WAN Optimization

WAN Optimization: Virtual vs. Physical Appliances

by |Published | 3 comments

Can virtual WAN optimization appliances be as effective as physical appliances? Well, in this video, we bust the myth that virtual WANop appliances cannot perform as effectively as physical appliances. This video compares Silver Peak’s VX virtual WAN optimization appliance to a competitive physical WAN optimization appliance in a simulated disaster recovery scenario.

You may also like

Overload-WANop-SilverPeak
Published

Danger! Overload!

“I have a theory about the human mind. A brain is a lot like a computer. It will only take so many […]

  • Zee

    Nice post!
    zee
    http://walisystems.com

  • Evan

    What does it mean that the data was “pre-compressed”?  And why would you test DR  recovery across a WAN link with just 5ms of latency?  Wouldn’t that be next door?

  • Anonymous

    Thanks Evan.
     
    You’re correct that 5ms of latency is “next door.” The reason that the video shows 5ms of latency is that it is a lab simulation and we didn’t configure the WAN emulator for higher latency. The test would be equally valid over much higher latencies that you would typically see in a WAN – 50ms, 100ms or even 300ms.

    We used pre-compressed data in the test as a way to simulate a backup/restore operation. Often, in the real world, when data is initially backed up, the WAN optimizer sees the data for the first time and it looks like “first pass” data. The restore operation, of course, looks like second pass. The first pass does not benefit from deduplication but the second pass does. A simple way to simulate this in the lab is to use pre-compressed data so that there is no repetitive (deduplicable) data in the first pass.

    While other vendors, such as Riverbed, cannot provide any optimization benefit in environments with less than 20ms latency, Silver Peak can.