468 inter-AS links. In order to inject a non-IGP enabled link into the 469 traffic-engineering database (TED) an implementation must support 470 configuration of static TE links. 472 4. Link to Path Aggregation 474 Distribution of all links available in the global Internet is 475 certainly possible, however not desirable from a scaling and privacy 476 point of view. Therefore an implementation may support link to path 477 aggregation. Rather than advertising all specific links of a domain, 478 an ASBR may advertise an "aggregate link" between a non-adjacent pair 479 of nodes. The "aggregate link" represents the aggregated set of link 480 properties between a pair of non-adjacent nodes. The actual methods 481 to compute the path properties (of bandwidth, metric) are outside the 482 scope of this document. The decision whether to advertise all 483 specific links or aggregated links is an operator's policy choice. 484 To highlight the varying levels of exposure, the following deployment 485 examples shall be discussed. 487 4.1. Example: No Link Aggregation 489 Consider Figure 8. Both AS1 and AS2 operators want to protect their 490 inter-AS {R1,R3}, {R2, R4} links using RSVP-FRR LSPs. If R1 wants to 491 compute its link-protection LSP to R3 it needs to "see" an alternate 492 path to R3. Therefore the AS2 operator exposes its topology. All 493 BGP TE enabled routers in AS1 "see" the full topology of AS and 494 therefore can compute a backup path. Note that the decision if the 495 direct link between {R3, R4} or the {R4, R5, R3) path is used is made 496 by the computing router. 498 AS1 : AS2 499 : 500 R1-------R3 501 | : | \ 502 | : | R5 503 | : | / 504 R2-------R4 505 : 506 : 508 Figure 8: no-link-aggregation 510 4.2. Example: ASBR to ASBR Path Aggregation 512 The brief difference between the "no-link aggregation" example and 513 this example is that no specific link gets exposed. Consider 514 Figure 9. The only link which gets advertised by AS2 is an 515 "aggregate" link between R3 and R4. This is enough to tell AS1 that 516 there is a backup path. However the actual links being used are 517 hidden from the topology. 519 AS1 : AS2 520 : 521 R1-------R3 522 | : | 523 | : | 524 | : | 525 R2-------R4 526 : 527 : 529 Figure 9: asbr-link-aggregation 531 4.3. Example: Multi-AS Path Aggregation 533 Service providers in control of multiple-ASes may even decide to not 534 expose their internal inter-AS links. Consider Figure 10. Rather 535 than exposing all specific R3 to R6 links, AS3 is modeled as a single 536 node which connects to the border routers of the aggregated domain. 538 AS1 : AS2 : AS3 539 : : 540 R1-------R3----- 541 | : : \ 542 | : : vR0 543 | : : / 544 R2-------R4----- 545 : : 546 : : 548 Figure 10: multi-as-aggregation 550 5. Originating the TED NLRI 552 A BGP Speaker must be configured to originate TED NLRIs. Usually 553 export of the TED database into BGP is enabled on ASBRs and ABRs. 555 The BGP Speaker shall throttle the rate of TED NLRI updates. An 556 implementation shall provide a configuration attribute for the 557 interval between updates. The minimum interval between updates is 30 558 seconds. 560 6. Receiving the TED NLRI 562 This section describes the processing of TED NLRIs at the receiving 563 BGP Speaker. 565 TE attributes for a link received from an IGP have higher priority 566 than TED NLRIs received via BGP. Multiple BGP Speakers may advertise 567 the same TED NLRI; the receiving BGP Speaker can individually choose 568 the source BGP Speaker for each NLRI. 570 The AS_PATH attribute is used both for loop detection and for NLRI 571 selection: the TED NLRI with shorter AS_PATH length is preferred. 