idnits 2.17.1 draft-vsperf-bmwg-vswitch-opnfv-00.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 (July 3, 2015) is 3221 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Unused Reference: 'RFC2330' is defined on line 763, but no explicit reference was found in the text == Unused Reference: 'RFC2680' is defined on line 773, but no explicit reference was found in the text == Unused Reference: 'RFC2681' is defined on line 776, but no explicit reference was found in the text == Unused Reference: 'RFC3393' is defined on line 782, but no explicit reference was found in the text == Unused Reference: 'RFC3432' is defined on line 786, but no explicit reference was found in the text == Unused Reference: 'RFC4689' is defined on line 793, but no explicit reference was found in the text == Unused Reference: 'RFC5357' is defined on line 801, but no explicit reference was found in the text == Unused Reference: 'RFC5905' is defined on line 805, but no explicit reference was found in the text == Unused Reference: 'RFC6049' is defined on line 825, but no explicit reference was found in the text == Unused Reference: 'RFC6248' is defined on line 828, but no explicit reference was found in the text == Unused Reference: 'RFC6390' is defined on line 832, but no explicit reference was found in the text ** Obsolete normative reference: RFC 2679 (Obsoleted by RFC 7679) ** Obsolete normative reference: RFC 2680 (Obsoleted by RFC 7680) == Outdated reference: A later version (-05) exists of draft-ietf-bmwg-virtual-net-00 Summary: 3 errors (**), 0 flaws (~~), 13 warnings (==), 1 comment (--).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: January 4, 2016 A. Morton 6 AT&T Labs 7 July 3, 2015 9 Benchmarking Virtual Switches in OPNFV 10 draft-vsperf-bmwg-vswitch-opnfv-00 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 January 4, 2016. 49 Copyright Notice 51 Copyright (c) 2015 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 . . . . . . . . . . . . . . . . . . . . . . . . 2 67 2. Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 68 3. Benchmarking Considerations . . . . . . . . . . . . . . . . . 3 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 . . . 6 74 4. VSWITCHPERF Specification Summary . . . . . . . . . . . . . . 8 75 5. 3x3 Matrix Coverage . . . . . . . . . . . . . . . . . . . . . 16 76 5.1. Speed of Activation . . . . . . . . . . . . . . . . . . . 16 77 5.2. Reliability of Activation . . . . . . . . . . . . . . . . 17 78 5.3. Scale of Activation . . . . . . . . . . . . . . . . . . . 17 79 5.4. Speed of Operation . . . . . . . . . . . . . . . . . . . 17 80 5.5. Accuracy of Operation . . . . . . . . . . . . . . . . . . 17 81 5.6. Reliability of Operation . . . . . . . . . . . . . . . . 17 82 5.7. Summary . . . . . . . . . . . . . . . . . . . . . . . . . 18 83 6. Security Considerations . . . . . . . . . . . . . . . . . . . 18 84 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 18 85 8. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 19 86 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 19 87 9.1. Normative References . . . . . . . . . . . . . . . . . . 19 88 9.2. Informative References . . . . . . . . . . . . . . . . . 20 89 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 20 91 1. Introduction 93 Benchmarking Methodology Working Group (BMWG) has traditionally 94 conducted laboratory characterization of dedicated physical 95 implementations of internetworking functions. The Black-box 96 Benchmarks of Throughput, Latency, Forwarding Rates and others have 97 served our industry for many years. Now, Network Function 98 Virtualization (NFV) has the goal to transform how internetwork 99 functions are implemented, and therefore has garnered much attention. 101 This memo describes the progress of the Open Platform for NFV (OPNFV) 102 project on virtual switch performance characterization, 103 "VSWITCHPERF". This project intends to build on the current and 104 completed work of the Benchmarking Methodology Working Group in IETF, 105 by referencing existing literature. For example, currently the most 106 referenced RFC is [RFC2544] (which depends on [RFC1242]) and 107 foundation of the benchmarking work in OPNFV is common and strong. 