idnits 2.17.1 draft-ietf-pce-inter-layer-frwk-03.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 19. -- Found old boilerplate from RFC 3978, Section 5.5, updated by RFC 4748 on line 906. -- Found old boilerplate from RFC 3979, Section 5, paragraph 1 on line 882. -- Found old boilerplate from RFC 3979, Section 5, paragraph 2 on line 889. -- Found old boilerplate from RFC 3979, Section 5, paragraph 3 on line 895. 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 ** The document seems to lack a 1id_guidelines paragraph about 6 months document validity -- however, there's a paragraph with a matching beginning. Boilerplate error? == The page length should not exceed 58 lines per page, but there was 1 longer page, the longest (page 1) being 928 lines Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust 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 (March 2007) is 6245 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- No issues found here. Summary: 4 errors (**), 0 flaws (~~), 2 warnings (==), 7 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 1 Network Working Group Eiji Oki 2 Internet Draft NTT 3 Category: Informational Jean-Louis Le Roux 4 Expires: September 2007 France Telecom 5 Adrian Farrel 6 Old Dog Consulting 7 March 2007 9 Framework for PCE-Based Inter-Layer MPLS and GMPLS Traffic 10 Engineering 12 draft-ietf-pce-inter-layer-frwk-03.txt 14 Status of this Memo 16 By submitting this Internet-Draft, each author represents that any 17 applicable patent or other IPR claims of which he or she is aware 18 have been or will be disclosed, and any of which he or she becomes 19 aware will be disclosed, in accordance with Section 6 of BCP 79. 21 Internet-Drafts are working documents of the Internet Engineering 22 Task Force (IETF), its areas, and its working groups. Note that 23 other groups may also distribute working documents as Internet- 24 Drafts. 26 Internet-Drafts are draft documents valid for a maximum of six 27 months and may be updated, replaced, or obsoleted by other documents 28 at any time. It is inappropriate to use Internet- Drafts as 29 reference material or to cite them other than as "work in progress." 31 The list of current Internet-Drafts can be accessed at 32 http://www.ietf.org/ietf/1id-abstracts.txt. 34 The list of Internet-Draft Shadow Directories can be accessed at 35 http://www.ietf.org/shadow.html. 37 Abstract 39 A network may comprise multiple layers. It is important to globally 40 optimize network resource utilization, taking into account all 41 layers, rather than optimizing resource utilization at each layer 42 independently. This allows better network efficiency to be achieved 43 through a process that we call inter-layer traffic engineering. The 44 Path Computation Element (PCE) can be a powerful tool to achieve 45 inter-layer traffic engineering. 47 This document describes a framework for applying the PCE-based 48 architecture to inter-layer Multiprotocol Label Switching (MPLS) and 49 Generalized MPLS (GMPLS) traffic engineering. It provides 50 suggestions for the deployment of PCE in support of multi-layer 51 networks. This document also describes network models where PCE 52 performs inter-layer traffic engineering, and the relationship 53 between PCE and a functional component called the Virtual Network 54 Topology Manager (VNTM). 56 Table of Contents 58 1. Terminology.....................................................2 59 2. Introduction....................................................2 60 3. Inter-Layer Path Computation....................................3 61 4. Inter-layer Path Computation Models.............................5 62 4.1. Single PCE Inter-Layer Path Computation......................5 63 4.2. Multiple PCE Inter-Layer Path Computation....................5 64 4.3. General Observations.........................................6 65 5. Inter-Layer Path Control........................................7 66 5.1. VNT Management...............................................7 67 5.2. Inter-Layer Path Control Models..............................7 68 5.2.1. Cooperation Model Between PCE and VNTM.....................7 69 5.2.2. Higher-Layer Signaling Trigger Model.......................9 70 5.2.3. Examples of Multi-Layer ERO...............................11 71 6. Choosing Between Inter-Layer Path Control Models...............11 72 6.1. VNTM Functions:.............................................11 73 6.2. Border LSR Functions:.......................................12 74 6.3. Complete Inter-Layer LSP Setup Time:........................12 75 6.4. Network Complexity..........................................12 76 6.5. Separation of Layer Management..............................13 77 7. Security Considerations........................................13 78 8. Acknowledgment.................................................14 79 9. References.....................................................14 80 9.1. Normative Reference.........................................14 81 9.2. Informative Reference.......................................14 82 10. Authors' Addresses...........................................15 83 11. Intellectual Property Statement..............................15 85 1. Terminology 87 This document uses terminology from the PCE-based path computation 88 architecture [RFC4655] and also common terminology from Multi 89 Protocol Label Switching (MPLS) [RFC3031], Generalized MPLS (GMPLS) 90 [RFC3945] and Multi-Layer Networks [MLN-REQ]. 