idnits 2.17.1 draft-ietf-ippm-alt-mark-08.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 (September 5, 2017) is 2424 days in the past. Is this intentional? Checking references for intended status: Experimental ---------------------------------------------------------------------------- == Outdated reference: A later version (-12) exists of draft-ietf-bier-mpls-encapsulation-07 == Outdated reference: A later version (-15) exists of draft-ietf-bier-pmmm-oam-02 == Outdated reference: A later version (-07) exists of draft-ietf-mpls-flow-ident-05 == Outdated reference: A later version (-10) exists of draft-ietf-mpls-rfc6374-sfl-00 == Outdated reference: A later version (-12) exists of draft-ietf-nvo3-encap-00 Summary: 0 errors (**), 0 flaws (~~), 6 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group G. Fioccola, Ed. 3 Internet-Draft A. Capello, Ed. 4 Intended status: Experimental M. Cociglio 5 Expires: March 9, 2018 L. Castaldelli 6 Telecom Italia 7 M. Chen, Ed. 8 L. Zheng, Ed. 9 Huawei Technologies 10 G. Mirsky, Ed. 11 ZTE 12 T. Mizrahi, Ed. 13 Marvell 14 September 5, 2017 16 Alternate Marking method for passive and hybrid performance monitoring 17 draft-ietf-ippm-alt-mark-08 19 Abstract 21 This document describes a method to perform packet loss, delay and 22 jitter measurements on live traffic. This method is based on 23 Alternate Marking (Coloring) technique. A report on the operational 24 experiment done at Telecom Italia is explained in order to give an 25 example and show the method applicability. This technique can be 26 applied in various situations as detailed in this document and could 27 be considered passive or hybrid depending on the application. 29 Requirements Language 31 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 32 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 33 "OPTIONAL" in this document are to be interpreted as described in BCP 34 14 [RFC2119] [RFC8174] when, and only when, they appear in all 35 capitals, as shown here. 37 Status of This Memo 39 This Internet-Draft is submitted in full conformance with the 40 provisions of BCP 78 and BCP 79. 42 Internet-Drafts are working documents of the Internet Engineering 43 Task Force (IETF). Note that other groups may also distribute 44 working documents as Internet-Drafts. The list of current Internet- 45 Drafts is at https://datatracker.ietf.org/drafts/current/. 47 Internet-Drafts are draft documents valid for a maximum of six months 48 and may be updated, replaced, or obsoleted by other documents at any 49 time. It is inappropriate to use Internet-Drafts as reference 50 material or to cite them other than as "work in progress." 52 This Internet-Draft will expire on March 9, 2018. 54 Copyright Notice 56 Copyright (c) 2017 IETF Trust and the persons identified as the 57 document authors. All rights reserved. 59 This document is subject to BCP 78 and the IETF Trust's Legal 60 Provisions Relating to IETF Documents 61 (https://trustee.ietf.org/license-info) in effect on the date of 62 publication of this document. Please review these documents 63 carefully, as they describe your rights and restrictions with respect 64 to this document. Code Components extracted from this document must 65 include Simplified BSD License text as described in Section 4.e of 66 the Trust Legal Provisions and are provided without warranty as 67 described in the Simplified BSD License. 69 Table of Contents 71 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 72 2. Overview of the method . . . . . . . . . . . . . . . . . . . 4 73 3. Detailed description of the method . . . . . . . . . . . . . 6 74 3.1. Packet loss measurement . . . . . . . . . . . . . . . . . 6 75 3.1.1. Coloring the packets . . . . . . . . . . . . . . . . 11 76 3.1.2. Counting the packets . . . . . . . . . . . . . . . . 11 77 3.1.3. Collecting data and calculating packet loss . . . . . 12 78 3.2. Timing aspects . . . . . . . . . . . . . . . . . . . . . 12 79 3.3. One-way delay measurement . . . . . . . . . . . . . . . . 14 80 3.3.1. Single marking methodology . . . . . . . . . . . . . 14 81 3.3.2. Double marking methodology . . . . . . . . . . . . . 16 82 3.4. Delay variation measurement . . . . . . . . . . . . . . . 17 83 4. Considerations . . . . . . . . . . . . . . . . . . . . . . . 18 84 4.1. Synchronization . . . . . . . . . . . . . . . . . . . . . 18 85 4.2. Data Correlation . . . . . . . . . . . . . . . . . . . . 18 86 4.3. Packet Re-ordering . . . . . . . . . . . . . . . . . . . 19 87 5. Implementation and deployment . . . . . . . . . . . . . . . . 20 88 5.1. Report on the operational experiment at Telecom Italia . 20 89 5.1.1. Metric transparency . . . . . . . . . . . . . . . . . 21 90 5.2. IP flow performance measurement (IPFPM) . . . . . . . . . 22 91 5.3. OAM Passive Performance Measurement . . . . . . . . . . . 22 92 5.4. RFC6374 Use Case . . . . . . . . . . . . . . . . . . . . 22 93 5.5. Application to active performance measurement . . . . . . 23 94 6. Hybrid measurement . . . . . . . . . . . . . . . . . . . . . 23 95 7. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 96 8. Compliance with RFC6390 guidelines . . . . . . . . . . . . . 24 97 9. Security Considerations . . . . . . . . . . . . . . . . . . . 25 98 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 27 99 11. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 27 100 12. References . . . . . . . . . . . . . . . . . . . . . . . . . 27 101 12.1. Normative References . . . . . . . . . . . . . . . . . . 27 102 12.2. Informative References . . . . . . . . . . . . . . . . . 28 103 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 30 105 1. Introduction 107 Nowadays, most of the traffic in Service Providers' networks carries 108 contents that are highly sensitive to packet loss [RFC7680], delay 109 [RFC7679], and jitter [RFC3393]. 111 In view of this scenario, Service Providers need methodologies and 112 tools to monitor and measure network performances with an adequate 113 accuracy, in order to constantly control the quality of experience 114 perceived by their customers. On the other hand, performance 115 monitoring provides useful information for improving network 116 management (e.g. isolation of network problems, troubleshooting, 117 etc.). 119 A lot of work related to OAM, that includes also performance 120 monitoring techniques, has been done by Standards Developing 121 Organizations(SDOs): [RFC7276] provides a good overview of existing 122 OAM mechanisms defined in IETF, ITU-T and IEEE. Considering IETF, a 123 lot of work has been done on fault detection and connectivity 124 verification, while a minor effort has been dedicated so far to 125 performance monitoring. The IPPM WG has defined standard metrics to 126 measure network performance; however, the methods developed in this 127 WG mainly refer to focus on active measurement techniques. More 128 recently, the MPLS WG has defined mechanisms for measuring packet 129 loss, one-way and two-way delay, and delay variation in MPLS 130 networks[RFC6374], but their applicability to passive measurements 131 has some limitations, especially for pure connection-less networks. 133 The lack of adequate tools to measure packet loss with the desired 134 accuracy drove an effort to design a new method for the performance 135 monitoring of live traffic, possibly easy to implement and deploy. 136 The effort led to the method described in this document: basically, 137 it is a passive performance monitoring technique, potentially 138 applicable to any kind of packet based traffic, including Ethernet, 139 IP, and MPLS, both unicast and multicast. The method addresses 140 primarily packet loss measurement, but it can be easily extended to 141 one-way delay and delay variation measurements as well. 143 The method has been explicitly designed for passive measurements but 144 it can also be used with active probes. Passive measurements are 145 usually more easily understood by customers and provide a much better 146 accuracy, especially for packet loss measurements. 148 RFC 7799 [RFC7799] defines passive and hybrid methods of measurement. 149 In particular, Passive Methods of Measurement are based solely on 150 observations of an undisturbed and unmodified packet stream of 151 interest; Hybrid Methods are Methods of Measurement that use a 152 combination of Active Methods and Passive Methods. 154 Taking into consideration these definitions, Alternate Marking Method 155 could be considered Hybrid or Passive depending on the case. In case 156 the marking field is obtained by changing existing field values of 157 the packets (e.g. DSCP field), the technique is Hybrid. In case the 158 marking field is dedicated, reserved and is included in the protocol 159 specification Alternate Marking technique can be considered as 160 Passive (e.g. RFC6374 Synonymous Flow Label or OAM Marking Bits in 161 BIER Header). 