idnits 2.17.1 draft-ietf-ccamp-inter-domain-framework-02.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** It looks like you're using RFC 3978 boilerplate. You should update this to the boilerplate described in the IETF Trust License Policy document (see https://trustee.ietf.org/license-info), which is required now. -- Found old boilerplate from RFC 3978, Section 5.1 on line 22. -- Found old boilerplate from RFC 3978, Section 5.5 on line 918. -- Found old boilerplate from RFC 3979, Section 5, paragraph 1 on line 783. -- Found old boilerplate from RFC 3979, Section 5, paragraph 2 on line 790. -- Found old boilerplate from RFC 3979, Section 5, paragraph 3 on line 796. ** Found boilerplate matching RFC 3978, Section 5.4, paragraph 1 (on line 910), which is fine, but *also* found old RFC 2026, Section 10.4C, paragraph 1 text on line 42. ** This document has an original RFC 3978 Section 5.4 Copyright Line, instead of the newer IETF Trust Copyright according to RFC 4748. ** This document has an original RFC 3978 Section 5.5 Disclaimer, instead of the newer disclaimer which includes the IETF Trust according to RFC 4748. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- ** Missing document type: Expected "INTERNET-DRAFT" in the upper left hand corner of the first page == No 'Intended status' indicated for this document; assuming Proposed Standard Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the RFC 3978 Section 5.4 Copyright Line does not match the current year -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (May 2005) is 6913 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) == Missing Reference: 'LMP' is mentioned on line 123, but not defined == Unused Reference: 'RFC3667' is defined on line 811, but no explicit reference was found in the text == Unused Reference: 'RFC3668' is defined on line 814, but no explicit reference was found in the text == Unused Reference: 'OVERLAY' is defined on line 869, but no explicit reference was found in the text ** Obsolete normative reference: RFC 3667 (Obsoleted by RFC 3978) ** Obsolete normative reference: RFC 3668 (Obsoleted by RFC 3979) -- Obsolete informational reference (is this intentional?): RFC 3784 (Obsoleted by RFC 5305) -- No information found for draft-otani-ccamp-interas-GMPLS-TE - is the name correct? Summary: 7 errors (**), 0 flaws (~~), 6 warnings (==), 9 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group Adrian Farrel 3 IETF Internet Draft Olddog Consulting 4 Proposed Status: Informational 5 Expires: November 2005 Jean-Philippe Vasseur 6 Cisco Systems, Inc. 8 Arthi Ayyangar 9 Juniper Networks 11 May 2005 13 A Framework for Inter-Domain MPLS Traffic Engineering 15 draft-ietf-ccamp-inter-domain-framework-02.txt 17 Status of this Memo 19 By submitting this Internet-Draft, each author represents that any 20 applicable patent or other IPR claims of which he or she is aware 21 have been or will be disclosed, and any of which he or she becomes 22 aware will be disclosed, in accordance with Section 6 of BCP 79. 24 Internet-Drafts are working documents of the Internet Engineering 25 Task Force (IETF), its areas, and its working groups. Note that 26 other groups may also distribute working documents as 27 Internet-Drafts. 29 Internet-Drafts are draft documents valid for a maximum of six months 30 and may be updated, replaced, or obsoleted by other documents at any 31 time. It is inappropriate to use Internet-Drafts as reference 32 material or to cite them other than as "work in progress." 34 The list of current Internet-Drafts can be accessed at 35 http://www.ietf.org/ietf/1id-abstracts.txt. 37 The list of Internet-Draft Shadow Directories can be accessed at 38 http://www.ietf.org/shadow.html. 40 Copyright Notice 42 Copyright (C) The Internet Society (2005). All Rights Reserved. 44 Abstract 46 This document provides a framework for establishing and controlling 47 Multiprotocol Label Switching (MPLS) and Generalized MPLS (GMPLS) 48 Label Switched Paths (LSPs) in multi-domain networks. 50 For the purposes of this document, a domain is considered to be any 51 collection of network elements within a common sphere of address 52 management or path computational responsibility. Examples of such 53 domains include IGP areas and Autonomous Systems. 55 Contents 57 1. Introduction ............................................... 3 58 1.1. Nested Domains ......................................... 3 59 1.2. Conventions used in this document ...................... 4 60 2. Signaling Options .......................................... 4 61 2.1. LSP Nesting ............................................ 4 62 2.2. Contiguous LSP ......................................... 5 63 2.3. LSP Stitching .......................................... 5 64 2.4. Hybrid Methods ......................................... 5 65 2.5. Control of Downstream Choice of Signaling Method ....... 6 66 3. Path Computation Techniques ................................ 6 67 3.1. Management Configuration ............................... 6 68 3.2. Head End Computation ................................... 6 69 3.2.1. Multi-Domain Visibility Computation ................ 7 70 3.2.2. Partial Visibility Computation ..................... 7 71 3.2.3. Local Domain Visibility Computation ................ 7 72 3.3. Domain Boundary Computation ............................ 8 73 3.4. Path Computation Element ............................... 8 74 3.4.1. Multi-Domain Visibility Computation ................ 9 75 3.4.2. Path Computation Use of PCE When Preserving 76 Confidentiality .................................... 