RFC 3116
Methodology for ATM Benchmarking
Pages: 127
Informational
Informational
Part 1 of 5 – Pages 1 to 17
Network Working Group J. Dunn Request for Comments: 3116 C. Martin Category: Informational ANC, Inc. June 2001 Methodology for ATM Benchmarking Status of this Memo This memo provides information for the Internet community. It does not specify an Internet standard of any kind. Distribution of this memo is unlimited. Copyright Notice Copyright (C) The Internet Society (2001). All Rights Reserved.Abstract
This document discusses and defines a number of tests that may be used to describe the performance characteristics of ATM (Asynchronous Transfer Mode) based switching devices. In addition to defining the tests this document also describes specific formats for reporting the results of the tests. This memo is a product of the Benchmarking Methodology Working Group (BMWG) of the Internet Engineering Task Force (IETF).Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 2. Background . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.1. Test Device Requirements . . . . . . . . . . . . . . . . . . 5 2.2. Systems Under Test (SUTs). . . . . . . . . . . . . . . . . . 5 2.3. Test Result Evaluation . . . . . . . . . . . . . . . . . . . 5 2.4. Requirements . . . . . . . . . . . . . . . . . . . . . . . . 5 2.5. Test Configurations for SONET. . . . . . . . . . . . . . . . 6 2.6. SUT Configuration. . . . . . . . . . . . . . . . . . . . . . 7 2.7. Frame Formats. . . . . . . . . . . . . . . . . . . . . . . . 8 2.8. Frame Sizes. . . . . . . . . . . . . . . . . . . . . . . . . 8 2.9. Verifying Received IP PDU's. . . . . . . . . . . . . . . . . 9 2.10. Modifiers . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.10.1. Management IP PDU's . . . . . . . . . . . . . . . . . . . 9 2.10.2. Routing Update IP PDU's . . . . . . . . . . . . . . . . . 10 2.11. Filters . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.11.1. Filter Addresses. . . . . . . . . . . . . . . . . . . . . 11 2.12. Protocol Addresses. . . . . . . . . . . . . . . . . . . . . 12
2.13. Route Set Up. . . . . . . . . . . . . . . . . . . . . . . . 12 2.14. Bidirectional Traffic . . . . . . . . . . . . . . . . . . . 12 2.15. Single Stream Path. . . . . . . . . . . . . . . . . . . . . 12 2.16. Multi-port. . . . . . . . . . . . . . . . . . . . . . . . . 13 2.17. Multiple Protocols. . . . . . . . . . . . . . . . . . . . . 14 2.18. Multiple IP PDU Sizes . . . . . . . . . . . . . . . . . . . 14 2.19. Testing Beyond a Single SUT . . . . . . . . . . . . . . . . 14 2.20. Maximum IP PDU Rate . . . . . . . . . . . . . . . . . . . . 15 2.21. Busty Traffic . . . . . . . . . . . . . . . . . . . . . . . 15 2.22. Trial Description . . . . . . . . . . . . . . . . . . . . . 16 2.23. Trial Duration. . . . . . . . . . . . . . . . . . . . . . . 16 2.24. Address Resolution. . . . . . . . . . . . . . . . . . . . . 16 2.25. Synchronized Payload Bit Pattern. . . . . . . . . . . . . . 16 2.26. Burst Traffic Descriptors . . . . . . . . . . . . . . . . . 17 3. Performance Metrics. . . . . . . . . . . . . . . . . . . . . . 17 3.1. Physical Layer-SONET . . . . . . . . . . . . . . . . . . . . 17 3.1.1. Pointer Movements. . . . . . . . . . . . . . . . . . . . . 17 3.1.1.1. Pointer Movement Propagation . . . . . . . . . . . . . . 17 3.1.1.2. Cell Loss due to Pointer Movement. . . . . . . . . . . . 19 3.1.1.3. IP Packet Loss due to Pointer Movement . . . . . . . . . 20 3.1.2. Transport Overhead (TOH) Error Count . . . . . . . . . . . 21 3.1.2.1. TOH Error Propagation. . . . . . . . . . . . . . . . . . 21 3.1.2.2. Cell Loss due to TOH Error . . . . . . . . . . . . . . . 22 3.1.2.3. IP Packet Loss due to TOH Error. . . . . . . . . . . . . 