109 See https://wiki.opnfv.org/ 110 characterize_vswitch_performance_for_telco_nfv_use_cases for more 111 background, and the OPNFV website for general information: 112 https://www.opnfv.org/ 114 The authors note that OPNFV distinguishes itself from other open 115 source compute and networking projects through its emphasis on 116 existing "telco" services as opposed to cloud-computing. There are 117 many ways in which telco requirements have different emphasis on 118 performance dimensions when compared to cloud computing: support for 119 and transfer of isochronous media streams is one example. 121 Note also that the move to NFV Infrastructure has resulted in many 122 new benchmarking initiatives across the industry, and the authors are 123 currently doing their best to maintain alignment with many other 124 projects, and this Internet Draft is evidence of the efforts. 126 2. Scope 128 The primary purpose and scope of the memo is to inform BMWG of work- 129 in-progress that builds on the body of extensive literature and 130 experience. Additionally, once the initial information conveyed here 131 is received, this memo may be expanded to include more detail and 132 commentary from both BMWG and OPNFV communities, under BMWG's 133 chartered work to characterize the NFV Infrastructure (a virtual 134 switch is an important aspect of that infrastructure). 136 3. Benchmarking Considerations 138 This section highlights some specific considerations (from 139 [I-D.ietf-bmwg-virtual-net])related to Benchmarks for virtual 140 switches. The OPNFV project is sharing its present view on these 141 areas, as they develop their specifications in the Level Test Design 142 (LTD) document. 144 3.1. Comparison with Physical Network Functions 146 To compare the performance of virtual designs and implementations 147 with their physical counterparts, identical benchmarks are needed. 148 BMWG has developed specifications for many network functions this 149 memo re-uses existing benchmarks through references, and expands them 150 during development of new methods. A key configuration aspect is the 151 number of parallel cores required to achieve comparable performance 152 with a given physical device, or whether some limit of scale was 153 reached before the cores could achieve the comparable level. 155 It's unlikely that the virtual switch will be the only application 156 running on the SUT, so CPU utilization, Cache utilization, and Memory 157 footprint should also be recorded for the virtual implementations of 158 internetworking functions. 160 3.2. Continued Emphasis on Black-Box Benchmarks 162 External observations remain essential as the basis for Benchmarks. 163 Internal observations with fixed specification and interpretation 164 will be provided in parallel to assist the development of operations 165 procedures when the technology is deployed. 167 3.3. New Configuration Parameters 169 A key consideration when conducting any sort of benchmark is trying 170 to ensure the consistency and repeatability of test results. When 171 benchmarking the performance of a vSwitch there are many factors that 172 can affect the consistency of results, one key factor is matching the 173 various hardware and software details of the SUT. This section lists 174 some of the many new parameters which this project believes are 175 critical to report in order to achieve repeatability. 177 Hardware details including: 179 o Platform details 181 o Processor details 183 o Memory information (type and size) 185 o Number of enabled cores 187 o Number of cores used for the test 189 o Number of physical NICs, as well as their details (manufacturer, 190 versions, type and the PCI slot they are plugged into) 192 o NIC interrupt configuration 194 o BIOS version, release date and any configurations that were 195 modified 197 o CPU microcode level 199 o Memory DIMM configurations (quad rank performance may not be the 200 same as dual rank) in size, freq and slot locations 202 o PCI configuration parameters (payload size, early ack option...) 