Fault Tolerance

Kind: quality-attribute

Layer: Correctness Core

Record: architecture:fault-tolerance

Severity: contextual

Scope: service, system, infrastructure

Canonical: Ontology

The degree to which a system keeps operating correctly when some of its components fail.

Listed in Architecture principles, after Graceful Degradation and before Resilience.

Requires

• Redundancy
• Error Handling

Reinforces

• Resilience

Enables

• Continued Operation Under Failure

Conflicts with

• Single Point of Failure

In tension with

• Cost

Tensions

• Fault Tolerance / Cost

Violated by

• Single Point of Failure

Refactored by

• Bounded Retry
• Failover
• Redundancy

Severity

• contextual

Category

• Error Handling / Resilience

Reinforced by

• Leader Election
• Consensus
• Graceful Degradation
• Retry Pattern
• Circuit Breaker Pattern
• Chaos Engineering
• RAID Redundancy

Linked from

• Control / Coordination / Centralization
• Error Handling / Resilience
• Observability / Auditability / Traceability
• Self-Healing / Recovery / Deployment Safety
• Transactions / State / Concurrency
• Core Vocabulary
• Error Handling / Resilience
• Severity levels
• The resolutions