23 3.1.3. Path Overhead (POH) Error Count. . . . . . . . . . . . . . 24 3.1.3.1. POH Error Propagation. . . . . . . . . . . . . . . . . . 24 3.1.3.2. Cell Loss due to POH Error . . . . . . . . . . . . . . . 25 3.1.3.3. IP Packet Loss due to POH Error. . . . . . . . . . . . . 26 3.2. ATM Layer. . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.2.1. Two-Point Cell Delay Variation (CDV) . . . . . . . . . . . 27 3.2.1.1. Test Setup . . . . . . . . . . . . . . . . . . . . . . . 27 3.2.1.2. Two-point CDV/Steady Load/One VCC. . . . . . . . . . . . 27 3.2.1.3. Two-point CDV/Steady Load/Twelve VCCs. . . . . . . . . . 28 3.2.1.4. Two-point CDV/Steady Load/Maximum VCCs . . . . . . . . . 30 3.2.1.5. Two-point CDV/Bursty VBR Load/One VCC. . . . . . . . . . 31 3.2.1.6. Two-point CDV/Bursty VBR Load/Twelve VCCs. . . . . . . . 32 3.2.1.7. Two-point CDV/Bursty VBR Load/Maximum VCCs . . . . . . . 34 3.2.1.8. Two-point CDV/Mixed Load/Three VCC's . . . . . . . . . . 35 3.2.1.9. Two-point CDV/Mixed Load/Twelve VCCs . . . . . . . . . . 36 3.2.1.10. Two-point CDV/Mixed Load/Maximum VCCs . . . . . . . . . 38 3.2.2. Cell Error Ratio (CER) . . . . . . . . . . . . . . . . . . 39 3.2.2.1. Test Setup . . . . . . . . . . . . . . . . . . . . . . . 39 3.2.2.2. CER/Steady Load/One VCC. . . . . . . . . . . . . . . . . 40 3.2.2.3. CER/Steady Load/Twelve VCCs. . . . . . . . . . . . . . . 41 3.2.2.4. CER/Steady Load/Maximum VCCs . . . . . . . . . . . . . . 42 3.2.2.5. CER/Bursty VBR Load/One VCC. . . . . . . . . . . . . . . 43 3.2.2.6. CER/Bursty VBR Load/Twelve VCCs. . . . . . . . . . . . . 44 3.2.2.7. CER/Bursty VBR Load/Maximum VCCs . . . . . . . . . . . . 46
3.2.3. Cell Loss Ratio (CLR). . . . . . . . . . . . . . . . . . . 47 3.2.3.1. CLR/Steady Load/One VCC. . . . . . . . . . . . . . . . . 47 3.2.3.2. CLR/Steady Load/Twelve VCCs. . . . . . . . . . . . . . . 48 3.2.3.3. CLR/Steady Load/Maximum VCCs . . . . . . . . . . . . . . 49 3.2.3.4. CLR/Bursty VBR Load/One VCC. . . . . . . . . . . . . . . 51 3.2.3.5. CLR/Bursty VBR Load/Twelve VCCs. . . . . . . . . . . . . 52 3.2.3.6. CLR/Bursty VBR Load/Maximum VCCs . . . . . . . . . . . . 53 3.2.4. Cell Misinsertion Rate (CMR) . . . . . . . . . . . . . . . 54 3.2.4.1. CMR/Steady Load/One VCC. . . . . . . . . . . . . . . . . 54 3.2.4.2. CMR/Steady Load/Twelve VCCs. . . . . . . . . . . . . . . 55 3.2.4.3. CMR/Steady Load/Maximum VCCs . . . . . . . . . . . . . . 57 3.2.4.4. CMR/Bursty VBR Load/One VCC. . . . . . . . . . . . . . . 58 3.2.4.5. CMR/Bursty VBR Load/Twelve VCCs. . . . . . . . . . . . . 59 3.2.4.6. CMR/Bursty VBR Load/Maximum VCCs . . . . . . . . . . . . 60 3.2.5. CRC Error Ratio (CRC-ER) . . . . . . . . . . . . . . . . . 62 3.2.5.1. CRC-ER/Steady Load/One VCC . . . . . . . . . . . . . . . 62 3.2.5.2. CRC-ER/Steady Load/Twelve VCCs . . . . . . . . . . . . . 63 3.2.5.3. CRC-ER/Steady Load/Maximum VCCs. . . . . . . . . . . . . 64 3.2.5.4. CRC-ER/Bursty VBR Load/One VCC . . . . . . . . . . . . . 65 3.2.5.5. CRC-ER/Bursty VBR Load/Twelve VCCs . . . . . . . . . . . 66 3.2.5.6. CRC-ER/Bursty VBR Load/Maximum VCCs. . . . . . . . . . . 68 3.2.5.7. CRC-ER/Bursty UBR Load/One VCC . . . . . . . . . . . . . 69 3.2.5.8. CRC-ER/Bursty UBR Load/Twelve VCCs . . . . . . . . . . . 70 3.2.5.9. CRC-ER/Bursty UBR Load/Maximum VCCs. . . . . . . . . . . 71 3.2.5.10. CRC-ER/Bursty Mixed Load/Three VCC. . . . . . . . . . . 73 3.2.5.11. CRC-ER/Bursty Mixed Load/Twelve VCCs. . . . . . . . . . 74 3.2.5.12. CRC-ER/Bursty Mixed Load/Maximum VCCs . . . . . . . . . 75 3.2.6. Cell Transfer Delay (CTD). . . . . . . . . . . . . . . . . 76 3.2.6.1. Test Setup . . . . . . . . . . . . . . . . . . . . . . . 76 3.2.6.2. CTD/Steady Load/One VCC. . . . . . . . . . . . . . . . . 