idnits 2.17.00 (12 Aug 2021) /tmp/idnits31060/draft-ietf-6man-ipv6-alt-mark-09.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 (August 27, 2021) is 267 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(See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-06) exists of draft-chen-pce-pcep-ifit-04 == Outdated reference: A later version (-02) exists of draft-fz-spring-srv6-alt-mark-01 == Outdated reference: A later version (-03) exists of draft-ietf-idr-sr-policy-ifit-02 Summary: 0 errors (**), 0 flaws (~~), 4 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 6MAN Working Group G. Fioccola 3 Internet-Draft T. Zhou 4 Intended status: Standards Track Huawei 5 Expires: February 28, 2022 M. Cociglio 6 Telecom Italia 7 F. Qin 8 China Mobile 9 R. Pang 10 China Unicom 11 August 27, 2021 13 IPv6 Application of the Alternate Marking Method 14 draft-ietf-6man-ipv6-alt-mark-09 16 Abstract 18 This document describes how the Alternate Marking Method can be used 19 as a passive performance measurement tool in an IPv6 domain. It 20 defines a new Extension Header Option to encode Alternate Marking 21 information in both the Hop-by-Hop Options Header and Destination 22 Options Header. 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 February 28, 2022. 41 Copyright Notice 43 Copyright (c) 2021 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 1.1. Terminology . . . . . . . . . . . . . . . . . . . . . . . 3 60 1.2. Requirements Language . . . . . . . . . . . . . . . . . . 3 61 2. Alternate Marking application to IPv6 . . . . . . . . . . . . 3 62 2.1. Controlled Domain . . . . . . . . . . . . . . . . . . . . 5 63 2.1.1. Alternate Marking Measurement Domain . . . . . . . . 6 64 3. Definition of the AltMark Option . . . . . . . . . . . . . . 7 65 3.1. Data Fields Format . . . . . . . . . . . . . . . . . . . 7 66 4. Use of the AltMark Option . . . . . . . . . . . . . . . . . . 8 67 5. Alternate Marking Method Operation . . . . . . . . . . . . . 10 68 5.1. Packet Loss Measurement . . . . . . . . . . . . . . . . . 10 69 5.2. Packet Delay Measurement . . . . . . . . . . . . . . . . 12 70 5.3. Flow Monitoring Identification . . . . . . . . . . . . . 13 71 5.3.1. Uniqueness of FlowMonID . . . . . . . . . . . . . . . 14 72 5.4. Multipoint and Clustered Alternate Marking . . . . . . . 15 73 5.5. Data Collection and Calculation . . . . . . . . . . . . . 15 74 6. Security Considerations . . . . . . . . . . . . . . . . . . . 16 75 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 19 76 8. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 19 77 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 19 78 9.1. Normative References . . . . . . . . . . . . . . . . . . 19 79 9.2. Informative References . . . . . . . . . . . . . . . . . 20 80 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 22 82 1. Introduction 84 [RFC8321] and [RFC8889] describe a passive performance measurement 85 method, which can be used to measure packet loss, latency and jitter 86 on live traffic. Since this method is based on marking consecutive 87 batches of packets, the method is often referred to as the Alternate 88 Marking Method. 90 This document defines how the Alternate Marking Method can be used to 91 measure performance metrics in IPv6. The rationale is to apply the 92 Alternate Marking methodology to IPv6 and therefore allow detailed 93 packet loss, delay and delay variation measurements both hop-by-hop 94 and end-to-end to exactly locate the issues in an IPv6 network. 96 The Alternate Marking is an on-path telemetry technique and consists 97 of synchronizing the measurements in different points of a network by 98 switching the value of a marking bit and therefore dividing the 99 packet flow into batches. Each batch represents a measurable entity 100 recognizable by all network nodes along the path. By counting the 101 number of packets in each batch and comparing the values measured by 102 different nodes, it is possible to precisely measure the packet loss. 103 Similarly, the alternation of the values of the marking bits can be 104 used as a time reference to calculate the delay and delay variation. 105 The Alternate Marking operation is further described in Section 5. 107 The format of IPv6 addresses is defined in [RFC4291] while [RFC8200] 108 defines the IPv6 Header, including a 20-bit Flow Label and the IPv6 109 Extension Headers. 111 This document introduces a new TLV (type-length-value) that can be 112 encoded in the Options Headers (Hop-by-Hop or Destination) for the 113 purpose of the Alternate Marking Method application in an IPv6 114 domain. 116 The threat model for the application of the Alternate Marking Method 117 in an IPv6 domain is reported in Section 6. As with all on-path 118 telemetry techniques, the only definitive solution is that this 119 methodology MUST be applied in a controlled domain. 121 1.1. Terminology 123 This document uses the terms related to the Alternate Marking Method 124 as defined in [RFC8321] and [RFC8889]. 126 1.2. Requirements Language 128 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 129 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 130 "OPTIONAL" in this document are to be interpreted as described in BCP 131 14 [RFC2119] [RFC8174] when, and only when, they appear in all 132 capitals, as shown here. 134 2. Alternate Marking application to IPv6 136 The Alternate Marking Method requires a marking field. Several 137 alternatives could be considered such as IPv6 Extension Headers, IPv6 138 Address and Flow Label. But, it is necessary to analyze the 139 drawbacks for all the available possibilities, more specifically: 141 Reusing existing Extension Header for Alternate Marking leads to a 142 non-optimized implementation; 143 Using the IPv6 destination address to encode the Alternate Marking 144 processing is very expensive; 146 Using the IPv6 Flow Label for Alternate Marking conflicts with the 147 utilization of the Flow Label for load distribution purpose 148 ([RFC6438]). 150 In the end, a new Hop-by-Hop or a new Destination Option is the best 151 choice. 