idnits 2.17.00 (12 Aug 2021) /tmp/idnits16702/draft-ietf-sfc-nsh-ecn-support-07.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 == Line 276 has weird spacing: '... sender doma...' == Line 277 has weird spacing: '...ingress v |...' -- The document date (October 6, 2021) is 227 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) No issues found here. Summary: 0 errors (**), 0 flaws (~~), 3 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 1 INTERNET-DRAFT D. Eastlake 2 Intended status: Proposed Standard Futurewei Technologies 3 B. Briscoe 4 Independent 5 Y. Li 6 Huawei Technologies 7 A. Malis 8 Malis Consulting 9 X. Wei 10 Huawei Technologies 11 Expires: April 5, 2022 October 6, 2021 13 Explicit Congestion Notification (ECN) and Congestion Feedback 14 Using the Network Service Header (NSH) and IPFIX 15 17 Abstract 19 Explicit congestion notification (ECN) allows a forwarding element to 20 notify downstream devices of the onset of congestion without having 21 to drop packets. Coupled with a means to feed information about 22 congestion back to upstream nodes, this can improve network 23 efficiency through better congestion control, frequently without 24 packet drops. This document specifies ECN and congestion feedback 25 support within a Service Function Chaining (SFC) architecture domain 26 through use of the Network Service Header (NSH, RFC 8300) and IP Flow 27 Information Export (IPFIX, RFC 7011). 29 Status of This Memo 31 This Internet-Draft is submitted in full conformance with the 32 provisions of BCP 78 and BCP 79. 34 Distribution of this document is unlimited. Comments should be sent 35 to the SFC Working Group mailing list or to the 36 authors. 38 Internet-Drafts are working documents of the Internet Engineering 39 Task Force (IETF), its areas, and its working groups. Note that 40 other groups may also distribute working documents as Internet- 41 Drafts. 43 Internet-Drafts are draft documents valid for a maximum of six months 44 and may be updated, replaced, or obsoleted by other documents at any 45 time. It is inappropriate to use Internet-Drafts as reference 46 material or to cite them other than as "work in progress." 47 The list of current Internet-Drafts can be accessed at 48 https://www.ietf.org/1id-abstracts.html. The list of Internet-Draft 49 Shadow Directories can be accessed at 50 https://www.ietf.org/shadow.html. 52 Table of Contents 54 1. Introduction............................................4 55 1.1 NSH Background.........................................4 56 1.2 ECN Background.........................................6 57 1.3 Tunnel Congestion Feedback Background..................6 58 1.4 Conventions Used in This Document......................8 60 2. The NSH ECN Field......................................10 62 3. ECN Support in the NSH.................................12 63 3.1 At The Ingress........................................13 64 3.2 At Transit Nodes......................................14 65 3.2.1 At NSH Transit Nodes................................14 66 3.2.2 At an SF/Proxy......................................15 67 3.2.3 At Other Forwarding Nodes...........................15 68 3.3 At Exit/Egress........................................16 69 3.4 Congestion Statistics and the Conservation of Packets.16 71 4. Tunnel Congestion Feedback Support.....................18 72 4.1 Congestion Level Measurements.........................18 73 4.3 Congestion Information Delivery.......................19 74 4.3 IPFIX Extensions......................................21 75 4.3.1 nshServicePathID....................................21 76 4.3.2 tunnelEcnCeCeByteTotalCount.........................21 77 4.3.3 tunnelEcnEctNectBytetTotalCount.....................22 78 4.3.4 tunnelEcnCeNectByteTotalCount.......................22 79 4.3.5 tunnelEcnCeEctByteTotalCount........................22 80 4.3.6 tunnelEcnEctEctByteTotalCount.......................23 81 4.3.7 tunnelEcnCEMarkedRatio..............................23 83 5. Example of Use.........................................24 85 6. IANA Considerations....................................27 86 6.1 SFC NSH Header ECN Bits...............................27 87 6.2 IPFIX Information Element IDs.........................27 89 7. Security Considerations................................29 90 8. Acknowledgements.......................................29 92 Normative References......................................30 93 Informative References....................................31 95 Authors' Addresses........................................32 97 1. Introduction 99 Explicit Congestion Notification (ECN [RFC3168]) allows a forwarding 100 element to notify downstream devices of the onset of congestion 101 without having to drop packets. Coupled with a means to feed 102 information about congestion back to upstream nodes, this can improve 103 network efficiency through better congestion control, frequently 104 without packet drops. This document specifies ECN and congestion 105 feedback support within a Service Function Chaining (SFC [RFC7665]) 106 architecture domain through use of the Network Service Header (NSH 107 [RFC8300]) and IP Flow Information Export (IPFIX [RFC7011]). 109 It requires that all ingress and egress nodes of the SFC domain 110 implement ECN. While congestion management will be the most effective 111 if all interior nodes of the SFC domain implement ECN, some benefit 112 is obtained even if some interior nodes do not implement ECN. 113 Congestion at any interior bottleneck where ECN marking is not 114 implemented will be unmanaged. 116 The subsections below in this section provide background information 117 on NSH, ECN, congestion feedback, and terminology used in this 118 document. 120 1.1 NSH Background 122 The Service Function Chaining (SFC [RFC7665]) architecture calls for 123 the encapsulation of traffic within a service function chaining 124 domain with a Network Service Header (NSH [RFC8300]) added by the 125 "Classifier" (ingress node) on entry to the domain and the NSH being 126 removed on exit from the domain at the egress node. The NSH is used 127 to control the path of a packet in an SFC domain. The NSH is a 128 natural place, in a domain where traffic is NSH encapsulated, to note 129 congestion, avoiding possible confusion due, for example, to changes 130 in the outer transport header in different parts of the domain. 132 | 133 v 134 +----------+ 135 . .|Classifier|. . . . . . . . . . . . . . 