163 This document is organized as follows: 165 o Section 2 gives an overview of the method, including a comparison 166 with different measurement strategies; 168 o Section 3 describes the method in detail; 170 o Section 4 reports considerations about synchronization, data 171 correlation and packet re-ordering; 173 o Section 5 reports examples of implementation and deployment of the 174 method. Furthermore the operational experiment done at Telecom 175 Italia is described; 177 o Section 6 introduces Hybrid measurement aspects; 179 o Section 7 is about the Compliance with RFC6390 guidelines; 181 o Section 8 includes some security aspects; 183 o Section 9 finally summarizes some concluding remarks. 185 2. Overview of the method 187 In order to perform packet loss measurements on a live traffic flow, 188 different approaches exist. The most intuitive one consists in 189 numbering the packets, so that each router that receives the flow can 190 immediately detect a packet missing. This approach, though very 191 simple in theory, is not simple to achieve: it requires the insertion 192 of a sequence number into each packet and the devices must be able to 193 extract the number and check it in real time. Such a task can be 194 difficult to implement on live traffic: if UDP is used as the 195 transport protocol, the sequence number is not available; on the 196 other hand, if a higher layer sequence number (e.g. in the RTP 197 header) is used, extracting that information from each packet and 198 process it in real time could overload the device. 200 An alternate approach is to count the number of packets sent on one 201 end, the number of packets received on the other end, and to compare 202 the two values. This operation is much simpler to implement, but 203 requires that the devices performing the measurement are in sync: in 204 order to compare two counters it is required that they refer exactly 205 to the same set of packets. Since a flow is continuous and cannot be 206 stopped when a counter has to be read, it could be difficult to 207 determine exactly when to read the counter. A possible solution to 208 overcome this problem is to virtually split the flow in consecutive 209 blocks by inserting periodically a delimiter so that each counter 210 refers exactly to the same block of packets. The delimiter could be 211 for example a special packet inserted artificially into the flow. 212 However, delimiting the flow using specific packets has some 213 limitations. First, it requires generating additional packets within 214 the flow and requires the equipment to be able to process those 215 packets. In addition, the method is vulnerable to out of order 216 reception of delimiting packets and, to a lesser extent, to their 217 loss. 219 The method proposed in this document follows the second approach, but 220 it doesn't use additional packets to virtually split the flow in 221 blocks. Instead, it "colors" the packets so that the packets 222 belonging to the same block will have the same color, whilst 223 consecutive blocks will have different colors. Each change of color 224 represents a sort of auto-synchronization signal that guarantees the 225 consistency of measurements taken by different devices along the 226 path. 228 Figure 1 represents a very simple network and shows how the method 229 can be used to measure packet loss on different network segments: by 230 enabling the measurement on several interfaces along the path, it is 231 possible to perform link monitoring, node monitoring or end-to-end 232 monitoring. The method is flexible enough to measure packet loss on 233 any segment of the network and can be used to isolate the faulty 234 element. 236 Traffic flow 237 ========================================================> 238 +------+ +------+ +------+ +------+ 239 ---<> R1 <>-----<> R2 <>-----<> R3 <>-----<> R4 <>--- 240 +------+ +------+ +------+ +------+ 241 . . . . . . 242 . . . . . . 243 . <------> <-------> . 244 . Node Packet Loss Link Packet Loss . 245 . . 246 <---------------------------------------------------> 247 End-to-End Packet loss 249 Figure 1: Available measurements 251 3. Detailed description of the method 253 This section describes in detail how the method operate. A special 254 emphasis is given to the measurement of packet loss, that represents 255 the core application of the method, but applicability to delay and 256 jitter measurements is also considered. 258 3.1. Packet loss measurement 260 The basic idea is to virtually split traffic flows into consecutive 261 blocks: each block represents a measurable entity unambiguously 262 recognizable by all network devices along the path. By counting the 263 number of packets in each block and comparing the values measured by 264 different network devices along the path, it is possible to measure 265 packet loss occurred in any single block between any two points. 267 As discussed in the previous section, a simple way to create the 268 blocks is to "color" the traffic (two colors are sufficient) so that 269 packets belonging to different consecutive blocks will have different 270 colors. Whenever the color changes, the previous block terminates 271 and the new one begins. Hence, all the packets belonging to the same 272 block will have the same color and packets of different consecutive 273 blocks will have different colors. The number of packets in each 274 block depends on the criterion used to create the blocks: 276 o if the color is switched after a fixed number of packets, then 277 each block will contain the same number of packets (except for any 278 losses); 280 o if the color is switched according to a fixed timer, then the 281 number of packets may be different in each block depending on the 282 packet rate. 284 The following figure shows how a flow looks like when it is split in 285 traffic blocks with colored packets. 287 A: packet with A coloring 288 B: packet with B coloring 290 | | | | | 291 | | Traffic flow | | 292 -------------------------------------------------------------------> 293 BBBBBBB AAAAAAAAAAA BBBBBBBBBBB AAAAAAAAAAA BBBBBBBBBBB AAAAAAA 294 -------------------------------------------------------------------> 295 ... | Block 5 | Block 4 | Block 3 | Block 2 | Block 1 296 | | | | | 298 Figure 2: Traffic coloring 300 Figure 3 shows how the method can be used to measure link packet loss 301 between two adjacent nodes. 303 Referring to the figure, let's assume we want to monitor the packet 304 loss on the link between two routers: router R1 and router R2. 305 According to the method, the traffic is colored alternatively with 306 two different colors, A and B. Whenever the color changes, the 307 transition generates a sort of square-wave signal, as depicted in the 308 following figure. 310 Color A ----------+ +-----------+ +---------- 311 | | | | 312 Color B +-----------+ +-----------+ 313 Block n ... Block 3 Block 2 Block 1 314 <---------> <---------> <---------> <---------> <---------> 316 Traffic flow 317 ===========================================================> 318 Color ...AAAAAAAAAAA BBBBBBBBBBB AAAAAAAAAAA BBBBBBBBBBB AAAAAAA... 319 ===========================================================> 321 Figure 3: Computation of link packet loss 323 Traffic coloring could be done by R1 itself or by an upward router. 324 R1 needs two counters, C(A)R1 and C(B)R1, on its egress interface: 325 C(A)R1 counts the packets with color A and C(B)R1 counts those with 326 color B. As long as traffic is colored A, only counter C(A)R1 will 327 be incremented, while C(B)R1 is not incremented; vice versa, when the 328 traffic is colored as B, only C(B)R1 is incremented. C(A)R1 and 329 C(B)R1 can be used as reference values to determine the packet loss 330 from R1 to any other measurement point down the path. Router R2, 331 similarly, will need two counters on its ingress interface, C(A)R2 332 and C(B)R2, to count the packets received on that interface and 333 colored with color A and B respectively. When an A block ends, it is 334 possible to compare C(A)R1 and C(A)R2 and calculate the packet loss 335 within the block; similarly, when the successive B block terminates, 336 it is possible to compare C(B)R1 with C(B)R2, and so on for every 337 successive block. 339 Likewise, by using two counters on R2 egress interface it is possible 340 to count the packets sent out of R2 interface and use them as 341 reference values to calculate the packet loss from R2 to any 342 measurement point down R2. 