153 The approach for the Alternate Marking application to IPv6 specified 154 in this memo is compliant with [RFC8200]. It involves the following 155 operations: 157 o The source node is the only one that writes the Option Header to 158 mark alternately the flow (for both Hop-by-Hop and Destination 159 Option). The intermediate nodes and destination node MUST only 160 read the marking values of the option without modifying the Option 161 Header. 163 o In case of Hop-by-Hop Option Header carrying Alternate Marking 164 bits, it is not inserted or deleted, but can be read by any node 165 along the path. The intermediate nodes may be configured to 166 support this Option or not and the measurement can be done only 167 for the nodes configured to read the Option. As further discussed 168 in Section 4, the presence of the hop-by-hop option should not 169 affect the traffic throughput both on nodes that do not recognize 170 this option and on the nodes that support it. However, it is 171 worth mentioning that there is a difference between theory and 172 practice. Indeed, in a real implementation it can happen that 173 packets with hop-by-hop option could also be skipped or processed 174 in the slow path. While some proposals are trying to address this 175 problem and make Hop-by-Hop Options more practical 176 ([I-D.peng-v6ops-hbh], [I-D.hinden-6man-hbh-processing]), these 177 aspects are out of the scope for this document. 179 o In case of Destination Option Header carrying Alternate Marking 180 bits, it is not processed, inserted, or deleted by any node along 181 the path until the packet reaches the destination node. Note 182 that, if there is also a Routing Header (RH), any visited 183 destination in the route list can process the Option Header. 185 Hop-by-Hop Option Header is also useful to signal to routers on the 186 path to process the Alternate Marking. However, as said, routers 187 will only examine this option if properly configured. 189 The optimization of both implementation and scaling of the Alternate 190 Marking Method is also considered and a way to identify flows is 191 required. The Flow Monitoring Identification field (FlowMonID), as 192 introduced in Section 5.3, goes in this direction and it is used to 193 identify a monitored flow. 195 The FlowMonID is different from the Flow Label field of the IPv6 196 Header ([RFC6437]). The Flow Label field in the IPv6 header is used 197 by a source to label sequences of packets to be treated in the 198 network as a single flow and, as reported in [RFC6438], it can be 199 used for load-balancing/equal cost multi-path (LB/ECMP). The reuse 200 of Flow Label field for identifying monitored flows is not considered 201 because it may change the application intent and forwarding behavior. 202 Also, the Flow Label may be changed en route and this may also 203 invalidate the integrity of the measurement. Furthermore, since the 204 Flow Label is pseudo-random, there is always a finite probability of 205 collision. Those reasons make the definition of the FlowMonID 206 necessary for IPv6. Indeed, the FlowMonID is designed and only used 207 to identify the monitored flow. Flow Label and FlowMonID within the 208 same packet are totally disjoint, have different scope, are used to 209 identify flows based on different criteria, and are intended for 210 different use cases. 212 The rationale for the FlowMonID is further discussed in Section 5.3. 213 This 20 bit field allows easy and flexible identification of the 214 monitored flow and enables improved measurement correlation and finer 215 granularity since it can be used in combination with the traditional 216 5-tuple to identify a flow. An important point that will be 217 discussed in Section 5.3.1 is the uniqueness of the FlowMonID and how 218 to allow disambiguation of the FlowMonID in case of collision. 220 The following section highlights an important requirement for the 221 application of the Alternate Marking to IPv6. The concept of the 222 controlled domain is explained and it is considered an essential 223 precondition, as also highlighted in Section 6. 225 2.1. Controlled Domain 227 [RFC8799] introduces the concept of specific limited domain solutions 228 and, in this regard, it is reported the IPv6 Application of the 229 Alternate Marking Method as an example. 231 IPv6 has much more flexibility than IPv4 and innovative applications 232 have been proposed, but for a number of reasons, such as the 233 policies, options supported, the style of network management and 234 security requirements, it is suggested to limit some of these 235 applications to a controlled domain. This is also the case of the 236 Alternate Marking application to IPv6 as assumed hereinafter. 238 Therefore, the IPv6 application of the Alternate Marking Method MUST 239 be deployed in a controlled domain. It is RECOMMENDED that an 240 implementation rejects packets that carry Alternate Marking data and 241 are entering or leaving the controlled domains. 243 A controlled domain is a managed network where it is required to 244 select, monitor and control the access to the network by enforcing 245 policies at the domain boundaries in order to discard undesired 246 external packets entering the domain and check the internal packets 247 leaving the domain. It does not necessarily mean that a controlled 248 domain is a single administrative domain or a single organization. A 249 controlled domain can correspond to a single administrative domain or 250 can be composed by multiple administrative domains under a defined 251 network management. Indeed, some scenarios may imply that the 252 Alternate Marking Method involves more than one domain, but in these 253 cases, it is RECOMMENDED that the multiple domains create a whole 254 controlled domain while traversing the external domain by employing 255 IPsec [RFC4301] authentication and encryption or other VPN technology 256 that provides full packet confidentiality and integrity protection. 257 In a few words, it must be possible to control the domain boundaries 258 and eventually use specific precautions if the traffic traverse the 259 Internet. 261 The security considerations reported in Section 6 also highlight this 262 requirement. 