9 77 3.4.3. Per-Domain Computation Servers ..................... 9 78 3.5. Optimal Path Computation ............................... 9 79 4. Distributing Reachability and TE Information .............. 10 80 5. Comments on Advanced Functions ............................ 11 81 5.1. LSP Re-Optimization ................................... 11 82 5.2. LSP Setup Failure ..................................... 12 83 5.3. LSP Repair ............................................ 12 84 5.4. Fast Reroute .......................................... 12 85 5.5. Comments on Path Diversity ............................ 13 86 5.6. Domain-Specific Constraints ........................... 14 87 5.7. Policy Control ........................................ 14 88 5.8. Inter-domain OAM ...................................... 15 89 5.9. Point-to-Multipoint ................................... 15 90 5.10. Applicability to Non-Packet Technologies ............. 15 91 6. Security Considerations ................................... 15 92 7. IANA Considerations ....................................... 16 93 8. Acknowledgements .......................................... 16 94 9. Intellectual Property Considerations ...................... 16 95 10. Normative References ..................................... 16 96 11. Informational References ................................. 17 97 12. Authors' Addresses ....................................... 18 98 13. Full Copyright Statement ................................. 19 100 1. Introduction 102 The Traffic Engineering Working Group has developed requirements for 103 inter-area and inter-AS MPLS Traffic Engineering in [INTER-AREA] and 104 [INTER-AS]. 106 Various proposals have subsequently been made to address some or all 107 of these requirements through extensions to the RSVP-TE and IGP 108 (ISIS, OSPF) protocols and procedures. 110 This document introduces the techniques for establishing TE LSPs 111 across multiple domains. The functional components of these 112 techniques are separated into the mechanisms for discovering 113 reachability and TE information, for computing the paths of LSPs, and 114 for signaling the LSPs. Note that the aim of this document is not to 115 detail each of those techniques which are covered in separate 116 documents, but rather to propose a framework for inter-domain MPLS 117 Traffic Engineering. 119 Note that in the remainder of this document, the term 'MPLS Traffic 120 Engineering' is used equally to apply to MPLS and GMPLS traffic. 121 Specific issues pertaining to the use of GMPLS in inter-domain 122 environments (for example, policy implications of the use of the Link 123 Management Protocol [LMP] on inter-domain links) these are covered in 124 a separate document. 126 For the purposes of this document, a domain is considered to be any 127 collection of network elements within a common sphere of address 128 management or path computational responsibility. Examples of such 129 domains include IGP areas and Autonomous Systems. However, domains of 130 computational responsibility may also exist as sub-domains of areas 131 or ASs. Wholly or partially overlapping domains are not within the 132 scope of this document. 134 1.1. Nested Domains 136 Nested domains are outside the scope of this document. It may be that 137 some domains that are nested administratively or for the purposes of 138 address space management can be considered as adjacent domains for 139 the purposes of this document, however the fact that the domains are 140 nested is then immaterial. 142 In the context of MPLS TE, domain A is considered to be nested within 143 domain B if domain A is wholly contained in Domain B, and domain B is 144 fully or partially aware of the TE characteristics and topology of 145 domain A. 147 For further consideration of nested domains see [MRN] 149 1.2. Conventions used in this document 151 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 152 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 153 document are to be interpreted as described in RFC 2119 [RFC2119]. 155 2. Signaling Options 157 Three distinct options for signaling TE LSPs across multiple domains 158 are identified. The choice of which options to use may be influenced 159 by the path computation technique used (see section 3), although some 160 path computation techniques may apply to multiple TE LSP types. The 161 choice may further depend on the application to which the TE LSPs are 162 put and the nature, topology and switching capabilities of the 163 network. 165 A comparison of the usages of the different signaling options is 166 beyond the scope of this document and should be the subject of a 167 separate applicability statement. 169 2.1. LSP Nesting 171 Forwarding Adjacencies (FAs) are introduced and explained in detail 172 in [HIER]. No further description is necessary in this document. 174 FAs can be used in support of inter-domain TE LSPs. In particular, an 175 FA may be used to achieve connectivity between any pair of LSRs 176 within a domain. The ingress and egress of the FA LSP could be the 177 edge nodes of the domain in which case connectivity is achieved 178 across the entire domain, or they could be any other pair of LSRs in 179 the domain. 181 The technique of carrying one TE LSP within another is termed LSP 182 nesting. An FA may provide a TE LSP tunnel to transport (i.e. nest) 183 multiple TE LSPs along a common part of their paths. Alternatively, a 184 TE LSP may carry (i.e. nest) a single LSP in a one-to-one mapping. 