77 3.2.6.3. CTD/Steady Load/Twelve VCCs. . . . . . . . . . . . . . . 78 3.2.6.4. CTD/Steady Load/Maximum VCCs . . . . . . . . . . . . . . 79 3.2.6.5. CTD/Bursty VBR Load/One VCC. . . . . . . . . . . . . . . 81 3.2.6.6. CTD/Bursty VBR Load/Twelve VCCs. . . . . . . . . . . . . 82 3.2.6.7. CTD/Bursty VBR Load/Maximum VCCs . . . . . . . . . . . . 83 3.2.6.8. CTD/Bursty UBR Load/One VCC. . . . . . . . . . . . . . . 85 3.2.6.9. CTD/Bursty UBR Load/Twelve VCCs. . . . . . . . . . . . . 86 3.2.6.10. CTD/Bursty UBR Load/Maximum VCCs. . . . . . . . . . . . 87 3.2.6.11. CTD/Mixed Load/Three VCC's. . . . . . . . . . . . . . . 88 3.2.6.12. CTD/Mixed Load/Twelve VCCs. . . . . . . . . . . . . . . 90 3.2.6.13. CTD/Mixed Load/Maximum VCCs . . . . . . . . . . . . . . 91 3.3. ATM Adaptation Layer (AAL) Type 5 (AAL5) . . . . . . . . . . 93 3.3.1. IP Packet Loss due to AAL5 Re-assembly Errors. . . . . . . 93 3.3.2. AAL5 Re-assembly Time. . . . . . . . . . . . . . . . . . . 94 3.3.3. AAL5 CRC Error Ratio . . . . . . . . . . . . . . . . . . . 95 3.3.3.1. Test Setup . . . . . . . . . . . . . . . . . . . . . . . 95 3.3.3.2. AAL5-CRC-ER/Steady Load/One VCC. . . . . . . . . . . . . 95 3.3.3.3. AAL5-CRC-ER/Steady Load/Twelve VCCs. . . . . . . . . . . 96
3.3.3.4. AAL5-CRC-ER/Steady Load/Maximum VCCs . . . . . . . . . . 97 3.3.3.5. AAL5-CRC-ER/Bursty VBR Load/One VCC. . . . . . . . . . . 99 3.3.3.6. AAL5-CRC-ER/Bursty VBR Load/Twelve VCCs. . . . . . . . .100 3.3.3.7. AAL5-CRC-ER/Bursty VBR Load/Maximum VCCs . . . . . . . .101 3.3.3.8. AAL5-CRC-ER/Mixed Load/Three VCC's . . . . . . . . . . .102 3.3.3.9. AAL5-CRC-ER/Mixed Load/Twelve VCCs . . . . . . . . . . .104 3.3.3.10. AAL5-CRC-ER/Mixed Load/Maximum VCCs . . . . . . . . . .105 3.4. ATM Service: Signaling . . . . . . . . . . . . . . . . . . .106 3.4.1. CAC Denial Time and Connection Establishment Time. . . . .106 3.4.2. Connection Teardown Time . . . . . . . . . . . . . . . . .107 3.4.3. Crankback Time . . . . . . . . . . . . . . . . . . . . . .108 3.4.4. Route Update Response Time . . . . . . . . . . . . . . . .109 3.5. ATM Service: ILMI. . . . . . . . . . . . . . . . . . . . . .110 3.5.1. MIB Alignment Time . . . . . . . . . . . . . . . . . . . .110 3.5.2. Address Registration Time. . . . . . . . . . . . . . . . .111 4. Security Considerations . . . . . . . . . . . . . . . . . . .112 5. Notices. . . . . . . . . . . . . . . . . . . . . . . . . . . .112 6. References . . . . . . . . . . . . . . . . . . . . . . . . . .113 7. Authors' Addresses . . . . . . . . . . . . . . . . . . . . . .113 APPENDIX A . . . . . . . . . . . . . . . . . . . . . . . . . . .114 APPENDIX B . . . . . . . . . . . . . . . . . . . . . . . . . . .114 APPENDIX C . . . . . . . . . . . . . . . . . . . . . . . . . . .116 Full Copyright Statement . . . . . . . . . . . . . . . . . . . .1271. Introduction
This document defines a specific set of tests that vendors can use to measure and report the performance characteristics of ATM network devices. The results of these tests will provide the user comparable data from different vendors with which to evaluate these devices. The methods defined in this memo are based on RFC 2544 "Benchmarking Methodology for Network Interconnect Devices". The document "Terminology for ATM Benchmarking" (RFC 2761), defines many of the terms that are used in this document. The terminology document should be consulted before attempting to make use of this document. The BMWG produces two major classes of documents: Benchmarking Terminology documents and Benchmarking Methodology documents. The Terminology documents present the benchmarks and other related terms. The Methodology documents define the procedures required to collect the benchmarks cited in the corresponding Terminology documents.