92 2. Introduction 94 A network may comprise multiple layers. These layers may represent 95 separations of technologies (e.g., packet switch capable (PSC), time 96 division multiplex (TDM), or lambda switch capable (LSC)) [RFC3945], 97 separation of data plane switching granularity levels (e.g., PSC-1, 98 PSC-2, VC4, or VC12) [MLN-REQ], or a distinction between client and 99 server networking roles. In this multi-layer network, Label Switched 100 Paths (LSPs) in a lower layer are used to carry higher-layer LSPs 101 across the lower-layer network. The network topology formed by 102 lower-layer LSPs and advertised to the higher layer is called a 103 Virtual Network Topology (VNT) [MLN-REQ]. 105 It may be effective to optimize network resource utilization 106 globally, i.e., taking into account all layers, rather than 107 optimizing resource utilization at each layer independently. This 108 allows better network efficiency to be achieved and is what we call 109 inter-layer traffic engineering. This includes mechanisms allowing 110 the computation of end-to-end paths across layers (known as inter- 111 layer path computation), and mechanisms for control and management 112 of the Virtual Network Topology (VNT) by setting up and releasing 113 LSPs in the lower layers [MLN-REQ]. 115 Oki et al Expires September 2007 2 116 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 118 Inter-layer traffic engineering is included in the scope of the Path 119 Computation Element (PCE)-based architecture [RFC4655], and PCE can 120 provide a suitable mechanism for resolving inter-layer path 121 computation issues. 123 PCE Communication Protocol requirements for inter-layer traffic 124 engineering are set forth in [PCE-INTER-LAYER-REQ]. 126 This document describes a framework for applying the PCE-based 127 architecture to inter-layer traffic engineering. It provides 128 suggestions for the deployment of PCE in support of multi-layer 129 networks. This document also describes network models where PCE 130 performs inter-layer traffic engineering, and the relationship 131 between PCE and a functional component in charge of the control and 132 management of the VNT, and called the Virtual Network Topology 133 Manager (VNTM). 135 3. Inter-Layer Path Computation 137 This section describes key topics of inter-layer path computation in 138 MPLS and GMPLS networks. 140 [RFC4206] defines a way to signal a higher-layer LSP, whose explicit 141 route includes hops traversed by LSPs in lower layers. The 142 computation of end-to-end paths across layers is called Inter-Layer 143 Path Computation. 145 A Label Switching Router (LSR) in the higher-layer might not have 146 information on the topology of the lower-layer, particularly in an 147 overlay or augmented model deployment, and hence may not be able to 148 compute an end-to-end path across layers. 150 PCE-based inter-layer path computation, consists of using one or 151 more PCEs to compute an end-to-end path across layers. This could be 152 achieved by a single PCE path computation where the PCE has topology 153 information about multiple layers and can directly compute an end- 154 to-end path across layers considering the topology of all of the 155 layers. Alternatively, the inter-layer path computation could be 156 performed as a multiple PCE computation where each member of a set 157 of PCEs has information about the topology of one or more layers 158 (but not all layers), and the PCEs collaborate to compute an end-to- 159 end path. 161 Consider, for instance, a two-layer network where the higher-layer 162 network is a packet-based IP/MPLS or GMPLS network, and the lower- 163 layer network is a GMPLS optical network. An ingress LSR in the 164 higher-layer network tries to set up an LSP to an egress LSR also in 165 the higher-layer network across the lower-layer network, and needs a 166 path in the higher-layer network. However, suppose that there is no 167 Traffic Engineering (TE) link in the higher-layer network between 168 border LSRs, which are located on the boundary between the higher- 169 layer and lower-layer networks, and that the ingress LSR does not 170 have topology visibility into the lower layer. If a single-layer 171 path computation is applied for the higher-layer, the path 172 computation fails because of the missing TE link. On the other hand, 173 inter-layer path computation is able to provide a route in the 174 higher-layer and a suggestion that a lower-layer LSP be set up 175 between border LSRs. 177 Oki et al Expires September 2007 3 178 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 180 Lower-layer LSPs that are advertised as TE links into the higher- 181 layer network form a Virtual Network Topology (VNT), which can be 182 used for routing higher-layer LSPs. Inter-layer path computation for 183 end-to-end LSPs in the higher-layer network that span the lower- 184 layer network may utilize the VNT, and PCE is a candidate for 185 computing the paths of such higher-layer LSPs within the higher- 186 layer network. Alternatively, the PCE-based path computation model 187 can: 189 - Perform a single computation on behalf of the ingress LSR using 190 information gathered from more than one layer. This mode is referred 191 to as Single PCE Computation in [RFC4655]. 193 - Compute a path on behalf of the ingress LSR through cooperation 194 with PCEs responsible for each layer. This mode is referred to as 195 Multiple PCE Computation with inter-PCE communication in [RFC4655]. 