186 The signaling trigger for the establishment of an FA LSP may be the 187 receipt of a signaling request for the TE LSP that it will carry, or 188 may be a management action to 'pre-engineer' a domain to be crossed 189 by TE LSPs that would be used as FAs by the traffic that has to 190 traverse the domain. Furthermore, the mapping (inheritance rules) 191 between attributes of the nested and FA LSPs (including bandwidth) 192 may be statically pre-configured or, for on-demand FA LSPs, may be 193 dynamic according to the properties of the nested LSPs. 195 Note that a hierarchical LSP may be constructed to span multiple 196 domains or parts of domains. However, how or whether such an LSP 197 could be advertised as an FA that spans domains is open for study. 198 The end points of a hierarchical LSP are not necessarily on domain 199 boundaries, so nesting is not limited to domain boundaries. 201 Note also that the IGP/EGP routing topology is maintained unaffected 202 by the LSP connectivity and TE links introduced by FA LSPs. That is, 203 the routing protocols do not exchange messages over the FA LSPs, and 204 such LSPs do not create routing adjacencies between routers. 206 During the operation of establishing a nested LSP that uses a 207 hierarchical LSP, the SENDER_TEMPLATE and SESSION objects remain 208 unchanged along the entire length of the nested LSP. 210 2.2. Contiguous LSP 212 A single contiguous LSP is established from ingress to egress in a 213 single signaling exchange. No further LSPs are required to be 214 established to support this LSP. Signaling occurs between adjacent 215 neighbors only (no tunneling), and hop-by-hop signaling is used. 217 2.3. LSP Stitching 219 LSP Stitching is described in [STITCH]. 221 In the LSP stitching model separate LSPs (referred to as a TE LSP 222 segments) are established and are "stitched" together in the data 223 plane so that a single end-to-end label switched path is achieved. 224 The distinction is that the component LSP segments are signaled as 225 distinct TE LSPs in the control plane. Each signaled TE LSP segment 226 has a different source and destination. 228 LSP stitching can be used in support of inter-domain TE LSPs. In 229 particular, an LSP segment may be used to achieve connectivity 230 between any pair of LSRs within a domain. The ingress and egress of 231 the LSP segment could be the edge nodes of the domain in which case 232 connectivity is achieved across the entire domain, or they could be 233 any other pair of LSRs in the domain. 235 The signaling trigger for the establishment of a TE LSP segment may 236 be the establishment of the previous TE LSP segment, the receipt of 237 setup request for TE LSP that it plans to stitch to a local TE LSP 238 segment, or may be a management action. 240 LSP segments may be managed as FAs and advertised as TE links. 242 2.4. Hybrid Methods 244 There is nothing to prevent the mixture of signaling methods 245 described above when establishing a single, end-to-end, inter-domain 246 TE LSP. It may be desirable in this case for the choice of the 247 various methods to be indicated along the path perhaps through the 248 RRO. 250 If there is a desire to restrict which methods are used, this MUST be 251 signaled as described in the next section. 253 2.5. Control of Downstream Choice of Signaling Method 255 Notwithstanding the previous section, an ingress LSR MAY wish to 256 restrict the signaling methods applied to a particular LSP at domain 257 boundaries across the network. Such control, where it is required, 258 may be achieved by the definition of appropriate new flags in the 259 SESSION-ATTRIBUTE object or the Attributes Flags TLV of the 260 LSP_ATTRIBUTES object [ATTRIB]. Before defining a mechanism to 261 provide this level of control, the functional requirement to control 262 the way in which the network delivers a service must be established 263 and due consideration must be given to the impact on 264 interoperability. 266 3. Path Computation Techniques 268 The discussion of path computation techniques within this document is 269 limited significantly to the determination of where computation may 270 take place and what components of the full path may be determined. 272 The techniques used are closely tied to the signaling methodologies 273 described in the previous section in that certain computation 274 techniques may require the use of particular signaling approaches and 275 vice versa. 277 Any discussion of the appropriateness of a particular path 278 computation technique in any given circumstance is beyond the scope 279 of this document and should be described in a separate applicability 280 statement. 282 Path computation algorithms are firmly out of scope of this document. 284 3.1. Management Configuration 286 Path computation may be performed by offline tools or by a network 287 planner. The resultant path may be supplied to the ingress LSR as 288 part of the TE LSP or service request, and encoded by the ingress LSR 289 as an ERO on the Path message that is sent out. 291 There is no reason why the path provided by the operator should not 292 span multiple domains if the relevant information is available to the 293 planner or the offline tool. The definition of what information is 294 needed to perform this operation and how that information is 295 gathered, is outside the scope of this document. 