idnits 2.17.1 draft-zhao-teas-pce-control-function-01.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (May 24, 2016) is 2894 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Outdated reference: A later version (-11) exists of draft-ietf-pce-pce-initiated-lsp-05 == Outdated reference: A later version (-18) exists of draft-ietf-pce-pceps-09 == Outdated reference: A later version (-21) exists of draft-ietf-pce-stateful-pce-14 == Outdated reference: A later version (-15) exists of draft-ietf-spring-segment-routing-08 == Outdated reference: A later version (-06) exists of draft-pkd-pce-pcep-yang-05 == Outdated reference: A later version (-08) exists of draft-zhao-pce-pcep-extension-for-pce-controller-03 == Outdated reference: A later version (-02) exists of draft-zhao-teas-pcecc-use-cases-00 Summary: 0 errors (**), 0 flaws (~~), 8 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 TEAS Working Group A. Farrel, Ed. 3 Internet-Draft Juniper Networks 4 Intended status: Informational Q. Zhao, Ed. 5 Expires: November 25, 2016 R. Li 6 Huawei Technologies 7 C. Zhou 8 Cisco Systems 9 May 24, 2016 11 An Architecture for Use of PCE and PCEP in a Network with Central 12 Control 13 draft-zhao-teas-pce-control-function-01 15 Abstract 17 The Path Computation Element (PCE) has become established as a core 18 component of Software Defined Networking (SDN) systems. It can 19 compute optimal paths for traffic across a network for any definition 20 of "optimal" and can also monitor changes in resource availability 21 and traffic demands to update the paths. 23 Conventionally, the PCE has been used to derive paths for MPLS Label 24 Switched Paths (LSPs). These paths are supplied using the Path 25 Computation Element Communication Protocol (PCEP) to the head end of 26 the LSP for signaling in the MPLS network. 28 SDN has a far broader applicability than just signaled MPLS traffic 29 engineered networks, and the PCE may be used to determine paths in a 30 wide range of use cases including static LSPs, segment routing, 31 service function chaining (SFC), and indeed any form of routed or 32 switched network. It is, therefore reasonable to consider PCEP as a 33 general southbound control protocol for use in these environments to 34 allow the PCE to be fully enabled as a central controller. 36 This document briefly introduces the architecture for PCE as a 37 central controller, examines the motivations and applicability for 38 PCEP as a southbound interface, and introduces the implications for 39 the protocol. This document does not describe the use cases in 40 detail and does not define protocol extensions: that work is left for 41 other documents. 43 Status of This Memo 45 This Internet-Draft is submitted in full conformance with the 46 provisions of BCP 78 and BCP 79. 48 Internet-Drafts are working documents of the Internet Engineering 49 Task Force (IETF). Note that other groups may also distribute 50 working documents as Internet-Drafts. The list of current Internet- 51 Drafts is at http://datatracker.ietf.org/drafts/current/. 53 Internet-Drafts are draft documents valid for a maximum of six months 54 and may be updated, replaced, or obsoleted by other documents at any 55 time. It is inappropriate to use Internet-Drafts as reference 56 material or to cite them other than as "work in progress." 58 This Internet-Draft will expire on November 25, 2016. 60 Copyright Notice 62 Copyright (c) 2016 IETF Trust and the persons identified as the 63 document authors. All rights reserved. 65 This document is subject to BCP 78 and the IETF Trust's Legal 66 Provisions Relating to IETF Documents 67 (http://trustee.ietf.org/license-info) in effect on the date of 68 publication of this document. Please review these documents 69 carefully, as they describe your rights and restrictions with respect 70 to this document. Code Components extracted from this document must 71 include Simplified BSD License text as described in Section 4.e of 72 the Trust Legal Provisions and are provided without warranty as 73 described in the Simplified BSD License. 75 Table of Contents 77 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 78 2. Architecture . . . . . . . . . . . . . . . . . . . . . . . . 4 79 2.1. Resilience and Scaling . . . . . . . . . . . . . . . . . 7 80 2.1.1. Partitioned Network . . . . . . . . . . . . . . . . . 8 81 2.1.2. Multiple Parallel Controllers . . . . . . . . . . . . 9 82 2.1.3. Hierarchical Controllers . . . . . . . . . . . . . . 10 83 3. Applicability . . . . . . . . . . . . . . . . . . . . . . . . 11 84 3.1. Technology-Oriented Applicability . . . . . . . . . . . . 12 85 3.1.1. Applicability to Control Plane Operated Networks . . 12 86 3.1.2. Static LSPs in MPLS . . . . . . . . . . . . . . . . . 12 87 3.1.3. MPLS Multicast . . . . . . . . . . . . . . . . . . . 13 88 3.1.4. Transport SDN . . . . . . . . . . . . . . . . . . . . 