2. Background
2.1. Test Device Requirements
This document is based on the requirement that a test device is available. The test device can either be off the shelf or can be easily built with current technologies. The test device must have a transmitting and receiving port for the interface type under test. The test device must be configured to transmit test PDUs and to analyze received PDUs. The test device should be able to transmit and analyze received data at the same time.2.2. Systems Under Test (SUTs)
There are a number of tests described in this document that do not apply to each SUT. Vendors should perform all of the tests that can be supported by a specific product type. It will take some time to perform all of the recommended tests under all of the recommended conditions.2.3. Test Result Evaluation
Performing all of the tests in this document will result in a great deal of data. The applicability of this data to the evaluation of a particular SUT will depend on its expected use and the configuration of the network in which it will be used. For example, the time required by a switch to provide ILMI services will not be a pertinent measurement in a network that does not use the ILMI protocol, such as an ATM WAN. Evaluating data relevant to a particular network installation may require considerable experience, which may not be readily available. Finally, test selection and evaluation of test results must be done with an understanding of generally accepted testing practices regarding repeatability, variance and the statistical significance of a small numbers of trials.2.4. Requirements
In this document, the words that are used to define the significance of each particular requirement are capitalized. These words are: * "MUST" This word, or the words "REQUIRED" and "SHALL" mean that the item is an absolute requirement of the specification. * "SHOULD" This word or the adjective "RECOMMENDED" means that there may exist valid reasons in particular circumstances to ignore this item, but the full implications should be understood and the case carefully weighed before choosing a different course.
* "MAY" This word or the adjective "OPTIONAL" means that this item
is truly optional. One vendor may choose to include the item
because a particular marketplace requires it or because it
enhances the product, for example; another vendor may omit the
same item.
An implementation is not compliant if it fails to satisfy one or more
of the MUST requirements for the protocols it implements. An
implementation that satisfies all the MUST and all the SHOULD
requirements for its protocols is said to be "unconditionally
compliant"; one that satisfies all the MUST requirements but not all
the SHOULD requirements for its protocols is said to be
"conditionally compliant".
2.5. Test Configurations for SONET
The test device can be connected to the SUT in a variety of
configurations depending on the test point. The following
configurations will be used for the tests described in this document.
1) Uni-directional connection: The test devices transmit port
(labeled Tx) is connected to the SUT receive port (labeled Rx).
The SUTs transmit port is connected to the test device receive
port (see Figure 1). In this configuration, the test device can
verify that all transmitted packets are acknowledged correctly.
Note that this configuration does not verify internal system
functions, but verifies one port on the SUT.
+-------------+ +-------------+
| Tx|-------------->|Rx |
| Test Rx|<--------------|Tx SUT |
| Device | | |
+-------------+ +-------------+
Figure 1
2) Bi-directional connection: The test devices first transmit port is
connected to the SUTs first receive port. The SUTs first transmit
port is connected to the test devices first receive port. The
test devices second transmit port is connected to the SUTs second
receive port. The SUTs second transmit port is connected to the
test devices second receive port (see Figure 2). In this
configuration, the test device can determine if all of the
transmitted packets were received and forwarded correctly. Note
that this configuration does verify internal system functions,
since it verifies two ports on the SUT.
