idnits 2.17.1 draft-ietf-bmwg-vswitch-opnfv-02.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 : ---------------------------------------------------------------------------- ** There are 17 instances of too long lines in the document, the longest one being 2 characters in excess of 72. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (April 17, 2017) is 2560 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- ** Obsolete normative reference: RFC 2679 (Obsoleted by RFC 7679) == Outdated reference: A later version (-03) exists of draft-huang-bmwg-virtual-network-performance-02 Summary: 2 errors (**), 0 flaws (~~), 2 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group M. Tahhan 3 Internet-Draft B. O'Mahony 4 Intended status: Informational Intel 5 Expires: October 19, 2017 A. Morton 6 AT&T Labs 7 April 17, 2017 9 Benchmarking Virtual Switches in OPNFV 10 draft-ietf-bmwg-vswitch-opnfv-02 12 Abstract 14 This memo describes the progress of the Open Platform for NFV (OPNFV) 15 project on virtual switch performance "VSWITCHPERF". This project 16 intends to build on the current and completed work of the 17 Benchmarking Methodology Working Group in IETF, by referencing 18 existing literature. The Benchmarking Methodology Working Group has 19 traditionally conducted laboratory characterization of dedicated 20 physical implementations of internetworking functions. Therefore, 21 this memo begins to describe the additional considerations when 22 virtual switches are implemented in general-purpose hardware. The 23 expanded tests and benchmarks are also influenced by the OPNFV 24 mission to support virtualization of the "telco" infrastructure. 26 Requirements Language 28 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 29 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 30 document are to be interpreted as described in RFC 2119 [RFC2119]. 32 Status of This Memo 34 This Internet-Draft is submitted in full conformance with the 35 provisions of BCP 78 and BCP 79. 37 Internet-Drafts are working documents of the Internet Engineering 38 Task Force (IETF). Note that other groups may also distribute 39 working documents as Internet-Drafts. The list of current Internet- 40 Drafts is at http://datatracker.ietf.org/drafts/current/. 42 Internet-Drafts are draft documents valid for a maximum of six months 43 and may be updated, replaced, or obsoleted by other documents at any 44 time. It is inappropriate to use Internet-Drafts as reference 45 material or to cite them other than as "work in progress." 47 This Internet-Draft will expire on October 19, 2017. 49 Copyright Notice 51 Copyright (c) 2017 IETF Trust and the persons identified as the 52 document authors. All rights reserved. 54 This document is subject to BCP 78 and the IETF Trust's Legal 55 Provisions Relating to IETF Documents 56 (http://trustee.ietf.org/license-info) in effect on the date of 57 publication of this document. Please review these documents 58 carefully, as they describe your rights and restrictions with respect 59 to this document. Code Components extracted from this document must 60 include Simplified BSD License text as described in Section 4.e of 61 the Trust Legal Provisions and are provided without warranty as 62 described in the Simplified BSD License. 64 Table of Contents 66 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 67 2. Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 68 3. Benchmarking Considerations . . . . . . . . . . . . . . . . . 4 69 3.1. Comparison with Physical Network Functions . . . . . . . 4 70 3.2. Continued Emphasis on Black-Box Benchmarks . . . . . . . 4 71 3.3. New Configuration Parameters . . . . . . . . . . . . . . 4 72 3.4. Flow classification . . . . . . . . . . . . . . . . . . . 6 73 3.5. Benchmarks using Baselines with Resource Isolation . . . 7 74 4. VSWITCHPERF Specification Summary . . . . . . . . . . . . . . 8 75 5. 3x3 Matrix Coverage . . . . . . . . . . . . . . . . . . . . . 16 76 5.1. Speed of Activation . . . . . . . . . . . . . . . . . . . 17 77 5.2. Accuracy of Activation section . . . . . . . . . . . . . 17 78 5.3. Reliability of Activation . . . . . . . . . . . . . . . . 