297 3.2. Head End Computation 299 The head end, or ingress, LSR may assume responsibility for path 300 computation when the operator supplies part or none of the explicit 301 path. The operator MUST, in any case, supply at least the destination 302 address (egress) of the LSP. 304 3.2.1. Multi-Domain Visibility Computation 306 If the ingress has sufficient visibility of the topology and TE 307 information for all of the domains across which it will route the LSP 308 to its destination then it may compute and provide the entire path. 309 The quality of this path (that is, its optimality as discussed in 310 section 3.5) is best if the ingress has full visibility into all 311 relevant domains rather than just sufficient visibility to provide 312 some path to the destination. 314 Extreme caution must be exercised in consideration of the 315 distribution of the requisite TE information. See section 4. 317 3.2.2. Partial Visibility Computation 319 It may be that the ingress does not have full visibility of the 320 topology of all domains, but does have information about the 321 connectedness of the domains and the TE resource availability across 322 the domains. In this case, the ingress is not able to provide a fully 323 specified strict explicit path from ingress to egress. However, the 324 ingress can supply an explicit path that comprises: 325 - explicit hops from ingress to the local domain boundary 326 - loose hops representing the domain entry points across the network 327 - a loose hop identifying the egress. 329 Alternatively, the explicit path may be expressed as: 330 - explicit hops from ingress to the local domain boundary 331 - strict hops giving abstract nodes representing each domain in turn 332 - a loose hop identifying the egress. 334 These two explicit path formats may be mixed. 336 This form of explicit path relies on some further computation 337 technique being applied at the domain boundaries. See section 3.3. 339 As with the multi-domain visibility option, extreme caution must be 340 exercised in consideration of the distribution of the requisite TE 341 information. See section 4. 343 3.2.3. Local Domain Visibility Computation 345 A final possibility for ingress-based computation is that the ingress 346 LSR has visibility only within its own domain, and connectivity 347 information only as far as determining one or more domain exit points 348 that may be suitable for carrying the LSP to its egress. 350 In this case the ingress builds an explicit path that comprises just: 351 - explicit hops from ingress to the local domain boundary 352 - a loose hop identifying the egress. 354 3.3. Domain Boundary Computation 356 If the partial explicit path methods described in sections 3.2.2 or 357 3.2.3 are applied then the LSR at each domain boundary is responsible 358 for ensuring that there is sufficient path information added to the 359 Path message to carry it at least to the next domain boundary (that 360 is, out of the new domain). 362 If the LSR at the domain boundary has full visibility to the egress 363 then it can supply the entire explicit path. Note however, that the 364 ERO processing rules of [RFC3209] state that it SHOULD only update 365 the ERO as far as the next specified hop (that is, the next domain 366 boundary if one was supplied in the original ERO) and, of course, 367 MUST NOT insert ERO subobjects immediately before a strict hop. 369 If the LSR at the domain boundary has only partial visibility (using 370 the definitions of section 3.2.2) it will fill in the path as far as 371 the next domain boundary, and will supply further domain/domain 372 boundary information if not already present in the ERO. 374 If the LSR at the domain boundary has only local visibility into the 375 immediate domain it will simply add information to the ERO to carry 376 the Path message as far as the next domain boundary. 378 3.4. Path Computation Element 380 The computation techniques in sections 3.2 and 3.3 rely on topology 381 and TE information being distributed to the ingress LSR and those 382 LSRs at domain boundaries. These LSRs are responsible for computing 383 paths. Note that there may be scaling concerns with distributing the 384 required information - see section 4. 386 An alternative technique places the responsibility for path 387 computation with a Path Computation Element (PCE) [PCE]. There may be 388 either a centralized PCE, or multiple PCEs (each having local 389 visibility and collaborating in a distributed fashion to compute an 390 end-to-end path) across the entire network and even within any one 391 domain. The PCE may collect topology and TE information from the same 392 sources as would be used by LSRs in the previous paragraph, or though 393 other means. 395 Each LSR called upon to perform path computation (and even the 396 offline management tools described in section 3.1) may abdicate the 397 task to a PCE of its choice. The selection of PCE(s) may be driven by 398 static configuration or the dynamic discovery by means of IGP or BGP 399 extensions. 401 3.4.1. Multi-Domain Visibility Computation 403 A PCE may have full visibility, perhaps through connectivity to 404 multiple domains. In this case it is able to supply a full explicit 405 path as in section 3.2.1. 407 3.4.2. Path Computation Use of PCE When Preserving Confidentiality 409 Note that although a centralized PCE or multiple collaborative PCEs 410 may have full visibility into one or more domains, it may be 411 desirable (e.g to preserve confidentiality) that the full path is not 412 provided to the ingress LSR. Instead, a partial path is supplied (as 413 in section 3.2.2 or 3.2.3) and the LSRs at each domain boundary are 414 required to make further requests for each successive segment of the 415 path. 417 In this way an end-to-end path may be computed using the full network 418 capabilities, but confidentiality between domains may be preserved. 