264 2.1.1. Alternate Marking Measurement Domain 266 The Alternate Marking measurement domain can overlap with the 267 controlled domain or may be a subset of the controlled domain. The 268 typical scenarios for the application of the Alternate Marking Method 269 depend on the controlled domain boundaries, in particular: 271 the user equipment can be the starting or ending node, only in 272 case it is fully managed and if it belongs to the controlled 273 domain. In this case the user generated IPv6 packets contain the 274 Alternate Marking data. But, in practice, this is not common due 275 to the fact that the user equipment cannot be totally secured in 276 the majority of cases. 278 the CPE (Customer Premises Equipment) is most likely to be the 279 starting or ending node since it connects the user's premises with 280 the service provider's network and therefore belongs to the 281 operator's controlled domain. Typically the CPE encapsulates a 282 received packet in an outer IPv6 header which contains the 283 Alternate Marking data. The CPE can also be able to filter and 284 drop packets from outside of the domain with inconsistent fields 285 to make effective the relevant security rules at the domain 286 boundaries, for example a simple security check can be to insert 287 the Alternate Marking data if and only if the destination is 288 within the controlled domain. 290 3. Definition of the AltMark Option 292 The definition of a new TLV for the Options Extension Headers, 293 carrying the data fields dedicated to the Alternate Marking method, 294 is reported below. 296 3.1. Data Fields Format 298 The following figure shows the data fields format for enhanced 299 Alternate Marking TLV (AltMark). This AltMark data can be 300 encapsulated in the IPv6 Options Headers (Hop-by-Hop or Destination 301 Option). 303 0 1 2 3 304 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 1 305 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 306 | Option Type | Opt Data Len | 307 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 308 | FlowMonID |L|D| Reserved | 309 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 311 where: 313 o Option Type: 8-bit identifier of the type of Option that needs to 314 be allocated. Unrecognized Types MUST be ignored on processing. 315 For Hop-by-Hop Options Header or Destination Options Header, 316 [RFC8200] defines how to encode the three high-order bits of the 317 Option Type field. The two high-order bits specify the action 318 that must be taken if the processing IPv6 node does not recognize 319 the Option Type; for AltMark these two bits MUST be set to 00 320 (skip over this Option and continue processing the header). The 321 third-highest-order bit specifies whether the Option Data can 322 change en route to the packet's final destination; for AltMark the 323 value of this bit MUST be set to 0 (Option Data does not change en 324 route). In this way, since the three high-order bits of the 325 AltMark Option are set to 000, it means that nodes can simply skip 326 this Option if they do not recognize and that the data of this 327 Option do not change en route, indeed the source is the only one 328 that can write it. 330 o Opt Data Len: 4. It is the length of the Option Data Fields of 331 this Option in bytes. 333 o FlowMonID: 20-bit unsigned integer. The FlowMon identifier is 334 described in Section 5.3. As further discussed below, it has been 335 picked as 20 bits since it is a reasonable value and a good 336 compromise in relation to the chance of collision if it is set 337 pseudo randomly by the source node or set by a centralized 338 controller. 340 o L: Loss flag for Packet Loss Measurement as described in 341 Section 5.1; 343 o D: Delay flag for Single Packet Delay Measurement as described in 344 Section 5.2; 346 o Reserved: is reserved for future use. These bits MUST be set to 347 zero on transmission and ignored on receipt. 349 4. Use of the AltMark Option 351 The AltMark Option is the best way to implement the Alternate Marking 352 method and it is carried by the Hop-by-Hop Options header and the 353 Destination Options header. In case of Destination Option, it is 354 processed only by the source and destination nodes: the source node 355 inserts and the destination node removes it. While, in case of Hop- 356 by-Hop Option, it may be examined by any node along the path, if 357 explicitly configured to do so. 359 It is important to highlight that the Option Layout can be used both 360 as Destination Option and as Hop-by-Hop Option depending on the Use 361 Cases and it is based on the chosen type of performance measurement. 362 In general, it is needed to perform both end to end and hop by hop 363 measurements, and the Alternate Marking methodology allows, by 364 definition, both performance measurements. In many cases the end-to- 365 end measurement is not enough and it is required the hop-by-hop 366 measurement, so the most complete choice can be the Hop-by-Hop 367 Options Header. 369 IPv6, as specified in [RFC8200], allows nodes to optionally process 370 Hop-by-Hop headers. Specifically the Hop-by-Hop Options header is 371 not inserted or deleted, but may be examined or processed by any node 372 along a packet's delivery path, until the packet reaches the node (or 373 each of the set of nodes, in the case of multicast) identified in the 374 Destination Address field of the IPv6 header. Also, it is expected 375 that nodes along a packet's delivery path only examine and process 376 the Hop-by-Hop Options header if explicitly configured to do so. 378 Another scenario that can be mentioned is the presence of a Routing 379 Header, in particular it is possible to consider SRv6. A new type of 380 Routing Header, referred as Segment Routing Header (SRH), has been 381 defined in [RFC8754] for SRv6. Like any other use case of IPv6, Hop- 382 by-Hop and Destination Options are usable when SRv6 header is 383 present. Because SRv6 is implemented through a Segment Routing 384 Header (SRH), Destination Options before the Routing Header are 385 processed by each destination in the route list, that means, in case 386 of SRH, by every SR node that is identified by the SR path. More 387 details about the SRv6 application are described in 388 [I-D.fz-spring-srv6-alt-mark]. 390 In summary, it is possible to list the alternative possibilities: 392 o Destination Option not preceding a Routing Header => measurement 393 only by node in Destination Address. 