NOM products enable users to manage their network by enforcing controlled changes to configuration of network devices apart from monitoring performance, fault and compliance of all devices in their network. While monitoring a network, it is always been a challenge to deduce and represent correlation, among anomalies across thousands of devices, in a format that users can consume, validate and act on. This paper proposes a method to deduce and visualize, possibly related anomalies on connected devices in the network based on user’s selection of an anomaly on a device of interest. While this paper takes network domain as an example to demonstrate the idea, it could be applied to any of the infrastructure and application management solutions in a datacenter. This paper proposes a unique approach to solve this problem by leveraging ML and network operator’s deployment/domain knowledge.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License