572 The Community and Extended Community path attributes are stored in 573 the RIB and may be used in operator-defined policies. Communities 574 can also be used to encode the IGP Area information. All other path 575 attributes are ignored. 577 7. Use Cases 579 7.1. MPLS TE 581 If a router wants to compute a MPLS TE path across IGP areas TED 582 lacks visibility of the complete topology. This is an issue for 583 large scale networks that need to segment their core networks into 584 distinct areas because inter-area TE cannot get deployed there. 585 Current solutions for inter area TE only compute the path for the 586 first area. The router only has full topological visibility for the 587 first area along the path, but not for subsequent areas. The best 588 practice is to use a technique called "loose-hop-expansion" which 589 uses the IGP computed shortest path topology for the remainder of the 590 path. Therefore no non-SPF based path setup is possible across 591 areas. This has disadvantages for path protection and path 592 engineering applications, as shown in Figure 11. 594 ............................... ................................... 595 : Area 51 : : Area 0 : 596 : +--------+ +--------+ +--------+ : 597 : ************************************************************ : 598 : * +-------| R1 |-----| ABR1 |-----| R3 |-------+ * : 599 : * | ######## | | # | | | | * : 600 : * | # +--------+ +----|-#-+ +--------+ | * : 601 : +-*-|-#-+ : :| # +---|-*-+ : 602 : | * # | : :| # | * | : 603 : | S # | : :| # | D | : 604 : | # | : :| # | | : 605 : +---|-#-+ : :| # +---|---+ : 606 : | # +--------+ +----|-#-+ +--------+ | : 607 : | ############################# | | | | : 608 : +-------| R2 |-----| ABR2 |-----| R4 |-------+ : 609 : | | | | | | : 610 : +--------+ +--------+ +--------+ : 611 : : : : 612 :.............................: :.................................: 614 ...... 615 **** Primary LSP : : Area Boundary 616 #### Bypass LSP :....: 618 Figure 11: MPLS TE Bypass LSP problem 620 Router S sets up an RSVP LSP from S to D. Although it has only 621 visibility into Area 51, the LSP setup ultimately succeeds, as 622 shortest path first routing from ABR1 onwards routes the RSVP message 623 towards destination D. What does not work is to setup a Link 624 Protection bypass LSP protection for the R1 to ABR1 link as shown in 625 the figure. The problem is that the TE database at Router R1 does 626 not have path visibility of the link between ABR1 and ABR2, such that 627 it can compute the Link Bypass LSP. 629 7.2. ALTO Server Network API 631 An ALTO Server is an entity that generates an abstracted network 632 topology and provides it to network-aware applications over a web 633 service based API. Example applications are p2p clients or trackers, 634 or CDNs. The abstracted network topology comes in the form of two 635 maps: the network map that specifies allocation of prefixes to PIDs, 636 and the cost map that specifies the cost between the PIDs. For more 637 details, see [I-D.ietf-alto-protocol]. 639 ALTO abstract network topologies can be auto-generated from the 640 physical topology of the underlying network. The generation would 641 typically be based on policies and rules set by the operator. Both 642 prefix and TE data are required: prefix data is required to generate 643 the network maps, TE (topology) data is required to generate the cost 644 maps. Prefix data is carried and originated in BGP, TE data is 645 originated and carried in an IGP. Without BGP TE NLRI the ALTO 646 Server would have to peer with both BGP Speakers and IGP in multiple 647 areas and/or ASes to obtain all the necessary network topology data. 648 The BGP TE NLRI allows for a single interface between the network and 649 the ALTO Server. 