395 o Hop-by-Hop Option => every router on the path with feature 396 enabled. 398 o Destination Option preceding a Routing Header => every destination 399 node in the route list. 401 In general, Hop-by-Hop and Destination Options are the most suitable 402 ways to implement Alternate Marking. 404 It is worth mentioning that new Hop-by-Hop Options are not strongly 405 recommended in [RFC7045] and [RFC8200], unless there is a clear 406 justification to standardize it, because nodes may be configured to 407 ignore the Options Header, drop or assign packets containing an 408 Options Header to a slow processing path. In case of the AltMark 409 data fields described in this document, the motivation to standardize 410 a new Hop-by-Hop Option is that it is needed for OAM (Operations, 411 Administration, and Maintenance). An intermediate node can read it 412 or not, but this does not affect the packet behavior. The source 413 node is the only one that writes the Hop-by-Hop Option to mark 414 alternately the flow, so, the performance measurement can be done for 415 those nodes configured to read this Option, while the others are 416 simply not considered for the metrics. 418 The Hop-by-Hop Option defined in this document is designed to take 419 advantage of the property of how Hop-by-Hop options are processed. 420 Nodes that do not support this Option SHOULD ignore them. This can 421 mean that, in this case, the performance measurement does not account 422 for all links and nodes along a path. The definition of the Hop-by- 423 Hop Options in this document is also designed to minimize throughput 424 impact both on nodes that do not recognize the Option and on node 425 that support it. Indeed, the three high-order bits of the Options 426 Header defined in this draft are 000 and, in theory, as per [RFC8200] 427 and [I-D.hinden-6man-hbh-processing], this means "skip if do not 428 recognize and data do not change en route". [RFC8200] also mentions 429 that the nodes only examine and process the Hop-by-Hop Options header 430 if explicitly configured to do so. For these reasons, this Hop-by- 431 Hop Option should not affect the throughput. However, in practice, 432 it is important to be aware that the things may be different in the 433 implementation and it can happen that packets with Hop-by-Hop are 434 forced onto the slow path, but this is a general issue, as also 435 explained in [I-D.hinden-6man-hbh-processing]. It is also worth 436 mentioning that the application to a controlled domain should avoid 437 the risk of arbitrary nodes dropping packets with Hop-by-Hop Options. 439 5. Alternate Marking Method Operation 441 This section describes how the method operates. [RFC8321] introduces 442 several applicable methods which are reported below, and a new field 443 is introduced to facilitate the deployment and improve the 444 scalability. 446 5.1. Packet Loss Measurement 448 The measurement of the packet loss is really straightforward in 449 comparison to the existing mechanisms, as detailed in [RFC8321]. The 450 packets of the flow are grouped into batches, and all the packets 451 within a batch are marked by setting the L bit (Loss flag) to a same 452 value. The source node can switch the value of the L bit between 0 453 and 1 after a fixed number of packets or according to a fixed timer, 454 and this depends on the implementation. The source node is the only 455 one that marks the packets to create the batches, while the 456 intermediate nodes only read the marking values and identify the 457 packet batches. By counting the number of packets in each batch and 458 comparing the values measured by different network nodes along the 459 path, it is possible to measure the packet loss occurred in any 460 single batch between any two nodes. Each batch represents a 461 measurable entity recognizable by all network nodes along the path. 463 Both fixed number of packets and fixed timer can be used by the 464 source node to create packet batches. But, as also explained in 465 [RFC8321], the timer-based batches are preferable because they are 466 more deterministic than the counter-based batches. There is no 467 definitive rule for counter-based batches, differently from timer- 468 based batches. Using a fixed timer for the switching offers better 469 control over the method, indeed the length of the batches can be 470 chosen large enough to simplify the collection and the comparison of 471 the measures taken by different network nodes. In the implementation 472 the counters can be sent out by each node to the controller that is 473 responsible for the calculation. It is also possible to exchange 474 this information by using other on-path techniques. But this is out 475 of scope for this document. 477 Packets with different L values may get swapped at batch boundaries, 478 and in this case, it is required that each marked packet can be 479 assigned to the right batch by each router. It is important to 480 mention that for the application of this method there are two 481 elements to consider: the clock error between network nodes and the 482 network delay. These can create offsets between the batches and out- 483 of-order of the packets. The mathematical formula on timing aspects, 484 explained in section 3.2 of [RFC8321], must be satisfied and it takes 485 into considerations the different causes of reordering such as clock 486 error and network delay. The assumption is to define the available 487 counting interval where to get stable counters and to avoid these 488 issues. Specifically, if the effects of network delay are ignored, 489 the condition to implement the methodology is that the clocks in 490 different nodes MUST be synchronized to the same clock reference with 491 an accuracy of +/- B/2 time units, where B is the fixed time duration 492 of the batch, which refers to the original marking interval at the 493 source node considering that this interval could fluctuate along the 494 path. In this way each marked packet can be assigned to the right 495 batch by each node. Usually the counters can be taken in the middle 496 of the batch period to be sure to take still counters. In a few 497 words this implies that the length of the batches MUST be chosen 498 large enough so that the method is not affected by those factors. 