204 o Power management at all levels (ACPI sleep states, processor 205 package, OS...) 207 Software details including: 209 o OS parameters and behavior (text vs graphical no one typing at the 210 console on one system) 212 o OS version (for host and VNF) 214 o Kernel version (for host and VNF) 216 o GRUB boot parameters (for host and VNF) 218 o Hypervisor details (Type and version) 220 o Selected vSwitch, version number or commit id used 222 o vSwitch launch command line if it has been parameterised 224 o Memory allocation to the vSwitch 226 o which NUMA node it is using, and how many memory channels 228 o DPDK or any other SW dependency version number or commit id used 230 o Memory allocation to a VM - if it's from Hugpages/elsewhere 232 o VM storage type: snapshot/independent persistent/independent non- 233 persistent 235 o Number of VMs 237 o Number of Virtual NICs (vNICs), versions, type and driver 239 o Number of virtual CPUs and their core affinity on the host 240 o Number vNIC interrupt configuration 242 o Thread affinitization for the applications (including the vSwitch 243 itself) on the host 245 o Details of Resource isolation, such as CPUs designated for Host/ 246 Kernel (isolcpu) and CPUs designated for specific processes 247 (taskset). - Test duration. - Number of flows. 249 Test Traffic Information: 251 o Traffic type - UDP, TCP, IMIX / Other 253 o Packet Sizes 255 o Deployment Scenario 257 3.4. Flow classification 259 Virtual switches group packets into flows by processing and matching 260 particular packet or frame header information, or by matching packets 261 based on the input ports. Thus a flow can be thought of a sequence 262 of packets that have the same set of header field values or have 263 arrived on the same port. Performance results can vary based on the 264 parameters the vSwitch uses to match for a flow. The recommended 265 flow classification parameters for any vSwitch performance tests are: 266 the input port, the source IP address, the destination IP address and 267 the ethernet protocol type field. It is essential to increase the 268 flow timeout time on a vSwitch before conducting any performance 269 tests that do not measure the flow setup time. Normally the first 270 packet of a particular stream will install the flow in the virtual 271 switch which adds an additional latency, subsequent packets of the 272 same flow are not subject to this latency if the flow is already 273 installed on the vSwitch. 275 3.5. Benchmarks using Baselines with Resource Isolation 277 This outline describes measurement of baseline with isolated 278 resources at a high level, which is the intended approach at this 279 time. 281 1. Baselines: 283 * Optional: Benchmark platform forwarding capability without a 284 vswitch or VNF for at least 72 hours (serves as a means of 285 platform validation and a means to obtain the base performance 286 for the platform in terms of its maximum forwarding rate and 287 latency). 289 Benchmark platform forwarding capability 291 __ 292 +--------------------------------------------------+ | 293 | +------------------------------------------+ | | 294 | | | | | 295 | | Simple Forwarding App | | Host 296 | | | | | 297 | +------------------------------------------+ | | 298 | | NIC | | | 299 +---+------------------------------------------+---+ __| 300 ^ : 301 | | 302 : v 303 +--------------------------------------------------+ 304 | | 305 | traffic generator | 306 | | 307 +--------------------------------------------------+ 309 * Benchmark VNF forwarding capability with direct connectivity 310 (vSwitch bypass, e.g., SR/IOV) for at least 72 hours (serves 311 as a means of VNF validation and a means to obtain the base 312 performance for the VNF in terms of its maximum forwarding 313 rate and latency). The metrics gathered from this test will 314 serve as a key comparison point for vSwitch bypass 315 technologies performance and vSwitch performance. 