13 89 3.1.5. Segment Routing . . . . . . . . . . . . . . . . . . . 13 90 3.1.6. Service Function Chaining . . . . . . . . . . . . . . 14 91 3.2. High-Level Applicability . . . . . . . . . . . . . . . . 14 92 3.2.1. Traffic Engineering . . . . . . . . . . . . . . . . . 14 93 3.2.2. Traffic Classification . . . . . . . . . . . . . . . 15 94 3.2.3. Service Delivery . . . . . . . . . . . . . . . . . . 15 95 4. Protocol Implications . . . . . . . . . . . . . . . . . . . . 16 96 5. Security Considerations . . . . . . . . . . . . . . . . . . . 16 97 6. Manageability Considerations . . . . . . . . . . . . . . . . 17 98 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 17 99 8. Contributors . . . . . . . . . . . . . . . . . . . . . . . . 17 100 9. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 18 101 10. References . . . . . . . . . . . . . . . . . . . . . . . . . 18 102 10.1. Normative References . . . . . . . . . . . . . . . . . . 18 103 10.2. Informative References . . . . . . . . . . . . . . . . . 18 104 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 20 106 1. Introduction 108 The Path Computation Element (PCE) [RFC4655] was developed to offload 109 path computation function from routers in an MPLS traffic engineered 110 network. Since then, the role and function of the PCE has grown to 111 cover a number of other uses (such as GMPLS [RFC7025]) and to allow 112 delegated control [I-D.ietf-pce-stateful-pce] and PCE-initiated use 113 of network resources [I-D.ietf-pce-pce-initiated-lsp]. 115 According to [RFC7399], Software Defined Networking (SDN) refers to a 116 separation between the control elements and the forwarding components 117 so that software running in a centralized system called a controller, 118 can act to program the devices in the network to behave in specific 119 ways. A required element in an SDN architecture is a component that 120 plans how the network resources will be used and how the devices will 121 be programmed. It is possible to view this component as performing 122 specific computations to place flows within the network given 123 knowledge of the availability of network resources, how other 124 forwarding devices are programmed, and the way that other flows are 125 routed. This is the function and purpose of a PCE, and the way that 126 a PCE integrates into a wider network control system including SDN is 127 presented in [RFC7491]. 129 In early PCE implementations, where the PCE was used to derive paths 130 for MPLS Label Switched Paths (LSPs), paths were requested by network 131 elements and the results of the path computations were supplied to 132 network elements using the Path Computation Element Communication 133 Protocol (PCEP) [RFC5440]. This protocol was later extended to allow 134 a PCE to send unsolicited requests to the network for LSP 135 establishment [I-D.ietf-pce-pce-initiated-lsp]. 137 SDN has a far broader applicability than just signaled MPLS or GMPLS 138 traffic engineered networks. The PCE component in an SDN system may 139 be used to determine paths in a wide range of use cases including 140 static LSPs, segment routing [I-D.ietf-spring-segment-routing], 141 service function chaining (SFC) [RFC7665], and indeed any form of 142 routed or switched network. It is, therefore reasonable to consider 143 PCEP as a general southbound control protocol for use in these 144 environments to allow the PCE to be fully enabled as a central 145 controller. 147 This document introduces the architecture for PCE as a central 148 controller, examines the motivations and applicability for PCEP as a 149 southbound interface, and introduces the implications for the 150 protocol. This document dos not describe the use cases in detail and 151 does not define protocol extensions: that work is left for other 152 documents. 154 2. Architecture 156 The architecture for the use of PCE within centralized control of a 157 network is based on the understanding that a PCE can determine how 158 connections should be placed and how resources should be used within 159 the network, and that the PCE can then cause those connections to be 160 established. Figure 1 shows how this control relationship works in a 161 network with an active control plane. This is a familiar view for 162 those who have read and understood [RFC4655] and 163 [I-D.ietf-pce-pce-initiated-lsp]. 165 In this mode of operation, the central controller is asked to create 166 connectivity by a network orchestrator, a service manager, an 167 Operations Support System (OSS), a Network Management Station (NMS), 168 or some other application. The PCE-based controller computes paths 169 with awareness of the network topology, the available resources, and 170 the other services supported in the network. This information is 171 held in the Traffic Engineering Database (TED) and other databases 172 available to the PCE. Then the PCE sends a request using PCEP to one 173 of the Network Elements (NEs), and that NE uses a control plane to 174 establish the requested connections and reserve the network 175 resources. 