344 Using a fixed timer for color switching offers a better control over 345 the method: the (time) length of the blocks can be chosen large 346 enough to simplify the collection and the comparison of measures 347 taken by different network devices. It's preferable to read the 348 value of the counters not immediately after the color switch: some 349 packets could arrive out of order and increment the counter 350 associated to the previous block (color), so it is worth waiting for 351 some time. A safe choice is to wait L/2 time units (where L is the 352 duration for each block) after the color switch, to read the still 353 counter of the previous color, so the possibility to read a running 354 counter instead of a still one is minimized. The drawback is that 355 the longer the duration of the block, the less frequent the 356 measurement can be taken. 358 The following table shows how the counters can be used to calculate 359 the packet loss between R1 and R2. The first column lists the 360 sequence of traffic blocks while the other columns contain the 361 counters of A-colored packets and B-colored packets for R1 and R2. 362 In this example, we assume that the values of the counters are reset 363 to zero whenever a block ends and its associated counter has been 364 read: with this assumption, the table shows only relative values, 365 that is the exact number of packets of each color within each block. 366 If the values of the counters were not reset, the table would contain 367 cumulative values, but the relative values could be determined simply 368 by difference from the value of the previous block of the same color. 370 The color is switched on the basis of a fixed timer (not shown in the 371 table), so the number of packets in each block is different. 373 +-------+--------+--------+--------+--------+------+ 374 | Block | C(A)R1 | C(B)R1 | C(A)R2 | C(B)R2 | Loss | 375 +-------+--------+--------+--------+--------+------+ 376 | 1 | 375 | 0 | 375 | 0 | 0 | 377 | | | | | | | 378 | 2 | 0 | 388 | 0 | 388 | 0 | 379 | | | | | | | 380 | 3 | 382 | 0 | 381 | 0 | 1 | 381 | | | | | | | 382 | 4 | 0 | 377 | 0 | 374 | 3 | 383 | | | | | | | 384 | ... | ... | ... | ... | ... | ... | 385 | | | | | | | 386 | 2n | 0 | 387 | 0 | 387 | 0 | 387 | | | | | | | 388 | 2n+1 | 379 | 0 | 377 | 0 | 2 | 389 +-------+--------+--------+--------+--------+------+ 391 Table 1: Evaluation of counters for packet loss measurements 393 During an A block (blocks 1, 3 and 2n+1), all the packets are 394 A-colored, therefore the C(A) counters are incremented to the number 395 seen on the interface, while C(B) counters are zero. Vice versa, 396 during a B block (blocks 2, 4 and 2n), all the packets are B-colored: 397 C(A) counters are zero, while C(B) counters are incremented. 399 When a block ends (because of color switching) the relative counters 400 stop incrementing and it is possible to read them, compare the values 401 measured on router R1 and R2 and calculate the packet loss within 402 that block. 404 For example, looking at the table above, during the first block 405 (A-colored), C(A)R1 and C(A)R2 have the same value (375), which 406 corresponds to the exact number of packets of the first block (no 407 loss). Also during the second block (B-colored) R1 and R2 counters 408 have the same value (388), which corresponds to the number of packets 409 of the second block (no loss). During blocks three and four, R1 and 410 R2 counters are different, meaning that some packets have been lost: 411 in the example, one single packet (382-381) was lost during block 412 three and three packets (377-374) were lost during block four. 414 The method applied to R1 and R2 can be extended to any other router 415 and applied to more complex networks, as far as the measurement is 416 enabled on the path followed by the traffic flow(s) being observed. 418 It's worth mentioning two different strategies that can be used when 419 implementing the method: 421 o flow-based: the flow-based strategy is used when only a limited 422 number of traffic flows need to be monitored. According to this 423 strategy, only a subset of the flows is colored. Counters for 424 packet loss measurements can be instantiated for each single flow, 425 or for the set as a whole, depending on the desired granularity. 426 A relevant problem with this approach is the necessity to know in 427 advance the path followed by flows that are subject to 428 measurement. Path rerouting and traffic load-balancing increase 429 the issue complexity, especially for unicast traffic. The problem 430 is easier to solve for multicast traffic where load balancing is 431 seldom used and static joins are frequently used to force traffic 432 forwarding and replication. 434 o link-based: measurements are performed on all the traffic on a 435 link by link basis. The link could be a physical link or a 436 logical link. Counters could be instantiated for the traffic as a 437 whole or for each traffic class (in case it is desired to monitor 438 each class separately), but in the second case a couple of 439 counters is needed for each class. 441 As mentioned, the flow-based measurement requires the identification 442 of the flow to be monitored and the discovery of the path followed by 443 the selected flow. It is possible to monitor a single flow or 444 multiple flows grouped together, but in this case measurement is 445 consistent only if all the flows in the group follow the same path. 446 Moreover if a measurement is performed by grouping many flows, it is 447 not possible to determine exactly which flow was affected by packets 448 loss. In order to have measures per single flow it is necessary to 449 configure counters for each specific flow. Once the flow(s) to be 450 monitored have been identified, it is necessary to configure the 451 monitoring on the proper nodes. Configuring the monitoring means 452 configuring the rule to intercept the traffic and configuring the 453 counters to count the packets. To have just an end-to-end 454 monitoring, it is sufficient to enable the monitoring on the first 455 and the last hop routers of the path: the mechanism is completely 456 transparent to intermediate nodes and independent from the path 457 followed by traffic flows. On the contrary, to monitor the flow on a 458 hop-by-hop basis along its whole path it is necessary to enable the 459 monitoring on every node from the source to the destination. In case 460 the exact path followed by the flow is not known a priori (i.e. the 461 flow has multiple paths to reach the destination) it is necessary to 462 enable the monitoring system on every path: counters on interfaces 463 traversed by the flow will report packet count, counters on other 464 interfaces will be null. 466 3.1.1. Coloring the packets 468 The coloring operation is fundamental in order to create packet 469 blocks. This implies choosing where to activate the coloring and how 470 to color the packets. 472 In case of flow-based measurements, it is desirable, in general, to 473 have a single coloring node because it is easier to manage and 474 doesn't rise any risk of conflict (consider the case where two nodes 475 color the same flow). Thus it is advantageous to color the flow as 476 close as possible to the source. In addition, coloring a flow close 477 to the source allows an end-to-end measure if a measurement point is 478 enabled on the last-hop router as well. The only requirement is that 479 the coloring must change periodically and every node along the path 480 must be able to identify unambiguously the colored packets. For 481 link-based measurements, all traffic needs to be colored when 482 transmitted on the link. If the traffic had already been colored, 483 then it has to be re-colored because the color must be consistent on 484 the link. This means that each hop along the path must (re-)color 485 the traffic; the color is not required to be consistent along 486 different links. 488 Traffic coloring can be implemented by setting a specific bit in the 489 packet header and changing the value of that bit periodically. How 490 to choose the marking field depends on the application and is out of 491 scope here. 493 3.1.2. Counting the packets 495 Assuming that the coloring of the packets is performed only by the 496 source node, the nodes between source and destination (included) have 497 to count the colored packets that they receive and forward: this 498 operation can be enabled on every router along the path or only on a 499 subset, depending on which network segment is being monitored (a 500 single link, a particular metro area, the backbone, the whole path). 502 Since the color switches periodically between two values, two 503 counters (one for each value) are needed: one counter for packets 504 with color A and one counter for packets with color B. For each flow 505 (or group of flows) being monitored and for every interface where the 506 monitoring is active, a couple of counters is needed. For example, 507 in order to monitor separately 3 flows on a router with 4 interfaces 508 involved, 24 counters are needed (2 counters for each of the 3 flows 509 on each of the 4 interfaces). 