419 Optionally, the PCE(s) may compute the entire path at the first 420 request and hold it in storage for subsequent requests, or it may 421 recompute the best path on each request or at regular intervals. 423 It may be the case that the centralized PCE or the collaboration 424 between PCEs may define a trust relationship greater than that 425 normally operational between domains. 427 3.4.3. Per-Domain Computation Elements 429 A third way that PCEs may be used is simply to have one (or more) per 430 domain. Each LSR within a domain that wishes to derive a path across 431 the domain may consult its local PCE. 433 This mechanism could be used for all path computations within the 434 domain, or specifically limited to computations for LSPs that will 435 leave the domain where external connectivity information can then be 436 restricted to just the PCE. 438 3.5. Optimal Path Computation 440 An optimal route might be defined as the route that would be computed 441 in the absence of domain boundaries. Another constraint to the 442 definition of 'optimal' might be to reduce or limit the number of 443 domains crossed by the LSP. It is easy to construct examples 444 that show that partitioning a network into domains, and the resulting 445 loss or aggregation of routing information may lead to the 446 computation of routes that are other than optimal. It is impossible 447 to guarantee optimal routing in the presence of aggregation / 448 abstraction / summarization of routing information. 450 It is beyond the scope of this document to define what is an optimum 451 path for an inter-domain TE LSP. This debate is abdicated in favor of 452 requirements documents and applicability statements. Note, however, 453 that the meaning of certain computation metrics may differ between 454 domains (see section 5.6). 456 4. Distributing Reachability and TE Information 458 Traffic Engineering information is collected into a TE Database (TED) 459 on which path computation algorithms operate either directly or by 460 first constructing a network graph. 462 The path computation techniques described in the previous section 463 make certain demands upon the distribution of reachability 464 information and the TE capabilities of nodes and links within domains 465 as well as the TE connectivity across domains. 467 Currently, TE information is distributed within domains by additions 468 to IGPs [RFC3630], [RFC3784]. 470 In cases where two domains are interconnected by one or more links 471 (that is, the domain boundary falls on a link rather than on a node), 472 there SHOULD be a mechanism to distribute the TE information 473 associated with the inter-domain links to the corresponding domains. 474 This would facilitate better path computation and reduce TE-related 475 crankbacks on these links. 477 Where a domain is a subset of an IGP area, filtering of TE 478 information may be applied at the domain boundary. This filtering may 479 be one way, or two way. 481 Where information needs to reach a PCE that spans multiple domains, 482 the PCE may snoop on the IGP traffic in each domain, or play an 483 active part as an IGP-capable node in each domain. The PCE might also 484 receive TED updates from a proxy within the domain. 486 It could be possible that an LSR that performs path computation (for 487 example, an ingress LSR) obtains the topology and TE information of 488 not just its own domain, but other domains as well. This information 489 may be subject to filtering applied by the advertising domain (for 490 example, the information may be limited to FAs across other domains, 491 or the information may be aggregated or abstracted). 493 Where any cross-domain reachability and TE information needs to be 494 advertised, consideration must be given to TE extensions to BGP, and 495 how these may be fed to the IGPs. Techniques for inter-domain TE 496 aggregation are also for further study. However, it must be noted 497 that any extensions that cause a significant increase in the amount 498 of processing (such as aggregation computation) at domain boundaries, 499 or a significant increase in the amount of information flooded (such 500 as detailed TE information) need to be treated with extreme caution 501 and compared carefully with the scaling requirements expressed in 502 [INTER-AREA] and [INTER-AS]. 504 5. Comments on Advanced Functions 506 This section provides some non-definitive comments on the constraints 507 placed on advanced MPLS TE functions by inter-domain MPLS. It does 508 not attempt to state the implications of using one inter-domain 509 technique or another. Such material is deferred to appropriate 510 applicability statements where statements about the capabilities of 511 existing or future signaling, routing and computation techniques to 512 deliver the functions listed should be made. 514 5.1. LSP Re-Optimization 516 Re-optimization is the process of moving a TE LSP from one path to 517 another, more preferable path (where no attempt is made in this 518 document to define 'preferable' as no attempt was made to define 519 'optimal'). Make-before-break techniques are usually applied to 520 ensure that traffic is disrupted as little as possible. The Shared 521 Explicit style is usually used to avoid double booking of network 522 resources. 524 Re-optimization may be available within a single domain. 525 Alternatively, re-optimization may involve a change in route across 526 several domains or might involve a choice of different transit 527 domains. 