177 -------------------------------------------- 178 | Orchestrator / Service Manager / OSS / NMS | 179 -------------------------------------------- 180 ^ 181 | 182 v 183 ------------ 184 | | ----- 185 | PCE-based |<---| TED | 186 | Controller | ----- 187 | | 188 ------------ 189 ^ 190 PCEP| 191 v 192 ---- ---- ---- ---- 193 | NE |<------->| NE |<--->| NE |<--->| NE | 194 ---- Control ---- ---- ---- 195 Plane 197 Figure 1: Architecture for Central Controller with Control Plane 199 Although the architecture shown in Figure 1 represents a form of SDN, 200 one objective of SDN in some environments is to remove the dependency 201 on a control plane. A transition architecture toward this goal is 202 presented in [RFC7491] and is shown in Figure 2. In this case, 203 services are still requested in the same way, and the PCE-based 204 controller still requests use of the network using PCEP. The main 205 difference is that the consumer of the PCEP messages is a Network 206 Controller that provisions the resources and instructs the data plane 207 using Southbound Interface (SBI) that provides an interface to each 208 NE. 210 -------------------------------------------- 211 | Orchestrator / Service Manager / OSS / NMS | 212 -------------------------------------------- 213 ^ 214 | 215 v 216 ------------ 217 | | ----- 218 | PCE-based |<---| TED | 219 | Controller | ----- 220 | | 221 ------------ 222 ^ 223 | PCEP 224 v 225 ------------ 226 | Network | 227 | Controller | 228 /------------\ 229 SBI / ^ ^ \ 230 / | | \ 231 / v v \ 232 ----/ ---- ---- \---- 233 | NE | | NE | | NE | | NE | 234 ---- ---- ---- ---- 236 Figure 2: Architecture Including a Network Controller 238 The approach in Figure 2 delivers the SDN functionality but is overly 239 complicated and insufficiently flexible. 241 o The complication is created by the use of two controllers in a 242 hierarchical organization, and the resultant use of two protocols 243 in a southbound direction. 245 o The lack of flexibility arises from the assumed or required lack 246 of a control plane. 248 This document describes an architecture that reduces the number of 249 components and is flexible to a number of deployment models and use 250 cases. In this hybrid approach (shown in Figure 3) the network 251 controller is PCE-enabled and can also speak PCEP as the SBI (i.e., 252 it can communicate with each node along the path using PCEP). That 253 means that the controller can communicate with a conventional control 254 plane-enabled NE using PCEP and can also use the same protocol to 255 program individual NEs. In this way the PCE-based controller can 256 control a wider range of networks and deliver many different 257 functions as described in Section 3. 259 PCEP is essentially already capable of acting as an SBI and only 260 small, use case- specific modifications to the protocol are needed to 261 support this architecture. The implications for the protocol are 262 discussed further in Section 4. 264 -------------------------------------------- 265 | Orchestrator / Service Manager / OSS / NMS | 266 -------------------------------------------- 267 ^ 268 | 269 v 270 ------------ 271 | | ----- 272 | PCE-based |<---| TED | 273 | Controller | ----- 274 | | 275 /------------\ 276 PCEP / ^ ^ \ 277 / | | \ 278 / v v \ 279 / ---- ---- \ 280 / | NE | | NE | \ 281 ----/ ---- ---- \---- 282 | NE | | NE | 283 ---- ---- 284 ^ ---- ---- ^ 285 :......>| NE |...| NE |<....: 286 Control Plane ---- ---- 288 Figure 3: Architecture for Node-by-Node Central Control 290 2.1. Resilience and Scaling 292 Systems with central controllers are vulnerable to two problems: 293 failure or overload of the single controller. These concerns are not 294 unique to the use of a PCE-based controller but need to be addressed 295 in this document before the PCE-based controller architecture can be 296 considered for use in all but the smallest networks. 298 There are three architectural mechanisms that can be applied to 299 address these issues. The mechanisms are described separately for 300 clarity, but a deployment use may any combination of the approaches. 302 For simplicity of illustration, these three approaches are shown in 303 the sections that follow without a control plane. However, the 304 general, hybrid approach of Figure 3 is applicable in each case. 306 2.1.1. Partitioned Network 308 The first and simplest approach to handling controller overload or 309 scalability is to use multiple controllers, each responsible for a 310 part of the network. We can call the resultant areas of control 311 "domains." 