How we did it

We tested the Cisco Catalyst 3550-12T running Cisco IOS, Version 12.1(4)EA1 software and the Cisco Catalyst 2950T-24, running Cisco IOS, Version 12.0(5.3)WC(1) software.

Testing was conducted using the Ixia 1600 (www.ixiacom.com) test tool to generate and monitor traffic through the switches. We used several of the scripts included in Ixia's ScriptMate software to measure throughput, latency and frame loss. All tests were performed repeatedly at Layer 2, and the results were averaged for the final score. On the 3550-12T, the tests were also performed at Layer 3 (IP routing) with static routes configured on the switch. The basic throughput, frame loss and latency tests were run using bidirectional traffic in port pairs at wire speed across all ports, and across all 24 Fast Ethernet ports on the Catalyst 2950T-24. Additional throughput tests were also run in a "many-to-one" configuration, with 10 ports on the 2950T-24 (each port delivering traffic at 100M bit/sec), which was then uplinked to a Gigabit port on the Catalyst 3550-12T.

We also ran a Gigabit Ethernet-over-copper cabling distance test to confirm that the Cisco switches handle the IEEE-standard 100-meter distance over Category 5 cabling. They do: We verified that, over 100-meter runs of Cat-5 cabling between Gigabit unshielded twisted pair ports on the 2960T-24 and 3550-12T, full-duplex traffic passed at maximum rates without errors.

Back to the main review