As businesses build new SD-WAN infrastructures, NetOps teams will need to be able to visualize these new, complex networks and how they impact and relate to legacy IT environments.
SD-WAN deployments are rapidly becoming a major focus for network operations (NetOps) teams around the world, and changing the face of networking as we know it. In fact, Gartner estimates that nearly two-thirds of all enterprises will have incorporated SD-WAN technology by 2021. One example of the acceleration of this market is Cisco’s recent move to upgrade a million routers with SD-WAN software. Now NetOps teams are tasked to assess the redesign and migration paths from on-premise licensing and configuration management to the Cisco SD-WAN cloud service.
As businesses focus on building new SD-WAN infrastructure like this over the next several years, NetOps teams will need to be able to visualize these new, complex networks and how they impact and relate to legacy IT environments. Here are eleven key visualizations every NetOps team would benefit from to successfully plan, deploy, verify and manage their SD-WAN projects.
NetOps teams need to see into every corner of the organization. This means end-to-end visibility across your organization – from endpoints, applications and sites to tunnels, VPNs and service provider transport SLA performance – all in one understandable and action-oriented view. With this level of insight, administrators gain real-time situational awareness and easy drill downs necessary for network tuning and issue resolution.
SD-WAN topology views – tunnel performance
Active topology maps should provide a layered set of filtered views to isolate applications, VPNs and policies in real time. These types of views allow NetOps teams to monitor key performance indicators and SLA KPIs and help you understand if they either remain within boundaries or are fluctuating for cloud applications and applications traversing service provider connections.
Current active alerts by site
Alerts or alarms are critical in determining when network conditions have changed from the baseline. In modern NPMD platforms, alerts are overlaid on top of the topology, site and summary views for cross team notifications, so you can engage in incident response together with the same information. These alerts should present the severity level of the issue, the trigger or cause, and semantic information such as location, time and description for automation of the service management ticket creation and distribution with popular enterprise services like ServiceNow or PagerDuty.
Proactive insights through machine learning
Machine Learning (ML) and advanced analytics are making rapid advances in modeling predictive conditions to avoid service affecting incidents like latency. Understanding potential or forecasted service performance issues and their implications ahead of time allows you to make the necessary adjustments and prevent the problem before it impacts the end user and the business.
With user experiences measured in seconds for response time, and quality based on fuzzy video or dropped calls, NetOps must be able to debug issues quickly. Seeing the application demands from services across the entire SD-WAN enables you with everything from a status summary to information about bandwidth consumption and quality of assigned service policy. All of this detail allows NetOps to manage and adjust network resources in order to keep demanding customers, partners and employees happy.
Flow performance views (audio/video)
When policies change – and they always do – it’s important to conduct path analysis. Some of the most useful SD-WAN visualizations show detailed path analytics, which can help IT to verify policy changes at scale and understand application path switching. As part of the ongoing management of SD-WAN deployments, NetOps teams also need the ability to visualize site-to-site tunnel performance correlated with service provider and policy thresholds.
Service provider performance views
When designing your SD-WAN, you could have several to many service provider agreements in place. Having the ability to analyze the performance, utilization, and paths associated with each of those services is critical to the success of the deployment. IT departments need visualizations that allow them to understand which telco infrastructure path application should take, evaluate how SD-WAN-enabled versus non-SD-WAN traffic should be handled during deployment, and more.
Site-level views are critical and practical to gain visibility into a particular site and better understand and manage the data interconnections and the service provider performance.
Having an application policy site-to-site view that interconnects multiple datasets to show the overlay fabric performance is key for SD-WAN deployments. By associating applications and their assigned Virtual Private Network (VPN) and DSCP class of service marking information, with the service provider performance between site pairs, you can validate the SLA performance of the path for situational awareness and problem resolution.
When managing performance or troubleshooting the SD-WAN, IT needs granular insight into device performance. These teams need to be able to assess each site and drill down into individual devices. This allows them to successfully verify application paths and service VPN topologies, as well as the ability to look at policy assignments and performance (for example with tunnel overlays, bandwidth, SLA class handing and more).
Detailed packet decoding is needed for complex network problems or serious security issues. Having packet capture and analysis is critical for NetOps teams to be able to drill down, find root cause, troubleshoot and resolve the issue quickly and confidently.