313 This approach is shown in Figure 4. It can clearly address some of 314 the scaling and overload concerns since each controller now only has 315 responsibility for a subset of the network elements. But this comes 316 at a cost because end-to-end connections require coordination between 317 the controllers. Furthermore, this technique does not remove the 318 single-point-of-failure concern even if it does reduce the impact on 319 the network of the failure of a single controller. 321 Note that PCEP is designed to work as a PCE-to-PCE protocol as well 322 as a PCE-to-PCC protocol, so it should be possible to use it to 323 coordinate between PCE-based controllers in this model. 325 -------------------------------------------- 326 | Orchestrator / Service Manager / OSS / NMS | 327 -------------------------------------------- 328 ^ ^ 329 | | 330 v v 331 ------------ Coord- ------------ 332 ----- | | ination | | ----- 333 | TED |--->| PCE-based |<-------->| PCE-based |<---| TED | 334 ----- | Controller | | Controller | ----- 335 | | | | 336 /------------ ------------\ 337 / ^ ^ ^ ^ \ 338 / | | | | \ 339 | | | | | | 340 v v v :: v v v 341 ---- ---- ---- :: ---- ---- ---- 342 | NE | | NE | | NE | :: | NE | | NE | | NE | 343 ---- ---- ---- :: ---- ---- ---- 344 :: 345 Domain 1 :: Domain 2 346 :: 348 Figure 4: Multiple Controllers on a Partitioned Network 350 2.1.2. Multiple Parallel Controllers 352 Multiple parallel controllers may be deployed as shown in Figure 5. 353 Each controller is capable of controlling all of the network elements 354 thus the failure of any one controller will not leave the network 355 unmanageable and, in normal circumstances, the load can be 356 distributed across the controllers. 358 To achieve full redundancy and to be able to continue to provide full 359 function in the event of the failure a controller, the controllers 360 must synchronize with each other. This is nominally a simple task if 361 there are just two controllers, but can actually be quite complex if 362 state changes in the network are not to be lost. Furthermore, if 363 there are more than two controllers, the synchronization between 364 controllers can become a hard problem. 366 Synchronization issues are often off-loaded as "database 367 synchronization" problems because distributed database packages have 368 already had to address these challenges. In networking the problem 369 may also be addressed by collecting the state from the network 370 (effectively using the network as a database) using normal routing 371 protocols such as OSPF, IS-IS, and BGP. 373 -------------------------------------------- 374 | Orchestrator / Service Manager / OSS / NMS | 375 -------------------------------------------- 376 ^ ^ 377 | ___________________ | 378 | | Synchronization | | 379 v v v v 380 ------------ ------------ 381 | | ----- | | 382 | PCE-based |<---| TED |--->| PCE-based | 383 | Controller | ----- | Controller | 384 | |__ ...........| | 385 ------------\ \_:__ :------------ 386 ^ ^ \___: \ .....: ^ ^ 387 | | .....:\ \_:___ ..: : 388 | |__:___ \___:_ \_:___ : 389 | ....: | .....: | ..: | : 390 | : | : | : 391 v v v v v v v v 392 ---- ---- ---- ---- 393 | NE | | NE | | NE | | NE | 394 ---- ---- ---- ---- 396 Figure 5: Multiple Redundant Controllers 398 2.1.3. Hierarchical Controllers 400 Figure 6 shows an approach with hierarchical controllers. This 401 approach was developed for PCEs in [RFC6805] and appears in various 402 SDN architectures where a "parent PCE", an "orchestrator", or "super 403 controller" takes responsibility for a high-level view of the network 404 before distributing tasks to lower level PCEs or controllers. 406 On its own, this approach does little to protect against the failure 407 of a controller, but it can make significant improvements in loading 408 and scaling of the individual controllers. It also offers a good way 409 to support end-to-end connectivity across multiple administrative or 410 technology-specific domains. 412 Note that this model can be arbitrarily recursive with one PCE-based 413 controller acting as the parent of of another set of PCE-based 414 controllers. 416 -------------------------------------------- 417 | Orchestrator / Service Manager / OSS / NMS | 418 -------------------------------------------- 419 ^ 420 | 421 v 422 ------------ 423 | Parent | ----- 424 | PCE-based |<---| TED | 425 | Controller | ----- 426 | | 427 ------------ 428 ^ ^ 429 | | 430 v v 431 ------------ ------------ 432 ----- | | | | ----- 433 | TED |--->| PCE-based | | PCE-based |<---| TED | 434 ----- | Controller | | Controller | ----- 435 /| | | |\ 436 / ------------ ------------ \ 437 / ^ ^ ^ ^ \ 438 / | | | | \ 439 / | | | | \ 440 | | | :: | | | 441 v v v :: v v v 442 ---- ---- ---- :: ---- ---- ---- 443 | NE | | NE | | NE | :: | NE | | NE | | NE | 444 ---- ---- ---- :: ---- ---- ---- 445 :: 446 Domain 1 :: Domain 2 447 :: 449 Figure 6: Hierarchical Controllers 451 3. Applicability 453 This section gives a very high-level introduction to the 454 applicability of a PCE-based centralized controller. There is no 455 attempt to explain each use case in detail, and the inclusion of a 456 use case is not intended to suggest that deploying a PCE-based 457 controller is a mandatory or recommended approach. The sections 458 below are provided as a stimulus to discussion of the applicability 459 of a PCE-based controller and it is expected that separate documents 460 will be written to develop the use cases in which there is interest 461 for implementation and deployment. As described in Section 4 462 specific enhancements to PCEP may be needed for some of these use 463 cases and it is expected that the documents that develop each use 464 case will also address any extensions to PCEP. 466 The rest of this section is divided into two sub-sections. The first 467 approaches the question of applicability from a consideration of the 468 network technology. The second looks at the high-level functions 469 that can be delivered by using a PCE-based controller. 471 As previously mentioned, this section is intended to just make 472 suggestions. Thus the material supplied is very brief. The omission 473 of a use case is in no way meant to imply some limit on the 474 applicability of PCE-based control. 476 3.1. Technology-Oriented Applicability 478 This section provides a list of use cases based on network 479 technology. 481 3.1.1. Applicability to Control Plane Operated Networks 483 This mode of operation is the common approach for an active, stateful 484 PCE to control a traffic engineered MPLS or GMPLS network 485 [I-D.ietf-pce-stateful-pce]. Note that the PCE-based controller 486 determines what LSPs are needed and where to place them. PCEP is 487 used to instruct the head end of each LSP, and the head end signals 488 in the control plane to set up the LSP. 490 3.1.2. Static LSPs in MPLS 492 Static LSPs are provisioned without the use of a control plane. This 493 means that they are established using management plane or "manual" 494 configuration. 496 Static LSPs can be provisioned as 1-hop, micro-LSPs at each node 497 along the path of an end-to-end path LSP. Each router along the path 498 must be told what label forwarding instructions to program and what 499 resources to reserve. The PCE-based controller keeps a view of the 500 network and determines the paths of the end-to-end LSPs just as it 501 does for the use case described in Section 3.1.1, but the controller 502 uses PCEP to communicate with each router along the path of the end- 503 to-end LSP. In this case the PCE-based controller will take 504 responsibility for managing some part of the MPLS label space for 505 each of the routers that it controls, and may taker wider 506 responsibility for partitioning the label space for each router and 507 allocating different parts for different uses communicating the 508 ranges to the router using PCEP. 510 3.1.3. MPLS Multicast 512 Multicast LSPs may be provisioned with a control plane or as static 513 LSPs. No extra considerations apply above those in Section 3.1.1 and 514 Section 3.1.2 except, of course, to note that the PCE must also 515 include the instructions about where the LSP branches, i.e., where 516 packets must be copied. 518 3.1.4. Transport SDN 520 Transport SDN (T-SDN) is the application of SDN techniques to 521 transport networks. In this respect a transport network is a network 522 built from any technology below the IP layer and designed to carry 523 traffic transparently in a connection-oriented way. Thus, an MPLS 524 traffic engineering network is a transport network although it is 525 more common to consider technologies such as Time Division 526 Multiplexing (TDM) and Optical Transport Networks (OTN). 528 Transport networks may be operated with or without a control plane 529 and may have point-to-point or point-to-multipoint connections. 530 Thus, all of the considerations in Section 3.1.1, Section 3.1.2, and 531 Section 3.1.3 apply. It may be the case that additional technology- 532 specific parameters are needed to configure the NEs and these 533 parameters will need to be carried in the PCEP messages. 535 3.1.5. Segment Routing 537 Segment routing is described in [I-D.ietf-spring-segment-routing]. 538 It relies on a series of forwarding instructions being placed in the 539 header or a packet: at each hop in the network a router looks at the 540 first instruction and may continue to forward the packet unchanged, 541 strip the top instruction and forward the packet, or strip the top 542 instruction, insert some additional instructions, and forward the 543 packet. 