511 In case of link-based measurements the behaviour is similar except 512 that coloring and counting operations are performed on a link by link 513 basis at each endpoint of the link. 515 Another important aspect to take into consideration is when to read 516 the counters: in order to count the exact number of packets of a 517 block the routers must perform this operation when that block has 518 ended: in other words, the counter for color A must be read when the 519 current block has color B, in order to be sure that the value of the 520 counter is stable. This task can be accomplished in two ways. The 521 general approach suggests to read the counters periodically, many 522 times during a block duration, and to compare these successive 523 readings: when the counter stops incrementing means that the current 524 block has ended and its value can be elaborated safely. 525 Alternatively, if the coloring operation is performed on the basis of 526 a fixed timer, it is possible to configure the reading of the 527 counters according to that timer: for example, reading the counter 528 for color A every period in the middle of the subsequent block with 529 color B is a safe choice. A sufficient margin should be considered 530 between the end of a block and the reading of the counter, in order 531 to take into account any out-of-order packets. 533 3.1.3. Collecting data and calculating packet loss 535 The nodes enabled to perform performance monitoring collect the value 536 of the counters, but they are not able to directly use this 537 information to measure packet loss, because they only have their own 538 samples. For this reason, an external Network Management System 539 (NMS) can be used to collect and elaborate data and to perform packet 540 loss calculation. The NMS compares the values of counters from 541 different nodes and can calculate if some packets were lost (even a 542 single packet) and also where packets were lost. 544 The value of the counters needs to be transmitted to the NMS as soon 545 as it has been read. This can be accomplished by using SNMP or FTP 546 and can be done in Push Mode or Polling Mode. In the first case, 547 each router periodically sends the information to the NMS, in the 548 latter case it is the NMS that periodically polls routers to collect 549 information. In any case, the NMS has to collect all the relevant 550 values from all the routers within one cycle of the timer. 552 If link-based measurement is used, it would be possible to use a 553 protocol to exchange values of counters between the two endpoints in 554 order to let them perform the packet loss calculation for each 555 traffic direction. A similar approach could be also applied to a 556 flow-based measurement. 558 3.2. Timing aspects 560 This document introduces two color switching method: one is based on 561 fixed number of packet, the other is based on fixed timer. But the 562 method based on fixed timer is preferable because is more 563 deterministic, and will be considered in the rest of the dcoument. 565 By considering the clock error between network devices R1 and R2, 566 they must be synchronized to the same clock reference with an 567 accuracy of +/- L/2 time units, where L is the time duration of the 568 block. So each colored packet can be assigned to the right batch by 569 each router. This is because the minimum time distance between two 570 packets of the same color but belonging to different batches is L 571 time units. 573 In practice, there are also out of order at batch boundaries, 574 strictly related to the delay between measurement points. This means 575 that, without considering clock error, we wait L/2 after color 576 switching to be sure to take a still counter. 578 In summary we need to take into account two contributions: clock 579 error between network devices and the interval we need to wait to 580 avoid out of order because of network delay. 582 The following figure explains both issues. 584 ...BBBBBBBBB | AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA | BBBBBBBBB... 585 |<======================================>| 586 | L | 587 ...=========>|<==================><==================>|<==========... 588 | L/2 L/2 | 589 |<===>| |<===>| 590 d | | d 591 |<==========================>| 592 available counting interval 594 Figure 4: Timing aspects 596 It is assumed that all network devices are synchronized to a common 597 reference time with an accuracy of +/- A/2. Thus, the difference 598 between the clock values of any two network devices is bounded by A. 600 The guardband d is given by: 602 d = A + D_max - D_min, 604 where A is the clock accuracy, D_max is an upper bound on the network 605 delay between the network devices, and D_min is a lower bound on the 606 delay. 608 The available counting interval is L - 2d that must be > 0. 610 The condition that must be satisfied and is a requirement on the 611 synchronization accuracy is: 613 d < L/2. 615 3.3. One-way delay measurement 617 The same principle used to measure packet loss can be applied also to 618 one-way delay measurement. There are three alternatives, as 619 described hereinafter. 621 3.3.1. Single marking methodology 623 The alternation of colors can be used as a time reference to 624 calculate the delay. Whenever the color changes (that means that a 625 new block has started) a network device can store the timestamp of 626 the first packet of the new block; that timestamp can be compared 627 with the timestamp of the same packet on a second router to compute 628 packet delay. Considering Figure 2, R1 stores a timestamp TS(A1)R1 629 when it sends the first packet of block 1 (A-colored), a timestamp 630 TS(B2)R1 when it sends the first packet of block 2 (B-colored) and so 631 on for every other block. R2 performs the same operation on the 632 receiving side, recording TS(A1)R2, TS(B2)R2 and so on. Since the 633 timestamps refer to specific packets (the first packet of each block) 634 we are sure that timestamps compared to compute delay refer to the 635 same packets. By comparing TS(A1)R1 with TS(A1)R2 (and similarly 636 TS(B2)R1 with TS(B2)R2 and so on) it is possible to measure the delay 637 between R1 and R2. In order to have more measurements, it is 638 possible to take and store more timestamps, referring to other 639 packets within each block. 641 In order to coherently compare timestamps collected on different 642 routers, the network nodes must be in sync. Furthermore, a 643 measurement is valid only if no packet loss occurs and if packet 644 misordering can be avoided, otherwise the first packet of a block on 645 R1 could be different from the first packet of the same block on R2 646 (f.i. if that packet is lost between R1 and R2 or it arrives after 647 the next one). 649 The following table shows how timestamps can be used to calculate the 650 delay between R1 and R2. The first column lists the sequence of 651 blocks while other columns contain the timestamp referring to the 652 first packet of each block on R1 and R2. The delay is computed as a 653 difference between timestamps. For the sake of simplicity, all the 654 values are expressed in milliseconds. 656 +-------+---------+---------+---------+---------+-------------+ 657 | Block | TS(A)R1 | TS(B)R1 | TS(A)R2 | TS(B)R2 | Delay R1-R2 | 658 +-------+---------+---------+---------+---------+-------------+ 659 | 1 | 12.483 | - | 15.591 | - | 3.108 | 660 | | | | | | | 661 | 2 | - | 6.263 | - | 9.288 | 3.025 | 662 | | | | | | | 663 | 3 | 27.556 | - | 30.512 | - | 2.956 | 664 | | | | | | | 665 | | - | 18.113 | - | 21.269 | 3.156 | 666 | | | | | | | 667 | ... | ... | ... | ... | ... | ... | 668 | | | | | | | 669 | 2n | 77.463 | - | 80.501 | - | 3.038 | 670 | | | | | | | 671 | 2n+1 | - | 24.333 | - | 27.433 | 3.100 | 672 +-------+---------+---------+---------+---------+-------------+ 674 Table 2: Evaluation of timestamps for delay measurements 676 The first row shows timestamps taken on R1 and R2 respectively and 677 referring to the first packet of block 1 (which is A-colored). Delay 678 can be computed as a difference between the timestamp on R2 and the 679 timestamp on R1. Similarly, the second row shows timestamps (in 680 milliseconds) taken on R1 and R2 and referring to the first packet of 681 block 2 (which is B-colored). Comparing timestamps taken on 682 different nodes in the network and referring to the same packets 683 (identified using the alternation of colors) it is possible to 684 measure delay on different network segments. 