idnits 2.17.1 draft-ietf-teas-pce-native-ip-12.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 (October 28, 2020) is 1274 days in the past. Is this intentional? Checking references for intended status: Experimental ---------------------------------------------------------------------------- ** Obsolete normative reference: RFC 7752 (Obsoleted by RFC 9552) == Outdated reference: A later version (-30) exists of draft-ietf-pce-pcep-extension-native-ip-09 Summary: 1 error (**), 0 flaws (~~), 2 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 TEAS Working Group A. Wang 3 Internet-Draft China Telecom 4 Intended status: Experimental B. Khasanov 5 Expires: May 1, 2021 Huawei Technologies 6 Q. Zhao 7 Etheric Networks 8 H. Chen 9 Futurewei 10 October 28, 2020 12 PCE in Native IP Network 13 draft-ietf-teas-pce-native-ip-12 15 Abstract 17 This document defines the architecture for traffic engineering within 18 native IP network, using multiple BGP sessions strategy and PCE 19 -based central control mechanism. It uses the Central Control 20 Dynamic Routing (CCDR) procedures described in this document, and the 21 Path Computation Element Communication Protocol (PCEP) extension 22 specified in draft ietf-pce-pcep-extension-native-ip. 24 Status of This Memo 26 This Internet-Draft is submitted in full conformance with the 27 provisions of BCP 78 and BCP 79. 29 Internet-Drafts are working documents of the Internet Engineering 30 Task Force (IETF). Note that other groups may also distribute 31 working documents as Internet-Drafts. The list of current Internet- 32 Drafts is at https://datatracker.ietf.org/drafts/current/. 34 Internet-Drafts are draft documents valid for a maximum of six months 35 and may be updated, replaced, or obsoleted by other documents at any 36 time. It is inappropriate to use Internet-Drafts as reference 37 material or to cite them other than as "work in progress." 39 This Internet-Draft will expire on May 1, 2021. 41 Copyright Notice 43 Copyright (c) 2020 IETF Trust and the persons identified as the 44 document authors. All rights reserved. 46 This document is subject to BCP 78 and the IETF Trust's Legal 47 Provisions Relating to IETF Documents 48 (https://trustee.ietf.org/license-info) in effect on the date of 49 publication of this document. Please review these documents 50 carefully, as they describe your rights and restrictions with respect 51 to this document. Code Components extracted from this document must 52 include Simplified BSD License text as described in Section 4.e of 53 the Trust Legal Provisions and are provided without warranty as 54 described in the Simplified BSD License. 56 Table of Contents 58 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 59 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3 60 3. CCDR Architecture in Simple Topology . . . . . . . . . . . . 4 61 4. CCDR Architecture in Large Scale Topology . . . . . . . . . . 5 62 5. CCDR Multiple BGP Sessions Strategy . . . . . . . . . . . . . 6 63 6. PCEP Extension for Key Parameters Delivery . . . . . . . . . 8 64 7. Deployment Consideration . . . . . . . . . . . . . . . . . . 9 65 7.1. Scalability . . . . . . . . . . . . . . . . . . . . . . . 9 66 7.2. High Availability . . . . . . . . . . . . . . . . . . . . 9 67 7.3. Incremental deployment . . . . . . . . . . . . . . . . . 10 68 8. Security Considerations . . . . . . . . . . . . . . . . . . . 10 69 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 10 70 10. Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . 11 71 11. References . . . . . . . . . . . . . . . . . . . . . . . . . 11 72 11.1. Normative References . . . . . . . . . . . . . . . . . . 11 73 11.2. Informative References . . . . . . . . . . . . . . . . . 