317 Benchmark VNF forwarding capability 319 __ 320 +--------------------------------------------------+ | 321 | +------------------------------------------+ | | 322 | | | | | 323 | | VNF | | | 324 | | | | | 325 | +------------------------------------------+ | | 326 | | Passthrough/SR-IOV | | Host 327 | +------------------------------------------+ | | 328 | | NIC | | | 329 +---+------------------------------------------+---+ __| 330 ^ : 331 | | 332 : v 333 +--------------------------------------------------+ 334 | | 335 | traffic generator | 336 | | 337 +--------------------------------------------------+ 339 * Benchmarking with isolated resources alone, with other 340 resources (both HW&SW) disabled Example, vSw and VM are SUT 342 * Benchmarking with isolated resources alone, leaving some 343 resources unused 345 * Benchmark with isolated resources and all resources occupied 347 2. Next Steps 349 * Limited sharing 351 * Production scenarios 353 * Stressful scenarios 355 4. VSWITCHPERF Specification Summary 357 The overall specification in preparation is referred to as a Level 358 Test Design (LTD) document, which will contain a suite of performance 359 tests. The base performance tests in the LTD are based on the pre- 360 existing specifications developed by BMWG to test the performance of 361 physical switches. These specifications include: 363 o [RFC2544] Benchmarking Methodology for Network Interconnect 364 Devices 366 o [RFC2889] Benchmarking Methodology for LAN Switching 368 o [RFC6201] Device Reset Characterization 370 o [RFC5481] Packet Delay Variation Applicability Statement 372 In addition to this, the LTD also re-uses the terminology defined by: 374 o [RFC2285] Benchmarking Terminology for LAN Switching Devices 376 o [RFC5481] Packet Delay Variation Applicability Statement 378 Specifications to be included in future updates of the LTD include: 380 o [RFC3918] Methodology for IP Multicast Benchmarking 382 o [RFC4737] Packet Reordering Metrics 384 As one might expect, the most fundamental internetworking 385 characteristics of Throughput and Latency remain important when the 386 switch is virtualized, and these benchmarks figure prominently in the 387 specification. 389 When considering characteristics important to "telco" network 390 functions, we must begin to consider additional performance metrics. 391 In this case, the project specifications have referenced metrics from 392 the IETF IP Performance Metrics (IPPM) literature. This means that 393 the [RFC2544] test of Latency is replaced by measurement of a metric 394 derived from IPPM's [RFC2679], where a set of statistical summaries 395 will be provided (mean, max, min, etc.). Further metrics planned to 396 be benchmarked include packet delay variation as defined by [RFC5481] 397 , reordering, burst behaviour, DUT availability, DUT capacity and 398 packet loss in long term testing at Throughput level, where some low- 399 level of background loss may be present and characterized. 401 Tests have been (or will be) designed to collect the metrics below: 403 o Throughput Tests to measure the maximum forwarding rate (in frames 404 per second or fps) and bit rate (in Mbps) for a constant load (as 405 defined by RFC1242) without traffic loss. 407 o Packet and Frame Delay Distribution Tests to measure average, min 408 and max packet and frame delay for constant loads. 410 o Packet Delay Tests to understand latency distribution for 411 different packet sizes and over an extended test run to uncover 412 outliers. 414 o Scalability Tests to understand how the virtual switch performs as 415 the number of flows, active ports, complexity of the forwarding 416 logic's configuration... it has to deal with increases. 418 o Stream Performance Tests (TCP, UDP) to measure bulk data transfer 419 performance, i.e. how fast systems can send and receive data 420 through the switch. 422 o Control Path and Datapath Coupling Tests, to understand how 423 closely coupled the datapath and the control path are as well as 424 the effect of this coupling on the performance of the DUT 425 (example: delay of the initial packet of a flow). 427 o CPU and Memory Consumption Tests to understand the virtual 428 switch's footprint on the system, usually conducted as auxiliary 429 measurements with benchmarks above. They include: CPU 430 utilization, Cache utilization and Memory footprint. 432 Future/planned test specs include: 434 o Request/Response Performance Tests (TCP, UDP) which measure the 435 transaction rate through the switch. 