Jun 21, 2013
The tech industry has gone through more change in the last decade than any other period of corporate IT. Virtualization has redefined the data center, the network access edge has become predominantly wireless, and consumers are now making decisions on what devices come into the business. However, one of the areas that hasn’t changed much at all is the corporate wide area network (WAN). Sure, we’ve talked about split tunnels and direct Internet access, but most companies still use the tried-and-true “hub and spoke” model where all WAN traffic is backhauled through the data center and out one ingress point.
Managing WAN performance is something network managers have struggled with, since the speed of the WAN is typically an order of magnitude lower than that of Local Area Networks (LANs). Today’s emerging computing models of mobile and cloud are network-centric, meaning that the old “best effort” network model doesn’t cut it any more. If companies are going to fully leverage these new computing models, the WAN must now evolve. If it doesn’t, businesses will not fully be able to utilize the network as a viable application delivery platform.
The main forces driving the need for WAN evolution today are:
Evolving the WAN is more than just upgrading speeds or buying new hardware. Companies need to shed the legacy thinking and move away from the hub and spoke model. The “trombone” effect of this older architecture is about as inefficient a way to use the network as there is today. Companies should embrace a meshed model where Internet access is direct from each location, creating an efficient way to access cloud and mobile applications. If this is too big a leap then at least move to some sort of hybrid mode where regional hubs are maintained, instead of one centralized location.
The other part of WAN evolution involves greater use of WAN optimization. Talk to any network manager than has deployed it and they’ll tell you that they’ll never run a non-optimized WAN again. One network manager referred to it as “network crack”, where once users get a taste of the optimized WAN, they’ll want more and more.
To date, I think WAN optimization has been treated as a tactical technology. Have a problem with Exchange, Windows, or storage? Deploy WAN optimization and the problem goes away. There’s really nothing wrong with this. However, I would rather see companies be proactive with it and consider WAN optimization to be a strategic technology, one that creates competitive differentiation.
The rate of change in IT certainly isn’t going to slow down any time soon. Companies that don’t address WAN evolution though will limit the value they get from many of the IT initiatives that are currently under way.
Image credit: Wikimedia Commons