136 . +----------+ . 137 . | +----+ . 138 . | --+ SF | Service . 139 . | / +----+ Function . 140 . v --- Chaining . 141 . +-----+/ +----+ domain . 142 . | SFF |--------+ SF | . 143 . +-----+\ +----+ . 144 . | --- . 145 . | \ +----+ . 146 . | --+ SF | . 147 . v +----+ . 148 . +-----+ +----+ . 149 . | SFF |-----------------+ SF | . 150 . +-----+ +----+ . 151 . | +----+ . 152 . | --+ SF | . 153 . | / +----+ . 154 . v --- . 155 . +-----+/ +----+ . 156 . | SFF |--------+ SF | . 157 . +-----+\ +----+ . 158 . | --- . 159 . | \ +----+ . 160 . | --+ SF | . 161 . v +----+ . 162 . +------+ . 163 . . .| Exit |. . . . . . . . . . . . . . . 164 +------+ 165 | 166 v 168 Figure 1. Example SFC Path Forwarding Nodes 170 Figure 1 shows an SFC domain for the purpose of illustrating the use 171 of the NSH. Traffic passes through a sequence of Service Function 172 Forwarders (SFFs) each of which sends the traffic to one or more 173 Service Functions (SFs). Each SF performs some operation on the 174 traffic, for example firewall or Network Address Translation (NAT) or 175 load balancer, and then returns it to the SFF from which it was 176 received. 178 Logically, during the transit of each SFF, the outer transport header 179 that got the packet to the SFF is stripped (see Figure 3), the SFF 180 decides on the next forwarding step, either adding a new transport 181 header or, if the SFF is the exit/egress, removing the NSH header. 182 The transport headers added may be different in different regions of 183 the SFC domain. For example, IP could be used for some SFF-to-SFF 184 communication and MPLS used for other such communication. 186 1.2 ECN Background 188 Explicit congestion notification (ECN [RFC3168]) allows a forwarding 189 element (such as a router or a Service Function Forwarder (SFF) or 190 Service Function (SF)) to notify downstream devices of the onset of 191 congestion without having to drop packets. This can be used as an 192 element in active queue management (AQM) [RFC7567] to improve network 193 efficiency through better traffic control without packet drops. The 194 forwarding element can explicitly mark some packets in an ECN field 195 instead of dropping the packet. For example, a two-bit field is 196 available for ECN marking in IP headers [RFC3168]. 198 1.3 Tunnel Congestion Feedback Background 200 Tunnels are widely deployed in various networks including data center 201 networks, enterprise network, and the public Internet. A tunnel 202 consists of ingress, egress, and a set of intermediate nodes 203 including routers. Tunnel Congestion Feedback (Section 4) is a 204 building block for congestion mitigation methods. It supports 205 feedback of congestion information from an egress node to an ingress 206 node. This document treats the SFC domain as a tunnel with the 207 initial Classifier node being the ingress; however, the Tunnel 208 Congestion Feedback facilities specified in this document MAY be used 209 in other contexts besides SFC domains. 211 Examples of actions that can be taken by an ingress node when it has 212 knowledge of downstream congestion include those listed below. 213 Details of implementing these traffic control methods, beyond those 214 given here, are outside the scope of this document. 216 Any action by a tunnel ingress to reduce congestion needs to allow 217 sufficient time for the end-to-end congestion control loop to respond 218 first, otherwise the system could go unstable. For instance by the 219 ingress taking a smoothed average of the level of congestion signaled 220 by feedback from the tunnel egress or delaying any action for at 221 least the worst case global round trip time (for example 100 222 milliseconds). 224 (1) Traffic throttling (policing), where the downstream traffic 225 flowing out of the ingress node is limited to reduce or eliminate 226 congestion. 228 (2) Upstream congestion feedback, where the ingress node sends 229 messages upstream to or towards the ultimate traffic source, a 230 function that can throttle traffic generation/transmission. 232 (3) Traffic re-direction, where the ingress node configures the NSH 233 of some future traffic so that it avoids congested paths. Great 234 care must be taken with this option to avoid (a) significant re- 235 ordering of traffic in flows that it is desirable to keep in 236 order and (b) oscillation/instability in traffic paths due to 237 alternate congestion of previously idle paths and the idling of 238 previously congested paths. For example, it is preferable to 239 classify traffic into flows of a sufficiently coarse granularity 240 that the flows are long lived and then use a stable path per 241 flow, sending only newly appearing flows on apparently 242 uncongested paths. 244 Figure 2 shows an example path from an original sender to a final 245 receiver passing through an example chain of service functions 246 between the ingress and egress of an SFC domain. The path is also 247 likely to pass through other network nodes outside the SFC domain 248 (not shown) before entering the SFC domain and after leaving the SFC 249 domain. 251 The figure shows typical congestion feedback that would be expected 252 from the final receiver to the origin sender, which controls the load 253 the origin sender applies to all elements on the path. The figure 254 also shows the congestion feedback from the egress to the ingress of 255 the SFC domain that is described in this document, to control or 256 balance load within the SFC domain. 258 .:= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = :. 259 _||_ End-to-End Congestion Feedback || 260 \ / || 261 \/ || 262 __ Inner Transport Header and Payload __ 263 | | ->- - - - - - - - - - - - - - ->- - - - - -- - - - - - ->- | | 264 | | | | 265 | | .:= = = = = = = = = = = = = = = = = = = = = =:. | | 266 | | _||_ Tunnel Congestion Feedback || | | 267 | | \ / || | | 268 | | \/ || | | 269 | | __ NSH __ | | 270 | | | |-------------------------->--------------| | | | 271 | |. . . | | ___ ___ ___ | |. . .| | 272 | | | | OT1 | | OT4 | | . . . | | OTn | | | | 273 | | | |-->--|SFF|--->---|SFF| |SFF|-->--| | | | 274 |__| |__| |___| |___| |___| |__| |__| 275 origin SFC | ^ | ^ SFC final 276 sender domain OT2| |OT3 OT6| |OT7 domain rcvr 277 ingress v | v | egress 278 +---+ +---+ 279 |SF | |SF | 280 +---+ +---+ 282 Figure 2. Congestion Feedback across an SFC Domain 284 SFC Domain congestion feedback in Figure 2 is shown within the 285 context of an end-to-end congestion feedback loop. Also shown is the 286 encapsulated layering of NSH headers within a series of outer 287 transport headers (OT1, OT2, ... OTn). 