197 - Perform separate path computations on behalf of the TE-LSP head- 198 end and each transit border LSR that is the entry point to a new 199 layer. This mode is referred to as Multiple PCE Computation (without 200 inter-PCE communication) in [RFC4655]. This option utilizes per- 201 layer path computation performed independently by successive PCEs. 203 The PCE invoked by the head-end LSR computes a path that the LSR can 204 use to signal an MPLS-TE or GMPLS LSP once the path information has 205 been converted to an Explicit Route Object (ERO) for use in RSVP-TE 206 signaling. There are two options. 208 - Option 1: Mono-layer path. 209 The PCE computes a "mono-layer" path, i.e., a path that includes 210 only TE links from the same layer. There are two cases for this 211 option. In the first case the PCE computes a path that includes 212 already established lower-layer LSPs or lower-layer LSPs to be 213 established on demand. That is, the resulting ERO includes sub- 214 object(s) corresponding to lower-layer hierarchical LSPs expressed 215 as the TE link identifiers of the hierarchical LSPs when advertised 216 as TE links in the higher-layer network. The TE link may be a 217 regular TE link that is actually established, or a virtual TE link 218 that is not established yet (see [MLN-REQ]). If it is a virtual TE 219 link, this triggers a setup attempt for a new lower-layer LSP when 220 signaling reaches the head-end of the lower-layer LSP. Note that the 221 path of a virtual TE link is not necessarily known in advance, and 222 this may require a further (lower-layer) path computation. 224 The second case is that the PCE computes a path that includes a 225 loose hop that spans the lower-layer network. The higher layer path 226 computation selects which lower layer network to use, and selects 227 the entry and exit points from that lower-layer network, but does 228 not select the path across the lower-layer network. A transit LSR 229 that is the entry point to the lower-layer network is expected to 230 expand the loose hop (either itself or relying on the services of a 231 PCE). The path expansion process on the border LSR may result either 232 in the selection of an existing lower-layer LSP, or in the 233 computation and setup of a new lower-layer LSP. 235 - Option 2: Multi-layer path. The PCE computes a "multi-layer" path, 236 i.e., a path that includes TE links from distinct layers [RFC4206]. 237 Such a path can include the complete path of one or more lower-layer 238 LSPs that already exist or are not yet established. In the latter 240 Oki et al Expires September 2007 4 241 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 243 case, the signaling of the higher-layer LSP will trigger the 244 establishment of the lower-layer LSPs. 246 4. Inter-layer Path Computation Models 248 As stated in Section 3, two PCE modes defined in the PCE 249 architecture can be used to perform inter-layer path computation. 250 They are discussed below. 252 4.1. Single PCE Inter-Layer Path Computation 254 In this model Inter-layer path computation is performed by a single 255 PCE that has topology visibility into all layers. Such a PCE is 256 called a multi-layer PCE. 258 In Figure 1, the network is comprised of two layers. LSRs H1, H2, H3, 259 and H4 belong to the higher layer, and LSRs H2, H3, L1, and L2 260 belong to the lower layer. The PCE is a multi-layer PCE that has 261 visibility into both layers. It can perform end-to-end path 262 computation across layers (single PCE path computation). For 263 instance, it can compute an optimal path H1-H2-L1-L2-H3-H4, for a 264 higher layer LSP from H1 to H4. This path includes the path of a 265 lower layer LSP from H2 to H3, already in existence or not yet 266 established. 268 ----- 269 | PCE | 270 ----- 271 ----- ----- ----- ----- 272 | LSR |--| LSR |................| LSR |--| LSR | 273 | H1 | | H2 | | H3 | | H4 | 274 ----- -----\ /----- ----- 275 \----- -----/ 276 | LSR |--| LSR | 277 | L1 | | L2 | 278 ----- ----- 280 Figure 1 : Multi-Layer PCE - 281 - A single PCE with multi-layer 282 visibility 284 4.2. Multiple PCE Inter-Layer Path Computation 286 In this model there is at least one PCE per layer, and each PCE has 287 topology visibility restricted to its own layer. Some providers may 288 want to keep the layer boundaries due to factors such as 289 organizational and/or service management issues. The choice for 290 multiple PCE computation instead of single PCE computation may also 291 be driven by scalability considerations, as in this mode a PCE only 292 needs to maintain topology information for one layer (resulting in a 293 size reduction for the Traffic Engineering Database (TED)). 295 These PCEs are called mono-layer PCEs. Mono-layer PCEs collaborate 296 to compute an end-to-end optimal path across layers. 298 Oki et al Expires September 2007 5 299 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 301 In Figure 2, there is one PCE in each layer. The PCEs from each 302 layer collaborate to compute an end-to-end path across layers. PCE 303 Hi is responsible for computations in the higher layer and may 304 "consult" with PCE Lo to compute paths across the lower layer. PCE 305 Lo is responsible for path computation in the lower layer. A simple 306 example of cooperation between the PCEs could be as follows: 307 - LSR H1 sends a request for a path H1-H4 to PCE Hi 308 - PCE Hi selects H2 as the entry point to the lower layer, and H3 as 309 the exit point. 