+-------------+ +-------------+ | Test Tx|-------------->|Rx | | Device Rx|<--------------|Tx SUT | | Tx Rx | | Tx Rx | +-------------+ +-------------+ | ^ | ^ | | | | | +------------------------+ | | | |---------------------------------| Figure 2 3) Uni-directional passthrough connection: The test devices first transmit port is connected to the SUT1 receive port. The SUT1 transmit port is connected to the test devices first receive port. The test devices second transmit port is connected to the SUT2 receive port. The SUT2 transmit port is connected to the test devices second receive port (see Figure 3). In this configuration, the test device can determine if all of the packets transmitted by SUT1 were correctly acknowledged by SUT2. Note that this configuration does not verify internal system functions, but verifies one port on each SUT. +-------------+ +-------------+ +-------------+ | Tx|---------->|Rx Tx|---------->|Rx | | SUT1 Rx|<----------|Tx Test Rx|<----------|Tx SUT2 | | | | Device | | | +-------------+ +-------------+ +-------------+ Figure 32.6. SUT Configuration
The SUT MUST be configured as described in the SUT users guide. Specifically, it is expected that all of the supported protocols will be configured and enabled. It is expected that all of the tests will be run without changing the configuration or setup of the SUT in any way other than that required to do the specific test. For example, it is not acceptable to disable all but one transport protocol when testing the throughput of that protocol. If PNNI or BISUP is used to initiate switched virtual connections (SVCs), the SUT configuration SHOULD include the normally recommended routing update intervals and keep alive frequency. The specific version of the software and the exact SUT configuration, including what functions are disabled and used during the tests MUST be included as part of the report of the results.
2.7. Frame formats
The formats of the test IP PDUs to use for TCP/IP and UPC/IP over ATM are shown in Appendix C: Test Frame Formats. Note that these IP PDUs are in accordance with RFC 2225. These exact IP PDU formats SHOULD be used in the tests described in this document for this protocol/media combination. These IP PDUs will be used as a template for testing other protocol/media combinations. The specific formats that are used to define the test IP PDUs for a particular test series MUST be included in the report of the results.2.8. Frame sizes
All of the described tests SHOULD be performed using a number of IP PDU sizes. Specifically, the sizes SHOULD include the maximum and minimum legitimate sizes for the protocol under test on the media under test and enough sizes in between to be able to get a full characterization of the SUT performance. Except where noted, at least five IP PDU sizes SHOULD be tested for each test condition. Theoretically the minimum size UDP Echo request IP PDU would consist of an IP header (minimum length 20 octets), a UDP header (8 octets), AAL5 trailer (8 octets) and an LLC/SNAP code point header (8 octets); therefore, the minimum size PDU will fit into one ATM cell. The theoretical maximum IP PDU size is determined by the size of the length field in the IP header. In almost all cases the actual maximum and minimum sizes are determined by the limitations of the media. In the case of ATM, the maximum IP PDU size SHOULD be the ATM MTU size, which is 9180 octets. In theory it would be ideal to distribute the IP PDU sizes in a way that would evenly distribute the theoretical IP PDU rates. These recommendations incorporate this theory but specify IP PDU sizes, which are easy to understand and remember. In addition, many of the same IP PDU sizes are specified on each of the media types to allow for easy performance comparisons. Note: The inclusion of an unrealistically small IP PDU size on some of the media types (i.e., with little or no space for data) is to help characterize the per-IP PDU processing overhead of the SUT. The IP PDU sizes that will be used are: 44, 64, 128, 256, 1024, 1518, 2048, 4472, 9180 The minimum size IP PDU for UDP on ATM is 44 octets, the minimum size of 44 is recommended to allow direct comparison to token ring performance. The IP PDU size of 4472 is recommended instead of the
theoretical FDDI maximum size of 4500 octets in order to permit the same type of comparison. An IP (i.e., not UDP) IP PDU may be used in addition if a higher data rate is desired, in which case the minimum IP PDU size is 28 octets.2.9. Verifying received IP PDUs
The test equipment SHOULD discard any IP PDUs received during a test run that are not actual forwarded test IP PDUs. For example, keep- alive and routing update IP PDUs SHOULD NOT be included in the count of received IP PDUs. In any case, the test equipment SHOULD verify the length of the received IP PDUs and check that they match the expected length. Preferably, the test equipment SHOULD include sequence numbers in the transmitted IP PDUs and check for these numbers on the received IP PDUs. If this is done, the reported results SHOULD include, in addition to the number of IP PDUs dropped, the number of IP PDUs that were received out of order, the number of duplicate IP PDUs received and the number of gaps in the received IP PDU numbering sequence. This functionality is required for some of the described tests.2.10. Modifiers