289 1.4 Conventions Used in This Document 291 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 292 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 293 "OPTIONAL" in this document are to be interpreted as described in BCP 294 14 [RFC2119] [RFC8174] when, and only when, they appear in all 295 capitals, as shown here. 297 Acronyms: 299 AQM - Active Queue Management [RFC7567] 301 CE - Congestion Experienced [RFC3168] 303 downstream - The direction from ingress to egress 304 ECN - Explicit Congestion Notification [RFC3168] 306 ECT - ECN Capable Transport [RFC3168] 308 IPFIX - IP Flow Information Export [RFC7011] 310 Not-ECT - Not ECN-Capable Transport [RFC3168] 312 NSH - Network Service Header [RFC8300] 314 SF - Service Function [RFC7665] 316 SFC - Service Function Chaining [RFC7665] 318 SFF - Service Function Forwarder [RFC7665] - A type of node that 319 forwards based on the NSH. 321 TLV - Type Length Value 323 upstream - The direction from egress to ingress 325 2. The NSH ECN Field 327 The NSH header is used to encapsulate traffic and control its 328 subsequent path (see Section 2 of [RFC8300]). The NSH also provides 329 for optional metadata inclusion, as shown in Figure 3. 331 +-----------------------------------+ 332 | Outer Transport Header | 333 +-----------------------------------+ 334 | Network Service Header (NSH) | 335 | +------------------------------+ | 336 | | Base Header | | 337 | +------------------------------+ | 338 | | Service Path Header | | 339 | +------------------------------+ | 340 | | Metadata (Context Header(s)) | | 341 | +------------------------------+ | 342 +-----------------------------------+ 343 | Original Packet / Frame / Payload | 344 +-----------------------------------+ 346 Figure 3. Data Encapsulation with the NSH 348 Two currently unused bits (indicated by "U") in the NSH Base Header 349 (Section 2.2 of [RFC8300]) are allocated for ECN indication as shown 350 in Figure 4. 352 0 1 2 3 353 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 354 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 355 |Ver|O|U| TTL | Length |U|U|U|U|MD Type| Next Protocol | 356 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 357 ^ ^ 358 | | 359 +-------+ 360 |NSH ECN| 361 | field | 362 +-------+ 364 Figure 4. NSH Base Header 366 RFC Editor NOTE: The above figure should be adjusted based on the 367 bits assigned by IANA (see Section 5) and this note deleted. 369 Table 1 shows the meaning of the code points in the NSH ECN field. 370 These have the same meaning as the ECN field code points in the IPv4 371 or IPv6 header as defined in [RFC3168]. 373 Binary Name Meaning 374 ------ ------- -------------------------------- 375 00 Not-ECT Not ECN-Capable Transport 376 01 ECT(1) ECN-Capable Transport 377 10 ECT(0) ECN-Capable Transport 378 11 CE Congestion Experienced 380 Table 1. ECN Field Code Points 382 3. ECN Support in the NSH 384 This section describes the required behavior to support ECN using the 385 NSH. There are two aspects to ECN support: 386 1. ECN propagation during encapsulation or decapsulation 387 2. ECN marking during congestion at bottlenecks. 389 While this section covers all combinations of ECN-aware and ECN- 390 unaware, it is expected that in most cases the NSH domain will be 391 uniform so that, if this document is applicable, all SFFs will 392 support ECN; however, some legacy SFs might not support ECN. 394 ECN Propagation: 396 The specification of ECN tunneling [RFC6040] explains that an 397 ingress must not propagate ECN support into an encapsulating 398 header unless the egress supports correct onward propagation of 399 the ECN field during decapsulation. We define Compliant ECN 400 Decapsulation here as decapsulation compliant with either 401 [RFC6040] or an earlier compatible equivalent ([RFC4301], or the 402 full functionality mode of [RFC3168]). 404 The procedures in Section 3.2.1 ensure that each ingress of the 405 large number of possible transport links within the SFC domain 406 does not propagate ECN support into the encapsulating outer 407 transport header unless the corresponding egress of that link 408 supports Compliant ECN Decapsulation. 410 Section 3.3 requires that all the egress nodes of the SFC domain 411 support Compliant ECN Decapsulation in conjunction with tunnel 412 congestion feedback, otherwise the scheme in this document will 413 not work. 415 ECN Marking: 417 At transit nodes the marking behavior specified in Section 3.2.1 418 is recommended and if not implemented at such transit nodes, there 419 may be unmanaged congestion. 421 Detection of congestion will be most effective if ECN marking is 422 supported by all potential bottlenecks inside the domain in which 423 NSH is being used to route traffic as well as at the ingress and 424 egress. Nodes that do not support ECN marking, or that support 425 AQM but not ECN, will naturally use drop to relieve congestion. 426 The gap in the end-to-end packet sequence will be detected as 427 congestion by the final receiving endpoint, but not by the NSH 428 egress (see Figure 2). 430 3.1 At The Ingress 432 When the ingress/Classifier encapsulates an incoming IP packet with 433 an NSH, it MUST set the NSH ECN field using the "Normal mode" 434 specified in [RFC6040] (i.e., copied from the incoming IP header). 436 Then, if the resulting NSH ECN field is Not-ECT, the ingress SHOULD 437 set it to ECT(0). This indicates that, even though the end-to-end 438 transport is not ECN-capable, the egress and ingress of the SFC 439 domain are acting as an ECN-capable transport. This approach will 440 inherently support all known variants of ECN, including the 441 experimental L4S capability [RFC8311] [ecnL4S]. 443 Packets arriving at the ingress might not use IP. If the protocol of 444 arriving packets supports an ECN field similar to IP, the procedures 445 for IP packets can be used. If arriving packets do not support an ECN 446 field similar to IP, they MUST be treated as if they are Not-ECT IP 447 packets. 449 Then, as the NSH encapsulated packet is further encapsulated with a 450 transport header, if ECN marking is available for that transport (as 451 it is for IP [RFC3168] and MPLS [RFC5129]), the ECN field of the 452 transport header MUST be set using the "Normal mode" specified in 453 [RFC6040] (i.e., copied from the NSH ECN field). 