310 - PCE Hi requests a path H2-H3 from PCE Lo. 311 - PCE Lo returns H2-L1-L2-H3 to PCE Hi. 312 - PEC Hi is able to compute the full path (H1-H2-L1-L2-H3-H4) and 313 return it to H1. 315 Of course more complex cooperation may be required if an optimal 316 end-to-end path is desired. 318 ----- 319 | PCE | 320 | Hi | 321 --+-- 322 | 323 ----- ----- | ----- ----- 324 | LSR |--| LSR |............|...........| LSR |--| LSR | 325 | H1 | | H2 | | | H3 | | H4 | 326 ----- -----\ --+-- /----- ----- 327 \ | PCE | / 328 \ | Lo | / 329 \ ----- / 330 \ / 331 \----- -----/ 332 | LSR |--| LSR | 333 | L1 | | L2 | 334 ----- ----- 336 Figure 2 : Cooperating Mono-Layer PCEs - Multiple PCEs with single- 337 layer visibility 339 4.3. General Observations 341 - Depending on implementation details, inter-layer path computation 342 time in the Single PCE inter-layer path computation model may be 344 Oki et al Expires September 2007 6 345 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 347 less than that of the Multiple PCE model with cooperating mono-layer 348 PCEs, because there is no requirement to exchange messages between 349 cooperating PCEs. 351 - When TE topology for all layer networks is visible within one 352 routing domain, the single PCE inter-layer path computation model 353 may be adopted because a PCE is able to collect all layers' TE 354 topologies by participating in only one routing domain. 356 - As the single PCE inter-layer path computation model uses more TE 357 topology information than is used by PCEs in the Multiple PCE path 358 computation model, it requires more computation power and memory. 360 When there are multiple candidate layer border nodes (we may say 361 that the higher layer is multi-homed), optimal path computation 362 requires that all the possible paths transiting different layer 363 border nodes or links be examined. This is relatively simple in the 364 single PCE inter-layer path computation model because the PCE has 365 full visibility - 366 - the computation is similar to the computation 367 within a single domain of a single layer. In the multiple PCE inter- 368 layer path computation model, backward recursive techniques 369 described in [BRPC] could be used, by considering layers as separate 370 domains. 372 5. Inter-Layer Path Control 374 5.1. VNT Management 376 As a result of inter-layer path computation, a PCE may determine 377 that there is insufficient bandwidth available in the higher-layer 378 network to support this or future higher-layer LSPs. The problem 379 might be resolved if new LSPs were provisioned across the lower- 380 layer network. Further, the modification, re-organization and new 381 provisioning of lower-layer LSPs may enable better utilization of 382 lower-layer network resources given the demands of the higher-layer 383 network. In other words, the VNT needs to be controlled or managed 384 in cooperation with inter-layer path computation. 386 A VNT Manager (VNTM) is defined as a network element that manages 387 and controls the VNT. PCE and VNT Managemer are distinct functional 388 elements that may or may not be co-located. 390 5.2. Inter-Layer Path Control Models 392 5.2.1. Cooperation Model Between PCE and VNTM 394 ----- ------ 395 | PCE |--->| VNTM | 396 ----- ------ 397 ^ : 398 : : 399 : : 400 v V 401 ----- ----- ----- ----- 402 | LSR |----| LSR |................| LSR |----| LSR | 403 | H1 | | H2 | | H3 | | H4 | 404 ----- -----\ /----- ----- 405 \----- -----/ 406 | LSR |--| LSR | 407 | L1 | | L2 | 409 Oki et al Expires September 2007 7 410 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 412 ----- ----- 414 Figure 3: Cooperation Model Between PCE and VNTM 416 A multi-layer network consists of higher-layer and lower-layer 417 networks. LSRs H1, H2, H3, and H4 belong to the higher-layer network, 418 LSRs H2, L1, L2, and H3 belong to the lower-layer network, as shown 419 in Figure 3. Consider that H1 requests PCE to compute an inter-layer 420 path between H1 and H4. There is no TE link in the higher-layer 421 between H2 and H3 before the path computation request fails. But the 422 PCE may provide information to the VNT Manager responsible for the 423 lower layer network that may help resolve the situation for future 424 higher-layer LSP setup. 426 The roles of PCE and VNTM are as follows. PCE performs inter-layer 427 path computation and is unable to supply a path because there is no 428 TE link between H2 and H3. The computation fails, but PCE suggests 429 to VNTM that a lower-layer LSP (H2-H3) could be established to 430 support future LSP requests. Messages from PCE to VNTM contain 431 information about the higher-layer demand (from H2 to H3). VNTM uses 432 local policy and possibly management/configuration input to 433 determine how to process the suggestion from PCE, and may request an 434 ingress LSR (e.g. H2) to establish a lower-layer LSP. VNTM or the 435 ingress LSR (H2) may themselves use a PCE with visibility into the 436 lower layer to compute the path of this new LSP. 438 When the higher-layer PCE fails to compute a path and notifies VNTM, 439 it may wait for the lower-layer LSP to be set up and advertised as a 440 TE link. It could then compute the complete end-to-end path for the 441 higher-layer LSP and return the result to the PCC. In this case, the 442 PCC may be kept waiting for some time, and it is important that the 443 PCC understands this. It is also important that the PCE and VNTM 444 have an agreement that the lower-layer LSP will be set up in a 445 timely manner, or that the PCE will be notified by VNTM that no new 446 LSP will become available. In any case, if the PCE decides to wait, 447 it must operates a timeout. An example of such a cooperative 448 procedure between PCE and VNTM is as follows using the exmaple 449 network in Figure 3. 