545 The segment routing architecture supports operations that can be used 546 to steer packet flows in a network thus providing a form of traffic 547 engineering. A PCE-based controller can be responsible for computing 548 the paths for packet flows in a segment routing network, for 549 configuring the forwarding actions on the routers, and for telling 550 the edge routers what instructions to attach to packets as they enter 551 the network. These last two operations can be achieved using PCEP 552 and the PCE-based controller will assume responsibility for managing 553 the space of labels or path identifiers used to determine how packets 554 are forwarded. 556 3.1.6. Service Function Chaining 558 Service Function Chaining (SFC) is described in [RFC7665]. It is the 559 process of directing traffic in a network such that it passes through 560 specific hardware devices or virtual machines (known as service 561 function nodes) that can perform particular desired functions on the 562 traffic. The set of functions to be performed and the locations at 563 which they are to be performed is known as service function chain. 564 Each packet is marked as belonging to a specific chain and that 565 marking lets each successive service function node know which 566 functions to perform and to which service function node to send the 567 packet next. 569 To operate an SFC network the service function nodes must be 570 configured to understand the packet markings and the edge nodes must 571 be told how to mark packets entering the network. Additionally it 572 may be necessary to establish tunnels between service function nodes 573 to carry the traffic. 575 Planning an SFC network requires load balancing between service 576 function nodes and traffic engineering across the network that 577 connects them. These are operations that can be performed by a PCE- 578 based controller, and that controller can use PCEP to program the 579 network and install the service function chains and any required 580 tunnels. 582 3.2. High-Level Applicability 584 This section provides a list of the high-level functions that can be 585 delivered by using a PCE-based controller. 587 3.2.1. Traffic Engineering 589 According to [RFC2702], Traffic Engineering (TE) is concerned with 590 performance optimization of operational networks. In general, it 591 encompasses the application of technology and scientific principles 592 to the measurement, modeling, characterization, control of Internet 593 traffic, and the application of such knowledge and techniques to 594 achieve specific performance objectives. 596 From a practical point of view this involves having an understanding 597 of the topology of the network, the characteristics of the nodes and 598 links in the network, and the traffic demands and flows across the 599 network. It also requires that actions can be taken to ensure that 600 traffic follows specific paths through the network. 602 PCE was specifically developed to address TE in an MPLS network, and 603 so a PCE-based controller is well suited to analyze TE problems and 604 supply answers that can be installed in the network using PCEP. PCEP 605 can be responsible for initiating paths across the network through a 606 control plane, or for installing state in the network node by node 607 such as in a Segment Routed network (see Section 3.1.5) or by 608 configuring IGP metrics. 610 3.2.2. Traffic Classification 612 Traffic classification is an important part of traffic engineering. 613 It is the process of looking at a packet to determine how it should 614 be treated as it is forwarded through the network. It applies in 615 many scenarios including MPLS traffic engineering (where it 616 determines what traffic is forwarded onto which LSPs), segment 617 routing (where it is used to select which set of forwarding 618 instructions to add to a packet), and service function chaining 619 (where it indicates along which service function chain a packet 620 should be forwarded). 622 Traffic classification is closely linked to the computational 623 elements of planning for the network functions just listed because it 624 determines how traffic load is balanced and distributed through the 625 network. Therefore, selecting what traffic classification should be 626 performed by a router is an important part of the work done by a PCE- 627 based controller. 629 Instructions can be passed from the controller to the routers using 630 PCEP. These instructions tell the routers how to map traffic to 631 paths or connections. The instructions may use the concept of a 632 Forwarding Equivalence Class (FEC). 634 3.2.3. Service Delivery 636 Various network services may be offered over a network. These 637 include protection services (including end-to-end protection 638 [RFC4427], restoration after failure, and fast reroute [RFC4090]), 639 Virtual Private Network (VPN) service (such as Layer 3 VPNs [RFC4364] 640 or Ethernet VPNs [RFC7432]), or Pseudowires [RFC3985]. 