* Results of Tolly Group’s QoS tests of Cisco, Extreme gear
Kevin Tolly recently let me know about another LAN-related test his firm performed. This one, commissioned by Avaya, pits Extreme Networks switches against Cisco switches.
The approach, Tolly says, was unconventional. Countless other tests of various equipment have measured the effectiveness of QoS controls on traffic that experiences congestion upon leaving the backplane and going out to the edge. That is, too much traffic is coming from the backplane and trying to squeeze onto one output port.
In this case, however, the Tolly Group measured the effectiveness of dealing with traffic when there is congestion going to the backplane. The example Tolly gave was that 40G bit/sec of traffic could be going to a backplane that has only 20G bit/sec of capacity. Tolly believes this is a first for QoS testing.
Tolly tested Extreme’s BlackDiamond 8810 and 10808 and Cisco’s Catalyst 4507 and 6509. The aim was to determine how these switches might do under a heavy load of voice, video and data. How would each device fare?
Tolly says:
“This test shows that Extreme’s QoS solution works effectively in either ‘egress’ or ‘ingress’ congestion situations. It also shows that while Cisco can perform ‘egress’ congestion control, its QoS simply does not work in ‘ingress’ congestion scenarios. Thus, in a converged ‘triple play’ enterprise, high-priority Cisco’s core/edge solutions cannot provide voice and video streams the ‘protection’ they need from being overwhelmed by low-priority traffic.”
When the background traffic load was turned up, the Extreme switches still managed to get 100% of voice calls through, while the Cisco switches “were unable to complete 100% of the calls, dropped high-priority data traffic and failed to deliver all of the IP video streams,” according to the report.
For the full report, go to: https://www.tolly.com/DocDetail.aspx?DocNumber=205121