686 For the sake of simplicity, in the above example a single measurement 687 is provided within a block, taking into account only the first packet 688 of each block. The number of measurements can be easily increased by 689 considering multiple packets in the block: for instance, a timestamp 690 could be taken every N packets, thus generating multiple delay 691 measurements. Taking this to the limit, in principle the delay could 692 be measured for each packet, by taking and comparing the 693 corresponding timestamps (possible but impractical from an 694 implementation point of view). 696 3.3.1.1. Mean delay 698 As mentioned before, the method previously exposed for measuring the 699 delay is sensitive to out of order reception of packets. In order to 700 overcome this problem, a different approach has been considered: it 701 is based on the concept of mean delay. The mean delay is calculated 702 by considering the average arrival time of the packets within a 703 single block. The network device locally stores a timestamp for each 704 packet received within a single block: summing all the timestamps and 705 dividing by the total number of packets received, the average arrival 706 time for that block of packets can be calculated. By subtracting the 707 average arrival times of two adjacent devices it is possible to 708 calculate the mean delay between those nodes. When computing the 709 mean delay, measurement error could be augmented by accumulating 710 measurement error of a lot of packets. This method is robust to out 711 of order packets and also to packet loss (only a small error is 712 introduced). Moreover, it greatly reduces the number of timestamps 713 (only one per block for each network device) that have to be 714 collected by the management system. On the other hand, it only gives 715 one measure for the duration of the block (f.i. 5 minutes), and it 716 doesn't give the minimum, maximum and median delay values (RFC 6703 717 [RFC6703]). This limitation could be overcome by reducing the 718 duration of the block (f.i. from 5 minutes to a few seconds), that 719 implicates an highly optimized implementation of the method. 721 By summing the mean delays of the two directions of a path, it is 722 also possible to measure the two-way mean delay (round-trip delay). 724 3.3.2. Double marking methodology 726 The Single marking methodology for one-way delay measurement is 727 sensitive to out of order reception of packets. The first approach 728 to overcome this problem is described before and is based on the 729 concept of mean delay. But the limitation of mean delay is that it 730 doesn't give information about the delay values distribution for the 731 duration of the block. Additionally it may be useful to have not 732 only the mean delay but also the minimum, maximum and median delay 733 values and, in wider terms, to know more about the statistic 734 distribution of delay values. So in order to have more information 735 about the delay and to overcome out of order issues, a different 736 approach can be introduced: it is based on double marking 737 methodology. 739 Basically, the idea is to use the first marking to create the 740 alternate flow and, within this colored flow, a second marking to 741 select the packets for measuring delay/jitter. The first marking is 742 needed for packet loss and mean delay measurement. The second 743 marking creates a new set of marked packets that are fully identified 744 over the network, so that a network device can store the timestamps 745 of these packets; these timestamps can be compared with the 746 timestamps of the same packets on a second router to compute packet 747 delay values for each packet. The number of measurements can be 748 easily increased by changing the frequency of the second marking. 749 But the frequency of the second marking must be not too high in order 750 to avoid out of order issues. Between packets with the second 751 marking there should be a security time gap (e.g. this gap could be, 752 at the minimum, the mean network delay calculated with the previous 753 methodology) to avoid out of order issues and also to have a number 754 of measurement packets that is rate independent. If a second marking 755 packet is lost, the delay measurement for the considered block is 756 corrupted and should be discarded. 758 Mean delay is calculated on all the packets of a sample and is a 759 simple computation to be performed for single marking method. In 760 some cases the mean delay measure is not sufficient to characterize 761 the sample, and more statistics of delay extent data are needed, e.g. 762 percentiles, variance and median delay values. The conventional 763 range (maximum-minimum) should be avoided for several reasons, 764 including stability of the maximum delay due to the influence by 765 outliers. RFC 5481 [RFC5481] section 6.5 highlights how the 99.9th 766 percentile of delay and delay variation is more helpful to 767 performance planners. To overcome this drawback the idea is to 768 couple the mean delay measure for the entire batch with double 769 marking method, where a subset of batch packets are selected for 770 extensive delay calculation by using a second marking. In this way 771 it is possible to perform a detailed analysis on these double marked 772 packets. Please note that there are classic algorithms for median 773 and variance calculation, but are out of the scope of this document. 774 The comparison between the mean delay for the entire batch and the 775 mean delay on these double marked packets gives an useful information 776 since it is possible to understand if the double marking measurements 777 are actually representative of the delay trends. 779 3.4. Delay variation measurement 781 Similarly to one-way delay measurement (both for single marking and 782 double marking), the method can also be used to measure the inter- 783 arrival jitter. We refer to the definition in RFC 3393 [RFC3393]. 784 The alternation of colors, for single marking method, can be used as 785 a time reference to measure delay variations. In case of double 786 marking, the time reference is given by the second marked packets. 787 Considering the example depicted in Figure 2, R1 stores a timestamp 788 TS(A)R1 whenever it sends the first packet of a block and R2 stores a 789 timestamp TS(B)R2 whenever it receives the first packet of a block. 790 The inter-arrival jitter can be easily derived from one-way delay 791 measurement, by evaluating the delay variation of consecutive 792 samples. 794 The concept of mean delay can also be applied to delay variation, by 795 evaluating the average variation of the interval between consecutive 796 packets of the flow from R1 to R2. 798 4. Considerations 800 This section highlights some considerations about the methodology. 802 4.1. Synchronization 804 The Alternate Marking technique does not require a strong 805 synchronization, especially for packet loss and two-way delay 806 measurement. Only one-way delay measurement requires network devices 807 to have synchronized clocks. 809 The color switching is the reference for all the network devices, and 810 the only requirement to be achieved is that all network devices have 811 to recognize the right batch along the path. 813 If the length of the measurement period is L time units, then all 814 network devices must be synchronized to the same clock reference with 815 an accuracy of +/- L/2 time units (without considering network 816 delay). This level of accuracy guarantees that all network devices 817 consistently match the color bit to the correct block. For example, 818 if the color is toggeled every second (L = 1 second), then clocks 819 must be synchronized with an accuracy of +/- 0.5 second to a common 820 time reference. 822 This synchronization requirement can be satisfied even with a 823 relatively inaccurate synchronization method. This is true for 824 packet loss and two-way delay measurement, instead, for one-way delay 825 measurement clock synchronization must be accurate. 827 Therefore, a system that uses only packet loss and two-way delay 828 measurement does not require synchronization. This is because the 829 value of the clocks of network devices does not affect the 830 computation of the two-way delay measurement. 832 4.2. Data Correlation 834 Data Correlation is the mechanism to compare counters and timestamps 835 for packet loss, delay and delay variation calculation. It could be 836 performed in several ways depending on the alternate marking 837 application and use case. 839 o A possibility is to use a centralized solution using Network 840 Management System (NMS) to correlate data; 842 o Another possibility is to define a protocol based distributed 843 solution, by defining a new protocol or by extending the existing 844 protocols (e.g. RFC6374, TWAMP, OWAMP) in order to communicate 845 the counters and timestamps between nodes. 847 In the following paragraphs an example data correlation mechanism is 848 explained and could be use independently of the adopted solutions. 