Event Prediction for Network Edge Devices using Log Analysis

Devices that control data flow at the boundary between two networks are known as Network Edge Devices and they are typically used as network entry or exit points. It is not possible to check and monitor these devices manually. Further monitoring these devices using a separate application creates unnecessary overhead. An effective approach is to automate monitoring of these devices by building intelligence into these devices to perform such tasks using Machine Learning. This paper proposes the usage of Elasticstack to ship all logs to a central server where they are analyzed to predict failures of the network edge devices. The research yielded good results while using application logs to predict failure of small cells.