455 A summary of these normative steps is given in Table 2. 457 +-----------------+---------------+ 458 | Incoming Header | Departing NSH | 459 | (also equal to | and Outer | 460 | departing Inner | Headers | 461 | Header) | | 462 +-----------------+---------------+ 463 | Not-ECT | ECT(0) | 464 | ECT(0) | ECT(0) | 465 | ECT(1) | ECT(1) | 466 | CE | CE | 467 +-----------------+---------------+ 469 Table 2. Setting of ECN fields by an ingress/Classifier 471 The requirements in this section apply to all ingress nodes for the 472 domain in which NSH is being used to route traffic. 474 3.2 At Transit Nodes 476 This section described behavior at nodes that forward based on the 477 NSH such as SFF and other forwarding nodes such as IP routers. Figure 478 5 shows a packet on the wire between forwarding nodes. 480 +-----------------+ 481 | Outer Header | 482 +-----------------+ 483 | NSH | 484 +-----------------+ 485 | Inner Header | 486 +-----------------+ 487 | Payload | 488 +-----------------+ 490 Figure 5. Packet in Transit 492 3.2.1 At NSH Transit Nodes 494 When a packet is received at an NSH based forwarding node such as an 495 SFF, say N1, the outer transport encapsulation is removed and its ECN 496 marking SHOULD be combined into the NSH ECN marking as specified in 497 [RFC6040]. If this is not done, any congestion encountered at non-NSH 498 transit nodes between N1 and the previous upstream NSH based 499 forwarding node will be lost and not transmitted downstream. 501 The NSH forwarding node SHOULD use a recognized AQM algorithm 502 [RFC7567] to detect congestion. If the NSH ECN field indicates ECT, 503 it will probabilistically set the NSH ECN field to the Congestion 504 Experienced (CE) value or, in cases of extreme congestion, drop the 505 packet. 507 When the NSH encapsulated packet is further encapsulated for 508 transmission to the next SFF or SF, ECN marking behavior depends on 509 whether or not the node that will decapsulate the outer header 510 supports Compliant ECN Decapsulation (see Section 3). If it does, 511 then the encapsulating node propagates the NSH ECN field to this 512 outer encapsulation using the "Normal Mode" of ECN encapsulation 513 [RFC6040] (the ECN field is copied). If it does not, then the 514 encapsulating node MUST clear ECN in the outer encapsulation to non- 515 ECT (the "Compatibility Mode" of [RFC6040]). 517 3.2.2 At an SF/Proxy 519 If the SF is NSH and ECN-aware, the processing is essentially the 520 same at the SF as at an SFF as discussed in Section 3.2.1. 522 If the SF is NSH-aware but ECN-unaware, then the SFF transmitting the 523 packet to the SF will use Compatibility Mode. Congestion encountered 524 in the SFF to SF and SF to SFF paths will be unmanaged. 526 If the SF is not NSH-aware, then an NSH proxy will be between the SFF 527 and the SF to avoid exposure of the SF that does not understand NSHs 528 to the NSH as shown in Figure 6. This is described in Section 4.6 of 529 [RFC7665]. The SF and proxy together look to the SFF like an NSH- 530 aware SF. The behavior at the proxy and SF in this case is as below: 532 If such a proxy is not ECN-aware then congestion in the entire 533 path from SFF to proxy to SF back to proxy to SFF will be 534 unmanaged. 536 | 537 v 538 +----------+ +---------+ 539 | | +-------+ | NSH | 540 | SFF +---->| NSH +---->|un-aware | 541 |(Service | | aware | | SF | 542 | Function |<----+ proxy |<----+(Service | 543 |Forwarder)| +-------+ |Function)| 544 +----------+ +---------+ 545 | 546 v 548 Figure 6. Proxy for NSH Un-aware SFF 550 If the proxy is ECN-aware, the proxy uses an AQM to indicate 551 congestion within the proxy in the NSH that it returns to the SFF. 552 The outer header used for the proxy-to-SF path uses Normal Mode. 553 The outer header used for the proxy-to-SFF path uses Normal Mode 554 based copying of the NSH ECN field to the outer header. Thus 555 congestion in the proxy will be managed. 557 Congestion in the SF will be managed only if the SF is ECN-aware 558 and implements an AQM. 560 3.2.3 At Other Forwarding Nodes 562 Other forwarding nodes, that is non-NSH forwarding nodes between NSH 563 forwarding nodes, such as IP or label switched routers, might also 564 contain potential bottlenecks. If so, they SHOULD implement an AQM 565 algorithm to update the ECN marking in the outer transport header as 566 specified in [RFC3168]. 568 3.3 At Exit/Egress 570 At the SFC domain egress node, first any actions are taken based on 571 Congestion Experienced or other values of ECN marking, such as 572 accumulating statistics to send back to the ingress (see Section 4) 573 or for other uses. If the packet being carried inside the NSH is IP, 574 when the NSH is removed the NSH ECN field MUST be combined with the 575 IP ECN field as specified in Table 3 that was extracted from 576 [RFC6040]. This requirement applies to all egress nodes for the 577 domain in which NSH is being used to route traffic. 579 +---------+---------------------------------------------+ 580 |Arriving | Arriving Outer Header | 581 | Inner +---------+-----------+-----------+-----------+ 582 | Header | Not-ECT | ECT(0) | ECT(1) | CE | 583 +---------+---------+-----------+-----------+-----------+ 584 | Not-ECT | Not-ECT | Not-ECT | Not-ECT | | 585 | ECT(0) | ECT(0) | ECT(0) | ECT(0) | CE | 586 | ECT(1) | ECT(1) | ECT(1) | ECT(1) | CE | 587 | CE | CE | CE | CE | CE | 588 +---------+---------+-----------+-----------+-----------+ 590 Table 3. Exit ECN Fields Merger 592 All the egress nodes of the SFC domain MUST support Compliant ECN 593 Decapsulation as specified in this section. If this is not the case, 594 the scheme described in this document will not work, and cannot be 595 used. 597 3.4 Congestion Statistics and the Conservation of Packets 599 The SFC specification permits an SF to absorb packets and to generate 600 new packets as well as simply processing and forwarding the packets 601 it receives. Such actions might appear to be packet loss due to 602 congestion or might mask the loss of packets by generating additional 603 packets. 605 The tunnel congestion feedback approach (Section 4) can detect 606 congestions in several ways. One way detects traffic loss by counting 607 payload packets and bytes in at the ingress and counting them out at 608 the egress. This does not work unless nodes conserve the number of 609 payload packets and/or bytes. Therefore, it will not be possible to 610 detect loss using this technique if traffic volume is not conserved 611 by the service function chain processing that traffic. 