499 The length of the batches can be determined based on the specific 500 deployment scenario. 502 L bit=1 ----------+ +-----------+ +---------- 503 | | | | 504 L bit=0 +-----------+ +-----------+ 505 Batch n ... Batch 3 Batch 2 Batch 1 506 <---------> <---------> <---------> <---------> <---------> 508 Traffic Flow 509 ===========================================================> 510 L bit ...1111111111 0000000000 11111111111 00000000000 111111111... 511 ===========================================================> 513 Figure 1: Packet Loss Measurement and Single-Marking Methodology 514 using L bit 516 It is worth mentioning that the duration of the batches is considered 517 stable over time in the previous figure. In theory, it is possible 518 to change the length of batches over time and among different flows 519 for more flexibility. But, in practice, it could complicate the 520 correlation of the information. 522 5.2. Packet Delay Measurement 524 The same principle used to measure packet loss can be applied also to 525 one-way delay measurement. Delay metrics MAY be calculated using the 526 two possibilities: 528 1. Single-Marking Methodology: This approach uses only the L bit to 529 calculate both packet loss and delay. In this case, the D flag 530 MUST be set to zero on transmit and ignored by the monitoring 531 points. The alternation of the values of the L bit can be used 532 as a time reference to calculate the delay. Whenever the L bit 533 changes and a new batch starts, a network node can store the 534 timestamp of the first packet of the new batch, that timestamp 535 can be compared with the timestamp of the first packet of the 536 same batch on a second node to compute packet delay. But this 537 measurement is accurate only if no packet loss occurs and if 538 there is no packet reordering at the edges of the batches. A 539 different approach can also be considered and it is based on the 540 concept of the mean delay. The mean delay for each batch is 541 calculated by considering the average arrival time of the packets 542 for the relative batch. There are limitations also in this case 543 indeed, each node needs to collect all the timestamps and 544 calculate the average timestamp for each batch. In addition, the 545 information is limited to a mean value. 547 2. Double-Marking Methodology: This approach is more complete and 548 uses the L bit only to calculate packet loss and the D bit (Delay 549 flag) is fully dedicated to delay measurements. The idea is to 550 use the first marking with the L bit to create the alternate flow 551 and, within the batches identified by the L bit, a second marking 552 is used to select the packets for measuring delay. The D bit 553 creates a new set of marked packets that are fully identified 554 over the network, so that a network node can store the timestamps 555 of these packets; these timestamps can be compared with the 556 timestamps of the same packets on a second node to compute packet 557 delay values for each packet. The most efficient and robust mode 558 is to select a single double-marked packet for each batch, in 559 this way there is no time gap to consider between the double- 560 marked packets to avoid their reorder. Regarding the rule for 561 the selection of the packet to be double-marked, the same 562 considerations in Section 5.1 apply also here and the double- 563 marked packet can be chosen within the available counting 564 interval that is not affected by factors such as clock errors. 565 If a double-marked packet is lost, the delay measurement for the 566 considered batch is simply discarded, but this is not a big 567 problem because it is easy to recognize the problematic batch and 568 skip the measurement just for that one. So in order to have more 569 information about the delay and to overcome out-of-order issues 570 this method is preferred. 572 In summary the approach with double marking is better than the 573 approach with single marking. Moreover, the two approaches provide 574 slightly different pieces of information and the data consumer can 575 combine them to have a more robust data set. 577 Similar to what said in Section 5.1 for the packet counters, in the 578 implementation the timestamps can be sent out to the controller that 579 is responsible for the calculation or could also be exchanged using 580 other on-path techniques. But this is out of scope for this 581 document. 583 L bit=1 ----------+ +-----------+ +---------- 584 | | | | 585 L bit=0 +-----------+ +-----------+ 587 D bit=1 + + + + + 588 | | | | | 589 D bit=0 ------+----------+----------+----------+------------+----- 591 Traffic Flow 592 ===========================================================> 593 L bit ...1111111111 0000000000 11111111111 00000000000 111111111... 595 D bit ...0000010000 0000010000 00000100000 00001000000 000001000... 596 ===========================================================> 598 Figure 2: Double-Marking Methodology using L bit and D bit 600 Likewise to packet delay measurement (both for Single Marking and 601 Double Marking), the method can also be used to measure the inter- 602 arrival jitter. 604 5.3. Flow Monitoring Identification 606 The Flow Monitoring Identification (FlowMonID) identifies the flow to 607 be measured and is required for some general reasons: 609 o First, it helps to reduce the per node configuration. Otherwise, 610 each node needs to configure an access-control list (ACL) for each 611 of the monitored flows. Moreover, using a flow identifier allows 612 a flexible granularity for the flow definition, indeed, it can be 613 used together with the 5-tuple. 615 o Second, it simplifies the counters handling. Hardware processing 616 of flow tuples (and ACL matching) is challenging and often incurs 617 into performance issues, especially in tunnel interfaces. 619 o Third, it eases the data export encapsulation and correlation for 620 the collectors. 