17 79 5.4. Scale of Activation . . . . . . . . . . . . . . . . . . . 17 80 5.5. Speed of Operation . . . . . . . . . . . . . . . . . . . 17 81 5.6. Accuracy of Operation . . . . . . . . . . . . . . . . . . 17 82 5.7. Reliability of Operation . . . . . . . . . . . . . . . . 17 83 5.8. Scalability of Operation . . . . . . . . . . . . . . . . 18 84 5.9. Summary . . . . . . . . . . . . . . . . . . . . . . . . . 18 85 6. Security Considerations . . . . . . . . . . . . . . . . . . . 18 86 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 19 87 8. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 19 88 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 19 89 9.1. Normative References . . . . . . . . . . . . . . . . . . 19 90 9.2. Informative References . . . . . . . . . . . . . . . . . 20 91 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 21 93 1. Introduction 95 Benchmarking Methodology Working Group (BMWG) has traditionally 96 conducted laboratory characterization of dedicated physical 97 implementations of internetworking functions. The Black-box 98 Benchmarks of Throughput, Latency, Forwarding Rates and others have 99 served our industry for many years. Now, Network Function 100 Virtualization (NFV) has the goal to transform how internetwork 101 functions are implemented, and therefore has garnered much attention. 103 This memo summarizes the progress of the Open Platform for NFV 104 (OPNFV) project on virtual switch performance characterization, 105 "VSWITCHPERF", through the Brahmaputra (second) release [BrahRel]. 106 This project intends to build on the current and completed work of 107 the Benchmarking Methodology Working Group in IETF, by referencing 108 existing literature. For example, currently the most often 109 referenced RFC is [RFC2544] (which depends on [RFC1242]) and 110 foundation of the benchmarking work in OPNFV is common and strong. 112 See [VSPERFhome] for more background, and the OPNFV website for 113 general information [OPNFV]. 115 The authors note that OPNFV distinguishes itself from other open 116 source compute and networking projects through its emphasis on 117 existing "telco" services as opposed to cloud-computing. There are 118 many ways in which telco requirements have different emphasis on 119 performance dimensions when compared to cloud computing: support for 120 and transfer of isochronous media streams is one example. 122 Note also that the move to NFV Infrastructure has resulted in many 123 new benchmarking initiatives across the industry. The authors are 124 currently doing their best to maintain alignment with many other 125 projects, and this Internet Draft is one part of the efforts. We 126 acknowledge the early work in 127 [I-D.huang-bmwg-virtual-network-performance], and useful discussion 128 with the authors. 130 2. Scope 132 The primary purpose and scope of the memo is to inform the industry 133 of work-in-progress that builds on the body of extensive BMWG 134 literature and experience, and describe the extensions needed for 135 benchmarking virtual switches. Inital feedback indicates that many 136 of these extensions may be applicable beyond the current scope (to 137 hardware switches in the NFV Infrastructure and to virtual routers, 138 for example). Additionally, this memo serves as a vehicle to include 139 more detail and commentary from BMWG and other Open Source 140 communities, under BMWG's chartered work to characterize the NFV 141 Infrastructure (a virtual switch is an important aspect of that 142 infrastructure). 144 The benchmarking covered in this memo should be applicable to many 145 types of vswitches, and remain vswitch-agnostic to great degree. 146 There has been no attempt to track and test all features of any 147 specific vswitch implementation. 149 3. Benchmarking Considerations 151 This section highlights some specific considerations (from 152 [I-D.ietf-bmwg-virtual-net])related to Benchmarks for virtual 153 switches. The OPNFV project is sharing its present view on these 154 areas, as they develop their specifications in the Level Test Design 155 (LTD) document. 