451 Step 1: H1 (PCC) requests PCE to compute a path between H1 and H4. 453 Step 2: The path computation fails because there is no TE link 454 across the lower-layer network. 456 Step 3: PCE suggests to VNTM that a new TE link connecting H2 and H3 457 would be useful. VNTM considers whether lower-layer LSPs should be 458 established if necessary and if acceptable within VNTM's policy 459 constraints. The PCE notifies VNTM that it will be waiting for the 460 TE link to be created. 462 Step 4: VNTM requests an ingress LSR in the lower-layer network 463 (e.g., H2) to establish a lower-layer LSP. The request message may 464 include a lower-layer LSP route obtained from the PCE responsible 465 for the lower-layer network. 467 Step 5: The ingress LSR signals to establish the lower-layer LSP. 469 Step 6: If the lower-layer LSP setup is successful, the ingress LSR 470 notifies VNTM that the LSP is complete and supplies the tunnel 471 information. 473 Oki et al Expires September 2007 8 474 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 476 Step 7: The ingress LSR (H2) advertises the new LSP as a TE link in 477 the higher-layer network routing instance. 479 Step 8: PCE notices the new TE link advertisement and recomputes the 480 requested path. 482 Step 9: PCE replies to H1 (PCC) with a computed higher-layer LSP 483 route. The computed path is categorized as a mono-layer path that 484 includes the already-established lower layer-LSP as a single hop in 485 the higher layer. The higher-layer route is specified as H1-H2-H3-H4, 486 where all hops are strict. 488 Step 9: H1 initiates signaling with the computed path H2-H3-H4 to 489 establish the higher-layer LSP. 491 5.2.2. Higher-Layer Signaling Trigger Model 493 ----- 494 | PCE | 495 ----- 496 ^ 497 : 498 : 499 v 500 ----- ----- ----- ----- 501 | LSR |----| LSR |................| LSR |--| LSR | 502 | H1 | | H2 | | H3 | | H4 | 503 ----- -----\ /----- ----- 504 \----- -----/ 505 | LSR |--| LSR | 506 | L1 | | L2 | 507 ----- ----- 509 Figure 4: Higher-layer Signaling Trigger Model 511 Figure 4 shows the higher-layer signaling trigger model. As in the 512 case described in Section 5.2.1, consider that H1 requests PCE to 513 compute a path between H1 and H4. There is no TE link in the higher- 514 layer between H2 and H3 before the path computation request. 516 PCE is unable to compute a mono-layer path, but may judge that the 517 establishment of a lower-layer LSP between H2 and H3 would provide 518 adequate connectivity. If the PCE has inter-layer visibility it may 519 return a path that includes hops in the lower layer (H1-H2-L1-L2-H3- 520 H4), but if it has no visiblity into the lower layer, it may return 521 a path with a loose hop from H2 to H3 (H1-H2-H3(loose)-H4). The 522 former is a multi-layer path, and the latter a mono-layer path that 523 includes loose hops. 525 In the higher-layer signaling trigger model with a multi-layer path, 526 the LSP route supplied by the PCE includes the route of a lower- 527 layer LSP that is not yet established. A border LSR that is located 528 at the boundary between the higher-layer and lower-layer networks 529 (H2 in this example) receives a higher-layer signaling message, 530 notices that the next hop is in the lower-layer network, starts to 531 setup the lower-layer LSP as described in [RFC4206]. Note that these 532 actions depends on a policy at the border LSR. An example procedure 533 of the signaling trigger model with a multi-layer path is as follows. 535 Oki et al Expires September 2007 9 536 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 538 Step 1: H1 (PCC) requests PCE to compute a path between H1 and H4. 539 The request indicates that inter-layer path computation is allowed. 541 Step 2: As a result of the inter-layer path computation, PCE judges 542 that a new lower-layer LSP needs to be established. 544 Step 3: PCE replies to H1 (PCC) with a computed multi-layer route 545 including higher-layer and lower-layer LSP routes. The route may be 546 specified as H1-H2-L1-L2-H3-H4, where all hops are strict. 548 Step 4: H1 initiates higher-layer signaling using the computed 549 explicit router of H2-L1-L2-H3-H4. 551 Step 5: The border LSR (H2) that receives the higher-layer signaling 552 message starts lower-layer signaling to establish a lower-layer LSP 553 along the specified lower-layer route of H2-L1-L2-H3. That is, the 554 border LSR recognizes the hops within the explicit route that apply 555 to the lower-layer network, verifies with local policy that a new 556 LSP is acceptable, and establishes the required lower-layer LSP. 557 Note that it is possible that a suitable lower-layer LSP has already 558 been established (or become available) between the time that the 559 computation was performed and the moment when the higher-layer 560 signaling message reached the border LSR. In this case, the border 561 LSR may select such a lower-layer LSP without the need to signal a 562 new LSP provided that the lower-layer LSP satisfies the explicit 563 route in the higher-layer signaling request. 