622 The FlowMon identifier field is to uniquely identify a monitored flow 623 within the measurement domain. The field is set at the source node. 624 The FlowMonID can be set in two ways: 626 * It can be assigned by the central controller. Since the 627 controller knows the network topology, it can set the value 628 properly to avoid or minimize ambiguity and guarantee the 629 uniqueness. In this regard, the controller can simply verify that 630 there is no ambiguity between different pseudo-randomly generated 631 FlowMonIDs on the same path. 633 * It can be algorithmically generated by the source node, that can 634 set it pseudo-randomly with some chance of collision. This 635 approach cannot guarantee the uniqueness of FlowMonID but it may 636 be preferred for local or private networks, where the conflict 637 probability is small due to the large FlowMonID space. 639 The value of 20 bits has been selected for the FlowMonID since it is 640 a good compromise and implies a low rate of ambiguous FlowMonIDs that 641 can be considered acceptable in most of the applications. Indeed, 642 with 20 bits the number of combinations is 1048576. 644 If the FlowMonID is set by the source node, the intermediate nodes 645 can read the FlowMonIDs from the packets in flight and act 646 accordingly. While, if the FlowMonID is set by the controller, both 647 possibilities are feasible for the intermediate nodes which can learn 648 by reading the packets or can be instructed by the controller. 650 When all values in the FlowMonID space are consumed, the centralized 651 controller can keep track and reassign the values that are not used 652 any more by old flows, while if the FlowMonID is pseudo randomly 653 generated by the source, conflicts and collisions are possible. 655 5.3.1. Uniqueness of FlowMonID 657 It is important to note that if the 20 bit FlowMonID is set 658 independently and pseudo randomly there is a chance of collision. 659 Indeed, by using the well-known birthday problem in probability 660 theory, if the 20 bit FlowMonID is set independently and pseudo 661 randomly without any additional input entropy, there is a 50% chance 662 of collision for 1206 flows. So, for more entropy, FlowMonID can 663 either be combined with other identifying flow information in a 664 packet (e.g. it is possible to consider the hashed 3-tuple Flow 665 Label, Source and Destination addresses). 667 This issue is more visible when the FlowMonID is pseudo randomly 668 generated by the source node and there needs to tag it with 669 additional flow information to allow disambiguation. While, in case 670 of a centralized controller, the controller should consider these 671 aspects and instruct the nodes properly in order to guarantee its 672 uniqueness. 674 It is worth highlighting that in most of the applications a low rate 675 of ambiguous FlowMonIDs can be acceptable, since this only affects 676 the measurement. For large scale measurements, where it is possible 677 to monitor a big number of flows, the disambiguation of the FlowMonID 678 field is something to take into account. 680 5.4. Multipoint and Clustered Alternate Marking 682 The Alternate Marking method can also be extended to any kind of 683 multipoint to multipoint paths, and the network clustering approach 684 allows a flexible and optimized performance measurement, as described 685 in [RFC8889]. 687 The Cluster is the smallest identifiable subnetwork of the entire 688 Network graph that still satisfies the condition that the number of 689 packets that goes in is the same that goes out. With network 690 clustering, it is possible to use the partition of the network into 691 clusters at different levels in order to perform the needed degree of 692 detail. So, for Multipoint Alternate Marking, FlowMonID can identify 693 in general a multipoint-to-multipoint flow and not only a point-to- 694 point flow. 696 5.5. Data Collection and Calculation 698 The nodes enabled to perform performance monitoring collect the value 699 of the packet counters and timestamps. There are several 700 alternatives to implement Data Collection and Calculation, but this 701 is not specified in this document. 703 There are documents on the control plane mechanisms of Alternate 704 Marking, e.g. [I-D.ietf-idr-sr-policy-ifit], 705 [I-D.chen-pce-pcep-ifit]. 707 6. Security Considerations 709 This document aims to apply a method to perform measurements that 710 does not directly affect Internet security nor applications that run 711 on the Internet. However, implementation of this method must be 712 mindful of security and privacy concerns. 714 There are two types of security concerns: potential harm caused by 715 the measurements and potential harm to the measurements. 717 Harm caused by the measurement: Alternate Marking implies 718 modifications on the fly to an Option Header of IPv6 packets by the 719 source node, but this must be performed in a way that does not alter 720 the quality of service experienced by the packets and that preserves 721 stability and performance of routers doing the measurements. As 722 already discussed in Section 4, it is RECOMMENDED that the AltMark 723 Option does not affect the throughput and therefore the user 724 experience. 726 Harm to the measurement: Alternate Marking measurements could be 727 harmed by routers altering the fields of the AltMark Option (e.g. 728 marking of the packets, FlowMonID) or by a malicious attacker adding 729 AltMark Option to the packets in order to consume the resources of 730 network devices and entities involved. As described above, the 731 source node is the only one that writes the Option Header while the 732 intermediate nodes and destination node only read it without 733 modifying the Option Header. But, for example, an on-path attacker 734 can modify the flags, whether intentionally or accidentally, or 735 deliberately insert a new option to the packet flow or delete the 736 option from the packet flow. The consequent effect could be to give 737 the appearance of loss or delay or invalidate the measurement by 738 modifying option identifiers, such as FlowMonID. The malicious 739 implication can be to cause actions from the network administrator 740 where an intervention is not necessary or to hide real issues in the 741 network. Since the measurement itself may be affected by network 742 nodes intentionally altering the bits of the AltMark Option or 743 injecting Options headers as a means for Denial of Service (DoS), the 744 Alternate Marking MUST be applied in the context of a controlled 745 domain, where the network nodes are locally administered and this 746 type of attack can be avoided. For this reason, the implementation 747 of the method is not done on the end node if it is not fully managed 748 and does not belong to the controlled domain. Packets generated 749 outside the controlled domain may consume router resources by 750 maliciously using the HbH Option, but this can be mitigated by 751 filtering these packets at the controlled domain boundary. This can 752 be done because, if the end node does not belong to the controlled 753 domain, it is not supposed to add the AltMark HbH Option, and it can 754 be easily recognized. 