157 3.1. Comparison with Physical Network Functions 159 To compare the performance of virtual designs and implementations 160 with their physical counterparts, identical benchmarks are needed. 161 BMWG has developed specifications for many network functions this 162 memo re-uses existing benchmarks through references, and expands them 163 during development of new methods. A key configuration aspect is the 164 number of parallel cores required to achieve comparable performance 165 with a given physical device, or whether some limit of scale was 166 reached before the cores could achieve the comparable level. 168 It's unlikely that the virtual switch will be the only application 169 running on the SUT, so CPU utilization, Cache utilization, and Memory 170 footprint should also be recorded for the virtual implementations of 171 internetworking functions. 173 3.2. Continued Emphasis on Black-Box Benchmarks 175 External observations remain essential as the basis for Benchmarks. 176 Internal observations with fixed specification and interpretation 177 will be provided in parallel to assist the development of operations 178 procedures when the technology is deployed. 180 3.3. New Configuration Parameters 182 A key consideration when conducting any sort of benchmark is trying 183 to ensure the consistency and repeatability of test results. When 184 benchmarking the performance of a vSwitch there are many factors that 185 can affect the consistency of results, one key factor is matching the 186 various hardware and software details of the SUT. This section lists 187 some of the many new parameters which this project believes are 188 critical to report in order to achieve repeatability. 190 Hardware details including: 192 o Platform details 194 o Processor details 196 o Memory information (type and size) 198 o Number of enabled cores 200 o Number of cores used for the test 202 o Number of physical NICs, as well as their details (manufacturer, 203 versions, type and the PCI slot they are plugged into) 205 o NIC interrupt configuration 207 o BIOS version, release date and any configurations that were 208 modified 210 o CPU microcode level 212 o Memory DIMM configurations (quad rank performance may not be the 213 same as dual rank) in size, freq and slot locations 215 o PCI configuration parameters (payload size, early ack option...) 217 o Power management at all levels (ACPI sleep states, processor 218 package, OS...) 220 Software details including: 222 o OS parameters and behavior (text vs graphical no one typing at the 223 console on one system) 225 o OS version (for host and VNF) 227 o Kernel version (for host and VNF) 229 o GRUB boot parameters (for host and VNF) 231 o Hypervisor details (Type and version) 233 o Selected vSwitch, version number or commit id used 235 o vSwitch launch command line if it has been parameterised 237 o Memory allocation to the vSwitch 238 o which NUMA node it is using, and how many memory channels 240 o DPDK or any other SW dependency version number or commit id used 242 o Memory allocation to a VM - if it's from Hugpages/elsewhere 244 o VM storage type: snapshot/independent persistent/independent non- 245 persistent 247 o Number of VMs 249 o Number of Virtual NICs (vNICs), versions, type and driver 251 o Number of virtual CPUs and their core affinity on the host 253 o Number vNIC interrupt configuration 255 o Thread affinitization for the applications (including the vSwitch 256 itself) on the host 258 o Details of Resource isolation, such as CPUs designated for Host/ 259 Kernel (isolcpu) and CPUs designated for specific processes 260 (taskset). - Test duration. - Number of flows. 