613 Nonetheless, if a bottleneck supports ECN marking, it will be 614 possible to detect the high level of CE markings that are associated 615 with congestion at that bottleneck by looking at the ratio of CE- 616 marked to non-CE-marked packets. However, it will not be possible for 617 the tunnel congestion feedback approach to detect any congestion, 618 whether slight or severe, if it occurs at a bottleneck that does not 619 support ECN marking. 621 4. Tunnel Congestion Feedback Support 623 The collection and storage of congestion information at the egress 624 may be useful for later analysis but, unless it can be fed back to a 625 point which can take action to reduce congestion, it will not be 626 useful in real time. Such congestion feedback to the ingress enables 627 it to take actions such as those listed in Section 1.3. 629 IP Flow Information Export (IPFIX [RFC7011]) provides a standard for 630 communicating traffic flow statistics. As extended by this document, 631 IPFIX messages from the egress to the ingress are used to communicate 632 the extent of congestion between an ingress and egress based on ECN 633 marking in the NSH. 635 4.1 Congestion Level Measurements 637 The congestion level measurements are based on ECN marking in the NSH 638 and packet drop. In particular the congestion information includes 639 the ratio of CE-marked packets to all packets and the ratio of 640 dropped packets to all packets. 642 If the congestion level is low enough, the packets are marked as CE 643 instead of being dropped, and then it is easy to calculate congestion 644 level according to the ratio of CE-marked packets. If the congestion 645 level is so high that ECT packets will be dropped, then the packet 646 loss ratio could be calculated by comparing total packets entering 647 ingress and total packets arriving at egress over the same span of 648 packets. If packet loss is detected for a flow that would preserve 649 the number of packets in the absence of congestion, then it can be 650 assumed that severe congestion has occurred in the tunnel. 652 The egress calculates the CE-marked packet ratio by counting packets 653 with different ECN markings. The CE-marked packet ratio will be used 654 as an indication of tunnel load level. It is assumed that nodes 655 between the ingress and egress will not drop packets biased towards 656 certain ECN codepoints, so calculating of CE-marked packet ratio is 657 not affect by packet drop. 659 The calculation of the fraction of packets droped is by comparing the 660 traffic volumes between ingress and egress. 662 Faked ECN-Capable Transport (ECT) is used at the ingress to defer 663 packet loss to the egress. The basic idea of faked ECT is that, when 664 encapsulating packets, the ingress first marks the tunnel outer 665 header (NSH for an SFC domain) according to [RFC6040], and then 666 remarks the outer header of Not-ECT packets as ECT. (ECT(0) and 667 ECT(1) are treated as the same.) Thus, as transmitted by the ingress 668 node, there will be one of three combinations of outer header ECN 669 field and inner header ECN field as follows: CE|CE, ECT|N-ECT, and 670 ECT|ECT (in the format of outer-ECN|inner-ECN); when decapsulating 671 packets at the egress, [RFC6040] defined decapsulation behavior is 672 used, and according to [RFC6040], the packets marked as CE|N-ECT will 673 be dropped. Faked-ECT is used to shift some drops to the egress in 674 order to allow the egress to calculate the CE-marked packet ratio 675 more precisely. 677 The ingress encapsulates packets and marks their outer header 678 according to faked ECT as described above. The ingress cumulatively 679 counts packet bytes for three types of ECN combination (CE|CE, ECT|N- 680 ECT, and ECT|ECT) and then the ingress regularly sends cumulative 681 bytes counts message of each type of ECN combination to the egress. 683 When each message arrives at the egress, (1) the egress calculates 684 the ratio of CE-marked packets; (2) the egress cumulatively counts 685 packet bytes coming from the ingress and adds its own bytes counts of 686 each type of ECN combination (CE|CE, ECT|N-ECT, CE|N-ECT, CE|ECT, and 687 ECT|ECT) to the message for ingress to calculate packet loss. The 688 egress feeds back the CE-marked packet ratio, packet loss ratio, 689 bytes counts information, and the like to the ingress as requested 690 for evaluating congestion level in the tunnel. 692 The statistics can be at the granularity of all traffic from the 693 ingress to the egress to learn about the overall congestion status of 694 the path between the ingress and the egress or at the granularity of 695 individual customer's traffic or a specific set of flows to learn 696 about their congestion contribution. 698 For example, the tunnelEcnCEMarkedRatio field (specified below) 699 indicates the fraction of traffic that has been marked in the ECN 700 field of the NSH as Congestion Experienced (CE). 702 4.3 Congestion Information Delivery 704 As described above, the tunnel ingress needs to send a messages 705 containing cumulative bytes counts of packets of each type of ECN 706 combination to the tunnel egress, and the tunnel egress also needs to 707 feed back messages with cumulative bytes counts of packets of each 708 type of ECN combination and the CE-marked packet ratio to the 709 ingress. This section specifies how the messages are conveyed. 711 IPFIX recommends, but does not require, use of SCTP [RFC4960] in 712 partial reliability mode [RFC3758] for the transport of its messages. 713 This mode allows loss of some packets, which is tolerable because 714 IPFIX communicates cumulative statistics. IPFIX over SCTP over IP 715 SHOULD be used directly where there is IP connectivity between the 716 ingress and egress; however, there might be different transport 717 protocols or address spaces used in different regions of an SFC 718 domain that make such direct IP connectivity problematic. The NSH 719 provides the general method of routing traffic within an SFC domain 720 so the encapsulation of the required IPFIX traffic in NSH MUST be 721 implemented and, when IP connectivity is not available, IPFIX over 722 NSH SHOULD be used along with configuration of appropriate SFC paths 723 for the IPFIX over NSH traffic. 