Two-Gig Fibre Channel: Party of One?
QLogic's 16-port switch, the SANbox2, was the only product submitted for the test, which was commissioned by Light Reading and conducted by independent benchmarking and network design consultancy Network Test Inc.
The test results are being announced today. See: Two Gigabits, One Vendor.
Four other vendors that boast of having 2-Gbit/s products were asked to participate in the test. All of them were no-shows. Vixel Corp. entered the test but withdrew early on. Gadzoox Networks Inc. and McData Corp. declined Light Reading's invitation, citing lack of resources. Brocade Communications Systems Inc. said it was "too busy" to take part and declined to help after its OEM partners failed to produce products for the test (see Greg Reyes, Chairman and CEO, Brocade).
According to David Newman, president of Network Test, QLogic not only showed up for the test, it passed it, becoming the first vendor to independently validate the performance of its 2-Gbit/s Fibre Channel technology.
"Fortunately for QLogic, its switches really do run at 2 Gbit/s," says Newman. "Until they’re able to show independent proof to the contrary, the number for all the other switch makers is zero."
Two-gig Fibre Channel has been touted as key to the future of high-speed data transfer and storage. While the market for 1-Gbit/s Fibre Channel switches has taken firm hold, thanks to burgeoning demand for storage area networks, the higher data rates and throughput of 2-Gbit/s are already being sought by users of applications calling for the fast transfer of enormous volumes of data, such as movie studios and Wall Street financial companies.
The test proved the call isn't coming too soon: "Enterprises now know that at least one vendor -- QLogic -- can deliver a high-quality 2-Gbit/s Fibre Channel fabric switch," says David Hill, research director at the Aberdeen Group Inc.
Besides validating the status of QLogic's 2-Gbit/s product, the test also highlighted key performance issues applicable to any switch. In the test lab, the SANbox2 was bombarded on all ports by frames of various sizes, using a SmartBits analyzer provided for the testbed by Spirent PLC (NYSE: SPM; London: SPT). Jitter, latency, forwarding rates, and throughput were recorded.
QLogic excelled in the baseline tests, particularly when it came to latency. "In general, QLogic put up excellent latency numbers," Newman says. "With large frames, the highest average latency we recorded was just 600 nanoseconds -- and even that worst-case number beats any latency number we’ve ever recorded in previous tests of other high-speed technologies."
The only downside, according to Newman, came when the switch was subjected to a battering of advanced, multiport "many-to-one and many-to-few" tests. While throughput held up to the onslaught, latency suffered under specific conditions.
The overall performance of QLogic's first product release, however, appears to have outshone any drawbacks. And its willingness to undergo public scrutiny gives QLogic a leg up on competitors that claim to have products but aren't ready to show them to prying eyes -- a clear indication, according to Newman, that the kinks probably haven't been worked out.
The test also is a valuable resource, analysts say, for those seeking real-world tirekicking. "Few enterprises have the resources to conduct product comparison testing," says Aberdeen's Hill. "Enterprises would rather rely on credible independent testing results than on marketing messages... Other 2-Gbit/s FC fabric switch vendors should get with the program as soon as possible."
Hill says he thinks the test methodology was sound. "I was very pleased with the professionalism and quality of the testing process," he says. "Enterprises should feel very confident in accepting the results of the tests that were conducted."
— Mary Jander, Senior Editor, Light Reading