262 Test Traffic Information: 264 o Traffic type - UDP, TCP, IMIX / Other 266 o Packet Sizes 268 o Deployment Scenario 270 3.4. Flow classification 272 Virtual switches group packets into flows by processing and matching 273 particular packet or frame header information, or by matching packets 274 based on the input ports. Thus a flow can be thought of a sequence 275 of packets that have the same set of header field values (5-tuple) or 276 have arrived on the same port. Performance results can vary based on 277 the parameters the vSwitch uses to match for a flow. The recommended 278 flow classification parameters for any vSwitch performance tests are: 279 the input port, the source IP address, the destination IP address and 280 the Ethernet protocol type field. It is essential to increase the 281 flow timeout time on a vSwitch before conducting any performance 282 tests that do not measure the flow setup time. Normally the first 283 packet of a particular stream will install the flow in the virtual 284 switch which adds an additional latency, subsequent packets of the 285 same flow are not subject to this latency if the flow is already 286 installed on the vSwitch. 288 3.5. Benchmarks using Baselines with Resource Isolation 290 This outline describes measurement of baseline with isolated 291 resources at a high level, which is the intended approach at this 292 time. 294 1. Baselines: 296 * Optional: Benchmark platform forwarding capability without a 297 vswitch or VNF for at least 72 hours (serves as a means of 298 platform validation and a means to obtain the base performance 299 for the platform in terms of its maximum forwarding rate and 300 latency). 302 Benchmark platform forwarding capability 304 __ 305 +--------------------------------------------------+ | 306 | +------------------------------------------+ | | 307 | | | | | 308 | | Simple Forwarding App | | Host 309 | | | | | 310 | +------------------------------------------+ | | 311 | | NIC | | | 312 +---+------------------------------------------+---+ __| 313 ^ : 314 | | 315 : v 316 +--------------------------------------------------+ 317 | | 318 | traffic generator | 319 | | 320 +--------------------------------------------------+ 322 * Benchmark VNF forwarding capability with direct connectivity 323 (vSwitch bypass, e.g., SR/IOV) for at least 72 hours (serves 324 as a means of VNF validation and a means to obtain the base 325 performance for the VNF in terms of its maximum forwarding 326 rate and latency). The metrics gathered from this test will 327 serve as a key comparison point for vSwitch bypass 328 technologies performance and vSwitch performance. 330 Benchmark VNF forwarding capability 332 __ 333 +--------------------------------------------------+ | 334 | +------------------------------------------+ | | 335 | | | | | 336 | | VNF | | | 337 | | | | | 338 | +------------------------------------------+ | | 339 | | Passthrough/SR-IOV | | Host 340 | +------------------------------------------+ | | 341 | | NIC | | | 342 +---+------------------------------------------+---+ __| 343 ^ : 344 | | 345 : v 346 +--------------------------------------------------+ 347 | | 348 | traffic generator | 349 | | 350 +--------------------------------------------------+ 352 * Benchmarking with isolated resources alone, with other 353 resources (both HW&SW) disabled Example, vSw and VM are SUT 355 * Benchmarking with isolated resources alone, leaving some 356 resources unused 358 * Benchmark with isolated resources and all resources occupied 360 2. Next Steps 362 * Limited sharing 364 * Production scenarios 366 * Stressful scenarios 368 4. VSWITCHPERF Specification Summary 370 The overall specification in preparation is referred to as a Level 371 Test Design (LTD) document, which will contain a suite of performance 372 tests. The base performance tests in the LTD are based on the pre- 373 existing specifications developed by BMWG to test the performance of 374 physical switches. These specifications include: 376 o [RFC2544] Benchmarking Methodology for Network Interconnect 377 Devices 379 o [RFC2889] Benchmarking Methodology for LAN Switching 381 o [RFC6201] Device Reset Characterization 383 o [RFC5481] Packet Delay Variation Applicability Statement 385 Some of the above/newer RFCs are being applied in benchmarking for 386 the first time, and represent a development challenge for test 387 equipment developers. Fortunately, many members of the testing 388 system community have engaged on the VSPERF project, including an 389 open source test system. 