850 When data is collected on the upstream and downstream node, e.g., 851 packet counts for packet loss measurement or timestamps for packet 852 delay measurement, and periodically reported to or pulled by other 853 nodes or NMS, a certain data correlation mechanism SHOULD be in use 854 to help the nodes or NMS to tell whether any two or more packet 855 counts are related to the same block of markers, or any two 856 timestamps are related to the same marked packet. 858 The alternate marking method described in this document literally 859 split the packets of the measured flow into different measurement 860 blocks, in addition a Block Number could be assigned to each of such 861 measurement block. The BN is generated each time a node reads the 862 data (packet counts or timestamps), and is associated with each 863 packet count and timestamp reported to or pulled by other nodes or 864 NMS. The value of BN could be calculated as the modulo of the local 865 time (when the data are read) and the interval of the marking time 866 period. 868 When the nodes or NMS see, for example, same BNs associated with two 869 packet counts from an upstream and a downstream node respectively, it 870 considers that these two packet counts corresponding to the same 871 block, i.e. that these two packet counts belong to the same block of 872 markers from the upstream and downstream node. The assumption of 873 this BN mechanism is that the measurement nodes are time 874 synchronized. This requires the measurement nodes to have a certain 875 time synchronization capability (e.g., the Network Time Protocol 876 (NTP) RFC 5905 [RFC5905], or the IEEE 1588 Precision Time Protocol 877 (PTP) [IEEE-1588]). Synchronization aspects are further discussed in 878 Section 4. 880 4.3. Packet Re-ordering 882 Due to ECMP, packet re-ordering is very common in IP network. The 883 accuracy of marking based PM, especially packet loss measurement, may 884 be affected by packet re-ordering. Take a look at the following 885 example: 887 Block : 1 | 2 | 3 | 4 | 5 |... 888 --------|---------|---------|---------|---------|---------|--- 889 Node R1 : AAAAAAA | BBBBBBB | AAAAAAA | BBBBBBB | AAAAAAA |... 890 Node R2 : AAAAABB | AABBBBA | AAABAAA | BBBBBBA | ABAAABA |... 892 Figure 5: Packet Reordering 894 In the following paragraphs an example of data correlation mechanism 895 is explained and could be use independently of the adopted solutions. 897 Most of the packet re-ordering occur at the edge of adjacent blocks, 898 and they are easy to handle if the interval of each block is 899 sufficient large. Then, it can assume that the packets with 900 different marker belong to the block that they are more close to. If 901 the interval is small, it is difficult and sometime impossible to 902 determine to which block a packet belongs. See above example, the 903 packet with the marker of "B" in block 3, there is no safe way to 904 tell whether the packet belongs to block 2 or block 4. 906 To choose a proper interval is important and how to choose a proper 907 interval is out of the scope of this document. But an implementation 908 SHOULD provide a way to configure the interval and allow a certain 909 degree of packet re-ordering. 911 5. Implementation and deployment 913 The methodology described in the previous sections can be applied in 914 various situations. Basically Alternate Marking technique could be 915 used in many cases for performance measurement. The only requirement 916 is to select and mark the flow to be monitored; in this way packets 917 are batched by the sender and each batch is alternately marked such 918 that can be easily recognized by the receiver. 920 An example of implementation and deployment is explained in the next 921 section, just to clarify how the method can work. 923 5.1. Report on the operational experiment at Telecom Italia 925 The method described in this document, also called PNPM (Packet 926 Network Performance Monitoring), has been invented and engineered in 927 Telecom Italia and it's currently being used in Telecom Italia's 928 network. The methodology has been applied by leveraging functions 929 and tools available on IP routers and it's currently being used to 930 monitor packet loss in some portions of Telecom Italia's network. 931 The application of the method to delay measurement is currently being 932 evaluated in Telecom Italia's labs. This section describes how the 933 features currently available on existing routing platforms can be 934 used to apply the method, in order to give an example of 935 implementation and deployment. 937 The current implementation in Telecom Italia uses the flow-based 938 strategy, as defined in section 3. The link-based strategy could be 939 applied to physical link or a logical link (e.g. Ethernet VLAN or a 940 MPLS PW). 942 The method is applied in Telecom Italia's network to multicast IPTV 943 channels or other specific traffic flows with high QoS requirements 944 (i.e. Mobile Backhauling traffic implemented with a VPN MPLS). 946 The implementation of the method by a Service Provider needs to use 947 the router features. With current router implementations, only QoS 948 related fields and features offer the required flexibility to set 949 bits in the packet header. In case a Service Provider only uses the 950 three most significant bits of the DSCP field (corresponding to IP 951 Precedence) for QoS classification and queuing, it is possible to use 952 the two less significant bits of the DSCP field (bit 0 and bit 1) to 953 implement the method without affecting QoS policies. One of the two 954 bits (bit 0) could be used to identify flows subject to traffic 955 monitoring (set to 1 if the flow is under monitoring, otherwise it is 956 set to 0), while the second (bit 1) can be used for coloring the 957 traffic (switching between values 0 and 1, corresponding to color A 958 and B) and creating the blocks. 960 In practice, coloring the traffic using the DSCP field can be 961 implemented by configuring on the router output interface an access 962 list that intercepts the flow(s) to be monitored and applies to them 963 a policy that sets the DSCP field accordingly. Since traffic 964 coloring has to be switched between the two values over time, the 965 policy needs to be modified periodically: an automatic script can be 966 used perform this task on the basis of a fixed timer. In Telecom 967 Italia's implementation this timer is set to 5 minutes: this value 968 showed to be a good compromise between measurement frequency and 969 stability of the measurement (i.e. possibility to collect all the 970 measures referring to the same block). 972 If traffic is colored using the DSCP field an access-list that 973 matches specific DSCP values can be used to count the packets of the 974 flow(s) being monitored. Also, a 5 minutes timer for color switching 975 is a safe choice for reading the counters. 977 5.1.1. Metric transparency 979 Since a Service Provider application is described here, the method 980 can be applied to end-to-end services supplied to Customers. So it 981 is important to highlight that the method SHOULD be transparent 982 outside the Service Provider domain. 984 In Telecom Italia's implementation the source node colors the packets 985 with a policy that is modified periodically via an automatic script 986 in order to alternate the DSCP field of the packets. The nodes 987 between source and destination (included) have to count with an 988 access-list the colored packets that they receive and forward. 990 Moreover the destination node has an important role: the colored 991 packets are intercepted and a policy restores and sets the DSCP field 992 of all the packets to the initial value. In this way the metric is 993 transparent because outside the section of the network under 994 monitoring the traffic flow is unchanged. 996 In such a case, thanks to this restoring technique, network elements 997 outside the Alternate Marking monitoring domain (e.g. the two 998 Provider Edge nodes of the Mobile Backhauling VPN MPLS) are totally 999 anaware that packets were marked. So this restoring technique makes 1000 Alternate Marking completely transparent outside its monitoring 1001 domain. 1003 5.2. IP flow performance measurement (IPFPM) 1005 This application of marking method is described in 1006 [I-D.chen-ippm-coloring-based-ipfpm-framework]. 1008 5.3. OAM Passive Performance Measurement 1010 In [I-D.ietf-bier-mpls-encapsulation] two OAM bits from Bit Index 1011 Explicit Replication (BIER) Header are reserved for the passive 1012 performance measurement marking method. [I-D.ietf-bier-pmmm-oam] 1013 details the measurement for multicast service over BIER domain. 1015 In addition, the alternate marking method could also be used in a 1016 Service Function Chaining (SFC) domain. 