12 74 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 12 76 1. Introduction 78 [RFC8735] describes the scenarios and simulation results for traffic 79 engineering in the native IP network to provide End-to-End (E2E) 80 performance assurance and QoS using PCE based centralized control, 81 referred to as Centralized Control Dynamic Routing (CCDR). Based on 82 the various scenarios and analysis as per [RFC8735], the solution for 83 traffic engineering in native IP network should meet the following 84 criteria: 86 o Same solution for native IPv4 and IPv6 traffic. 88 o Support for intra-domain and inter-domain scenarios. 90 o Achieve End to End traffic assurance, with determined QoS 91 behavior. 93 o No changes in routers forwarding behavior. 95 o Capability to use the power of centrally control and the 96 flexibility/robustness of distributed network control plane. 98 o Different network requirements such as large traffic amount and 99 prefix scale. 101 o Adjusting the optimal path dynamically upon the change of network 102 status. No need for physical links resources reservation in 103 advance. 105 Stateful PCE [RFC8231] specifies a set of extensions to PCEP to 106 enable stateful control of paths such as MPLS-TE Label Switched 107 Paths(LSP)s between and across PCEP sessions in compliance with 108 [RFC4657]. It includes mechanisms to achieve state synchronization 109 between Path Computation Clients(PCCs) and PCEs, delegation of 110 control of LSPs to PCEs, and PCE control of timing and sequence of 111 path computations within and across PCEP sessions. Furthermore, 112 [RFC8281] specifies a mechanism to dynamically instantiate LSPs on a 113 PCC based on the requests from a stateful PCE or a controller using 114 stateful PCE. [RFC8283] introduces the architecture for PCE as a 115 central controller as an extension of the architecture described in 116 [RFC4655] and assumes the continued use of PCEP as the protocol used 117 between PCE and PCC.[RFC8283] further examines the motivations and 118 applicability for PCEP as a Southbound Interface (SBI), and 119 introduces the implications for the protocol. 121 This document defines the architecture for traffic engineering within 122 native IP network, using multiple BGP session strategy, to meet the 123 above criteria in dynamical and centrally control mode. The 124 architecture is referred as CCDR architecture. It depends on the 125 central control (PCE) element to compute the optimal path for 126 selected traffic, and utilizes the dynamic routing behavior of 127 traditional IGP/BGP protocols to forward such traffic. 129 The control messages between PCE and underlying network node are 130 transmitted via Path Computation Element Communications Protocol 131 (PCEP) protocol. The related PCEP extensions are provided in draft 132 [I-D.ietf-pce-pcep-extension-native-ip]. 134 2. Terminology 136 This document uses the following terms defined in [RFC5440]: 138 o PCE 140 o PCEP 142 o PCC 144 Other terms are defined in this document: 146 o CCDR: Central Control Dynamic Routing 148 o E2E: End to End 150 o ECMP: Equal-Cost Multipath 152 o RR: Route Reflector 154 o SDN: Software Defined Network 156 3. CCDR Architecture in Simple Topology 158 Figure 1 illustrates the CCDR architecture for traffic engineering in 159 simple topology. The topology is comprised by four devices which are 160 SW1, SW2, R1, R2. There are multiple physical links between R1 and 161 R2. Traffic between prefix PF11(on SW1) and prefix PF21(on SW2) is 162 normal traffic, traffic between prefix PF12(on SW1) and prefix 163 PF22(on SW2) is priority traffic that should be treated accordingly. 165 +-----+ 166 +----------+ PCE +--------+ 167 | +-----+ | 168 | | 169 | BGP Session 1(lo11/lo21)| 170 +-------------------------+ 171 | | 172 | BGP Session 2(lo12/lo22)| 173 +-------------------------+ 174 PF12 | | PF22 175 PF11 | | PF21 176 +---+ +-----+-----+ +-----+-----+ +---+ 177 |SW1+---------+(lo11/lo12)+-------------+(lo21/lo22)+--------------+SW2| 178 +---+ | R1 +-------------+ R2 | +---+ 179 +-----------+ +-----------+ 181 Figure 1: CCDR architecture in simple topology 183 In Intra-AS scenario, IGP and BGP are deployed between R1 and R2. In 184 inter-AS scenario, only native BGP protocol is deployed. The traffic 185 between each address pair may change in real time and the 186 corresponding source/destination addresses of the traffic may also 187 change dynamically. 