391 In addition to this, the LTD also re-uses the terminology defined by: 393 o [RFC2285] Benchmarking Terminology for LAN Switching Devices 395 o [RFC5481] Packet Delay Variation Applicability Statement 397 Specifications to be included in future updates of the LTD include: 399 o [RFC3918] Methodology for IP Multicast Benchmarking 401 o [RFC4737] Packet Reordering Metrics 403 As one might expect, the most fundamental internetworking 404 characteristics of Throughput and Latency remain important when the 405 switch is virtualized, and these benchmarks figure prominently in the 406 specification. 408 When considering characteristics important to "telco" network 409 functions, we must begin to consider additional performance metrics. 410 In this case, the project specifications have referenced metrics from 411 the IETF IP Performance Metrics (IPPM) literature. This means that 412 the [RFC2544] test of Latency is replaced by measurement of a metric 413 derived from IPPM's [RFC2679], where a set of statistical summaries 414 will be provided (mean, max, min, etc.). Further metrics planned to 415 be benchmarked include packet delay variation as defined by [RFC5481] 416 , reordering, burst behaviour, DUT availability, DUT capacity and 417 packet loss in long term testing at Throughput level, where some low- 418 level of background loss may be present and characterized. 420 Tests have been (or will be) designed to collect the metrics below: 422 o Throughput Tests to measure the maximum forwarding rate (in frames 423 per second or fps) and bit rate (in Mbps) for a constant load (as 424 defined by [RFC1242]) without traffic loss. 426 o Packet and Frame Delay Distribution Tests to measure average, min 427 and max packet and frame delay for constant loads. 429 o Packet Delay Tests to understand latency distribution for 430 different packet sizes and over an extended test run to uncover 431 outliers. 433 o Scalability Tests to understand how the virtual switch performs as 434 the number of flows, active ports, complexity of the forwarding 435 logic's configuration... it has to deal with increases. 437 o Stream Performance Tests (TCP, UDP) to measure bulk data transfer 438 performance, i.e. how fast systems can send and receive data 439 through the switch. 441 o Control Path and Datapath Coupling Tests, to understand how 442 closely coupled the datapath and the control path are as well as 443 the effect of this coupling on the performance of the DUT 444 (example: delay of the initial packet of a flow). 446 o CPU and Memory Consumption Tests to understand the virtual 447 switch's footprint on the system, usually conducted as auxiliary 448 measurements with benchmarks above. They include: CPU 449 utilization, Cache utilization and Memory footprint. 451 o The so-called "Soak" tests, where the selected test is conducted 452 over a long period of time (with an ideal duration of 24 hours, 453 but only long enough to determine that stability issues exist when 454 found; there is no requirement to continue a test when a DUT 455 exhibits instability over time). The key performance 456 characteristics and benchmarks for a DUT are determined (using 457 short duration tests) prior to conducting soak tests. The purpose 458 of soak tests is to capture transient changes in performance which 459 may occur due to infrequent processes, memory leaks, or the low 460 probability coincidence of two or more processes. The stability 461 of the DUT is the paramount consideration, so performance must be 462 evaluated periodically during continuous testing, and this results 463 in use of [RFC2889] Frame Rate metrics instead of [RFC2544] 464 Throughput (which requires stopping traffic to allow time for all 465 traffic to exit internal queues), for example. 467 Future/planned test specs include: 469 o Request/Response Performance Tests (TCP, UDP) which measure the 470 transaction rate through the switch. 472 o Noisy Neighbour Tests, to understand the effects of resource 473 sharing on the performance of a virtual switch. 475 o Tests derived from examination of ETSI NFV Draft GS IFA003 476 requirements [IFA003] on characterization of acceleration 477 technologies applied to vswitches. 