fault-injection-testing

Circuit breaker state transitions are tested by simulating deterministic fault scenarios across storage and network dependencies. This provides a state machine to verify end-to-end recovery workflows and ensure boundary faults do not violate system invariants.

Network failures are simulated using a controlled fault injector that creates deterministic timing and boundary-only fault scenarios. This exercises resilience patterns across external dependencies and validates recovery workflows without leaking data.

Retry policies with exponential backoff and jitter are implemented as built-in testing utilities to validate recovery from transient faults. You apply these policies during fault injection to verify external dependency calls recover under simulated failure conditions.

Queue preservation and data integrity are testable during fault injection using built-in queue preservation assertions. These utilities verify that boundary faults do not leak data or violate invariants across distributed system recovery workflows.

Fault injection testing works for external storage dependencies by applying deterministic fault scenarios to simulate failures. It supports boundary-only injections to ensure storage calls handle errors gracefully and maintain data invariants during recovery.

Intermittent resilience test failures often stem from non-deterministic timing in fault scenarios. This tool solves that by providing deterministic timing and boundary-only injections, ensuring consistent and repeatable fault injection for reliable test results.