529 Re-optimization requires that all or part of the path of the LSP be 530 re-computed. The techniques used may be selected as described in 531 section 3, and this will influence whether the whole or part of the 532 path is re-optimized. 534 The trigger for path computation and re-optimization may be an 535 operator request, a timer, information about a change in 536 availability of network resources, or a change in operational 537 parameters (for example bandwidth) of an LSP. This trigger MUST be 538 applied to the point in the network that requests re-computation and 539 controls re-optimization and may require additional signaling. 541 Note also that where multiple diverse paths are applied end-to-end 542 (i.e. not simply within protection domains - see section 5.5) the 543 point of calculation for re-optimization (whether it is PCE, ingress, 544 or domain entry point) needs to know all such paths before attempting 545 re-optimization of any one path. 547 It may be the case that re-optimization is best achieved by 548 recomputing the paths of multiple LSPs at once. Indeed, this can be 549 shown to be most efficient when the paths of all LSPs are known, not 550 simply those LSPs that originate at a particular ingress. While this 551 problem is inherited from single domain re-optimization and is out of 552 scope within this document, it should be noted that the problem grows 553 in complexity when LSPs wholly within one domain affect the 554 re-optimization path calculations performed in another domain. 556 5.2. LSP Setup Failure 558 When an inter-domain LSP setup fails in some domain other than the 559 first, various options are available for reporting and retrying the 560 LSP. 562 In the first instance, a retry may be attempted within the domain 563 that contains the failure. That retry may be attempted by nodes 564 wholly within the domain, or the failure may be referred back to the 565 LSR at the domain boundary. 567 If the failure cannot be bypassed within the domain where the failure 568 occurred (perhaps there is no suitable alternate route, perhaps 569 rerouting is not allowed by domain policy, or perhaps the Path 570 message specifically bans such action), the error MUST be reported 571 back to the previous or head-end domain. 573 Subsequent repair attempts may be made by domains further upstream, 574 but will only be properly effective if sufficient information about 575 the failure and other failed repair attempts is also passed back 576 upstream [CRANKBACK]. Note that there is a tension between this 577 requirement and that of confidentiality although crankback 578 aggregation may be applicable at domain boundaries. 580 Further attempts to signal the failed LSP may apply the information 581 about the failures as constraints to path computation, or may signal 582 them as specific path exclusions [EXCLUDE]. 584 When requested by signaling, the failure may also be systematically 585 reported to the head-end LSR. 587 5.3. LSP Repair 589 An LSP that fails after it has been established may be repaired 590 dynamically by re-routing. The behavior in this case is either like 591 that for re-optimization, or for handling setup failures (see 592 previous two sections). 594 Fast Reroute may also be used (see below). 596 5.4. Fast Reroute 598 MPLS Traffic Engineering Fast Reroute ([FRR]) defines local 599 protection schemes intended to provide fast recovery (in 10s of 600 msecs) of fast-reroutable TE LSPs upon link/SRLG/Node failure. A 601 backup TE LSP is configured and signaled at each hop, and activated 602 upon detecting or being informed of a network element failure. The 603 node immediately upstream of the failure (called the PLR - Point of 604 Local Repair) reroutes the set of protected TE LSPs onto the 605 appropriate backup tunnel(s) and around the failed resource. 607 In the context of inter-domain TE, there are several different 608 failure scenarios that must be analyzed. Provision of suitable 609 solutions may be further complicated by the fact that [FRR] specifies 610 two distinct modes of operation referred to as the "one to one mode" 611 and the "facility back-up mode". 613 The failure scenarios specific to inter-domain TE are as follows: 615 - Failure of a domain edge node that is present in both domains. 616 There are two sub-cases: 618 - The PLR and the MP are in the same domain 620 - The PLR and the MP are in different domains. 622 - Failure of a domain edge node that is only present in one of the 623 domains. 625 - Failure of an inter-domain link. 627 The techniques that must be employed to use Fast Reroute for the 628 different methods of signaling LSPs identified in section 2 differ 629 considerably. These should be explained further in applicability 630 statements of, in the case, of a change in base behavior, in 631 implementation guidelines specific to the signaling techniques. 633 Note that after local repair has been performed, it may be desirable 634 to re-optimize the LSP (see section 5.1). If the point of 635 re-optimization (for example the ingress LSR) lies in a different 636 domain to the failure, it may rely on the delivery of a PathErr or 637 Notify message to inform it of the local repair event. 639 It is important to note that Fast Reroute techniques are only 640 applicable to packet switching networks because other network 641 technologies cannot apply label stacking within the same switching 642 type. Segment protection [SEG-PROT] provides a suitable alternative 643 that is applicable to packet and non-packet networks. 