479 The flexibility of deployment of a virtual switch within a network 480 means that the BMWG IETF existing literature needs to be used to 481 characterize the performance of a switch in various deployment 482 scenarios. The deployment scenarios under consideration include: 484 Physical port to virtual switch to physical port 486 __ 487 +--------------------------------------------------+ | 488 | +--------------------+ | | 489 | | | | | 490 | | v | | Host 491 | +--------------+ +--------------+ | | 492 | | phy port | vSwitch | phy port | | | 493 +---+--------------+------------+--------------+---+ __| 494 ^ : 495 | | 496 : v 497 +--------------------------------------------------+ 498 | | 499 | traffic generator | 500 | | 501 +--------------------------------------------------+ 503 Physical port to virtual switch to VNF to virtual switch to physical 504 port 506 __ 507 +---------------------------------------------------+ | 508 | | | 509 | +-------------------------------------------+ | | 510 | | Application | | | 511 | +-------------------------------------------+ | | 512 | ^ : | | 513 | | | | | Guest 514 | : v | | 515 | +---------------+ +---------------+ | | 516 | | logical port 0| | logical port 1| | | 517 +---+---------------+-----------+---------------+---+ __| 518 ^ : 519 | | 520 : v __ 521 +---+---------------+----------+---------------+---+ | 522 | | logical port 0| | logical port 1| | | 523 | +---------------+ +---------------+ | | 524 | ^ : | | 525 | | | | | Host 526 | : v | | 527 | +--------------+ +--------------+ | | 528 | | phy port | vSwitch | phy port | | | 529 +---+--------------+------------+--------------+---+ __| 530 ^ : 531 | | 532 : v 533 +--------------------------------------------------+ 534 | | 535 | traffic generator | 536 | | 537 +--------------------------------------------------+ 539 Physical port to virtual switch to VNF to virtual switch to VNF to 540 virtual switch to physical port 542 __ 543 +----------------------+ +----------------------+ | 544 | Guest 1 | | Guest 2 | | 545 | +---------------+ | | +---------------+ | | 546 | | Application | | | | Application | | | 547 | +---------------+ | | +---------------+ | | 548 | ^ | | | ^ | | | 549 | | v | | | v | | Guests 550 | +---------------+ | | +---------------+ | | 551 | | logical ports | | | | logical ports | | | 552 | | 0 1 | | | | 0 1 | | | 553 +---+---------------+--+ +---+---------------+--+__| 554 ^ : ^ : 555 | | | | 556 : v : v _ 557 +---+---------------+---------+---------------+--+ | 558 | | 0 1 | | 3 4 | | | 559 | | logical ports | | logical ports | | | 560 | +---------------+ +---------------+ | | 561 | ^ | ^ | | | Host 562 | | |-----------------| v | | 563 | +--------------+ +--------------+ | | 564 | | phy ports | vSwitch | phy ports | | | 565 +---+--------------+----------+--------------+---+_| 566 ^ : 567 | | 568 : v 569 +--------------------------------------------------+ 570 | | 571 | traffic generator | 572 | | 573 +--------------------------------------------------+ 575 Physical port to virtual switch to VNF 577 __ 578 +---------------------------------------------------+ | 579 | | | 580 | +-------------------------------------------+ | | 581 | | Application | | | 582 | +-------------------------------------------+ | | 583 | ^ | | 584 | | | | Guest 585 | : | | 586 | +---------------+ | | 587 | | logical port 0| | | 588 +---+---------------+-------------------------------+ __| 589 ^ 590 | 591 : __ 592 +---+---------------+------------------------------+ | 593 | | logical port 0| | | 594 | +---------------+ | | 595 | ^ | | 596 | | | | Host 597 | : | | 598 | +--------------+ | | 599 | | phy port | vSwitch | | 600 +---+--------------+------------ -------------- ---+ __| 601 ^ 602 | 603 : 604 +--------------------------------------------------+ 605 | | 606 | traffic generator | 607 | | 608 +--------------------------------------------------+ 610 VNF to virtual switch to physical port 612 __ 613 +---------------------------------------------------+ | 614 | | | 615 | +-------------------------------------------+ | | 616 | | Application | | | 617 | +-------------------------------------------+ | | 618 | : | | 619 | | | | Guest 620 | v | | 621 | +---------------+ | | 622 | | logical port | | | 623 +-------------------------------+---------------+---+ __| 624 : 625 | 626 v __ 627 +------------------------------+---------------+---+ | 628 | | logical port | | | 629 | +---------------+ | | 630 | : | | 631 | | | | Host 632 | v | | 633 | +--------------+ | | 634 | vSwitch | phy port | | | 635 +-------------------------------+--------------+---+ __| 636 : 637 | 638 v 639 +--------------------------------------------------+ 640 | | 641 | traffic generator | 642 | | 643 +--------------------------------------------------+ 645 VNF to virtual switch to VNF 647 __ 648 +----------------------+ +----------------------+ | 649 | Guest 1 | | Guest 2 | | 650 | +---------------+ | | +---------------+ | | 651 | | Application | | | | Application | | | 652 | +---------------+ | | +---------------+ | | 653 | | | | ^ | | 654 | v | | | | | Guests 655 | +---------------+ | | +---------------+ | | 656 | | logical ports | | | | logical ports | | | 657 | | 0 | | | | 0 | | | 658 +---+---------------+--+ +---+---------------+--+__| 659 : ^ 660 | | 661 v : _ 662 +---+---------------+---------+---------------+--+ | 663 | | 1 | | 1 | | | 664 | | logical ports | | logical ports | | | 665 | +---------------+ +---------------+ | | 666 | | ^ | | Host 667 | L-----------------+ | | 668 | | | 669 | vSwitch | | 670 +------------------------------------------------+_| 672 A set of Deployment Scenario figures is available on the VSPERF Test 673 Methodology Wiki page [TestTopo]. 675 5. 3x3 Matrix Coverage 677 This section organizes the many existing test specifications into the 678 "3x3" matrix (introduced in [I-D.ietf-bmwg-virtual-net]). Because 679 the LTD specification ID names are quite long, this section is 680 organized into lists for each occupied cell of the matrix (not all 681 are occupied, also the matrix has grown to 3x4 to accommodate scale 682 metrics when displaying the coverage of many metrics/benchmarks). 683 The current version of the LTD specification is available [LTD]. 685 The tests listed below assess the activation of paths in the data 686 plane, rather than the control plane. 688 A complete list of tests with short summaries is available on the 689 VSPERF "LTD Test Spec Overview" Wiki page [LTDoverV]. 691 5.1. Speed of Activation 693 o Activation.RFC2889.AddressLearningRate 695 o PacketLatency.InitialPacketProcessingLatency 697 5.2. Accuracy of Activation section 699 o CPDP.Coupling.Flow.Addition 701 5.3. Reliability of Activation 703 o Throughput.RFC2544.SystemRecoveryTime 705 o Throughput.RFC2544.ResetTime 707 5.4. Scale of Activation 709 o Activation.RFC2889.AddressCachingCapacity 711 5.5. Speed of Operation 713 o Throughput.RFC2544.PacketLossRate 715 o CPU.RFC2544.0PacketLoss 717 o Throughput.RFC2544.PacketLossRateFrameModification 719 o Throughput.RFC2544.BackToBackFrames 721 o Throughput.RFC2889.MaxForwardingRate 723 o Throughput.RFC2889.ForwardPressure 725 o Throughput.RFC2889.BroadcastFrameForwarding 727 5.6. Accuracy of Operation 729 o Throughput.RFC2889.ErrorFramesFiltering 731 o Throughput.RFC2544.Profile 733 5.7. Reliability of Operation 735 o Throughput.RFC2889.Soak 737 o Throughput.RFC2889.SoakFrameModification 738 o PacketDelayVariation.RFC3393.Soak 740 5.8. Scalability of Operation 742 o Scalability.RFC2544.0PacketLoss 744 o MemoryBandwidth.RFC2544.0PacketLoss.Scalability 746 5.9. Summary 748 |------------------------------------------------------------------------| 749 | | | | | | 750 | | SPEED | ACCURACY | RELIABILITY | SCALE | 751 | | | | | | 752 |------------------------------------------------------------------------| 753 | | | | | | 754 | Activation | X | X | X | X | 755 | | | | | | 756 |------------------------------------------------------------------------| 757 | | | | | | 758 | Operation | X | X | X | X | 759 | | | | | | 760 |------------------------------------------------------------------------| 761 | | | | | | 762 | De-activation | | | | | 763 | | | | | | 764 |------------------------------------------------------------------------| 766 6. Security Considerations 768 Benchmarking activities as described in this memo are limited to 769 technology characterization of a Device Under Test/System Under Test 770 (DUT/SUT) using controlled stimuli in a laboratory environment, with 771 dedicated address space and the constraints specified in the sections 772 above. 