Anomaly-based Fault Detection with Interaction Analysis Using State Interface
Byoung Uk Kim · BiblioBoard Library Catalog (Open Research Library) · 2009
In this paper, we developed an effective rule-based fault detection algorithm to detect any type of faults for a distributed computing environment. And we evaluate the false alarm rates of our approach for four different fault scenarios and for different data sizes with varying levels of noise. Our analysis show that our approach is superior when compared to other techniques. For example, the precision value that is trained with Tranining with PN equals to 0.998 in 50% noise value and the missed alarm and false alarm rate is near 0%. We are currently extending our approach to not only detect the fault s once they occur, but also perform root-cause analysis and automatic fault recovery.