Rethinking the Network with Today’s End-user in Mind


By Lou Najdzin, Vertical Dev. Director, Global Enterprise, Equinix


This is Part 1 in Lou Najdzin’s three-part blog series “Rethinking the network”

More, more, more: more apps, more data, more devices. Mobile, mobile, mobile: mobile workers, mobile apps, mobile customers.

The one constant is change. If you’re like many enterprise CIOs today, you are challenged to ensure that end-user performance expectations are met or exceeded regardless of their location. While end-users measure satisfaction through interacting with the applications and data, the CIO must deal with the need for continuous, high-speed connectivity and the associated logistical, technological, and cost challenges. And it’s only going to get tougher. A recent Cisco study offers some thought provoking statistics:

• From 2011 – 2016, global IP traffic will more than triple
• Over the same period, global mobile traffic increase 18x
• Approximately 15 billion devices will be added to the global network over the next 3 – 5 years

So why are you deploying and purchasing network services like it’s business as usual and a static user population? Consider some of the limitations of the typical legacy architecture:

• Long, expensive routes
• Choke points
• Low reliability
• Unpredictable service
• Variable quality of experience

What are your options? Sure, you could install WAN optimization appliances. They will improve network performance, but they don’t address underlying network architecture and implementation issues. Or you might consider remote display services (RDS) or virtual desktop infrastructure (VDI). These can reduce traffic across the WAN, but they really don’t address the underlying network design limitations. You might even outsource to an Application Delivery Network (ADN). While an ADN will probably increase performance, you may still miss application availability and operational visibility targets. In addition, this solution can be expensive for heavy users.

When all else fails, you can throw bandwidth at the problem. While network evolution has made bandwidth more affordable and available, capacity alone may not solve performance issues. A recent Google study entitled “More Bandwidth Doesn’t Matter (much)” compares the impact of bandwidth and latency on web page load times. The bottom line? After about 5 Mbps, throwing more bandwidth at the problem adds little, BUT decreasing latency shows linear improvement on page load times.

So while the above network performance techniques all can help improve the end-user experience, they do not fully address the fundamental challenges of delivering a high-quality user experience anytime, anywhere on any device. Enterprises need a new connectivity strategy for optimizing network performance. Enter the Network Performance Hub (NPH).

An NPH is built by moving common, off-the-shelf enterprise network equipment from the corporate location (offices and/or data centers) to carrier-neutral, shared data centers where the Network Service Providers (NSPs) come to peer. Why move your network equipment from the corporate data center to a carrier-neutral facility? Consider what these carrier-neutral data centers offer:

• Access to undersea cable landing station termination points
• Access to major metros globally with significant carrier choice
• Access to local loop and long haul provider backbones
• Internet and Ethernet exchanges
• Availability of cloud service providers and Content Delivery Networks

With NPHs in key locations you can transform that maze of long, expensive, bandwidth-limited circuits into an optimized regional WAN topology. What would your business gain?

• A simpler network
• Better performance
• Lower TCO because of supplier choice
• Increased capacity
• Reduced time to market for new services
• Direct access to cloud service providers

Perhaps you are thinking “We build our own data centers and outsource our network requirements to a carrier. Why would we work with an external data center provider?” The answer is simple: the game has changed. If you need to transform your network to keep up with anytime, anywhere, any device world you need to consider the carrier-neutral data center. This does not mean razing the corporate data center and moving all applications and data.

But carrier-neutral data centers can be used in partnership with existing enterprise data centers to develop a connectivity plan that ensures satisfied end-users in an anytime, anywhere, any device world. And because this type of connectivity plan will not only improve performance, but will more than likely reduce total cost of ownership you might even “have your cake and eat it too”.

Be sure to checkout our whitepaper “Network Optimization: The Value of Network Performance Hubs” for a more detailed look into the world of NPH.