It is useful to characterize the SUTs performance under a number of conditions. Some of these conditions are noted below. The reported results SHOULD include as many of these conditions as the test equipment is able to generate. The suite of tests SHOULD be run first without any modifying conditions, then repeated under each of the modifying conditions separately. To preserve the ability to compare the results of these tests, any IP PDUs that are required to generate the modifying conditions (excluding management queries) will be included in the same data stream as that of the normal test IP PDUs and in place of one of the test IP PDUs. They MUST not be supplied to the SUT on a separate network port.2.10.1. Management IP PDUs
Most ATM data networks now make use of ILMI, signaling and OAM. In many environments, there can be a number of management stations sending queries to the same SUT at the same time. Management queries MUST be made in accordance with the applicable specification, e.g., ILMI sysUpTime getNext requests will be made in accordance with ILMI 4.0. The response to the query MUST be verified by the test equipment. Note that, for each management protocol in
use, this requires that the test equipment implement the associated protocol state machine. One example of the specific query IP PDU (ICMP) that should be used is shown in Appendix C.2.10.2. Routing update IP PDUs
The processing of PNNI updates could have a significant impact on the ability of a switch to forward cells and complete calls. If PNNI is configured on the SUT, one routing update MUST be transmitted before the first test IP PDU is transmitted during the trial. The test SHOULD verify that the SUT has properly processed the routing update. PNNI routing update IP PDUs SHOULD be sent at the rate specified in Appendix B. Appendix C defines one routing update PDU for the TCP/IP over ATM example. The routing updates are designed to change the routing on a number of networks that are not involved in the forwarding of the test data. The first IP PDU sets the routing table state to "A", the second one changes the state to "B". The IP PDUs MUST be alternated during the trial. The test SHOULD verify that the SUT has properly processed the routing update.2.11. Filters
Filters are added to switches to selectively inhibit the forwarding of cells that would normally be forwarded. This is usually done to implement security controls on the data that is accepted between one area and another. Different products have different capabilities to implement filters. Filters are applicable only if the SUT supports the filtering feature. The SUT SHOULD be first configured to add one filter condition and the tests performed. This filter SHOULD permit the forwarding of the test data stream. This filter SHOULD be of the form as described in the SUT Users Guide. The SUT SHOULD be then reconfigured to implement a total of 25 filters. The first 24 of these filters SHOULD be based on 24 separate ATM NSAP Network Prefix addresses. The 24 ATM NSAP Network Prefix addresses SHOULD not be any that are represented in the test data stream. The last filter SHOULD permit the forwarding of the test data stream. By "first" and "last" we mean to ensure that in the second case, 25 conditions must be checked before the data IP over ATM PDUs will match the conditions that permit the forwarding of the IP PDU. Of course, if the SUT reorders the filters or does not use a linear scan of the filter rules the effect of the sequence in which the filters are input is properly lost.
The exact filters configuration command lines used SHOULD be included with the report of the results.2.11.1. Filter Addresses
Two sets of filter addresses are required, one for the single filter case and one for the 25 filter case. The single filter case should permit traffic from ATM address [Switch Network Prefix] 00 00 00 00 00 01 00 to ATM address [Switch Network Prefix] 00 00 00 00 00 02 00 and deny all other traffic. Note that the 13 octet Switch Network Prefix MUST be configured before this test can be run. The 25 filter case should follow the following sequence. deny [Switch Network Prefix] 00 00 00 00 00 01 00 to [Switch Network Prefix] 00 00 00 00 00 03 00 deny [Switch Network Prefix] 00 00 00 00 00 01 00 to [Switch Network Prefix] 00 00 00 00 00 04 00 deny [Switch Network Prefix] 00 00 00 00 00 01 00 to [Switch Network Prefix] 00 00 00 00 00 05 00 ... deny [Switch Network Prefix] 00 00 00 00 00 01 00 to [Switch Network Prefix] 00 00 00 00 00 0C 00 deny [Switch Network Prefix] 00 00 00 00 00 01 00 to [Switch Network Prefix] 00 00 00 00 00 0D 00 allow [Switch Network Prefix] 00 00 00 00 00 01 00 to [Switch Network Prefix] 00 00 00 00 00 02 00 deny [Switch Network Prefix] 00 00 00 00 00 01 00 to [Switch Network Prefix] 00 00 00 00 00 0E 00 deny [Switch Network Prefix] 00 00 00 00 00 01 00 to [Switch Network Prefix] 00 00 00 00 00 0F 00 ... deny [Switch Network Prefix] 00 00 00 00 00 01 00 to [Switch Network Prefix] 00 00 00 00 00 18 00 deny all else All previous filter conditions should be cleared from the switch before this sequence is entered. The sequence is selected to test to see if the switch sorts the filter conditions or accepts them in the order that they were entered. Both of these procedures will result in a greater impact on performance than will some form of hash coding.
