* App Performance View by Peter Sevcik and Rebecca Wetzel
If you haven’t seen Network World’s App Performance View blog by Peter Sevcik and Rebecca Wetzel, you should check it out, if only for their recent sobering look at the claims of application-acceleration vendors.
Entitled “Beware of Bogus Math – Ignore 10X, 100X WAN Acceleration Claims!” the post explains how some vendors will claim that their product improves response time by a factor of, say, 100 or 400, but that number can be misleading.
In a nutshell, if a vendor’s equipment reduces response time from 100 seconds to 10 seconds, the vendor may claim that the improvement in response time is 10X, since 10 is one-tenth of 100. If the vendor can get the response time from 100 seconds to 5 seconds, the vendor would claim the improvement is 20X – even though the actual difference between a 10X improvement and a 20X improvement is a pretty small percentage of the total.
A better way to look at it is in percentages – a change in response time from 100 seconds to 10 seconds would be a 90% improvement, while a reduction from 100 seconds to 5 seconds would be a 95% improvement.
Some of the best discussions that Sevcik and Wetzel have gotten started have come from their calls for feedback on different vendors’ equipment. For example, there is some great feedback from users of start-up Exinda’s x800, Blue Coat’s ProxySG, Juniper’s WX, Riverbed’s Steelhead, and particularly Cisco’s WAAS.
They’re even reaching out with a survey – and if you take it, you could win yourself an iPod Touch. Check it out today.
