virhekesto

Virhekesto, literally “fault endurance” in Finnish, is the property of a system to continue operating, or degrade gracefully, in the presence of faults. In Finnish technical usage, it corresponds to the concept of fault tolerance in English.

In computing and engineering, virhekesto describes a system's ability to withstand hardware failures, software bugs, or transient faults without catastrophic failure. It encompasses the capacity to maintain essential functions, recover quickly, and prevent cascading errors.

Common approaches include redundancy (duplication of components), error detection and correction (ECC memory, checksums), failover and hot-swapping, isolation of faults, and diversity in design to avoid common-mode failures. Graceful degradation and self-healing systems are also part of virhekesto, allowing systems to continue operating at reduced capacity rather than failing entirely.

In distributed systems and communications, virhekesto is achieved with replication, consensus protocols (Paxos, Raft), and Byzantine fault tolerance to handle arbitrary faults. These techniques help ensure consistency and availability even when some nodes act erratically or fail.

Metrics used to assess virhekesto include availability, mean time between failures (MTBF), mean time to repair (MTTR), and resilience under load. Standards and practices in reliability engineering, safety certifications, and design guidelines inform how virhekesto is implemented in critical systems.

Examples include RAID levels, ECC RAM, mirrored storage, hot standby power supplies, Kubernetes self-healing, and distributed databases with replication. Trade-offs involve increased cost, complexity, and potential performance penalties, requiring careful balancing of reliability, latency, and resource use.