565 Step 6: After the lower-layer LSP is established, the higher-layer 566 signaling continues along the specified higher-layer route of H2-H3- 567 H4 using hierarchical signaling [RFC4206]. 569 On the other hand, in the signaling trigger model with a mono-layer 570 path, a higher-layer LSP route includes a loose hop to traverse the 571 lower-layer network between the two border LSRs. A border LSR that 572 receives a higher-layer signaling message needs to determine a path 573 for a new lower-layer LSP. It applies local policy to verify that a 574 new LSP is acceptable and then either consults a PCE with 575 responsibility for the lower-layer network or computes the path by 576 itself, and initiates signaling to establish the lower-layer LSP. 577 Again, it is possible that a suitable lower-layer LSP has already 578 been established (or become available). In this case, the border LSR 579 may select such a lower-layer LSP without the need to signal a new 580 LSP provided that the lower-layer LSP satisfies the explicit route 581 in the higher-layer signaling request. Since the higher-layer 582 signaling request used a loose hop without specifying any specifics 583 of the path within the lower-layer network, the border LSR has 584 greater freedom to choose a lower-layer LSP than in the previous 585 example. 587 The difference between procedures of the signaling trigger model 588 with a multi-layer path and a mono-layer path is Step 5. Step 5 of 589 the signaling trigger model with a mono layer path is as follows: 591 Step 5': The border LSR (H2) that receives the higher-layer 592 signaling message applies local policy to verify that a new LSP is 593 acceptable and then initiates establishment of a lower-layer LSP. It 594 either consults a PCE with responsibility for the lower-layer 595 network or computes the route by itself to expand the loose hop 596 route in the higher-layer path. 598 Oki et al Expires September 2007 10 599 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 601 Finally, note that a virtual TE link may have been advertised into 602 the higher-layer network. This causes the PCE to return a path H1- 603 H2-H3-H4 where all the hops are strict. But when the higher-layer 604 signaling message reaches the layer border node H2 (that was 605 responsible for advertising the virtual TE link) it realizes that 606 the TE link does not exist yet, and signals the necessary LSP across 607 the lower-layer network using its own path determination (just as 608 for a loose hop in the higher layer) before continuing with the 609 higher-layer signaling. 611 5.2.3. Examples of Multi-Layer ERO 613 PCE 614 ^ 615 : 616 : 617 V 618 H1--H2 H3--H4 619 \ / 620 L1==L2==L3--L4--L5 621 | 622 | 623 L6--L7 624 \ 625 H5--H6 627 Figure 5: Example of a Multi-Layer Network 629 This section describes how lower-layer LSP setup is performed in the 630 higher-layer signaling trigger model using an ERO that can include 631 subobjects in both the higher and lower layers. It gives rise to 632 several options for the ERO when it reaches the last LSR in the 633 higher layer network (H2). 634 1. The next subobject is a loose hop to H3 (mono layer ERO). 635 2. The next subobject is a strict hop to L1 followed by a loose hop 636 to H3. 637 3. The next subobjects are a series of hops (strict or loose) in the 638 lower-layer network followed by H3. For example, {L1(strict), 639 L3(loose), L5(loose), H3(strict)} 641 In the first example, the lower layer can utilize any LSP tunnel 642 that will deliver the end-to-end LSP to H3. In the third case, the 643 lower layer must select an LSP tunnel that traverses L3 and L5. 644 However, this does not mean that the lower layer can or should use 645 an LSP from L1 to L3 and another from L3 to L5. 647 6. Choosing Between Inter-Layer Path Control Models 649 This section compares the cooperation model between PCE and VNTM, 650 and the higher-layer signaling trigger model, in terms of VNTM 651 functions, border LSR functions, higher-layer signaling time, and 652 complexity (in terms of number of states and messages). An 653 appropriate model may be chosen by a network operator in different 654 deployment scenarios taking all these considerations into account. 656 6.1. VNTM Functions: 658 In the cooperation model, VNTM functions are required. In this model, 659 communications are required between PCE and VNTM, and between VNTM 660 and a border LSR. VNTM-LSR communication can rely on existing 661 GMPLS-TE MIB modules. PCE-VNTM communication will be detailed in 662 further revisions of this document. 664 Oki et al Expires September 2007 11 665 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 667 In the higher-layer signaling trigger model, no VNTM functions are 668 required, and no such communications are required. 670 If VNTM functions are not supported in a multi-layer network, the 671 higher-layer signaling trigger model has to be chosen. 673 The inclusion of VNTM functionality allows better coordination of 674 cross-network LSP tunnels and application of network-wide policy 675 that is far harder to apply in the trigger model since it requires 676 the coordination of policy between multiple border LSRs. 678 6.2. Border LSR Functions: 680 In the higher-layer signaling trigger model, a border LSR must have 681 some additional functions. It needs to trigger lower-layer signaling 682 when a higher-layer path message suggests that lower-layer LSP setup 683 is necessary. Note that, if virtual TE links are used, the border 684 LSRs must be capable of triggered signaling. 