Conservatively, 60% of network service outages are caused when planned changes don’t work as intended. While undesirable, this is not surprising given (1) the complexity of today’s dynamic IP networks, multiple AS’s, VPNs, TE tunnels, and various classes of service, and (2) the frequency of maintenance windows required to support different applications, traffic growth, and new customers. Engineers, armed only with aggregated statistical data and static network models, have little insight to the impacts that network maintenance will have on existing services. The unintended consequences can be costly.
The Explorer Suite’s unique combination of route, traffic, and performance analytics offers deep, predictive insight to IP network behavior under different conditions. Using the live network’s routing topology, traffic matrices, and performance baselines, network engineers can interactively model changes to accurately predict their impact. By doing this immediately before maintenance windows, iterating as necessary, expected outcomes can be assured. After the maintenance is complete, before-and-after comparison reports provide immediate confirmation. For example, it is simple to assess the impact of adding, removing or reconfiguring a router before the change is made. Similarly, failure scenarios can be simulated to show how the current network will re-converge and whether or not there is sufficient fail-over capacity to maintain service levels.
As operators embrace SDN in the WAN to increase agility, they need an analytics-driven orchestration solution that enables intelligent service provisioning while mitigating the risk from rapid, automated network changes. Static network maps and point-in-time device metrics must give way to always-current models that are fed by real-time routing, traffic, and performance telemetry. Packet Design’s Explorer Suite is designed for SDN and can help operators make a smooth transition to self-healing, self-optimizing networks.