SD-WAN (software-defined wide area network) is probably the most successful enterprise network trend that is continuously growing, IDC forecast it to reach $4.5 billion by 2022. As with many new technologies, SD-WAN deployment also comes with new challenges, especially for IT teams.
While there is a general consensus that SD-WAN has many advantages over an expensive MPLS network, IT teams are required to solve new operational challenges. Smoother integration with existing networking monitoring solutions becomes a concern where SD-WAN is not replacing MPLS but being added on top.
SD-WAN monitoring strategies can help to overcome some of these issues.
SD-WAN is best understood as WAN with network functionality controlled by software. You can overlay SD-WAN over different combinations like LTE, MPLS, and internet broadband. Network admins can centrally control and monitor SD-WAN functions, allowing traffic to flow seamlessly irrespective of the type of underlying connection. Intelligent direction of traffic allows users to connect to applications securely, efficiently and cheaply.
SD-WAN delivered over the cloud is a compelling and practical technology that can be deployed by any organization on a tight budget. It is also highly scalable and capable of delivering complex management.
SD-WAN implementations could negatively affect end-to-end network performance. For example, the router may route traffic across a link with lower speed, slowing down the connection. In the case of packet duplication, the overall bandwidth available to users is reduced. As a result, applications may perform slower than before causing users to complain. Troubleshooting these sorts of issues is very difficult without the right information.
Network professionals have to deal with a lack of comprehensive network visibility while trying to proactively identify performance issues. SD-WAN products, create more virtual components in an existing network making it complex to manage, troubleshoot and triage.
Common management challenges include:
Application performance monitoring implemented by SD-WAN routers is passive in nature. They are primarily programmed to:
Such monitoring fails to provide an impactful measure of the end-user experience, especially with respect to application performance. It can become a challenge to enforce correct SLA (service level agreement) in the absence of data pertaining to network latency, DNS resolution time, packet loss, etc.
Split tunnel implementation is the basic premise of an SD-WAN architecture. In a typical split-tunnel scenario, the remote site or a branch location has both- a direct broadband connection for reaching the internet and a private connection for the intranet. In such a setup the central network monitoring solution:
Even if the problem is related to the network, there can be numerous factors causing the network to slow down. These issues can be related to the Wi-Fi network, the client, the LAN, the WAN links, the configuration of the SD-WAN appliance, etc.
It is very important to not monitor a new technology just for the sake of it, but to add context to the monitored data. This context gives more clarity regarding application delivery over SD-WAN bundles. You can then take network improvement decisions with greater confidence while constantly improving end user experience. While there are multiple choices for metrics when it comes SD-WAN monitoring, the following are our top three recommendations.
It is essential to include an all-encompassing, end-to-end topology visualization pertaining to all network components. The network and IT operation teams are concerned with 2 main types of topology.
The traditional physical topology includes all the underlying connected switches, servers, SD-WAN edge devices, routers, etc. You can monitor physical topology by conducting periodic SNMP polling of the enterprise-wide infrastructure followed by data correlation.
Flow topology sits on top of the underlying physical topology. This adds more value to an SD-WAN set-up than the first type. The flow topology allows operation management teams to visualize the traffic path of the application. It is also capable of displaying a path change in real-time. You can use link aggregation for granular monitoring.
Most SD-WAN vendors offer in-built support for collecting and reporting network metrics like packet loss, jitter and latency. However, instead of working solely based on these metrics, you can combine application path analysis for better context.
If the operations monitoring team can view both the tunnels and the applications that use these tunnels, then they are in a better position to determine the performance issues. Additionally, the teams can also correlate these issues in the context of business-critical applications. In other words, including such capacity analysis and a predictive component can enhance the strength of the overall monitoring process. The operations team can preemptively respond to applications which may suffer from performance degradation.
Active path testing is effective when it can emulate real application traffic. If your SD-WAN vendor has built-in support for this feature, then you can leverage it and correlate the test results with the recommendations above. Active path testing as a proactive monitoring strategy can:
Any SD-WAN monitoring should take into account the whole end-to-end experience, from the user layer to the far end destination. Smart monitoring solutions use both passive and active network monitoring agents. End-to-End network tests are run continuously, and results are retrieved in real-time and stored for historical review.
CarrierBid Professional Services offers comprehensive telecom consulting services for large enterprise and small business customers. Reach out for solutions to all your network management and deployment needs.