Faults are not uncommon in modern complex enterprise systems and these tend to
propagate throughout the system. It is important to diagnose the observed failure
symptoms to determine the root cause of the failures to initiate proper recovery
actions. The project aims to provide probabilistic root cause analysis for distributed
enterprise systems using the failure dependency graph of the application. Bayesian
Belief Networks(BBNs) are used to encode the application failure model based on the
topology of the application.
. Proactive network anomaly detection.M.Tech. Seminar
July, 2004.November, 2004