725 IPFIX messages could travel along the same path as network data 726 traffic. In any case, an IPFIX message packet may get lost in case of 727 network congestion. Even though the missing information could be 728 recovered because of the use of cumulative counts, the message SHOULD 729 be transmitted at a higher priority than users' traffic flows to 730 improve the promptness of congestion information feedback. 732 The ingress node can do congestion management at different 733 granularity which means both the overall aggregated inner tunnel 734 congestion level and congestion level contributed by certain traffic 735 flows could be measured for different congestion management purposes. 736 For example, if the ingress only wants to limit congestion volume 737 caused by certain traffic flows, such as UDP-based traffic, then 738 congestion volume for that traffic can be fed back; or if the ingress 739 is doing overall congestion management, the aggregated congestion 740 volume can be fed back. 742 When sending IPFIX messages from ingress to egress, the ingress acts 743 as IPFIX exporter and the egress acts as IPFIX collector; When 744 feeding back congestion level information from egress to ingress, 745 then the egress acts as IPFIX exporter and ingress acts as IPFIX 746 collector. 748 The combination of congestion level measurement and congestion 749 information delivery procedures are as following: 751 o The ingress node determines the IPFIX template record to be used. 752 The template record can be pre-configured or determined at 753 runtime, the content of the template record will be determined 754 according to the granularity of congestion management; if the 755 ingress wants to limit congestion volume contributed by specific 756 traffic flows then the elements such as source IP address, 757 destination IP address, flow ID and CE-marked packet volume of the 758 flows, etc., will be included in the template record. 760 o Metering at the ingress measures traffic volume according to the 761 template record chosen and then the measurement records are sent 762 to the egress. 764 o Metering on the egress measures congestion level information 765 according to template record which SHOULD be the same as the 766 template record sent by the ingress. 768 o The egress sends its measurement records together with the 769 measurement records of the ingress back to the ingress. 771 4.3 IPFIX Extensions 773 This section specifies the new IPFIX Information Elements needed. It 774 conforms to [RFC7013]. 776 4.3.1 nshServicePathID 778 In order to identify SFC flows, so that congestion can be measured 779 and reported at that granularity, it is necessary for IPFIX to be 780 able to classify traffic based on the Service Path Identifier field 781 of the NSH [RFC8300]. Thus an NSH Service Path Identifier 782 (nshServicePathID) IPFIX Information Element [RFC7012] is specified. 784 Name: nshServicePathID 786 Description: Network Service Header [RFC8300] Service Path 787 Identifier. This is a 24-bit value which is left justified in 788 the Information Element. The low order byte MUST be sent as 789 zero and ignored on receipt. 791 Abstract Data Type: unsigned32 793 Data Type Semantics: identifier 795 ElementId: TBD0 797 Status: current 799 4.3.2 tunnelEcnCeCeByteTotalCount 801 Description: The total number of bytes of incoming packets with 802 the CE|CE ECN marking combination at the Observation Point 803 since the Metering Process (re-)initialization for this 804 Observation Point. 806 Abstract Data Type: unsigned64 808 Data Type Semantics: totalCounter 810 ElementId: TBD1 811 Statues: current 813 Units: bytes 815 4.3.3 tunnelEcnEctNectBytetTotalCount 817 Description: The total number of bytes of incoming packets with 818 the ECT|N-ECT ECN marking combination (ECT(0) and ECT(1) are 819 treated the same as each other) at the Observation Point since 820 the Metering Process (re-)initialization for this Observation 821 Point. 823 Abstract Data Type: unsigned64 825 Data Type Semantics: totalCounter 827 ElementId: TBD2 829 Statues: current 831 Units: bytes 833 4.3.4 tunnelEcnCeNectByteTotalCount 835 Description: The total number of bytes of incoming packets with 836 the CE|N-ECT ECN marking combination at the Observation Point 837 since the Metering Process (re-)initialization for this 838 Observation Point. 840 Abstract Data Type: unsigned64 842 Data Type Semantics: totalCounter 844 ElementId: TBD3 846 Statues: current 848 Units: bytes 850 4.3.5 tunnelEcnCeEctByteTotalCount 852 Description: The total number of bytes of incoming packets with 853 the CE|ECT ECN marking combination (ECT(0) and ECT(1) are 854 treated the same as each other) at the Observation Point since 855 the Metering Process (re-)initialization for this Observation 856 Point. 858 Abstract Data Type: unsigned64 860 Data Type Semantics: totalCounter 862 ElementId: TBD4 864 Statues: current 866 Units: bytes 868 4.3.6 tunnelEcnEctEctByteTotalCount 870 Description: The total number of bytes of incoming packets with 871 the ECT|ECT ECN marking combination (ECT(0) and ECT(1) are 872 treated the same as each other) at the Observation Point since 873 the Metering Process (re-)initialization for this Observation 874 Point. 876 Abstract Data Type: unsigned64 878 Data Type Semantics: totalCounter 880 ElementId: TBD5 882 Statues: current 884 Units: bytes 886 4.3.7 tunnelEcnCEMarkedRatio 888 Description: The ratio of CE-marked packets at the Observation 889 Point. 891 Abstract Data Type: float32 893 ElementId: TBD6 895 Statues: current 897 5. Example of Use 899 This section provides an example of the solution described in this 900 document. 902 First, IPFIX template records are exchanged between ingress and 903 egress to negotiate the format of the data records to be exchanged. 904 The example here is to measure the congestion level for the overall 905 tunnel caused by all the traffic. After the negotiation is finished, 906 the ingress sends in-band messages to the egress containing the 907 number of each kind of ECN-marked packets (i.e., CE|CE, ECT|N-ECT and 908 ECT|ECT) received before it sent the message. 910 After the egress receives the message, the egress calculates the CE- 911 marked packet ratio and counts the number of different kinds of ECN- 912 marking packets received before it received the message. Then the 913 egress sends a feedback message containing the counts together with 914 the information in the ingress's message back to the ingress. 916 Figures 7 to 10 below illustrate the example procedure between 917 ingress and egress. 