189 The key ideas of the CCDR architecture for this simple topology are 190 the followings: 192 o Build two BGP sessions between R1 and R2, via the different 193 loopback addresses on these routers. 195 o Set the explicit peer route on R1 and R2 respectively for BGP next 196 hop to different physical link addresses between R1 and R2. Such 197 explicit peer route can be set in the format of static route to 198 BGP peer address, which is different from the route learned from 199 the IGP protocol. 201 o Send different prefixes via the established BGP sessions. For 202 example, PF11/PF21 via the BGP session 1 and PF12/PF22 via the BGP 203 session 2. 205 After the above actions, the bi-direction traffic between the PF11 206 and PF21, and the bi-direction traffic between PF12 and PF22 will go 207 through different physical links between R1 and R2. 209 If there is more traffic between PF12 and PF22 that needs to be 210 assured , one can add more physical links between R1 and R2 to reach 211 the the next hop for BGP session 2. In this cases the prefixes that 212 advertised by the BGP peers need not be changed. 214 If, for example, there is bi-directional traffic from another address 215 pair that needs to be assured (for example prefix PF13/PF23), and the 216 total volume of assured traffic does not exceed the capacity of the 217 previously provisioned physical links, one need only to advertise the 218 newly added source/destination prefixes via the BGP session 2. The 219 bi-direction traffic between PF13/PF23 will go through the assigned 220 dedicated physical links as the traffic between PF12/PF22. 222 Such decouple philosophy achieves the flexible control capability for 223 the network traffic, to achieve the determined QoS assurance effect 224 to meet the application's requirement. The router needs only support 225 native IP and multiple BGP sessions setup via different loopback 226 addresses. 228 4. CCDR Architecture in Large Scale Topology 230 When the assured traffic spans across the large scale network, as 231 that illustrated in Figure 2, the multiple BGP sessions cannot be 232 established hop by hop, especially for the iBGP within one AS. 234 For such scenario, we should consider using the Route Reflector (RR) 235 [RFC4456] to achieve the similar effect. Every edge router will 236 establish two BGP sessions with the RR via different loopback 237 addresses respectively. The other steps for traffic differentiation 238 are same as that described in the CCDR architecture for simple 239 topology. 241 As shown in Figure 2, if we select R3 as the RR, every edge router(R1 242 and R7 in this example) will build two BGP session with the RR. If 243 the PCE selects the dedicated path as R1-R2-R4-R7, then the operator 244 should set the explicit peer routes via PCEP protocol on these 245 routers respectively, pointing to the BGP next hop (loopback 246 addresses of R1 and R7, which are used to send the prefix of the 247 assured traffic) to the selected forwarding address. 249 +-----+ 250 +----------------+ PCE +------------------+ 251 | +--+--+ | 252 | | | 253 | | | 254 | ++-+ | 255 +------------------+R3+-------------------+ 256 PF12 | +--+ | PF22 257 PF11 | | PF21 258 +---+ ++-+ +--+ +--+ +-++ +---+ 259 |SW1+-------+R1+----------+R5+----------+R6+---------+R7+--------+SW2| 260 +---+ ++-+ +--+ +--+ +-++ +---+ 261 | | 262 | | 263 | +--+ +--+ | 264 +------------+R2+----------+R4+-----------+ 265 +--+ +--+ 266 Figure 2: CCDR architecture in large scale network 268 5. CCDR Multiple BGP Sessions Strategy 270 In general situation, different applications may require different 271 QoS criteria, which may include: 273 o Traffic that requires low latency and is not sensitive to packet 274 loss. 276 o Traffic that requires low packet loss and can endure higher 277 latency. 