756 The flow identifier (FlowMonID) composes the AltMark Option together 757 with the two marking bits (L and D). As explained in Section 5.3.1, 758 there is a chance of collision if the FlowMonID is set pseudo 759 randomly and a solution exists. In general this may not be a problem 760 and a low rate of ambiguous FlowMonIDs can be acceptable, since this 761 does not cause significant harm to the operators or their clients and 762 this harm may not justify the complications of avoiding it. But, for 763 large scale measurements, a big number of flows could be monitored 764 and the probability of a collision is higher, thus the disambiguation 765 of the FlowMonID field can be considered. 767 The privacy concerns also need to be analyzed even if the method only 768 relies on information contained in the Option Header without any 769 release of user data. Indeed, from a confidentiality perspective, 770 although AltMark Option does not contain user data, the metadata can 771 be used for network reconnaissance to compromise the privacy of users 772 by allowing attackers to collect information about network 773 performance and network paths. AltMark Option contains two kinds of 774 metadata: the marking bits (L and D bits) and the flow identifier 775 (FlowMonID). 777 The marking bits are the small information that is exchanged 778 between the network nodes. Therefore, due to this intrinsic 779 characteristic, network reconnaissance through passive 780 eavesdropping on data-plane traffic is difficult. Indeed, an 781 attacker cannot gain information about network performance from a 782 single monitoring point. The only way for an attacker can be to 783 eavesdrop on multiple monitoring points at the same time, because 784 they have to do the same kind of calculation and aggregation as 785 Alternate Marking requires. 787 The FlowMonID field is used in the AltMark Option as the 788 identifier of the monitored flow. It represents a more sensitive 789 information for network reconnaissance and may allow a flow 790 tracking type of attack because an attacker could collect 791 information about network paths. 793 Furthermore, in a pervasive surveillance attack, the information that 794 can be derived over time is more. But, as further described 795 hereinafter, the application of the Alternate Marking to a controlled 796 domain helps to mitigate all the above aspects of privacy concerns. 798 At the management plane, attacks can be set up by misconfiguring or 799 by maliciously configuring AltMark Option. Thus, AltMark Option 800 configuration MUST be secured in a way that authenticates authorized 801 users and verifies the integrity of configuration procedures. 802 Solutions to ensure the integrity of AltMark Option are outside the 803 scope of this document. Also, attacks on the reporting of the 804 statistics between the monitoring points and the network management 805 system (e.g. centralized controller) can interfere with the proper 806 functioning of the system. Hence, the channels used to report back 807 flow statistics MUST be secured. 809 As stated above, the precondition for the application of the 810 Alternate Marking is that it MUST be applied in specific controlled 811 domains, thus confining the potential attack vectors within the 812 network domain. [RFC8799] analyzes and discusses the trend towards 813 network behaviors that can be applied only within a limited domain. 814 This is due to the specific set of requirements especially related to 815 security, network management, policies and options supported which 816 may vary between such limited domains. A limited administrative 817 domain provides the network administrator with the means to select, 818 monitor and control the access to the network, making it a trusted 819 domain. In this regard it is expected to enforce policies at the 820 domain boundaries to filter both external packets with AltMark Option 821 entering the domain and internal packets with AltMark Option leaving 822 the domain. Therefore, the trusted domain is unlikely subject to 823 hijacking of packets since packets with AltMark Option are processed 824 and used only within the controlled domain. 826 Additionally, it is to be noted that the AltMark Option is carried by 827 the Options Header and it may have some impact on the packet sizes 828 for the monitored flow and on the path MTU, since some packets might 829 exceed the MTU. However, the relative small size (48 bit in total) 830 of these Option Headers and its application to a controlled domain 831 help to mitigate the problem. 833 It is worth mentioning that the security concerns may change based on 834 the specific deployment scenario and related threat analysis, which 835 can lead to specific security solutions that are beyond the scope of 836 this document. As an example, the AltMark Option can be used as Hop- 837 by-Hop or Destination Option and, in case of Destination Option, 838 multiple administrative domains may be traversed by the AltMark 839 Option that is not confined to a single administrative domain. In 840 this case, the user, aware of the kind of risks, may still want to 841 use Alternate Marking for telemetry and test purposes but the 842 controlled domain must be composed by more than one administrative 843 domains. To this end, the inter-domain links need to be secured 844 (e.g., by IPsec, VPNs) in order to avoid external threats and realize 845 the whole controlled domain. 