Traditional networks have typically relied on multiprotocol label switching (MPLS) circuits to connect sites to data centers – it’s a classic routing technique in private telecommunications networks that directs data from one node to the next based on labels. However, the rapid adoption of cloud-based applications creates an MPLS bottleneck as network traffic is increasingly distributed and Internet-bound, rather than destined for the corporate data center.

To mitigate this problem, IT organizations are now moving towards a hybrid WAN architecture, utilizing both MPLS and Internet-based connections, or sometimes Internet-only connections. Using this architecture, sites and users are directly connected to the Internet instead of being backhauled through data centers, whatever their location is.

Most SD-WAN platforms have analytics capability to monitor some degree of application and network performance … but only from network edge to network edge. However, capabilities differ from one vendor to another, leading to separate, fragmented view of the network performance. The demand is increasing for getting visibility all the way from the end-user to the actual application or service. Here are three key reasons:

Your SD-WAN Does Not Control the Internet

Network professionals attempting to deliver a global network using the Internet must confront the problem of unpredictable performance. Unlike MPLS links, the Internet is a diverse and unreliable transport mechanism. It comprises a vast, complex web of service providers with continuously changing connections. As the distance grows, latency rises, and an extended latency delay is something that cannot be tolerated in enterprise-grade communication. The lack of visibility beyond the edge of the enterprise network can leave network operations entirely out of control. 

Your SD-WAN Network Paths Can Change Unpredictably

Additionally, SD-WAN traffic policies and path remediation can introduce dynamic routing changes that may have a positive impact but can also deteriorate the overall end-to-end performance. As a result, applications can run slower than before the corrective action, making troubleshooting very difficult without additional insight. This variance in performance creates a situation where services can work optimally sometimes but perform poorly at other times. This inconsistency is disruptive for end-users, diminishes productivity, impacts the overall digital experience, and ultimately damages network operations’ reputation – even if they don’t own the impacted infrastructure.

Your SD-WAN Does Not Measure the End-User Experience

Today, end-users must be able to access whichever applications and data they need to complete their task on any device from any branch office or network segment. As employees use on-prem services, unified communication, SaaS applications, or access data in the cloud, the digital experience should stay consistent. SD-WAN solutions are specifically geared to address these challenges. However, too many fail to deliver that universally consistent, omnichannel experience. This is because they only focus on edge-to-edge and device performance – not the entire end-to-end user experience.

Enterprise networks are shifting. The pandemic and resulting societal disruptions have changed the way people work. The accelerated cloud transformation has distributed resources everywhere outside the data center. The Internet has become the new enterprise network, and network operations have responsibility for services over which they have less control than ever. In response, IT organizations need to continuously validate the level of performance they delivery by correlating it to the actual user experience. Forward-thinking strategies are turning to continuous, end-to-end SD-WAN performance monitoring that ensures site availability, application performance, and policy compliance related to the end-user experience.

As native SD-WAN monitoring capabilities fill some basic needs, they also leave major gaps in multi-vendor and end-user experience visibility. Network professionals, in an effort to rethink their network monitoring strategies, can take an approach to Experience-Driven NetOps to ensure network connections are experience-proven and network operations teams are experience-driven. Modern network monitoring should employ end-user experience metrics to determine if the network is in a good state or not—for any user, on any device, on any network, anywhere. If you can prove the user experience is reliable, then you can more effectively ensure resilient network delivery.

Discover more on how to bring controller and overlay perspectives together with the underlay performance and the user experience in this white paper, “Continuous, End-to-End Validation of SD-WAN Performance from the End User Perspective”.