279 o Traffic that requires low jitter. 281 These different traffic requirements can be summarized in the 282 following table: 284 +----------------+-------------+---------------+-----------------+ 285 | Prefix Set No. | Latency | Packet Loss | Jitter | 286 +----------------+-------------+---------------+-----------------+ 287 | 1 | Low | Normal | Don't care | 288 +----------------+-------------+---------------+-----------------+ 289 | 2 | Normal | Low | Dont't care | 290 +----------------+-------------+---------------+-----------------+ 291 | 3 | Normal | Normal | Low | 292 +----------------+-------------+---------------+-----------------+ 293 Table 1. Traffic Requirement Criteria 295 For Prefix Set No.1, we can select the shortest distance path to 296 carry the traffic; for Prefix Set No.2, we can select the path that 297 has end to end under-loading links; For Prefix Set No.3, we can let 298 all assured traffic pass the determined single path, no Equal Cost 299 Multipath (ECMP) distribution on the parallel links is desired. 301 It is almost impossible to provide an End-to-End (E2E) path 302 efficiently with latency, jitter, packet loss constraints to meet the 303 above requirements in large scale IP-based network via the 304 distributed routing protocol, but these requirements can be solved 305 with the assistance of PCE, as that described in [RFC4655] and 306 [RFC8283] because the PCE has the overall network view, can collect 307 real network topology and network performance information about the 308 underlying network, select the appropriate path to meet various 309 network performance requirements of different traffics. 311 The architecture to implement the CCDR Multiple BGP sessions strategy 312 is the followings: 314 Here PCE is the main component of the Software Definition Network 315 (SDN) controller and is responsible for optimal path computation for 316 priority traffic. 318 o SDN controller gets topology via BGP-LS [RFC7752] and link 319 utilization information via existing Network Monitor System (NMS) 320 from the underlying network. 322 o PCE calculates the appropriate path upon application's 323 requirements, sends the key parameters to edge/RR routers(R1, R7 324 and R3 in Fig.3) to establish multiple BGP sessions. The loopback 325 addresses used for BGP sessions should be planned in advance and 326 distributed in the domain. 328 o PCE sends the route information to the routers (R1,R2,R4,R7 in 329 Fig.3) on forwarding path via PCEP 330 [I-D.ietf-pce-pcep-extension-native-ip] , to build the path to the 331 BGP next-hop of the advertised prefixes. 333 o PCE send the prefixes information to the PCC to let them 334 advertises different prefixes via the specified BGP session. 336 o If the assured traffic prefixes were changed but the total volume 337 of assured traffic does not exceed the physical capacity of the 338 previous E2E path, PCE needs only change the prefixed advertised 339 via the edge routers (R1,R7 in Fig.3). 341 o If the volume of assured traffic exceeds the capacity of previous 342 calculated path, PCE can recalculate and add the appropriate paths 343 to accommodate the exceeding traffic. After that, PCE needs to 344 update on-path routers to build the forwarding path hop by hop. 346 +------------+ 347 | Application| 348 +------+-----+ 349 | 350 +--------+---------+ 351 +----------+SDN Controller/PCE+-----------+ 352 | +--------^---------+ | 353 | | | 354 | | | 355 PCEP | BGP-LS|PCEP | PCEP 356 | | | 357 | +v-+ | 358 +------------------+R3+-------------------+ 359 PF12 | +--+ | PF22 360 PF11 | | PF21 361 +---+ +v-+ +--+ +--+ +-v+ +---+ 362 |SW1+-------+R1+----------+R5+----------+R6+---------+R7+--------+SW2| 363 +---+ ++-+ +--+ +--+ +-++ +---+ 364 | | 365 | | 366 | +--+ +--+ | 367 +------------+R2+----------+R4+-----------+ 369 Figure 3: CCDR architecture for Multi-BGP deployment 371 6. PCEP Extension for Key Parameters Delivery 373 The PCEP protocol needs to be extended to transfer the following key 374 parameters: 376 o Peer information that is used to build the BGP session 378 o Explicit route information to BGP next hop of advertised prefixes 380 o Advertised prefixes and their associated BGP session. 