651 7.3. Path Computation Element (PCE) TED Synchronization Protocol 653 RFC4655, Section 5.2, Figure 2 [RFC4655] describes a Path Computation 654 Element (PCE) which synchronizes its traffic engineering database 655 (TED) by use of a routing protocol. This memo describes the first 656 standardized protocol for PCE to learn about inter-AS or inter-area 657 TE information. 659 8. IANA Considerations 661 This document requests a code point from the registry of Address 662 Family Numbers 664 This document requests creation of a new registry for node anchor, 665 link descriptor and link attribute TLVs. The range of Codepoints in 666 the registry is 0-65535. Values 0-255 will shadow Codepoints of the 667 IANA Protocol Registry for IS-IS, sub-TLV Codepoints for TLV 22. 668 Values 256-65535 will be used for Codepoints that are specific to the 669 BGP TE NLRI. The registry will be initialized as shown in Table 2 670 and Table 3. Allocations within the registry will require 671 documentation of the proposed use of the allocated value and approval 672 by the Designated Expert assigned by the IESG (see [RFC5226]). 674 Note to RFC Editor: this section may be removed on publication as an 675 RFC. 677 9. Security Considerations 679 This draft does not affect the BGP security model. 681 10. Acknowledgements 683 We would like to thank Alia Atlas, David Ward, John Scudder, Kaliraj 684 Vairavakkalai, Nischal Sheth and Yakov Rekhter from Juniper Networks, 685 Inc. and Richard Woundy from Comcast for their invaluable input and 686 comments. 688 11. References 690 11.1. Normative References 692 [IANA-ISIS] 693 "IS-IS TLV Codepoint, Sub-TLVs for TLV 22", . 697 [RFC1918] Rekhter, Y., Moskowitz, R., Karrenberg, D., Groot, G., and 698 E. Lear, "Address Allocation for Private Internets", 699 BCP 5, RFC 1918, February 1996. 701 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 702 Requirement Levels", BCP 14, RFC 2119, March 1997. 704 [RFC3630] Katz, D., Kompella, K., and D. Yeung, "Traffic Engineering 705 (TE) Extensions to OSPF Version 2", RFC 3630, 706 September 2003. 708 [RFC4202] Kompella, K. and Y. Rekhter, "Routing Extensions in 709 Support of Generalized Multi-Protocol Label Switching 710 (GMPLS)", RFC 4202, October 2005. 712 [RFC4271] Rekhter, Y., Li, T., and S. Hares, "A Border Gateway 713 Protocol 4 (BGP-4)", RFC 4271, January 2006. 715 [RFC4655] Farrel, A., Vasseur, J., and J. Ash, "A Path Computation 716 Element (PCE)-Based Architecture", RFC 4655, August 2006. 718 [RFC4760] Bates, T., Chandra, R., Katz, D., and Y. Rekhter, 719 "Multiprotocol Extensions for BGP-4", RFC 4760, 720 January 2007. 722 [RFC4893] Vohra, Q. and E. Chen, "BGP Support for Four-octet AS 723 Number Space", RFC 4893, May 2007. 725 [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an 726 IANA Considerations Section in RFCs", BCP 26, RFC 5226, 727 May 2008. 729 [RFC5305] Li, T. and H. Smit, "IS-IS Extensions for Traffic 730 Engineering", RFC 5305, October 2008. 732 [RFC5307] Kompella, K. and Y. Rekhter, "IS-IS Extensions in Support 733 of Generalized Multi-Protocol Label Switching (GMPLS)", 734 RFC 5307, October 2008. 736 [RFC6119] Harrison, J., Berger, J., and M. Bartlett, "IPv6 Traffic 737 Engineering in IS-IS", RFC 6119, February 2011. 739 11.2. Informative References 741 [I-D.ietf-alto-protocol] 742 Alimi, R., Penno, R., and Y. Yang, "ALTO Protocol", 743 draft-ietf-alto-protocol-06 (work in progress), 744 October 2010. 746 Authors' Addresses 748 Hannes Gredler 749 Juniper Networks, Inc. 750 1194 N. Mathilda Ave. 751 Sunnyvale, CA 94089 752 US 754 Email: hannes@juniper.net 756 Jan Medved 757 Juniper Networks, Inc. 758 1194 N. Mathilda Ave. 759 Sunnyvale, CA 94089 760 US 762 Email: jmedved@juniper.net