1018 The application of the marking method to Network Virtualization 1019 Overlays (NVO3) protocols is a work in progress (see 1020 [I-D.ietf-nvo3-encap]). 1022 5.4. RFC6374 Use Case 1024 RFC6374 [RFC6374] uses the LM packet as the packet accounting 1025 demarcation point. Unfortunately this gives rise to a number of 1026 problems that may lead to significant packet accounting errors in 1027 certain situations. [I-D.ietf-mpls-flow-ident] discusses the desired 1028 capabilities for MPLS flow identification in order to perform a 1029 better in-band performance monitoring of user data packets. A method 1030 of accomplishing identification is Synonymous Flow Labels (SFL) 1031 introduced in [I-D.bryant-mpls-sfl-framework], while 1032 [I-D.ietf-mpls-rfc6374-sfl] describes RFC6374 performance 1033 measurements with SFL. 1035 5.5. Application to active performance measurement 1037 [I-D.fioccola-ippm-alt-mark-active] describes how to extend the 1038 existing Active Measurement Protocol, in order to implement alternate 1039 marking methodology. [I-D.fioccola-ippm-rfc6812-alt-mark-ext] 1040 describes an extension to the Cisco SLA Protocol Measurement-Type 1041 UDP-Measurement. 1043 6. Hybrid measurement 1045 The method has been explicitly designed for passive measurements but 1046 it can also be used with active measurements. In order to have both 1047 end to end measurements and intermediate measurements (hybrid 1048 measurements) two end points can exchanges artificial traffic flows 1049 and apply alternate marking over these flows. In the intermediate 1050 points artificial traffic is managed in the same way as real traffic 1051 and measured as specified before. So the application of marking 1052 method can simplify also the active measurement, as explained in 1053 [I-D.fioccola-ippm-alt-mark-active]. 1055 7. Summary 1057 The advantages of the method described in this document are: 1059 o easy implementation: it can be implemented using features already 1060 available on major routing platforms; 1062 o low computational effort: the additional load on processing is 1063 negligible; 1065 o accurate packet loss measurement: single packet loss granularity 1066 is achieved with a passive measurement; 1068 o potential applicability to any kind of packet/frame -based 1069 traffic: Ethernet, IP, MPLS, etc., both unicast and multicast; 1071 o robustness: the method can tolerate out of order packets and it's 1072 not based on "special" packets whose loss could have a negative 1073 impact; 1075 o no interoperability issues: the features required to implement the 1076 method are available on all current routing platforms. 1078 The method doesn't raise any specific need for protocol extension, 1079 but it could be further improved by means of some extension to 1080 existing protocols. Specifically, the use of DiffServ bits for 1081 coloring the packets could not be a viable solution in some cases: a 1082 standard method to color the packets for this specific application 1083 could be beneficial. 1085 8. Compliance with RFC6390 guidelines 1087 RFC6390 [RFC6390] defines a framework and a process for developing 1088 Performance Metrics for protocols above and below the IP layer (such 1089 as IP-based applications that operate over reliable or datagram 1090 transport protocols). 1092 This document doesn't aim to propose a new Performance Metric but a 1093 new method of measurement for a few Performance Metrics that have 1094 already been standardized. Nevertheless, it's worth applying 1095 [RFC6390] guidelines to the present document, in order to provide a 1096 more complete and coherent description of the proposed method. We 1097 used a subset of the Performance Metric Definition template defined 1098 by [RFC6390]. 1100 o Metric name and description: as already stated, this document 1101 doesn't propose any new Performance Metric. On the contrary, it 1102 describes a novel method for measuring packet loss [RFC7680]. The 1103 same concept, with small differences, can also be used to measure 1104 delay [RFC7679], and jitter [RFC3393]. The document mainly 1105 describes the applicability to packet loss measurement. 1107 o Method of Measurement or Calculation: according to the method 1108 described in the previous sections, the number of packets lost is 1109 calculated by subtracting the value of the counter on the source 1110 node from the value of the counter on the destination node. Both 1111 counters must refer to the same color. The calculation is 1112 performed when the value of the counters is in a steady state. 1114 o Units of Measurement: the method calculates and reports the exact 1115 number of packets sent by the source node and not received by the 1116 destination node. 1118 o Measurement Points: the measurement can be performed between 1119 adjacent nodes, on a per-link basis, or along a multi-hop path, 1120 provided that the traffic under measurement follows that path. In 1121 case of a multi-hop path, the measurements can be performed both 1122 end-to-end and hop-by-hop. 1124 o Measurement Timing: the method have a constraint on the frequency 1125 of measurements. In order to perform a measure, the counter must 1126 be in a steady state: this happens when the traffic is being 1127 colored with the alternate color; for example in the Telecom 1128 Italia application of the method the time interval is set to 5 1129 minutes. 1131 o Implementation: the Telecom Italia application of the method uses 1132 two encodings of the DSCP field to color the packets; this enables 1133 the use of policy configurations on the router to color the 1134 packets and accordingly configure the counter for each color. The 1135 path followed by traffic being measured should be known in advance 1136 in order to configure the counters along the path and be able to 1137 compare the correct values. 1139 o Use and Applications: the method can be used to measure packet 1140 loss with high precision on live traffic; moreover, by combining 1141 end-to-end and per-link measurements, the method is useful to 1142 pinpoint the single link that is experiencing loss events. 1144 o Reporting Model: the value of the counters has to be sent to a 1145 centralized management system that perform the calculations; such 1146 samples must contain a reference to the time interval they refer 1147 to, so that the management system can perform the correct 1148 correlation; the samples have to be sent while the corresponding 1149 counter is in a steady state (within a time interval), otherwise 1150 the value of the sample should be stored locally. 1152 o Dependencies: the values of the counters have to be correlated to 1153 the time interval they refer to; moreover, as far the Telecom 1154 Italia application of the method is based on DSCP values, there 1155 are significant dependencies on the usage of the DSCP field: it 1156 must be possible to rely on unused DSCP values without affecting 1157 QoS-related configuration and behavior; moreover, the intermediate 1158 nodes must not change the value of the DSCP field not to alter the 1159 measurement. 1161 o Organization of Results: the method of measurement produces 1162 singletons. 1164 o Parameters: currently, the main parameter of the method is the 1165 time interval used to alternate the colors and read the counters. 1167 9. Security Considerations 1169 This document specifies a method to perform measurements in the 1170 context of a Service Provider's network and has not been developed to 1171 conduct Internet measurements, so it does not directly affect 1172 Internet security nor applications which run on the Internet. 1173 However, implementation of this method must be mindful of security 1174 and privacy concerns. 1176 There are two types of security concerns: potential harm caused by 1177 the measurements and potential harm to the measurements. 1179 o Harm caused by the measurement: the measurements described in this 1180 document are passive, so there are no new packets injected into 1181 the network causing potential harm to the network itself and to 1182 data traffic. Nevertheless, the method implies modifications on 1183 the fly to the IP header of data packets: this must be performed 1184 in a way that doesn't alter the quality of service experienced by 1185 packets subject to measurements and that preserve stability and 1186 performance of routers doing the measurements. One of the main 1187 security threats in OAM protocols is network reconnaissance; an 1188 attacker can gather information about the network performance by 1189 passively eavesdropping to OAM messages. The advantage of the 1190 methods described in this document is that the marking bits are 1191 the only information that is exchanged between the network 1192 devices. Therefore, passive eavesdropping to data plane traffic 1193 does not allow attackers to gain information about the network 1194 performance. 