437 o Noisy Neighbour Tests, to understand the effects of resource 438 sharing on the performance of a virtual switch. 440 The flexibility of deployment of a virtual switch within a network 441 means that the BMWG IETF existing literature needs to be used to 442 characterize the performance of a switch in various deployment 443 scenarios. The deployment scenarios under consideration include: 445 Physical port to virtual switch to physical port 447 __ 448 +--------------------------------------------------+ | 449 | +--------------------+ | | 450 | | | | | 451 | | v | | Host 452 | +--------------+ +--------------+ | | 453 | | phy port | vSwitch | phy port | | | 454 +---+--------------+------------+--------------+---+ __| 455 ^ : 456 | | 457 : v 458 +--------------------------------------------------+ 459 | | 460 | traffic generator | 461 | | 462 +--------------------------------------------------+ 464 Physical port to virtual switch to VNF to virtual switch to physical 465 port 467 __ 468 +---------------------------------------------------+ | 469 | | | 470 | +-------------------------------------------+ | | 471 | | Application | | | 472 | +-------------------------------------------+ | | 473 | ^ : | | 474 | | | | | Guest 475 | : v | | 476 | +---------------+ +---------------+ | | 477 | | logical port 0| | logical port 1| | | 478 +---+---------------+-----------+---------------+---+ __| 479 ^ : 480 | | 481 : v __ 482 +---+---------------+----------+---------------+---+ | 483 | | logical port 0| | logical port 1| | | 484 | +---------------+ +---------------+ | | 485 | ^ : | | 486 | | | | | Host 487 | : v | | 488 | +--------------+ +--------------+ | | 489 | | phy port | vSwitch | phy port | | | 490 +---+--------------+------------+--------------+---+ __| 491 ^ : 492 | | 493 : v 494 +--------------------------------------------------+ 495 | | 496 | traffic generator | 497 | | 498 +--------------------------------------------------+ 500 Physical port to virtual switch to VNF to virtual switch to VNF to 501 virtual switch to physical port 503 __ 504 +----------------------+ +----------------------+ | 505 | Guest 1 | | Guest 2 | | 506 | +---------------+ | | +---------------+ | | 507 | | Application | | | | Application | | | 508 | +---------------+ | | +---------------+ | | 509 | ^ | | | ^ | | | 510 | | v | | | v | | Guests 511 | +---------------+ | | +---------------+ | | 512 | | logical ports | | | | logical ports | | | 513 | | 0 1 | | | | 0 1 | | | 514 +---+---------------+--+ +---+---------------+--+__| 515 ^ : ^ : 516 | | | | 517 : v : v _ 518 +---+---------------+---------+---------------+--+ | 519 | | 0 1 | | 3 4 | | | 520 | | logical ports | | logical ports | | | 521 | +---------------+ +---------------+ | | 522 | ^ | ^ | | | Host 523 | | |-----------------| v | | 524 | +--------------+ +--------------+ | | 525 | | phy ports | vSwitch | phy ports | | | 526 +---+--------------+----------+--------------+---+_| 527 ^ : 528 | | 529 : v 530 +--------------------------------------------------+ 531 | | 532 | traffic generator | 533 | | 534 +--------------------------------------------------+ 536 Physical port to virtual switch to VNF 538 __ 539 +---------------------------------------------------+ | 540 | | | 541 | +-------------------------------------------+ | | 542 | | Application | | | 543 | +-------------------------------------------+ | | 544 | ^ | | 545 | | | | Guest 546 | : | | 547 | +---------------+ | | 548 | | logical port 0| | | 549 +---+---------------+-------------------------------+ __| 550 ^ 551 | 552 : __ 553 +---+---------------+------------------------------+ | 554 | | logical port 0| | | 555 | +---------------+ | | 556 | ^ | | 557 | | | | Host 558 | : | | 559 | +--------------+ | | 560 | | phy port | vSwitch | | 561 +---+--------------+------------ -------------- ---+ __| 562 ^ 563 | 564 : 565 +--------------------------------------------------+ 566 | | 567 | traffic generator | 568 | | 569 +--------------------------------------------------+ 571 VNF to virtual switch to physical port 573 __ 574 +---------------------------------------------------+ | 575 | | | 576 | +-------------------------------------------+ | | 577 | | Application | | | 578 | +-------------------------------------------+ | | 579 | : | | 580 | | | | Guest 581 | v | | 582 | +---------------+ | | 583 | | logical port | | | 584 +-------------------------------+---------------+---+ __| 585 : 586 | 587 v __ 588 +------------------------------+---------------+---+ | 589 | | logical port | | | 590 | +---------------+ | | 591 | : | | 592 | | | | Host 593 | v | | 594 | +--------------+ | | 595 | vSwitch | phy port | | | 596 +-------------------------------+--------------+---+ __| 597 : 598 | 599 v 600 +--------------------------------------------------+ 601 | | 602 | traffic generator | 603 | | 604 +--------------------------------------------------+ 606 VNF to virtual switch to VNF 608 __ 609 +----------------------+ +----------------------+ | 610 | Guest 1 | | Guest 2 | | 611 | +---------------+ | | +---------------+ | | 612 | | Application | | | | Application | | | 613 | +---------------+ | | +---------------+ | | 614 | | | | ^ | | 615 | v | | | | | Guests 616 | +---------------+ | | +---------------+ | | 617 | | logical ports | | | | logical ports | | | 618 | | 0 | | | | 0 | | | 619 +---+---------------+--+ +---+---------------+--+__| 620 : ^ 621 | | 622 v : _ 623 +---+---------------+---------+---------------+--+ | 624 | | 1 | | 1 | | | 625 | | logical ports | | logical ports | | | 626 | +---------------+ +---------------+ | | 627 | | ^ | | Host 628 | L-----------------+ | | 629 | | | 630 | vSwitch | | 631 +------------------------------------------------+_| 633 5. 3x3 Matrix Coverage 635 This section organizes the many existing test specifications into the 636 "3x3" matrix (introduced in [I-D.ietf-bmwg-virtual-net]). Because 637 the LTD specification ID names are quite long, this section is 638 organized into lists for each occupied cell of the matrix (not all 639 are occupied, also the matrix has grown to 3x4 to accommodate scale 640 metrics). 642 The tests listed below assess the activation of paths in the data 643 plane, rather than the control plane. 645 (Editor's Note: a complete list of tests is available here: 646 https://wiki.opnfv.org/wiki/vswitchperf_test_spec_review ) 648 5.1. Speed of Activation 650 o Throughput.RFC2889.AddressLearningRate 652 o Throughput.RFC2889.AddressCachingCapacity 653 o PacketLatency.InitialPacketProcessingLatency 655 o 657 5.2. Reliability of Activation 659 o Throughput.RFC2544.SystemRecoveryTime 661 o Throughput.RFC2544.ResetTime 663 5.3. Scale of Activation 665 o Throughput.RFC2889.AddressCachingCapacity 667 o 669 5.4. Speed of Operation 671 o Throughput.RFC2544.PacketLossRate 673 o Throughput.RFC2544.PacketLossRateFrameModification 675 o Throughput.RFC2544.BackToBackFrames 677 o Throughput.RFC2889.ForwardingRate 679 o Throughput.RFC2889.ForwardPressure 681 o Throughput.RFC2889.BroadcastFrameForwarding 683 o RFC2889 Broadcast Frame Latency test 685 5.5. Accuracy of Operation 687 o Throughput.RFC2889.ErrorFramesFiltering 689 o 691 5.6. Reliability of Operation 693 o Throughput.RFC2544.Soak 695 o Throughput.RFC2544.SoakFrameModification 697 o 699 5.7. Summary 701 |------------------------------------------------------------------------| 702 | | | | | | 703 | | SPEED | ACCURACY | RELIABILITY | SCALE | 704 | | | | | | 705 |------------------------------------------------------------------------| 706 | | | | | | 707 | Activation | X | | X | X | 708 | | | | | | 709 |------------------------------------------------------------------------| 710 | | | | | | 711 | Operation | X | X | X | | 712 | | | | | | 713 |------------------------------------------------------------------------| 714 | | | | | | 715 | De-activation | | | | | 716 | | | | | | 717 |------------------------------------------------------------------------| 719 6. Security Considerations 721 Benchmarking activities as described in this memo are limited to 722 technology characterization of a Device Under Test/System Under Test 723 (DUT/SUT) using controlled stimuli in a laboratory environment, with 724 dedicated address space and the constraints specified in the sections 725 above. 727 The benchmarking network topology will be an independent test setup 728 and MUST NOT be connected to devices that may forward the test 729 traffic into a production network, or misroute traffic to the test 730 management network. 