Method and system of fault diagnosis in a data network. Specifically, one embodiment of the present invention discloses a method for automating the process of determining faults in a data network. A network management station (NMS) receives filtered fault data from the various subnetworks of computer systems that comprise the data network. Each of the subnetworks have an associated performance manager that monitors, collects and filters fault data from the subnetwork. Each of the performance managers send their respective filtered fault data to the NMS. The NMS further analyses the filtered fault data to identify the fault and locate a source of the fault within the topology of the data network. The NMS can further query the network for additional fault data as necessary. The NMS displays the fault, cause, and location of the cause to a network engineer for correction.