Silver Peak named as a Leader in Gartner Magic Quadrant for WAN Optimization Controllers for 2012

Gartner has published their 2010 Magic Quadrant for WAN Optimization Controllers, authored by Joe Skorupa, Andy Rolfe, and Severine Real. Silver Peak has been positioned as a “Leader” in the January 2012 Magic Quadrant report.

Silver Peak’s position in the Magic Quadrant closes out a banner year that included being named to the Wall Street Journal’s list of top 50 venture backed companies, broad international expansion, and continued data center partnerships with Avaya, Dell, EMC, Hitachi Data Systems (HDS), HP, NetApp, and Xiotech. Leading customers from around the globe have deployed Silver Peaks award-winning NX and VX appliances, and the recently released VRX data center class virtual appliances, which scale from Mbps to Gbps of WAN capacity. The products are designed to optimize all IP traffic regardless of transport protocol and application version.

According to the report, “WAN optimization is about improving the performance of business applications over WAN connections. Most networks carry a variety of traffic types of differing characteristics and importance. Many organizations are striving to manage this traffic to optimize the response times of critical applications and reduce costs, given that bandwidth continues to represent a significant proportion of operating expenditure for wide-area data networks. But the cost of bandwidth isn’t the only consideration—matching the allocation of WAN resources to business needs is also important. And as resources are increasingly centralized, minimizing the effect of latency on application response times is becoming a critical requirement. In addition, virtualization and new application environments such as cloud computing and Web services can put an unexpected strain on the network.”

Bandwidth vs Optimisation

When assessing the performance impact of increasing WAN Bandwidth against implementing WAN Optimisation, you really do need to consider ALL of the variables.

Speed of light is a major factor. When calculating the effective throughput of a WAN connection, you need to consider the time it takes for a typical IP packet to get from point A to point B. The theoretical maximum throughput of a WAN connection is determined by dividing Windows Size of a packet by the Round Trip Time (RTT) of that connection (Throughput = Windows Size / RTT).

Let’s look at a typical MPLS WAN Connection between the UK and Japan. The average RTT during May 2011 between Verizon’s hubs in the UK and Japan was 272.80ms. If we take a standard TCP packet of 65,535bytes and divide this by an RTT of 272.80ms, we find that the maximum theoretical throughput of the connection, regardless of the amount of bandwidth we have, is just 240Kbps.

We can increase packet window sizes to improve this, but done manually this would be pretty unfeasible – unless we are talking small-scale single application stream.

WAN Optimisation solutions can deliver significant performance increases on low latency WAN connections, but when you’re contending with high latency connections they can be the difference between having to deploy a remote Data Centre and being able to centralise your key IT service infrastructure.

Packet loss and Out of Order packets also have a significant impact on the effective throughput of a WAN Connection. I’ve burnt enough of your time on the topic of latency here, so I’ll leave that for another day.

TCP Acceleration, latency mitigation, loss correction, packet order correction and, last but by no means least, data de-duplication are standard features of the leading WAN Optimisation solutions of today. All of these features considered alongside the key variables of WAN connection performance add up to a very compelling case for WAN Optimisation.

Response Data Communications Ltd has been delivering WAN Optimisation solutions for over 8 years now. From the results we have seen, both through Proof of Concepts and Live Deployments, WAN Optimisation should be an integral part of today’s enterprise infrastructure strategy.