Method and system for fault diagnosis in a data network

A method of fault diagnosis comprising:

• a) receiving a plurality of fault data pertaining to said data network;
• b) filtering said plurality of fault data to eliminate extraneous data down to a core of fault data;
• i) querying said data network to generate additional fault data different from said plurality of fault data when said core of fault data is insufficient to identify faults;
• ii) receiving said additional fault data;
• c) determining whether said core of fault data and said additional fault data is due to a broken link or congestion in said data network;
• d) performing a ping walk to isolate a source of said core of fault data and additional fault data and to determine a location of said source, if said core of fault data and additional fault data is due to said broken link;
• and
• e) using deductive reasoning to isolate said source of said core of fault data and additional fault data and identify said cause of said core of fault data and additional fault data, if said core of fault data and additional fault data is due to said congestion.

A data network that is capable of fault diagnosis comprising:

• a plurality of subnetworks that generate a plurality of fault data, each of said plurality of subnetworks comprising network components that are coupled together via a distributing component;
• a plurality of performance managers coupled to said plurality of subnetworks for monitoring said plurality of subnetworks for said plurality of fault data, for filtering said plurality of fault data, and for querying said plurality of subnets to generate additional fault data different from said plurality of fault data when said plurality of fault data is insufficient to identify faults, each of said plurality of network performance managers coupled to and associated with one of said plurality of subnetworks;
• and
• a single network management station coupled to each of said plurality of performance managers for analyzing said plurality of fault data that is filtered and said additional fault data to identify faults and isolate sources of said faults.

A computer system comprising:

• a processor;
• a display coupled to said processor;
• a computer readable memory coupled to said processor and containing program instructions that, when executed, implement the steps of:
• a) receiving a plurality of fault data pertaining to said data network;
• b) filtering said plurality of fault data to obtain a core of fault data;
• i) querying said data network to generate additional fault data different from said plurality of fault data when said core of fault data is insufficient to identify faults;
• ii) receiving said additional fault data;
• and
• c) analyzing said core of fault data and said additional fault data to identify a fault associated with said core of fault data.