2.12. Protocol addresses
It is easier to implement these tests using a single logical stream of data, with one source ATM address and one destination ATM address, and for some conditions like the filters described above, a practical requirement. Networks in the real world are not limited to single streams of data. The test suite SHOULD be first run with a single ATM source and destination address pair. The tests SHOULD then be repeated with using a random destination address. In the case of testing single switches, the addresses SHOULD be random and uniformly distributed over a range of 256 seven octet user parts. In the case of testing multiple interconnected switches, the addresses SHOULD be random and uniformly distributed over the 256 network prefixes, each of which should support 256 seven octet user parts. The specific address ranges to use for ATM are shown in Appendix A. IP to ATM address mapping MUST be accomplished as described in RFC 2225.2.13. Route Set Up
It is not reasonable that all of the routing information necessary to forward the test stream, especially in the multiple address case, will be manually set up. If PNNI and/or ILMI are running, at the start of each trial a routing update MUST be sent to the SUT. This routing update MUST include all of the ATM addresses that will be required for the trial. This routing update will have to be repeated at the interval required by PNNI or ILMI. An example of the format and repetition interval of the update IP PDUs is given in Appendix B (interval and size) and Appendix C (format).2.14. Bidirectional traffic
Bidirectional performance tests SHOULD be run with the same data rate being offered from each direction. The sum of the data rates should not exceed the theoretical limit for the media.2.15. Single stream path
The full suite of tests SHOULD be run with the appropriate modifiers for a single receive and transmit port on the SUT. If the internal design of the SUT has multiple distinct pathways, for example, multiple interface cards each with multiple network ports, then all possible permutations of pathways SHOULD be tested separately. If multiple interconnected switches are tested, the test MUST specify routes, which allow only one path between source and destination ATM addresses.
2.16. Multi-port
Many switch products provide several network ports on the same interface module. Each port on an interface module SHOULD be stimulated in an identical manner. Specifically, half of the ports on each module SHOULD be receive ports and half SHOULD be transmit ports. For example if a SUT has two interface module each of which has four ports, two ports on each interface module be receive ports and two will be transmit ports. Each receive port MUST be offered the same data rate. The addresses in the input data streams SHOULD be set so that an IP PDU will be directed to each of the transmit ports in sequence. That is, all transmit ports will receive an identical distribution of IP PDUs from a particular receive port. Consider the following 6 port SUT: -------------- ---------| Rx A Tx X|-------- ---------| Rx B Tx Y|-------- ---------| Rx C Tx Z|-------- -------------- The addressing of the data streams for each of the inputs SHOULD be: stream sent to Rx A: IP PDU to Tx X, IP PDU to Tx Y, IP PDU to Tx Z stream sent to Rx B: IP PDU to Tx X, IP PDU to Tx Y, IP PDU to Tx Z stream sent to Rx C IP PDU to Tx X, IP PDU to Tx Y, IP PDU to Tx Z Note: Each stream contains the same sequence of IP destination addresses; therefore, each transmit port will receive 3 IP PDUs simultaneously. This procedure ensures that the SUT will have to process multiple IP PDUs addressed to the same transmit port simultaneously. The same configuration MAY be used to perform a bi-directional multi-stream test. In this case all of the ports are considered both receive and transmit ports. Each data stream MUST consist of IP PDUs whose addresses correspond to the ATM addresses all of the other ports.
2.17. Multiple protocols
This document does not address the issue of testing the effects of a mixed protocol environment other than to suggest that if such tests are wanted then PDUs SHOULD be distributed between all of the test protocols. The distribution MAY approximate the conditions on the network in which the SUT would be used.2.18. Multiple IP PDU sizes
This document does not address the issue of testing the effects of a mixed IP PDU size environment other than to suggest that, if such tests are required, then IP PDU size SHOULD be evenly distributed among all of the PDU sizes listed in this document. The distribution MAY approximate the conditions on the network in which the SUT would be used.2.19. Testing beyond a single SUT
In the performance testing of a single SUT, the paradigm can be described as applying some input to a SUT and monitoring the output. The results of which can be used to form a basis of characterization of that device under those test conditions. This model is useful when the test input and output are homogeneous (e.g., 64-byte IP, AAL5 PDUs into the SUT; 64 byte IP, AAL5 PDUs out). By extending the single SUT test model, reasonable benchmarks regarding multiple SUTs or heterogeneous environments may be collected. In this extension, the single SUT is replaced by a system of interconnected network SUTs. This test methodology would support the benchmarking of a variety of device/media/service/protocol combinations. For example, a configuration for a LAN-to-WAN-to-LAN test might be: (1) ATM UNI -> SUT 1 -> BISUP -> SUT 2 -> ATM UNI Or an extended LAN configuration might be: (2) ATM UNI -> SUT 1 -> PNNI Network -> SUT 2 -> ATM UNI In both examples 1 and 2, end-to-end benchmarks of each system could be empirically ascertained. Other behavior may be characterized through the use of intermediate devices. In example 2, the configuration may be used to give an indication of the effect of PNNI routing on IP throughput.