847 It might be theoretically possible to modulate the marking or the 848 other fields of the AltMark Option to serve as a covert channel to be 849 used by an on-path observer. This may affect both the data and 850 management plane, but, here too, the application to a controlled 851 domain helps to reduce the effects. 853 The Alternate Marking application described in this document relies 854 on a time synchronization protocol. Thus, by attacking the time 855 protocol, an attacker can potentially compromise the integrity of the 856 measurement. A detailed discussion about the threats against time 857 protocols and how to mitigate them is presented in [RFC7384]. 858 Network Time Security (NTS), described in [RFC8915], is a mechanism 859 that can be employed. Also, the time, which is distributed to the 860 network nodes through the time protocol, is centrally taken from an 861 external accurate time source, such as an atomic clock or a GPS 862 clock. By attacking the time source it can be possible to compromise 863 the integrity of the measurement as well. There are security 864 measures that can be taken to mitigate the GPS spoofing attacks and a 865 network administrator should certainly employ solutions to secure the 866 network domain. 868 7. IANA Considerations 870 The Option Type should be assigned in IANA's "Destination Options and 871 Hop-by-Hop Options" registry. 873 This draft requests the following IPv6 Option Type assignment from 874 the Destination Options and Hop-by-Hop Options sub-registry of 875 Internet Protocol Version 6 (IPv6) Parameters 876 (https://www.iana.org/assignments/ipv6-parameters/). 878 Hex Value Binary Value Description Reference 879 act chg rest 880 ---------------------------------------------------------------- 881 TBD 00 0 tbd AltMark [This draft] 883 8. Acknowledgements 885 The authors would like to thank Bob Hinden, Ole Troan, Martin Duke, 886 Lars Eggert, Roman Danyliw, Alvaro Retana, Eric Vyncke, Warren 887 Kumari, Benjamin Kaduk, Stewart Bryant, Christopher Wood, Yoshifumi 888 Nishida, Tom Herbert, Stefano Previdi, Brian Carpenter, Greg Mirsky, 889 Ron Bonica for the precious comments and suggestions. 891 9. References 893 9.1. Normative References 895 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 896 Requirement Levels", BCP 14, RFC 2119, 897 DOI 10.17487/RFC2119, March 1997, 898 . 900 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 901 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 902 May 2017, . 904 [RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6 905 (IPv6) Specification", STD 86, RFC 8200, 906 DOI 10.17487/RFC8200, July 2017, 907 . 909 9.2. Informative References 911 [I-D.chen-pce-pcep-ifit] 912 Yuan, H., Zhou, T., Li, W., Fioccola, G., and Y. Wang, 913 "Path Computation Element Communication Protocol (PCEP) 914 Extensions to Enable IFIT", draft-chen-pce-pcep-ifit-04 915 (work in progress), July 2021. 917 [I-D.fz-spring-srv6-alt-mark] 918 Fioccola, G., Zhou, T., and M. Cociglio, "Segment Routing 919 Header encapsulation for Alternate Marking Method", draft- 920 fz-spring-srv6-alt-mark-01 (work in progress), July 2021. 922 [I-D.hinden-6man-hbh-processing] 923 Hinden, R. M. and G. Fairhurst, "IPv6 Hop-by-Hop Options 924 Processing Procedures", draft-hinden-6man-hbh- 925 processing-01 (work in progress), June 2021. 927 [I-D.ietf-idr-sr-policy-ifit] 928 Qin, F., Yuan, H., Zhou, T., Fioccola, G., and Y. Wang, 929 "BGP SR Policy Extensions to Enable IFIT", draft-ietf-idr- 930 sr-policy-ifit-02 (work in progress), July 2021. 932 [I-D.peng-v6ops-hbh] 933 Peng, S., Li, Z., Xie, C., Qin, Z., and G. Mishra, 934 "Processing of the Hop-by-Hop Options Header", draft-peng- 935 v6ops-hbh-06 (work in progress), August 2021. 937 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 938 Architecture", RFC 4291, DOI 10.17487/RFC4291, February 939 2006, . 941 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 942 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 943 December 2005, . 945 [RFC6437] Amante, S., Carpenter, B., Jiang, S., and J. Rajahalme, 946 "IPv6 Flow Label Specification", RFC 6437, 947 DOI 10.17487/RFC6437, November 2011, 948 . 950 [RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label 951 for Equal Cost Multipath Routing and Link Aggregation in 952 Tunnels", RFC 6438, DOI 10.17487/RFC6438, November 2011, 953 . 955 [RFC7045] Carpenter, B. and S. Jiang, "Transmission and Processing 956 of IPv6 Extension Headers", RFC 7045, 957 DOI 10.17487/RFC7045, December 2013, 958 . 960 [RFC7384] Mizrahi, T., "Security Requirements of Time Protocols in 961 Packet Switched Networks", RFC 7384, DOI 10.17487/RFC7384, 962 October 2014, . 964 [RFC8321] Fioccola, G., Ed., Capello, A., Cociglio, M., Castaldelli, 965 L., Chen, M., Zheng, L., Mirsky, G., and T. Mizrahi, 966 "Alternate-Marking Method for Passive and Hybrid 967 Performance Monitoring", RFC 8321, DOI 10.17487/RFC8321, 968 January 2018, . 970 [RFC8754] Filsfils, C., Ed., Dukes, D., Ed., Previdi, S., Leddy, J., 971 Matsushima, S., and D. Voyer, "IPv6 Segment Routing Header 972 (SRH)", RFC 8754, DOI 10.17487/RFC8754, March 2020, 973 . 975 [RFC8799] Carpenter, B. and B. Liu, "Limited Domains and Internet 976 Protocols", RFC 8799, DOI 10.17487/RFC8799, July 2020, 977 . 979 [RFC8889] Fioccola, G., Ed., Cociglio, M., Sapio, A., and R. Sisto, 980 "Multipoint Alternate-Marking Method for Passive and 981 Hybrid Performance Monitoring", RFC 8889, 982 DOI 10.17487/RFC8889, August 2020, 983 . 985 [RFC8915] Franke, D., Sibold, D., Teichel, K., Dansarie, M., and R. 986 Sundblad, "Network Time Security for the Network Time 987 Protocol", RFC 8915, DOI 10.17487/RFC8915, September 2020, 988 . 990 Authors' Addresses 992 Giuseppe Fioccola 993 Huawei 994 Riesstrasse, 25 995 Munich 80992 996 Germany 998 Email: giuseppe.fioccola@huawei.com 1000 Tianran Zhou 1001 Huawei 1002 156 Beiqing Rd. 1003 Beijing 100095 1004 China 1006 Email: zhoutianran@huawei.com 1008 Mauro Cociglio 1009 Telecom Italia 1010 Via Reiss Romoli, 274 1011 Torino 10148 1012 Italy 1014 Email: mauro.cociglio@telecomitalia.it 1016 Fengwei Qin 1017 China Mobile 1018 32 Xuanwumenxi Ave. 1019 Beijing 100032 1020 China 1022 Email: qinfengwei@chinamobile.com 1024 Ran Pang 1025 China Unicom 1026 9 Shouti South Rd. 1027 Beijing 100089 1028 China 1030 Email: pangran@chinaunicom.cn