382 Once the router receives such information, it should establish the 383 BGP session with the peer appointed in the PCEP message, build the 384 end to end dedicated path hop by hop and advertise the prefixes that 385 contained in the corresponding PCEP message. 387 The dedicated path is preferred by making sure that the explicit 388 route created by PCE has the higher priority (lower route preference) 389 than the route information created by other dynamic protocols. 391 All above dynamically created states (BGP sessions, Explicit route, 392 Prefix advertised prefix, ) will be cleared on the expiration of 393 state timeout interval which is based on the existing Stateful PCE 394 [RFC8231] and PCECC [RFC8283] mechanism. 396 Details of communications between PCEP and BGP subsystems in router's 397 control plane are out of scope of this draft and will be described in 398 separate draft [I-D.ietf-pce-pcep-extension-native-ip] . 400 7. Deployment Consideration 402 7.1. Scalability 404 In CCDR architecture, PCE needs only influence the edge routers for 405 the prefixes advertisement via the multiple BGP sessions deployment. 406 The route information for these prefixes within the on-path routers 407 were distributed via the BGP protocol. 409 For multiple domains deployment, the PCE or the pool of PCEs that 410 responsible for these domains need only control the edge router to 411 build multiple EBGP sessions, all other procedures are the same that 412 in one domain. 414 Unlike the solution from BGP Flowspec, the on-path router need only 415 keep the specific policy routes for the BGP next-hop of the 416 differentiate prefixes, not the specific routes to the prefixes 417 themselves. This can lessen the burden from the table size of policy 418 based routes for the on-path routers, and has more expandability when 419 comparing with BGP flowspec or Openflow solution. For example, if we 420 want to differentiate 1000 prefixes from the normal traffic, CCDR 421 needs only one explicit peer route in every on-path router, but the 422 BGP flowspec or Openflow needs 1000 policy routes on them. 424 7.2. High Availability 426 The CCDR architecture is based on the distributed IP protocol. If 427 the PCE failed, the forwarding plane will not be impacted, as the BGP 428 session between all devices will not flap, and the forwarding table 429 will remain unchanged. 431 If one node on the optimal path is failed, the priority traffic will 432 fall over to the best-effort forwarding path. One can even design 433 several assurance paths to load balance/hot-standby the priority 434 traffic to meet the path failure situation. 436 For high availability of PCE/SDN-controller, operator should rely on 437 existing high availability solutions for SDN controller, such as 438 clustering technology and deployment. 440 7.3. Incremental deployment 442 Not every router within the network will support the PCEP extension 443 that defined in [I-D.ietf-pce-pcep-extension-native-ip] 444 simultaneously. 446 For such situations, router on the edge of domain can be upgraded 447 first, and then the traffic can be assured between different domains. 448 Within each domain, the traffic will be forwarded along the best- 449 effort path. Service provider can selectively upgrade the routers on 450 each domain in sequence. 