Glossary term
Single Point of Failure (SPOF)
Subject: Technology, Cyber and Data, IT Resilience and Third-Party Risk
Also known as: Single point of failure
A single point of failure is any one component, person, supplier or process whose breakdown would stop an entire service or system because no backup or alternative route exists.
Explanation
Resilient design avoids relying on one thing. A SPOF can be a server, a network link, a power feed, a database, a software licence server, a key employee who alone knows a process, or one supplier with no substitute. Teams find them by drawing the path a service depends on and asking of each element: if this failed now, what stops? The usual remedy is redundancy, such as duplicate hardware, separate network routes, clustered databases, trained backups for staff, or a second supplier. Redundancy has a cost, so firms weigh it against how critical the service is. Hidden SPOFs are common, for example two supposedly separate links that share one cable duct. The idea applies beyond technology, to people, suppliers and locations. Resilience reviews hunt for these because large outages so often trace back to one dependency nobody had mapped.
Example
An airline's check-in gateway runs on two servers, so it looks safe. During a dependency mapping exercise, the team discovers both servers rely on one authentication service, which has no standby. If it fails, check-in stops. The airline adds a second instance at another site and tests automatic switch-over.
Not to be confused with
- Concentration Risk
- Concentration risk is heavy exposure to one counterparty, sector or provider; a SPOF is a specific component whose loss halts a service.