RFC 2544
acronymnetworking
The 1999 benchmarking methodology that defined how a network device's throughput, latency, frame loss and burst tolerance are to be measured.
Its introduction is unusually blunt for a standards document: vendors engage in specsmanship, it says, involving smoke and mirrors to confuse the people buying their products. The whole methodology exists to make performance claims comparable, by fixing the frame sizes, the trial durations and the reporting so that two vendors' numbers mean the same thing. It also has a sequel worth knowing about: the IETF later published an applicability statement titled Use on Production Networks Considered Harmful, because a test that deliberately drives a device to its loss threshold does exactly that to whatever else is running.
Also known as: benchmarking methodology, bradner mcquaid
Sources
- RFC 2544, Benchmarking Methodology for Network Interconnect Devices, Bradner and McQuaid, March 1999, whose introduction states that vendors often engage in specsmanship involving smoke and mirrors to confuse potential users, and which defines throughput, latency, frame loss and back-to-back frame tests
- RFC 2544 obsoletes RFC 1944 and is updated by RFC 6815, Applicability Statement for RFC 2544: Use on Production Networks Considered Harmful