1196 o Harm to the measurement: the measurements could be harmed by 1197 routers altering the marking of the packets, or by an attacker 1198 injecting artificial traffic. Authentication techniques, such as 1199 digital signatures, may be used where appropriate to guard against 1200 injected traffic attacks. Since the measurement itself may be 1201 affected by routers (or other network devices) along the path of 1202 IP packets intentionally altering the value of marking bits of 1203 packets, as mentioned above, the mechanism specified in this 1204 document can be applied just in the context of a controlled 1205 domain, and thus the routers (or other network devices) are 1206 locally administered and this type of attack can be avoided. In 1207 addition, an attacker can't gain information about network 1208 performance from a single monitoring point, and must use 1209 synchronized monitoring points at multiple points on the path, 1210 because they have to do the same kind of measurement and 1211 aggregation that Service Providers using Alternate Marking must 1212 do. 1214 The privacy concerns of network measurement are limited because the 1215 method only relies on information contained in the IP header without 1216 any release of user data. 1218 Delay attacks are another potential threat in the context of this 1219 document. Delay measurement is performed using a specific packet in 1220 each block, marked by a dedicated color bit. Therefore, a man-in- 1221 the-middle attacker can selectively induce synthetic delay only to 1222 delay-colored packets, causing systematic error in the delay 1223 measurements. As discussed in previous sections, the methods 1224 described in this document rely on an underlying time synchronization 1225 protocol. Thus, by attacking the time protocol an attacker can 1226 potentially compromise the integrity of the measurement. A detailed 1227 discussion about the threats against time protocols and how to 1228 mitigate them is presented in RFC 7384 [RFC7384]. 1230 10. IANA Considerations 1232 There are no IANA actions required. 1234 11. Acknowledgements 1236 The previous IETF drafts about this technique were: 1237 [I-D.cociglio-mboned-multicast-pm] and [I-D.tempia-opsawg-p3m]. 1239 The authors would like to thank Alberto Tempia Bonda, Domenico 1240 Laforgia, Daniele Accetta and Mario Bianchetti for their contribution 1241 to the definition and the implementation of the method. 1243 12. References 1245 12.1. Normative References 1247 [IEEE-1588] 1248 IEEE 1588-2008, "IEEE Standard for a Precision Clock 1249 Synchronization Protocol for Networked Measurement and 1250 Control Systems", July 2008. 1252 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1253 Requirement Levels", BCP 14, RFC 2119, 1254 DOI 10.17487/RFC2119, March 1997, 1255 . 1257 [RFC3393] Demichelis, C. and P. Chimento, "IP Packet Delay Variation 1258 Metric for IP Performance Metrics (IPPM)", RFC 3393, 1259 DOI 10.17487/RFC3393, November 2002, 1260 . 1262 [RFC5905] Mills, D., Martin, J., Ed., Burbank, J., and W. Kasch, 1263 "Network Time Protocol Version 4: Protocol and Algorithms 1264 Specification", RFC 5905, DOI 10.17487/RFC5905, June 2010, 1265 . 1267 [RFC7679] Almes, G., Kalidindi, S., Zekauskas, M., and A. Morton, 1268 Ed., "A One-Way Delay Metric for IP Performance Metrics 1269 (IPPM)", STD 81, RFC 7679, DOI 10.17487/RFC7679, January 1270 2016, . 1272 [RFC7680] Almes, G., Kalidindi, S., Zekauskas, M., and A. Morton, 1273 Ed., "A One-Way Loss Metric for IP Performance Metrics 1274 (IPPM)", STD 82, RFC 7680, DOI 10.17487/RFC7680, January 1275 2016, . 1277 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 1278 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 1279 May 2017, . 1281 12.2. Informative References 1283 [I-D.bryant-mpls-sfl-framework] 1284 Bryant, S., Chen, M., Li, Z., Swallow, G., Sivabalan, S., 1285 and G. Mirsky, "Synonymous Flow Label Framework", draft- 1286 bryant-mpls-sfl-framework-05 (work in progress), June 1287 2017. 1289 [I-D.chen-ippm-coloring-based-ipfpm-framework] 1290 Chen, M., Zheng, L., Mirsky, G., Fioccola, G., and T. 1291 Mizrahi, "IP Flow Performance Measurement Framework", 1292 draft-chen-ippm-coloring-based-ipfpm-framework-06 (work in 1293 progress), March 2016. 1295 [I-D.cociglio-mboned-multicast-pm] 1296 Cociglio, M., Capello, A., Bonda, A., and L. Castaldelli, 1297 "A method for IP multicast performance monitoring", draft- 1298 cociglio-mboned-multicast-pm-01 (work in progress), 1299 October 2010. 1301 [I-D.fioccola-ippm-alt-mark-active] 1302 Fioccola, G., Clemm, A., Bryant, S., Cociglio, M., 1303 Chandramouli, M., and A. Capello, "Alternate Marking 1304 Extension to Active Measurement Protocol", draft-fioccola- 1305 ippm-alt-mark-active-01 (work in progress), March 2017. 1307 [I-D.fioccola-ippm-rfc6812-alt-mark-ext] 1308 Fioccola, G., Clemm, A., Cociglio, M., Chandramouli, M., 1309 and A. Capello, "Alternate Marking Extension to Cisco SLA 1310 Protocol RFC6812", draft-fioccola-ippm-rfc6812-alt-mark- 1311 ext-01 (work in progress), March 2016. 1313 [I-D.ietf-bier-mpls-encapsulation] 1314 Wijnands, I., Rosen, E., Dolganow, A., Tantsura, J., 1315 Aldrin, S., and I. Meilik, "Encapsulation for Bit Index 1316 Explicit Replication in MPLS and non-MPLS Networks", 1317 draft-ietf-bier-mpls-encapsulation-07 (work in progress), 1318 June 2017. 1320 [I-D.ietf-bier-pmmm-oam] 1321 Mirsky, G., Zheng, L., Chen, M., and G. Fioccola, 1322 "Performance Measurement (PM) with Marking Method in Bit 1323 Index Explicit Replication (BIER) Layer", draft-ietf-bier- 1324 pmmm-oam-02 (work in progress), July 2017. 1326 [I-D.ietf-mpls-flow-ident] 1327 Bryant, S., Pignataro, C., Chen, M., Li, Z., and G. 1328 Mirsky, "MPLS Flow Identification Considerations", draft- 1329 ietf-mpls-flow-ident-05 (work in progress), July 2017. 1331 [I-D.ietf-mpls-rfc6374-sfl] 1332 Bryant, S., Chen, M., Li, Z., Swallow, G., Sivabalan, S., 1333 Mirsky, G., and G. Fioccola, "RFC6374 Synonymous Flow 1334 Labels", draft-ietf-mpls-rfc6374-sfl-00 (work in 1335 progress), June 2017. 1337 [I-D.ietf-nvo3-encap] 1338 Boutros, S., Ganga, I., Garg, P., Manur, R., Mizrahi, T., 1339 Mozes, D., and E. Nordmark, "NVO3 Encapsulation 1340 Considerations", draft-ietf-nvo3-encap-00 (work in 1341 progress), June 2017. 1343 [I-D.tempia-opsawg-p3m] 1344 Capello, A., Cociglio, M., Castaldelli, L., and A. Bonda, 1345 "A packet based method for passive performance 1346 monitoring", draft-tempia-opsawg-p3m-04 (work in 1347 progress), February 2014. 1349 [RFC5481] Morton, A. and B. Claise, "Packet Delay Variation 1350 Applicability Statement", RFC 5481, DOI 10.17487/RFC5481, 1351 March 2009, . 1353 [RFC6374] Frost, D. and S. Bryant, "Packet Loss and Delay 1354 Measurement for MPLS Networks", RFC 6374, 1355 DOI 10.17487/RFC6374, September 2011, 1356 . 1358 [RFC6390] Clark, A. and B. Claise, "Guidelines for Considering New 1359 Performance Metric Development", BCP 170, RFC 6390, 1360 DOI 10.17487/RFC6390, October 2011, 1361 . 1363 [RFC6703] Morton, A., Ramachandran, G., and G. Maguluri, "Reporting 1364 IP Network Performance Metrics: Different Points of View", 1365 RFC 6703, DOI 10.17487/RFC6703, August 2012, 1366 . 1368 [RFC7276] Mizrahi, T., Sprecher, N., Bellagamba, E., and Y. 1369 Weingarten, "An Overview of Operations, Administration, 1370 and Maintenance (OAM) Tools", RFC 7276, 1371 DOI 10.17487/RFC7276, June 2014, 1372 . 1374 [RFC7384] Mizrahi, T., "Security Requirements of Time Protocols in 1375 Packet Switched Networks", RFC 7384, DOI 10.17487/RFC7384, 1376 October 2014, . 1378 [RFC7799] Morton, A., "Active and Passive Metrics and Methods (with 1379 Hybrid Types In-Between)", RFC 7799, DOI 10.17487/RFC7799, 1380 May 2016, . 1382 Authors' Addresses 1384 Giuseppe Fioccola (editor) 1385 Telecom Italia 1386 Via Reiss Romoli, 274 1387 Torino 10148 1388 Italy 1390 Email: giuseppe.fioccola@telecomitalia.it 1392 Alessandro Capello (editor) 1393 Telecom Italia 1394 Via Reiss Romoli, 274 1395 Torino 10148 1396 Italy 1398 Email: alessandro.capello@telecomitalia.it 1400 Mauro Cociglio 1401 Telecom Italia 1402 Via Reiss Romoli, 274 1403 Torino 10148 1404 Italy 1406 Email: mauro.cociglio@telecomitalia.it 1407 Luca Castaldelli 1408 Telecom Italia 1409 Via Reiss Romoli, 274 1410 Torino 10148 1411 Italy 1413 Email: luca.castaldelli@telecomitalia.it 1415 Mach(Guoyi) Chen (editor) 1416 Huawei Technologies 1418 Email: mach.chen@huawei.com 1420 Lianshu Zheng (editor) 1421 Huawei Technologies 1423 Email: vero.zheng@huawei.com 1425 Greg Mirsky (editor) 1426 ZTE 1427 USA 1429 Email: gregimirsky@gmail.com 1431 Tal Mizrahi (editor) 1432 Marvell 1433 6 Hamada st. 1434 Yokneam 1435 Israel 1437 Email: talmi@marvell.com