Test dependency failures before they happen. Exercises show how user flows and server policies behave under uncertainty.
Evaluation approach
Treat timeouts, invalid responses and network absence as separate scenarios. Check messages, transaction state and operational alerts together.
Application example
Verify that an unreachable verification service does not duplicate a payment and that pending work follows the intended safe policy.
Limits and considerations
A complete shutdown does not reproduce every slowdown or partial failure.
Test recovery too
Making verification unreachable is only the beginning. Examine waiting, errors and retries, then determine how old events and operations are handled when service returns.
Retries must not amplify load or duplicate transactions. The user-visible state should agree with the server's actual outcome. Use the exercise to refine default failure policy and support flows.
Checks and decisions
- Test partial failures
- Limit retries
- Measure recovery
Use findings to improve policies, support procedures and alerts.
Sources
The primary references above provide the technical basis. Example workflows and evaluation suggestions are this publication’s explanations, not independent test results for a particular product.