732 Further, benchmarking is performed on a "black-box" basis, relying 733 solely on measurements observable external to the DUT/SUT. 735 Special capabilities SHOULD NOT exist in the DUT/SUT specifically for 736 benchmarking purposes. Any implications for network security arising 737 from the DUT/SUT SHOULD be identical in the lab and in production 738 networks. 740 7. IANA Considerations 742 No IANA Action is requested at this time. 744 8. Acknowledgements 746 The authors acknowledge 748 9. References 750 9.1. Normative References 752 [NFV.PER001] 753 "Network Function Virtualization: Performance and 754 Portability Best Practices", Group Specification ETSI GS 755 NFV-PER 001 V1.1.1 (2014-06), June 2014. 757 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 758 Requirement Levels", BCP 14, RFC 2119, March 1997. 760 [RFC2285] Mandeville, R., "Benchmarking Terminology for LAN 761 Switching Devices", RFC 2285, February 1998. 763 [RFC2330] Paxson, V., Almes, G., Mahdavi, J., and M. Mathis, 764 "Framework for IP Performance Metrics", RFC 2330, May 765 1998. 767 [RFC2544] Bradner, S. and J. McQuaid, "Benchmarking Methodology for 768 Network Interconnect Devices", RFC 2544, March 1999. 770 [RFC2679] Almes, G., Kalidindi, S., and M. Zekauskas, "A One-way 771 Delay Metric for IPPM", RFC 2679, September 1999. 773 [RFC2680] Almes, G., Kalidindi, S., and M. Zekauskas, "A One-way 774 Packet Loss Metric for IPPM", RFC 2680, September 1999. 776 [RFC2681] Almes, G., Kalidindi, S., and M. Zekauskas, "A Round-trip 777 Delay Metric for IPPM", RFC 2681, September 1999. 779 [RFC2889] Mandeville, R. and J. Perser, "Benchmarking Methodology 780 for LAN Switching Devices", RFC 2889, August 2000. 782 [RFC3393] Demichelis, C. and P. Chimento, "IP Packet Delay Variation 783 Metric for IP Performance Metrics (IPPM)", RFC 3393, 784 November 2002. 786 [RFC3432] Raisanen, V., Grotefeld, G., and A. Morton, "Network 787 performance measurement with periodic streams", RFC 3432, 788 November 2002. 790 [RFC3918] Stopp, D. and B. Hickman, "Methodology for IP Multicast 791 Benchmarking", RFC 3918, October 2004. 793 [RFC4689] Poretsky, S., Perser, J., Erramilli, S., and S. Khurana, 794 "Terminology for Benchmarking Network-layer Traffic 795 Control Mechanisms", RFC 4689, October 2006. 797 [RFC4737] Morton, A., Ciavattone, L., Ramachandran, G., Shalunov, 798 S., and J. Perser, "Packet Reordering Metrics", RFC 4737, 799 November 2006. 801 [RFC5357] Hedayat, K., Krzanowski, R., Morton, A., Yum, K., and J. 802 Babiarz, "A Two-Way Active Measurement Protocol (TWAMP)", 803 RFC 5357, October 2008. 805 [RFC5905] Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network 806 Time Protocol Version 4: Protocol and Algorithms 807 Specification", RFC 5905, June 2010. 809 [RFC6201] Asati, R., Pignataro, C., Calabria, F., and C. Olvera, 810 "Device Reset Characterization", RFC 6201, March 2011. 812 9.2. Informative References 814 [I-D.ietf-bmwg-virtual-net] 815 Morton, A., "Considerations for Benchmarking Virtual 816 Network Functions and Their Infrastructure", draft-ietf- 817 bmwg-virtual-net-00 (work in progress), June 2015. 819 [RFC1242] Bradner, S., "Benchmarking terminology for network 820 interconnection devices", RFC 1242, July 1991. 822 [RFC5481] Morton, A. and B. Claise, "Packet Delay Variation 823 Applicability Statement", RFC 5481, March 2009. 825 [RFC6049] Morton, A. and E. Stephan, "Spatial Composition of 826 Metrics", RFC 6049, January 2011. 828 [RFC6248] Morton, A., "RFC 4148 and the IP Performance Metrics 829 (IPPM) Registry of Metrics Are Obsolete", RFC 6248, April 830 2011. 832 [RFC6390] Clark, A. and B. Claise, "Guidelines for Considering New 833 Performance Metric Development", BCP 170, RFC 6390, 834 October 2011. 836 Authors' Addresses 837 Maryam Tahhan 838 Intel 840 Email: maryam.tahhan@intel.com 842 Billy O'Mahony 843 Intel 845 Email: billy.o.mahony@intel.com 847 Al Morton 848 AT&T Labs 849 200 Laurel Avenue South 850 Middletown,, NJ 07748 851 USA 853 Phone: +1 732 420 1571 854 Fax: +1 732 368 1192 855 Email: acmorton@att.com 856 URI: http://home.comcast.net/~acmacm/