774 The benchmarking network topology will be an independent test setup 775 and MUST NOT be connected to devices that may forward the test 776 traffic into a production network, or misroute traffic to the test 777 management network. 779 Further, benchmarking is performed on a "black-box" basis, relying 780 solely on measurements observable external to the DUT/SUT. 782 Special capabilities SHOULD NOT exist in the DUT/SUT specifically for 783 benchmarking purposes. Any implications for network security arising 784 from the DUT/SUT SHOULD be identical in the lab and in production 785 networks. 787 7. IANA Considerations 789 No IANA Action is requested at this time. 791 8. Acknowledgements 793 The authors appreciate and acknowledge comments from Scott Bradner, 794 Marius Georgescu, Ramki Krishnan, Doug Montgomery, Martin Klozik, 795 Christian Trautman, and others for their reviews. 797 9. References 799 9.1. Normative References 801 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 802 Requirement Levels", BCP 14, RFC 2119, 803 DOI 10.17487/RFC2119, March 1997, 804 . 806 [RFC2285] Mandeville, R., "Benchmarking Terminology for LAN 807 Switching Devices", RFC 2285, DOI 10.17487/RFC2285, 808 February 1998, . 810 [RFC2544] Bradner, S. and J. McQuaid, "Benchmarking Methodology for 811 Network Interconnect Devices", RFC 2544, 812 DOI 10.17487/RFC2544, March 1999, 813 . 815 [RFC2679] Almes, G., Kalidindi, S., and M. Zekauskas, "A One-way 816 Delay Metric for IPPM", RFC 2679, DOI 10.17487/RFC2679, 817 September 1999, . 819 [RFC2889] Mandeville, R. and J. Perser, "Benchmarking Methodology 820 for LAN Switching Devices", RFC 2889, 821 DOI 10.17487/RFC2889, August 2000, 822 . 824 [RFC3918] Stopp, D. and B. Hickman, "Methodology for IP Multicast 825 Benchmarking", RFC 3918, DOI 10.17487/RFC3918, October 826 2004, . 828 [RFC4737] Morton, A., Ciavattone, L., Ramachandran, G., Shalunov, 829 S., and J. Perser, "Packet Reordering Metrics", RFC 4737, 830 DOI 10.17487/RFC4737, November 2006, 831 . 833 [RFC6201] Asati, R., Pignataro, C., Calabria, F., and C. Olvera, 834 "Device Reset Characterization", RFC 6201, 835 DOI 10.17487/RFC6201, March 2011, 836 . 838 9.2. Informative References 840 [BrahRel] "Brahmaputra, Second OPNFV Release 841 https://wiki.opnfv.org/display/SWREL/Brahmaputra". 843 [I-D.huang-bmwg-virtual-network-performance] 844 Huang, L., Rong, G., Mandeville, B., and B. Hickman, 845 "Benchmarking Methodology for Virtualization Network 846 Performance", draft-huang-bmwg-virtual-network- 847 performance-02 (work in progress), March 2017. 849 [I-D.ietf-bmwg-virtual-net] 850 Morton, A., "Considerations for Benchmarking Virtual 851 Network Functions and Their Infrastructure", draft-ietf- 852 bmwg-virtual-net-05 (work in progress), March 2017. 854 [IFA003] "https://docbox.etsi.org/ISG/NFV/Open/Drafts/ 855 IFA003_Acceleration_-_vSwitch_Spec/". 857 [LTD] "LTD Test Specification 858 http://artifacts.opnfv.org/vswitchperf/brahmaputra/docs/ 859 requirements/index.html". 861 [LTDoverV] 862 "LTD Test Spec Overview 863 https://wiki.opnfv.org/display/vsperf/ 864 LTD+Test+Spec+Overview". 866 [OPNFV] "OPNFV Home https://www.opnfv.org/". 868 [RFC1242] Bradner, S., "Benchmarking Terminology for Network 869 Interconnection Devices", RFC 1242, DOI 10.17487/RFC1242, 870 July 1991, . 872 [RFC5481] Morton, A. and B. Claise, "Packet Delay Variation 873 Applicability Statement", RFC 5481, DOI 10.17487/RFC5481, 874 March 2009, . 876 [TestTopo] 877 "Test Topologies https://wiki.opnfv.org/display/vsperf/ 878 Test+Methodology". 880 [VSPERFhome] 881 "VSPERF Home https://wiki.opnfv.org/display/vsperf/ 882 VSperf+Home". 884 Authors' Addresses 886 Maryam Tahhan 887 Intel 889 Email: maryam.tahhan@intel.com 891 Billy O'Mahony 892 Intel 894 Email: billy.o.mahony@intel.com 896 Al Morton 897 AT&T Labs 898 200 Laurel Avenue South 899 Middletown,, NJ 07748 900 USA 902 Phone: +1 732 420 1571 903 Fax: +1 732 368 1192 904 Email: acmorton@att.com 905 URI: http://home.comcast.net/~acmacm/