645 5.5. Comments on Path Diversity 647 Diverse paths may be required in support of load sharing and/or 648 protection. Such diverse paths may be required to be node diverse, 649 link diverse, fully path diverse (that is, link and node diverse), or 650 SRLG diverse. 652 Diverse path computation is a classic problem familiar to all graph 653 theory majors. The problem is compounded when there are areas of 654 'private knowledge' such as when domains do not share topology 655 information. The problem is generally considered to be easier and 656 more efficient when the diverse paths can be computed 657 'simultaneously' on the fullest set of information. That being said, 658 various techniques (out of the scope of this document) exist to 659 ensure end-to-end path diversity across multiple domains. 661 Many network technologies utilize 'protection domains' because they 662 fit well with the capabilities of the technology. As a result, many 663 domains are operated as protection domains. In this model, protection 664 paths converge at domain boundaries. 666 Note that the question of SRLG identification is not yet fully 667 answered. There are two classes of SRLG: 669 - those that indicate resources that are all contained witin one 670 domain 672 - those that span domains. 674 The former might be identified using a combination of domain ID and 675 an SRLG ID that is administered by the domain. The latter requires 676 a wider scope to the SRLG ID, and it is not clear how this would be 677 administered. 679 5.6. Domain-Specific Constraints 681 While the meaning of certain constraints, like bandwidth, can be 682 assumed to be constant across different domains, other TE constraints 683 (such as resource affinity, color, metric, priority, etc.) may have 684 different meanings in different domains and this may impact the 685 ability to support DiffServ-aware MPLS, or to manage pre-emption. 687 In order to achieve consistent meaning and LSP establishment, this 688 fact must be considered when performing constraint-based path 689 computation or when signaling across domain boundaries. 691 A mapping function can be derived for most constraints based on 692 policy agreements between the Domain administrators. The details of 693 such a mapping function are outside the scope of this document, but 694 it is important to note that the default behavior MUST either be 695 that a constant mapping is applied or that any requirement to apply 696 these constraints across a domain boundary must fail in the absence 697 of explicit mapping rules. 699 5.7. Policy Control 701 Domain boundaries are natural points for policy control. There is 702 little to add on this subject except to note that a TE LSP that 703 cannot be established on a path through one domain because of a 704 policy applied at the domain boundary, may be satisfactorily 705 established using a path that avoids the demurring domain. In any 706 case, when a TE LSP signaling attempt is rejected due to 707 non-compliance with some policy constraint, this SHOULD be reflected 708 to the ingress LSR. 710 5.8. Inter-domain OAM 712 Some elements of OAM may be intentionally confined within a domain. 713 Others (such as end-to-end liveness and connectivity testing) clearly 714 need to span the entire multi-domain TE LSP. Where issues of 715 confidentiality are strong, collaboration between PCEs or domain 716 boundary nodes might be required in order to provide end-to-end OAM. 718 The different signaling mechanisms described above may need 719 refinements to [LSPPING], and [BFD-MPLS], etc., to gain full 720 end-to-end visibility. These protocols should, however, be considered 721 in the light of confidentiality requirements. 723 Route recording is a commonly used feature of signaling that provides 724 OAM information about the path of an established LSP. When an LSP 725 traverses a domain boundary, the border node may remove or aggregate 726 some of the recorded information for confidentiality or other policy 727 reasons. 729 5.9. Point-to-Multipoint 731 Inter-domain point-to-multipoint (P2MP) requirements are explicitly 732 out of scope of this document. They may be covered by other documents 733 dependent on the details of MPLS TE P2MP solutions. 735 5.10. Applicability to Non-Packet Technologies 737 Non-packet switching technologies may present particular issues for 738 inter-domain LSPs. While packet switching networks may utilize 739 control planes built on MPLS or GMPLS technology, non-packet networks 740 are limited to GMPLS. 742 The specific architectural considerations and requirements for 743 inter-domain LSP setup in non-packet networks are covered in a 744 separate document [GMPLS-AS]. 746 6. Security Considerations 748 Requirements for security within domains are unchanged from [RFC3209] 749 and [RFC3473], but requirements for inter-domain security are, if 750 anything, more significant. 752 Authentication techniques identified for use with RSVP-TE can only 753 operate across domain boundaries if there is coordination between the 754 administrators of those domains. 756 Confidentiality may also be considered to be security factors. 758 Applicability statements for particular combinations of signaling, 759 routing and path computation techniques are expected to contain 760 detailed security sections. 762 7. IANA Considerations 764 This document makes no requests for any IANA action. 766 8. Acknowledgements 768 The authors would like to extend their warmest thanks to Kireeti 769 Kompella for convincing them to expend effort on this document. 