Benchmarking Methodology (bmwg) Charter Chair(s): Scott Bradner, sob@harvard.edu Mailing Lists: General Discussion: bmwg@harvisr.harvard.edu To Subscribe: bmwg-request@harvisr.harvard.edu Archive: Description of Working Group: The major goal of the Benchmark Methodology Working Group is to make a series of recommendations concerning the measurement of the performance characteristics of different classes of network equipment and software services. Each recommendation will describe the class of equipment or service, discuss the performance characteristics that are pertinent to that class, specify a suite of performance benchmarks that test the described characteristics, as well as specify the requirements for common reporting of benchmark results. Classes of network equipment can be broken down into two broad categories. The first deals with stand-alone network devices such as routers, bridges, repeaters, and LAN wiring concentrators. The second category includes host dependent equipment and services, such as network interfaces or TCP/IP implementations. Once benchmarking methodologies for stand-alone devices have matured sufficiently, the group plans to focus on methodologies for testing system-wide performance, including issues such as the responsiveness of routing algorithms to topology changes. Goals and Milestones: Done Issue a document that provides a common set of definitions for performance criteria, such as latency and throughput. Done The document will also define various classes of stand-alone network devices such as repeaters, bridges, routers, and LAN wiring concentrators as well as detail the relative importance of various performance criteria within each class. 1 TBD Once the community has had time to comment on the definitions of devices and performance criteria, a second document will be issued. This document will make specific recommendations regarding the suite of benchmark performance tests for each of the defined classes of network devices. 2