686 If the ERO in the higher-layer Path message uses a mono-layer path 687 or specifies a loose hop, the border LSR receiving the Path message 688 must obtain a lower-layer route either by consulting a PCE or by 689 using its own computation engine. If the ERO in the higher-layer 690 Path message uses a multi-layer path, the border LSR must judge 691 whether lower-layer signaling is needed. 693 In the cooperation model, no additional function for triggered 694 signaling is required in border LSRs except when virtual TE links 695 are used. Therefore, if these additional functions are not supported 696 in border LSRs, where a border LSR is controlled by VNTM to set up a 697 lower-layer LSP, the cooperation model has to be chosen. 699 6.3. Complete Inter-Layer LSP Setup Time: 701 Complete inter-layer LSP setup time includes inter-layer path 702 computation, signaling, and communication time between PCC and PCE, 703 PCE and VNTM, and VNTM and LSR. In the cooperation model, the 704 additional communication steps are required compared with the 705 higher-layer signaling trigger model. On the other hand, the 706 cooperation model provides better control at the cost of a longer 707 service setup time. 709 Note that, in terms of higher-layer signaling time, in the higher- 710 layer signaling trigger model, the required time from when higher- 711 layer signaling starts to when it is completed, is more than that of 712 the cooperation model except when a virtual TE link is included. 713 This is because the former model requires lower-layer signaling to 714 take place during the higher-layer signaling. A higher-layer ingress 715 LSR has to wait for more time until the higher-layer signaling is 716 completed. A higher-layer ingress LSR is required to be tolerant of 717 longer path setup times. 719 6.4. Network Complexity 721 If the higher and lower layer networks have multiple interconnects 722 then optimal path computation for end-to-end LSPs that cross the 723 layer boundaries is non-trivial. The higher layer LSP must be routed 724 to the correct layer border nodes to achieve optimality in both 725 layers. 727 Oki et al Expires September 2007 12 728 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 730 Where the lower layer LSPs are advertised into the higher layer 731 network as TE links, the computation can be resolved in the higher 732 layer network. Care needs to be taken in the allocation of TE 733 metrics (i.e., costs) to the lower layer LSPs as they are advertised 734 as TE links into the higher layer network, and this might be a 735 function for a VNT Manager component. Similarly, attention should be 736 given to the fact that the LSPs crossing the lower-layer network 737 might share points of common failure (e.g., they might traverse the 738 same link in the lower-layer network) and the shared risk link 739 groups (SRLGs) for the TE links advertised in the higher-layer must 740 be set accordingly. 742 In the single PCE model an end-to-end path can be found in a single 743 computation because there is full visibility into both layers and 744 all possible paths through all layer interconnects can be considered. 746 Where PCEs cooperate to determine a path, an iterative computation 747 model such as [BRPC] can be used to select an optimal path across 748 layers. 750 When non-cooperating mono-layer PCEs, each of which is in a separate 751 layer, are used with the triggered LSP model, it is not possible to 752 determine the best border LSRs, and connectivity cannot even be 753 guaranteed. In this case, signaling crankback techniques [CRANK] can 754 be used to eventually achieve connectivity, but optimality is far 755 harder to achieve. In this model, a PCE that is requested by an 756 ingress LSR to compute a path expects a border LSR to setup a lower- 757 layer path triggered by high-layer signaling when there is no TE 758 link between border LSRs. 760 6.5. Separation of Layer Management 762 Many network operators may want to provide a clear separation 763 between the management of the different layer networks. In some 764 cases, the lower layer network may come from a separate commercial 765 arm of an organization or from a different corporate body entirely. 766 In these cases, the policy applied to the establishment of LSPs in 767 the lower-layer network and to the advertisement of these LSPs as TE 768 links in the higher-layer network will reflect commercial agreements 769 and security concerns (see next section). Since the capacity of the 770 LSPs in the lower-layer network are likely to be significantly 771 larger than those in the client higher-layer network (multiplex- 772 server model), the administrator of the lower-layer network may want 773 to exercise caution before allowing a single small demand in the 774 higher layer to tie up valuable resources in the lower layer. 776 The necessary policy points for this separation of administration 777 and management are more easily achieved through the VNTM approach 778 than by using triggered signaling. In effect, the VNTM is the 779 coordination point for all lower layer LSPs and can be closely tied 780 to a human operator as well as to policy and billing. Such a model 781 can also be achieved using triggered signaling. 783 7. Security Considerations 785 Inter-layer traffic engineering with PCE raises new security issues 786 in both inter-layer path control models. 