The network industry has largely been focused on network transformation over the past few years. Most of the vendors, though, have been geared towards the evolution of the data center network. It’s time that businesses started looking at evolving the wide area network (WAN) as this is often where the biggest pain points is for application performance.
The WAN fundamentally hasn’t changed at all in the past 30 years, as most companies still use the traditional “hub and spoke” design with a private network technology, such as MPLS. Often the WAN has a backup connection that becomes active when the primary fails. This model has worked well for decades now, so living by the “if it ain’t broke, don’t fix it” credo has meant that most companies just leave well enough alone and haven’t done anything to evolve the WAN.
I think it’s fair to say that most network managers understand why this architecture is inefficient. It was really designed for client/server traffic, and all Internet traffic is backhauled through a central location. Also, much of the traffic “trombones” up and down the WAN links through a central hub, moving from branch to branch or even Internet to branch. This is one of the reasons we’ve been talking about WAN re-design for years now. In my opinion, though, I think it’s time to take this seriously.
So, after all these years, why evolve the WAN now? It’s because mobile and cloud computing had significantly changed WAN traffic patterns. Prior to the cloud era, most business applications were located in the data center and Internet traffic was but a small amount of WAN traffic. Today, more and more companies are shifting applications to the cloud, making the Internet account for the majority of traffic while also increasing the importance of the WAN. Applications moving to the cloud means the Internet edge should shift from the data center to the branch. Also, the rise of collaborative applications generates more peer-to-peer traffic, requiring more branch-to-branch traffic. Lastly, the active-passive design that is widely used today must be replaced with an active-active architecture for higher uptime and better performance.
One possible solution to WAN challenges is for companies to use Cisco’s IWAN technology. For those that don’t know IWAN, it’s a feature available on Cisco’s ISR-AX branch router that allows companies to replace their private WAN links with Internet connections. The technology effectively bonds the two connections together, creating one bigger pipe and determining which connection to send the traffic down based on price or performance policies. IWAN also applies security at the branch edge so customers can be confident that the network is as or more secure than a private WAN link. I believe IWAN can give Cisco customers better performance as well as enabling them to load balance traffic across the two connections at a much lower price point than traditional networks.
However, while IWAN solves many WAN challenges, it could cause some new problems. The automation capabilities of IWAN can actually mask network problems, making them invisible to network managers. For example, one of the Internet connections could fail, but this could go unnoticed since all of the traffic would continue to go down the other pipe. Some users may experience some slowness when it comes to the application, but the WAN would still operate.
LiveAction, one of Cisco’s developer partners, now has the capability to help troubleshoot IWAN deployments. LiveAction visualization and granular QoS control improves visibility into the WAN and helps identify problems even if none are being reported by a user. The tool also has the capability, through a GUI interface, to resolve the problem without requiring the services of a high-level Cisco engineer. In addition to QoS monitoring, LiveAction provides visualization and reporting of Cisco’s PfR and AVC flows to uncover any other issues that IWAN may have masked.
For Cisco ISR-AX customers, IWAN is a great leap forward in WAN evolution. Just don’t forget about the management tools that can be used to find what IWAN hides.