919 +---------------------------------+----------------------+ 920 |Set ID=2 Length=40 | 921 |---------------------------------|----------------------| 922 |Template ID=256 Field Count=8 | 923 |---------------------------------|----------------------| 924 |tunnelEcnCeCeByteTotalCount Field Length=8 | 925 |---------------------------------|----------------------| 926 |tunnelEcnEctNectByteTotalCount Field Length=8 | 927 |---------------------------------|----------------------| 928 |tunnelEcnEctEctByteTotalCount Field Length=8 | 929 |---------------------------------|----------------------| 930 |tunnelEcnCeNectByteTotalCount Field Length=8 | 931 |---------------------------------|----------------------| 932 |tunnelEcnCeEctByteTotalCount Field Length=8 | 933 +---------------------------------|----------------------+ 934 |tunnelEcnCEMarkedRatio Field Length=4 | 935 +---------------------------------+----------------------+ 937 Figure 7. Template Record Sent From Egress to Ingress 939 +---------------------------------+----------------------+ 940 |Set ID=2 Length=28 | 941 |---------------------------------|----------------------| 942 |Template ID=257 Field Count=3 | 943 |---------------------------------|----------------------| 944 |tunnelEcnCeCeByteTotalCount Field Length=8 | 945 |---------------------------------|----------------------| 946 |tunnelEcnEctNectByteTotalCount Field Length=8 | 947 |---------------------------------|----------------------| 948 |tunnelEcnEctEctByteTotalCount Field Length=8 | 949 |---------------------------------+----------------------| 951 Figure 8. Template Record Sent From Ingress to Egress 953 +-------+ +-+ +-+ +-+ +-+ +-+ +-+ +-+ +-------+ 954 | | |M| |P| |P| |P| |M| |P| |P| | | 955 | | +-+ +-+ +-+ +-+ +-+ +-+ +-+ | | 956 | |<---------------------------------------| | 957 | | | | 958 | | | | 959 |egress | +-+ +-+ |ingress| 960 | | |M| |M| | | 961 | | +-+ +-+ | | 962 | |--------------------------------------->| | 963 | | | | 964 | | | | 965 +-------+ +-------+ 967 +-+ 968 |M| : Message Packet 969 +-+ 971 +-+ 972 |P| : User Packet 973 +-+ 975 Figure 9. Traffic flow Between Ingress and Egress 976 Set ID=257, Length=28 977 +------+ A1 +-------+ 978 | | B1 | | 979 | | C1 | | 980 | | <----------------------------- | | 981 | | | | 982 | | | | 983 | | SetID=256, Length=72 | | 984 | | A1 | | 985 | | B1 | | 986 |egress| C1 |ingress| 987 | | A2 | | 988 | | B2 | | 989 | | C2 | | 990 | | D | | 991 | | E | | 992 | | R | | 993 | | ----------------------------> | | 994 | | | | 995 +------+ +-------+ 997 Figure 10. Messages Between Ingress and Egress 999 The following provides an example of how the tunnel congestion level 1000 can be calculated (see Figure 10): 1002 The congestion Level could be divided into two categories: (1) 1003 slight congestion (no packets dropped); (2) serious congestion 1004 (packets are being dropped). 1006 For slight congestion, the congestion level is indicated by the 1007 ratio of CE-marked packets: 1009 ce_marked = R; 1011 For serious congestion, the congestion level is indicated as the 1012 volume of traffic loss: 1014 total_ingress = (A1 + B1 + C1) 1016 total_egress = (A2 + B2 + C2 + D + E) 1018 volume_loss = (total_ingress - total_egress) 1020 6. IANA Considerations 1022 The following subsections provide IANA assignment considerations. 1024 6.1 SFC NSH Header ECN Bits 1026 IANA is requested to assign two contiguous bits in the NSH Base 1027 Header Bits registry for ECN (bits 16 and 17 suggested) and note this 1028 assignment as follows: 1030 Bit Description Reference 1031 ---------- ----------- ----------------- 1032 tbd(16-17) NSH ECN [this document] 1034 6.2 IPFIX Information Element IDs 1036 IANA is requested to assign IPFIX Information Element IDs as follows: 1038 ElementID: TBD0 1039 Name: nshServicePathID 1040 Data Type: unsigned32 1041 Data Type Semantics: identifier 1042 Status: current 1043 Description: The Network Service Header [RFC8300] Service Path 1044 Identifier. 1046 ElementID: TBD1 1047 Name: tunnelEcnCeCePacketTotalCount 1048 Data Type: unsigned64 1049 Data Type Semantics: totalCounter 1050 Status: current 1051 Description: The total number of bytes of incoming packets with 1052 the CE|CE ECN marking combination at the Observation Point 1053 since the Metering Process (re-)initialization for this 1054 Observation Point. 1055 Units: octets 1057 ElementID: TBD2 1058 Name: tunnelEcnEctNectPacketTotalCount 1059 Data Type: unsigned64 1060 Data Type Semantics: totalCounter 1061 Status: current 1062 Description: The total number of bytes of incoming packets with 1063 the ECT|N-ECT ECN marking combination at the Observation Point 1064 since the Metering Process (re-)initialization for this 1065 Observation Point. 1067 Units: octets 1069 ElementID: TBD3 1070 Name: tunnelEcnCeNectPacketTotalCount 1071 Data Type: unsigned64 1072 Data Type Semantics: totalCounter 1073 Status: current 1074 Description: The total number of bytes of incoming packets with 1075 the CE|N-ECT ECN marking combination at the Observation Point 1076 since the Metering Process (re-)initialization for this 1077 Observation Point. 1078 Units: octets 1080 ElementID: TBD4 1081 Name: tunnelEcnCeEctPacketTotalCount 1082 Data Type: unsigned64 1083 Data Type Semantics: totalCounter 1084 Status: current 1085 Description: The total number of bytes of incoming packets with 1086 the CE|ECT ECN marking combination at the Observation Point 1087 since the Metering Process (re-)initialization for this 1088 Observation Point. 1089 Units: octets 1091 ElementID: TBD5 1092 Name: tunnelEcnEctEctPacketTotalCount 1093 Data Type: unsigned64 1094 Data Type Semantics: totalCounter 1095 Status: current 1096 Description: The total number of bytes of incoming packets with 1097 the CE|ECT(0) ECN marking combination at the Observation Point 1098 since the Metering Process (re-)initialization for this 1099 Observation Point. 1100 Units: octets 1102 ElementID: TBD6 1103 Name: tunnelEcnCEMarkedRatio 1104 Data Type: float32 1105 Status: current 1106 Description: The ratio of CE-marked Packet at the Observation 1107 Point. 1109 7. Security Considerations 1111 For general NSH security considerations, see [RFC8300]. 1113 For security considerations concerning tampering with ECN signaling, 1114 see [RFC3168]. For security considerations concerning ECN and 1115 encapsulation, see [RFC6040]. 