771 Grateful thanks to Dimitri Papadimitriou and Tomohiro Otani for their 772 review and suggestions on the text. 774 9. Intellectual Property Considerations 776 The IETF takes no position regarding the validity or scope of any 777 Intellectual Property Rights or other rights that might be claimed to 778 pertain to the implementation or use of the technology described in 779 this document or the extent to which any license under such rights 780 might or might not be available; nor does it represent that it has 781 made any independent effort to identify any such rights. Information 782 on the procedures with respect to rights in RFC documents can be 783 found in BCP 78 and BCP 79. 785 Copies of IPR disclosures made to the IETF Secretariat and any 786 assurances of licenses to be made available, or the result of an 787 attempt made to obtain a general license or permission for the use of 788 such proprietary rights by implementers or users of this 789 specification can be obtained from the IETF on-line IPR repository at 790 http://www.ietf.org/ipr. 792 The IETF invites any interested party to bring to its attention any 793 copyrights, patents or patent applications, or other proprietary 794 rights that may cover technology that may be required to implement 795 this standard. Please address the information to the IETF at 796 ietf-ipr@ietf.org. 798 10. Normative References 800 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 801 Requirement Levels", BCP 14, RFC 2119, March 1997. 803 [RFC3209] Awduche, et al, "Extensions to RSVP for LSP Tunnels", 804 RFC 3209, December 2001. 806 [RFC3473] Berger, L., Editor "Generalized Multi-Protocol Label 807 Switching (GMPLS) Signaling - Resource ReserVation 808 Protocol-Traffic Engineering (RSVP-TE) Extensions", 809 RFC 3473, January 2003. 811 [RFC3667] Bradner, S., "IETF Rights in Contributions", BCP 78, 812 RFC 3667, February 2004. 814 [RFC3668] Bradner, S., Ed., "Intellectual Property Rights in IETF 815 Technology", BCP 79, RFC 3668, February 2004. 817 11. Informational References 819 [RFC3630] Katz, D., Yeung, D., Kompella, K., "Traffic Engineering 820 Extensions to OSPF Version 2", RFC 3630, September 2003 822 [RFC3784] Li, T., Smit, H., "IS-IS extensions for Traffic 823 Engineering", RFC 3784, June 2004. 825 [ATTRIB] A. Farrel, D. Papadimitriou, JP. Vasseur, "Encoding of 826 Attributes for Multiprotocol Label Switching (MPLS) 827 Label Switched Path (LSP) Establishment Using RSVP-TE", 828 draft-ietf-mpls-rsvpte-attributes, work in progress. 830 [BFD-MPLS] R. Aggarwal and K. Kompella, "BFD For MPLS LSPs", work 831 in progress. 833 [CRANKBACK] Farrel, A., et al., "Crankback Signaling Extensions for 834 MPLS Signaling", draft-ietf-ccamp-crankback, 835 work in progress. 837 [EXCLUDE] Lee et all, Exclude Routes - Extension to RSVP-TE, 838 draft-ietf-ccamp-rsvp-te-exclude-route, work in 839 progress. 841 [FRR] Ping Pan, et al, "Fast Reroute Extensions to RSVP-TE 842 for LSP Tunnels", draft-ietf-mpls-rsvp-lsp-fastreroute, 843 work in progress. 845 [GMPLS-AS] Otani, T., Kumaki, K., and Okamoto, S., "GMPLS Inter-AS 846 Traffic Engineering Requirements", 847 draft-otani-ccamp-interas-GMPLS-TE, work in progress. 849 [HIER] Kompella K., Rekhter Y., "LSP Hierarchy with 850 Generalized MPLS TE", draft-ietf-mpls-lsp-hierarchy, 851 work in progress. 853 [INTER-AREA] Le Roux, Vasseur et Boyle, "Requirements for support of 854 Inter-Area and Inter-AS MPLS Traffic Engineering", 855 draft-ietf-tewg-interarea-mpls-te-req, work in 856 progress. 858 [INTER-AS] Zhang, R., Vasseur, JP. et al, "MPLS Inter-AS Traffic 859 Engineering requirements", 860 draft-ietf-tewg-interas-mpls-te-req, work in progress. 862 [LSPPING] Kompella, K., et al., " Detecting Data Plane Liveliness 863 in MPLS", draft-ietf-mpls-lsp-ping, work in progress. 865 [MRN] K. Shiomoto, et al., "Requirements for GMPLS-based 866 multi-region and multi-layer networks", 867 draft-shiomoto-ccamp-gmpls-mrn-reqs, work in progress. 869 [OVERLAY] G. Swallow et al, "GMPLS RSVP Support for the Overlay 870 Model", draft-ietf-ccamp-gmpls-overlay, work in 871 progress. 873 [PCE] Ash, G., Farrel, A., and Vasseur, JP., "Path 874 Computation Element (PCE) Architecture", 875 draft-ietf-pce-architecture, work in progress. 877 [SEG-PROT] Berger, L., Bryskin, I., Papadimitriou, D. and Farrel, 878 A., "GMPLS Based Segment Recovery", 879 draft-ietf-ccamp-gmpls-segment-recovery, work in 880 progress. 882 [STITCH] Ayyangar, A., and Vasseur, JP., "LSP Stitching with 883 Generalized MPLS TE", 884 draft-ietf-ccamp-lsp-stitching, work in progress. 886 12. Authors' Addresses 888 Adrian Farrel 889 Old Dog Consulting 890 EMail: adrian@olddog.co.uk 892 Jean-Philippe Vasseur 893 Cisco Systems, Inc. 894 300 Beaver Brook Road 895 Boxborough , MA - 01719 896 USA 897 Email: jpv@cisco.com 899 Arthi Ayyangar 900 Juniper Networks, Inc 901 1194 N.Mathilda Ave 902 Sunnyvale, CA 94089 903 USA 904 Email: arthi@juniper.net 906 13. Full Copyright Statement 908 Copyright (C) The Internet Society (2005). This document is subject 909 to the rights, licenses and restrictions contained in BCP 78, and 910 except as set forth therein, the authors retain all their rights. 912 This document and the information contained herein are provided on an 913 "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS 914 OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY AND THE INTERNET 915 ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED, 916 INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE 917 INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED 918 WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.