788 Oki et al Expires September 2007 13 789 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 791 In the cooperation model between PCE and VNTM, when the PCE judges a 792 new lower-layer LSP, communications between PCE and VNTM and between 793 VNTM and a border LSR are needed. In this case, there are some 794 security concerns that need to be addressed for these communications. 795 These communications should have some security mechanisms to ensure 796 authenticity, privacy and integrity. In particular, it is important 797 to protect against false triggers for LSP setup in the lower-layer 798 network. 800 In the higher-layer signaling trigger model, there are several 801 security concerns. First, PCE may inform PCC, which is located in 802 the higher-layer network, of multi-layer path information that 803 includes an ERO in the lower-layer network, while the PCC may not 804 have TE topology visibility into the lower-layer network. This 805 raises a security concern, where lower-layer hop information is 806 known to transit LSRs supporting a higher-layer LSP. Some security 807 mechanisms to ensure authenticity, privacy and integrity may be used. 809 Security issues may also exist when a single PCE is granted full 810 visibility of TE information that applies to multiple layers. 812 8. Acknowledgment 814 We would like to thank Kohei Shiomoto, Ichiro Inoue, Julien Meuric, 815 Jean-Francois Peltier, Young Lee, and Ina Minei for their useful 816 comments. 818 9. References 820 9.1. Normative Reference 822 [RFC3031] Rosen, E., Viswanathan, A., and R. Callon, "Multiprotocol 823 Label Switching Architecture", RFC 3031, January 2001. 824 [RFC3945] Mannie, E., "Generalized Multi-Protocol Label Switching 825 Architecture", RFC 3945, October 2004. 827 [RFC4206] Kompella, K., and Rekhter, Y., "Label Switched Paths (LSP) 828 Hierarchy with Generalized Multi-Protocol Label Switching (GMPLS) 829 Traffic Engineering (TE)", RFC 4206, October 2005. 831 [RFC4655] A. Farrel, JP. Vasseur and J. Ash, "A Path Computation 832 Element (PCE)-Based Architecture", RFC 4655, August 2006. 834 9.2. Informative Reference 836 [MLN-REQ] K. Shiomoto et al., "Requirements for GMPLS-based multi- 837 region networks (MRN)", draft-ietf-ccamp-gmpls-mln-reqs (work in 838 progress). 840 [PCE-INTER-LAYER-REQ] E. Oki et al., "PCC-PCE Communication 841 Requirements for Inter-Layer Traffic Engineering", draft-ietf-pce- 842 inter-layer-req (work in progress). 844 [BRPC] JP. Vasseur et al., "A Backward Recursive PCE-based 845 Computation (BRPC) procedure to compute shortest inter-domain 846 Traffic Engineering Label Switched Paths", draft-ietf-pce-brpc (work 847 in progress). 849 [CRANK] A. Farrel et al., "Crankback Signaling Extensions for MPLS 850 and GMPLS RSVP-TE", draft-ietf-ccamp-crankback (work in progress). 852 Oki et al Expires September 2007 14 853 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 855 10. Authors' Addresses 857 Eiji Oki 858 NTT 859 3-9-11 Midori-cho, 860 Musashino-shi, Tokyo 180-8585, Japan 861 Email: oki.eiji@lab.ntt.co.jp 863 Jean-Louis Le Roux 864 France Telecom R&D, 865 Av Pierre Marzin, 866 22300 Lannion, France 867 Email: jeanlouis.leroux@orange-ftgroup.com 869 Adrian Farrel 870 Old Dog Consulting 871 Email: adrian@olddog.co.uk 873 11. Intellectual Property Statement 875 The IETF takes no position regarding the validity or scope of any 876 Intellectual Property Rights or other rights that might be claimed 877 to pertain to the implementation or use of the technology described 878 in this document or the extent to which any license under such 879 rights might or might not be available; nor does it represent that 880 it has made any independent effort to identify any such rights. 881 Information on the procedures with respect to rights in RFC 882 documents can be found in BCP 78 and BCP 79. 884 Copies of IPR disclosures made to the IETF Secretariat and any 885 assurances of licenses to be made available, or the result of an 886 attempt made to obtain a general license or permission for the use 887 of such proprietary rights by implementers or users of this 888 specification can be obtained from the IETF on-line IPR repository 889 at http://www.ietf.org/ipr. 891 The IETF invites any interested party to bring to its attention any 892 copyrights, patents or patent applications, or other proprietary 893 rights that may cover technology that may be required to implement 894 this standard. Please address the information to the IETF at ietf- 895 ipr@ietf.org. 897 Disclaimer of Validity 899 This document and the information contained herein are provided on 900 an "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE 901 REPRESENTS OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE 902 IETF TRUST AND THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL 903 WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY 904 WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE 905 ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS 906 FOR A PARTICULAR PURPOSE. 908 Copyright Statement 910 Copyright (C) The IETF Trust (2007). 912 Oki et al Expires September 2007 15 913 draft-ietf-pce-inter-layer-frwk-03.txt March 2007 915 This document is subject to the rights, licenses and restrictions 916 contained in BCP 78, and except as set forth therein, the authors 917 retain all their rights. 919 Oki et al Expires September 2007 16