Analysis of a network fault detection system to support decision making

Abstract

When a network fault occurs, administrators spend a lot of time finding the causes as well as solving the problem. Network fault localization is a process that aims to find the exact source of the fault. This process consists of a diagnosis process that considers the symptoms ranging from end-to-end connectivity fault to more sophisticated symptoms such as SLA violations. To solve network faults, it is necessary a great amount of information that allows network administrators to analyze and determine symptoms, and then reduce the number of possible solutions. Considering the above discussed, it is necessary a decision support system that allows network administrators to diagnose a network fault and solve it. In this work, we present a fault detection system for LAN networks. This system is based on a set of rules that allows determining the specific problem as well as to obtain possible solutions to them. The rules were defined based on the data collected from network monitoring process as well as the expertise of network administrators. The system was evaluated to measure its effectiveness regarding network fault detection, obtaining encouraging results.