452 8. Security Considerations 454 A PCE needs to assure calculation of E2E path based on the status of 455 network and the service requirements in real-time. 457 The PCE needs consider the explicit route deployment order (for 458 example, from tail router to head router) to eliminate the possible 459 transient traffic loop. 461 The setup of BGP session, prefix advertisement and explicit peer 462 route establishment are all controlled by the PCE. To prevent the 463 bogus PCE to send harmful messages to the network nodes, the network 464 devices should authenticate the validity of PCE and keep secures 465 communication channel between them. Mechanism described in [RFC8253] 466 should be used to avoid such situation. 468 CCDR architecture does not require the change of forward behavior on 469 the underlay devices, then there will no additional security impact 470 on the devices. 472 9. IANA Considerations 474 This document does not require any IANA actions. 476 10. Acknowledgement 478 The author would like to thank Deborah Brungard, Adrian Farrel, 479 Vishnu Beeram, Lou Berger, Dhruv Dhody, Raghavendra Mallya , Mike 480 Koldychev, Haomian Zheng, Penghui Mi, Shaofu Peng and Jessica Chen 481 for their supports and comments on this draft. 483 11. References 485 11.1. Normative References 487 [RFC4456] Bates, T., Chen, E., and R. Chandra, "BGP Route 488 Reflection: An Alternative to Full Mesh Internal BGP 489 (IBGP)", RFC 4456, DOI 10.17487/RFC4456, April 2006, 490 . 492 [RFC4655] Farrel, A., Vasseur, J., and J. Ash, "A Path Computation 493 Element (PCE)-Based Architecture", RFC 4655, 494 DOI 10.17487/RFC4655, August 2006, 495 . 497 [RFC4657] Ash, J., Ed. and J. Le Roux, Ed., "Path Computation 498 Element (PCE) Communication Protocol Generic 499 Requirements", RFC 4657, DOI 10.17487/RFC4657, September 500 2006, . 502 [RFC5440] Vasseur, JP., Ed. and JL. Le Roux, Ed., "Path Computation 503 Element (PCE) Communication Protocol (PCEP)", RFC 5440, 504 DOI 10.17487/RFC5440, March 2009, 505 . 507 [RFC7752] Gredler, H., Ed., Medved, J., Previdi, S., Farrel, A., and 508 S. Ray, "North-Bound Distribution of Link-State and 509 Traffic Engineering (TE) Information Using BGP", RFC 7752, 510 DOI 10.17487/RFC7752, March 2016, 511 . 513 [RFC8231] Crabbe, E., Minei, I., Medved, J., and R. Varga, "Path 514 Computation Element Communication Protocol (PCEP) 515 Extensions for Stateful PCE", RFC 8231, 516 DOI 10.17487/RFC8231, September 2017, 517 . 519 [RFC8253] Lopez, D., Gonzalez de Dios, O., Wu, Q., and D. Dhody, 520 "PCEPS: Usage of TLS to Provide a Secure Transport for the 521 Path Computation Element Communication Protocol (PCEP)", 522 RFC 8253, DOI 10.17487/RFC8253, October 2017, 523 . 525 [RFC8281] Crabbe, E., Minei, I., Sivabalan, S., and R. Varga, "Path 526 Computation Element Communication Protocol (PCEP) 527 Extensions for PCE-Initiated LSP Setup in a Stateful PCE 528 Model", RFC 8281, DOI 10.17487/RFC8281, December 2017, 529 . 531 [RFC8283] Farrel, A., Ed., Zhao, Q., Ed., Li, Z., and C. Zhou, "An 532 Architecture for Use of PCE and the PCE Communication 533 Protocol (PCEP) in a Network with Central Control", 534 RFC 8283, DOI 10.17487/RFC8283, December 2017, 535 . 537 [RFC8735] Wang, A., Huang, X., Kou, C., Li, Z., and P. Mi, 538 "Scenarios and Simulation Results of PCE in a Native IP 539 Network", RFC 8735, DOI 10.17487/RFC8735, February 2020, 540 . 542 11.2. Informative References 544 [I-D.ietf-pce-pcep-extension-native-ip] 545 Wang, A., Khasanov, B., Fang, S., Tan, R., and C. Zhu, 546 "PCEP Extension for Native IP Network", draft-ietf-pce- 547 pcep-extension-native-ip-09 (work in progress), October 548 2020. 550 Authors' Addresses 552 Aijun Wang 553 China Telecom 554 Beiqijia Town, Changping District 555 Beijing 102209 556 China 558 Email: wangaj3@chinatelecom.cn 560 Boris Khasanov 561 Huawei Technologies 562 Moskovskiy Prospekt 97A 563 St.Petersburg 196084 564 Russia 566 Email: bhassanov@yahoo.com 567 Quintin Zhao 568 Etheric Networks 569 1009 S CLAREMONT ST 570 SAN MATEO, CA 94402 571 USA 573 Email: qzhao@ethericnetworks.com 575 Huaimo Chen 576 Futurewei 577 Boston, MA 578 USA 580 Email: huaimo.chen@futurewei.com