Because multiple SUTs are treated as a single system, there are limitations to this methodology. For instance, this methodology may yield an aggregate benchmark for a tested system. That benchmark alone, however, may not necessarily reflect asymmetries in behavior between the SUTs, latencies introduced by other apparatus (e.g., CSUs/DSUs, switches), etc. Further, care must be used when comparing benchmarks of different systems by ensuring that the SUTs' features and configuration of the tested systems have the appropriate common denominators to allow comparison.2.20. Maximum IP PDU rate
The maximum IP PDU rates that should be used when testing LAN connections SHOULD be the listed theoretical maximum rate for the IP PDU size on the media. The maximum IP PDU rate that should be used when testing WAN connections SHOULD be greater than the listed theoretical maximum rate for the IP PDU size on that speed connection. The higher rate for WAN tests is to compensate for the fact that some vendors employ various forms of header compression. A list of maximum IP PDU rates for LAN connections is included in Appendix B.2.21. Bursty traffic
It is convenient to measure the SUT performance under steady state load; however, this is an unrealistic way to gauge the functioning of a SUT. Actual network traffic normally consists of bursts of IP PDUs. Some of the tests described below SHOULD be performed with both constant bit rate, bursty Unspecified Bit Rate (UBR) Best Effort [AF-TM4.1] and Variable Bit Rate Non-real Time (VBR-nrt) Best Effort [AF-TM4.1]. The IP PDUs within a burst are transmitted with the minimum legitimate inter-IP PDU gap. The objective of the test is to determine the minimum interval between bursts that the SUT can process with no IP PDU loss. Tests SHOULD be run with burst sizes of 10% of Maximum Burst Size (MBS), 20% of MBS, 50% of MBS and 100% MBS. Note that the number of IP PDUs in each burst will depend on the PDU size. For UBR, the MBS refers to the associated VBR traffic parameters.
2.22. Trial description
A particular test consists of multiple trials. Each trial returns one piece of information, for example the loss rate at a particular input IP PDU rate. Each trial consists of five of phases: a) If the SUT is a switch supporting PNNI, send the routing update to the SUT receive port and wait two seconds to be sure that the routing has settled. b) Send an ATM ARP PDU to determine the ATM address corresponding to the destination IP address. The formats of the ATM ARP PDU that should be used are shown in the Test Frame Formats document and MUST be in accordance with RFC 2225. c) Stimulate SUT with traffic load. d) Wait for two seconds for any residual IP PDUs to be received. e) Wait for at least five seconds for the SUT to restabilize.2.23. Trial duration
The objective of the tests defined in this document is to accurately characterize the behavior of a particular piece of network equipment under varying traffic loads. The choice of test duration must be a compromise between this objective and keeping the duration of the benchmarking test suite within reasonable bounds. The SUT SHOULD be stimulated for at least 60 seconds. If this time period results in a high variance in the test results, the SUT SHOULD be stimulated for at least 300 seconds.2.24. Address resolution
The SUT MUST be able to respond to address resolution requests sent by another SUT, an ATM ARP server or the test equipment in accordance with RFC 2225.2.25. Synchronized Payload Bit Pattern.
Some measurements assume that both the transmitter and receiver payload information is synchronized. Synchronization MUST be achieved by supplying a known bit pattern to both the transmitter and receiver. This bit pattern MUST be one of the following: PRBS-15, PRBS-23, 0xFF00, or 0xAA55.
2.26. Burst Traffic Descriptors.
Some measurements require busty traffic patterns. These patterns MUST conform to one of the following traffic descriptors: 1) PCR=100% allotted line rate, SCR=50% allotted line rate, and MBS=8192 2) PCR=100% allotted line rate, SCR=50% allotted line rate, and MBS=4096 3) PCR=90% allotted line rate, SCR=50% allotted line rate, and MBS=8192 4) PCR=90% allotted line rate, SCR=50% allotted line rate, and MBS=4096 5) PCR=90% allotted line rate, SCR=45% allotted line rate, and MBS=8192 6) PCR=90% allotted line rate, SCR=45% allotted line rate, and MBS=4096 7) PCR=80% allotted line rate, SCR=40% allotted line rate, and MBS=65536 8) PCR=80% allotted line rate, SCR=40% allotted line rate, and MBS=32768 The allotted line rate refers to the total available line rate divided by the number of VCCs in use.
(page 17 continued on part 2)