642 Delivering services over a network in an optimal way requires 643 coordination in the way that network resources are allocated to 644 support the services. A PCE-based central controller can consider 645 the whole network and all components of a service at once when 646 planning how to deliver the service. It can then use PCEP to manage 647 the network resources and to install the necessary associations 648 between those resources. 650 4. Protocol Implications 652 PCEP is push-pull protocol that is designed to move requests and 653 responses between a server (the PCE) and Path Computation Clients 654 (PCCs - the network elements). In particular, it has a message 655 (PCInitiate [I-D.ietf-pce-pce-initiated-lsp]) that can be sent by the 656 PCE to install state or cause actions at the PCC, and a response 657 message (PCRpt) that is used to confirm the request. 659 As such, there is an expectation that only relatively minor changes 660 to PCEP are required to support the concept of a PCE-based 661 controller. The only work expected to be needed is small extensions 662 to carry additional or specific information elements for the 663 individual use cases. Where possible, consistent with the general 664 principles of how protocols are extended, any additions to the 665 protocol should be made in a generic way such that they are open to 666 use in a range of applications. 668 It is anticipated that new documents will be produced for each use 669 case dependent on support and demand. Such documents will explain 670 the use case and define the necessary protocol extensions. 672 5. Security Considerations 674 Security considerations for a PCE-based controller are little 675 different from those for any other PCE system. That is, the 676 operation relies heavily on the use and security of PCEP and so 677 consideration should be given to the security features discussed in 678 [RFC5440] and the additional mechanisms described in 679 [I-D.ietf-pce-pceps]. 681 It should be observed that the trust model of a network that operates 682 with out a control plane is different from one with a control plane. 683 The conventional "chain of trust" used with a control plane is 684 replaced by individual trust relationships between the controller and 685 each individual NE. This model may be considerably easier to manage 686 and so is more likely to be operated with a high level of security. 687 However debate will rage over overall system security and the 688 opportunity for attacks in an architecture with a central controller 689 since the network can be vulnerable to denial of service attacks on 690 the controller, and the forwarding system may be harmed by attacks on 691 the messages sent to individual routers. In short, while the 692 interactions with a PCE-based controller are not substantially 693 different from those in any other SDN architecture, the security 694 implications of SDN are still open for discussion. The IRTF's SDN 695 Research Group (SDNRG) continues to discuss this topic. 697 It is expected that each new document that is produced for a specific 698 use case will also include considerations of the security impacts of 699 the use of a PCE-based central controller on the network type and 700 services being managed. 702 6. Manageability Considerations 704 The architecture described in this document is a management 705 architecture: the PCE-based controller is a management component that 706 controls the network through a southbound management protocol (PCEP). 708 RFC 5440 [RFC5440] contains a substantive manageability 709 considerations section that examines how a PCE-based system and a 710 PCE-enabled system may be managed. A MIB module for PCEP was 711 published as RFC 7420 [RFC7420] and a YANG module for PCEP has also 712 been proposed [I-D.pkd-pce-pcep-yang]. 714 7. IANA Considerations 716 This document makes no requests for IANA action. 718 8. Contributors 720 The following people contributed to discussions that led to the 721 development of this document: 723 Cyril Margaria 724 Email: cmargaria@juniper.net 726 Sudhir Cheruathur 727 Email: scheruathur@juniper.net 729 Dhruv Dhody 730 Email: dhruv.dhody@huawei.com 732 Daniel King 733 Email: daniel@olddog.co.uk 735 Iftekhar Hussain 736 Email: IHussain@infinera.com 738 Anurag Sharma 739 Email: AnSharma@infinera.com 741 Eric Wu 742 Email: eric.wu@huawei.com 744 9. Acknowledgements 746 The ideas in this document owe a lot to the work started by the 747 authors of [I-D.zhao-teas-pcecc-use-cases] and 748 [I-D.zhao-pce-pcep-extension-for-pce-controller]. The authors of 749 this document fully acknowledge the prior work and thank those 750 involved for opening the discussion. The individuals concerned are: 751 King Ke, Luyuan Fang, Chao Zhou, Boris Zhang, Zhenbin Li. 753 This document has benefited from the discussions within a small ad 754 hoc design team the members of which are listed as document 755 contributors. 