Faculty Advisor

Gerstenfeld, Arthur

Center

WS / Wall Street MQP

Abstract

The project ‘Predictive Analysis for Network Data Storm’ involves the analysis of big data in Splunk, which indexes machine-generated big data and allows efficient querying and visualization, to develop a set of thresholds to predict a network meltdown, or commonly known as a data storm. The WPI team analyzed multiple datasets to spot patterns and determine the major differences between the normal state and the storm state of the network. A set of rules and thresholds were fully developed for the Fixed Income Transversal Tools team in BNP Paribas, who implemented the model in their internal real-time monitoring tool ‘SCADA’ to predict and prevent network data storms.

Publisher

Worcester Polytechnic Institute

Date Accepted

January 2014

Major

Industrial Engineering

Project Type

Major Qualifying Project

Accessibility

Unrestricted

Advisor Department

Business

Share

COinS