1117 For general IPFIX security considerations, see [RFC7011]. If deployed 1118 in an untrusted environment, the signaling traffic between ingress 1119 and egress can be protected utilizing the security mechanisms 1120 provided by IPFIX (see Section 11 in [RFC7011]). The tunnel 1121 endpoints (the ingress and egress for an SFC domain) are assumed to 1122 be in the same administrative domain, so they will trust each other. 1124 The solution in this document does not introduce any greater 1125 potential to invade privacy than would have been available without 1126 the solution. 1128 8. Acknowledgements 1130 Most of the material on Tunnel Congestion Feedback was originally in 1131 draft-ietf-tsvwg-tunnel-congestion-feedback. After discussion with 1132 the authors of that draft, the authors of this draft, and the Chairs 1133 of the TSVWG and SFC Working Groups, the Tunnel Congestion Feedback 1134 draft was merged into this draft. 1136 The authors wish to thank the following for their comments, 1137 suggestions, and reviews: 1139 David Black, Sami Boutros, Anthony Chan, Lingli Deng, Liang Geng, 1140 Joel Halpern, Jake Holland, John Kaippallimalil, Tal Mizrahi, 1141 Vincent Roca, Lei Zhu 1143 Normative References 1145 [RFC2119] - Bradner, S., "Key words for use in RFCs to Indicate 1146 Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, 1147 March 1997, . 1149 [RFC3168] - Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 1150 of Explicit Congestion Notification (ECN) to IP", RFC 3168, DOI 1151 10.17487/RFC3168, September 2001, . 1154 [RFC3758] - Stewart, R., Ramalho, M., Xie, Q., Tuexen, M., and P. 1155 Conrad, "Stream Control Transmission Protocol (SCTP) Partial 1156 Reliability Extension", RFC 3758, DOI 10.17487/RFC3758, May 1157 2004, . 1159 [RFC5129] - Davie, B., Briscoe, B., and J. Tay, "Explicit Congestion 1160 Marking in MPLS", RFC 5129, DOI 10.17487/RFC5129, January 2008, 1161 . 1163 [RFC6040] - Briscoe, B., "Tunnelling of Explicit Congestion 1164 Notification", RFC 6040, DOI 10.17487/RFC6040, November 2010, 1165 . 1167 [RFC7011] - Claise, B., Ed., Trammell, B., Ed., and P. Aitken, 1168 "Specification of the IP Flow Information Export (IPFIX) 1169 Protocol for the Exchange of Flow Information", STD 77, RFC 1170 7011, DOI 10.17487/RFC7011, September 2013, . 1173 [RFC7013] - Trammell, B. and B. Claise, "Guidelines for Authors and 1174 Reviewers of IP Flow Information Export (IPFIX) Information 1175 Elements", BCP 184, RFC 7013, DOI 10.17487/RFC7013, September 1176 2013, . 1178 [RFC7567] - Baker, F., Ed., and G. Fairhurst, Ed., "IETF 1179 Recommendations Regarding Active Queue Management", BCP 197, 1180 RFC 7567, DOI 10.17487/RFC7567, July 2015, . 1183 [RFC8174] - Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 1184 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, May 1185 2017, 1187 [RFC8300] - Quinn, P., Ed., Elzur, U., Ed., and C. Pignataro, Ed., 1188 "Network Service Header (NSH)", RFC 8300, DOI 10.17487/RFC8300, 1189 January 2018, . 1191 Informative References 1193 [RFC4301] - Kent, S. and K. Seo, "Security Architecture for the 1194 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, December 1195 2005, . 1197 [RFC4960] - Stewart, R., Ed., "Stream Control Transmission Protocol", 1198 RFC 4960, DOI 10.17487/RFC4960, September 2007, 1199 . 1201 [RFC7012] - Claise, B., Ed., and B. Trammell, Ed., "Information Model 1202 for IP Flow Information Export (IPFIX)", RFC 7012, DOI 1203 10.17487/RFC7012, September 2013, . 1206 [RFC7665] - Halpern, J., Ed., and C. Pignataro, Ed., "Service 1207 Function Chaining (SFC) Architecture", RFC 7665, DOI 1208 10.17487/RFC7665, October 2015, . 1211 [RFC8311] - Black, D., "Relaxing Restrictions on Explicit Congestion 1212 Notification (ECN) Experimentation", RFC 8311, DOI 1213 10.17487/RFC8311, January 2018, . 1216 [ecnL4S] - De Schepper, K., and B. Briscoe, "Identifying Modified 1217 Explicit Congestion Notification (ECN) Semantics for Ultra-Low 1218 Queuing Delay (L4S)", draft-ietf-tsvwg-ecn-l4s-id, work in 1219 progress. 1221 Authors' Addresses 1223 Donald E. Eastlake, 3rd 1224 Futurewei Technologies 1225 2386 Panoramic Circle 1226 Apopka, FL 32703 USA 1228 Tel: +1-508-333-2270 1229 Email: d3e3e3@gmail.com 1231 Bob Briscoe 1232 Independent 1233 UK 1235 Email: ietf@bobbriscoe.net 1236 URI: http://bobbriscoe.net/ 1238 Yizhou Li 1239 Huawei Technologies 1240 101 Software Avenue, 1241 Nanjing 210012, P. R China 1243 Phone: +86-25-56624584 1244 EMail: liyizhou@huawei.com 1246 Andrew G. Malis 1247 Malis Consulting 1249 Email: agmalis@gmail.com 1251 Xinpeng Wei 1252 Huawei Technologies 1253 Beiqing Rd. Z-park No.156, Haidian District, 1254 Beijing, 100095, P. R. China 1256 EMail: weixinpeng@huawei.com 1258 Copyright and IPR Provisions 1260 Copyright (c) 2021 IETF Trust and the persons identified as the 1261 document authors. All rights reserved. 1263 This document is subject to BCP 78 and the IETF Trust's Legal 1264 Provisions Relating to IETF Documents 1265 (http://trustee.ietf.org/license-info) in effect on the date of 1266 publication of this document. Please review these documents 1267 carefully, as they describe your rights and restrictions with respect 1268 to this document. Code Components extracted from this document must 1269 include Simplified BSD License text as described in Section 4.e of 1270 the Trust Legal Provisions and are provided without warranty as 1271 described in the Simplified BSD License. The definitive version of 1272 an IETF Document is that published by, or under the auspices of, the 1273 IETF. Versions of IETF Documents that are published by third parties, 1274 including those that are translated into other languages, should not 1275 be considered to be definitive versions of IETF Documents. The 1276 definitive version of these Legal Provisions is that published by, or 1277 under the auspices of, the IETF. Versions of these Legal Provisions 1278 that are published by third parties, including those that are 1279 translated into other languages, should not be considered to be 1280 definitive versions of these Legal Provisions. For the avoidance of 1281 doubt, each Contributor to the IETF Standards Process licenses each 1282 Contribution that he or she makes as part of the IETF Standards 1283 Process to the IETF Trust pursuant to the provisions of RFC 5378. No 1284 language to the contrary, or terms, conditions or rights that differ 1285 from or are inconsistent with the rights and licenses granted under 1286 RFC 5378, shall have any effect and shall be null and void, whether 1287 published or posted by such Contributor, or included with or in such 1288 Contribution.