A cascading failure is failure in a system of interconnected parts, where the service provided depends on the operation of a preceding part, and the failure of a preceding part can trigger the failure of successive parts. Redundant parts can lessen the impact of, but not prevent, a failure. Monitoring the operation of a system, in real-time, and judicious disconnection of parts can stop a cascade.

In a system with a single point of failure, loss of service can be catastrophic when that part fails.

Some classic examples of system failure due to cascading failures can be found in wide-area power outages.

External links

• A. E. Motter and Y.-C. Lai, Cascade-based attacks on complex networks, Physical Review E (Rapid Communications) 66, 065102 (2002).
• Protection Strategies for Cascading Grid Failures — A Shortcut Approach
• Ian Dobson, Benjamin A. Carreras, and David E. Newman, preprint A loading-dependent model of probabilistic cascading failure, Probability in the Engineering and Informational Sciences, vol. 19, no. 1, January 2005, pp. 15-32.