757 Thanks to Michael Scharf and Andy Malis for a lively discussion of 758 this document. 760 10. References 762 10.1. Normative References 764 [RFC4655] Farrel, A., Vasseur, J., and J. Ash, "A Path Computation 765 Element (PCE)-Based Architecture", RFC 4655, 766 DOI 10.17487/RFC4655, August 2006, 767 . 769 10.2. Informative References 771 [I-D.ietf-pce-pce-initiated-lsp] 772 Crabbe, E., Minei, I., Sivabalan, S., and R. Varga, "PCEP 773 Extensions for PCE-initiated LSP Setup in a Stateful PCE 774 Model", draft-ietf-pce-pce-initiated-lsp-05 (work in 775 progress), October 2015. 777 [I-D.ietf-pce-pceps] 778 Lopez, D., Dios, O., Wu, W., and D. Dhody, "Secure 779 Transport for PCEP", draft-ietf-pce-pceps-09 (work in 780 progress), March 2016. 782 [I-D.ietf-pce-stateful-pce] 783 Crabbe, E., Minei, I., Medved, J., and R. Varga, "PCEP 784 Extensions for Stateful PCE", draft-ietf-pce-stateful- 785 pce-14 (work in progress), March 2016. 787 [I-D.ietf-spring-segment-routing] 788 Filsfils, C., Previdi, S., Decraene, B., Litkowski, S., 789 and R. Shakir, "Segment Routing Architecture", draft-ietf- 790 spring-segment-routing-08 (work in progress), May 2016. 792 [I-D.pkd-pce-pcep-yang] 793 Dhody, D., Hardwick, J., Beeram, V., and J. Tantsura, "A 794 YANG Data Model for Path Computation Element 795 Communications Protocol (PCEP)", draft-pkd-pce-pcep- 796 yang-05 (work in progress), January 2016. 798 [I-D.zhao-pce-pcep-extension-for-pce-controller] 799 Zhao, Q., Li, Z., Dhody, D., and C. Zhou, "PCEP Procedures 800 and Protocol Extensions for Using PCE as a Central 801 Controller (PCECC) of LSPs", draft-zhao-pce-pcep- 802 extension-for-pce-controller-03 (work in progress), March 803 2016. 805 [I-D.zhao-teas-pcecc-use-cases] 806 Zhao, Q., Li, Z., Ke, Z., Fang, L., Zhou, C., and T. 807 Communications, "The Use Cases for Using PCE as the 808 Central Controller(PCECC) of LSPs", draft-zhao-teas-pcecc- 809 use-cases-00 (work in progress), March 2016. 811 [RFC2702] Awduche, D., Malcolm, J., Agogbua, J., O'Dell, M., and J. 812 McManus, "Requirements for Traffic Engineering Over MPLS", 813 RFC 2702, DOI 10.17487/RFC2702, September 1999, 814 . 816 [RFC3985] Bryant, S., Ed. and P. Pate, Ed., "Pseudo Wire Emulation 817 Edge-to-Edge (PWE3) Architecture", RFC 3985, 818 DOI 10.17487/RFC3985, March 2005, 819 . 821 [RFC4090] Pan, P., Ed., Swallow, G., Ed., and A. Atlas, Ed., "Fast 822 Reroute Extensions to RSVP-TE for LSP Tunnels", RFC 4090, 823 DOI 10.17487/RFC4090, May 2005, 824 . 826 [RFC4364] Rosen, E. and Y. Rekhter, "BGP/MPLS IP Virtual Private 827 Networks (VPNs)", RFC 4364, DOI 10.17487/RFC4364, February 828 2006, . 830 [RFC4427] Mannie, E., Ed. and D. Papadimitriou, Ed., "Recovery 831 (Protection and Restoration) Terminology for Generalized 832 Multi-Protocol Label Switching (GMPLS)", RFC 4427, 833 DOI 10.17487/RFC4427, March 2006, 834 . 836 [RFC5440] Vasseur, JP., Ed. and JL. Le Roux, Ed., "Path Computation 837 Element (PCE) Communication Protocol (PCEP)", RFC 5440, 838 DOI 10.17487/RFC5440, March 2009, 839 . 841 [RFC6805] King, D., Ed. and A. Farrel, Ed., "The Application of the 842 Path Computation Element Architecture to the Determination 843 of a Sequence of Domains in MPLS and GMPLS", RFC 6805, 844 DOI 10.17487/RFC6805, November 2012, 845 . 847 [RFC7025] Otani, T., Ogaki, K., Caviglia, D., Zhang, F., and C. 848 Margaria, "Requirements for GMPLS Applications of PCE", 849 RFC 7025, DOI 10.17487/RFC7025, September 2013, 850 . 852 [RFC7399] Farrel, A. and D. King, "Unanswered Questions in the Path 853 Computation Element Architecture", RFC 7399, 854 DOI 10.17487/RFC7399, October 2014, 855 . 857 [RFC7420] Koushik, A., Stephan, E., Zhao, Q., King, D., and J. 858 Hardwick, "Path Computation Element Communication Protocol 859 (PCEP) Management Information Base (MIB) Module", 860 RFC 7420, DOI 10.17487/RFC7420, December 2014, 861 . 863 [RFC7432] Sajassi, A., Ed., Aggarwal, R., Bitar, N., Isaac, A., 864 Uttaro, J., Drake, J., and W. Henderickx, "BGP MPLS-Based 865 Ethernet VPN", RFC 7432, DOI 10.17487/RFC7432, February 866 2015, . 868 [RFC7491] King, D. and A. Farrel, "A PCE-Based Architecture for 869 Application-Based Network Operations", RFC 7491, 870 DOI 10.17487/RFC7491, March 2015, 871 . 873 [RFC7665] Halpern, J., Ed. and C. Pignataro, Ed., "Service Function 874 Chaining (SFC) Architecture", RFC 7665, 875 DOI 10.17487/RFC7665, October 2015, 876 . 878 Authors' Addresses 880 Adrian Farrel (editor) 881 Juniper Networks 883 Email: adrian@olddog.co.uk 884 Quintin Zhao (editor) 885 Huawei Technologies 886 125 Nagog Technology Park 887 Acton, MA 01719 888 USA 890 Email: quintin.zhao@huawei.com 892 Robin Li 893 Huawei Technologies 894 Huawei Bld., No.156 Beiqing Road 895 Beijing 100095 896 China 898 Email: lizhenbin@huawei.com 900 Chao Zhou 901 Cisco Systems 903 Email: chao.zhou@cisco.com