Files
git.stella-ops.org/docs/features/unchecked/tests/chaos-failure-testing-infrastructure.md

25 lines
1.9 KiB
Markdown

# Chaos/Failure Testing Infrastructure
## Module
__Tests
## Status
IMPLEMENTED
## Description
A chaos testing library exists for failure choreography and integration testing scenarios.
## Implementation Details
- **Failure Choreographer**: `src/__Tests/__Libraries/StellaOps.Testing.Chaos/FailureChoreographer.cs` -- orchestrates failure injection sequences across distributed services, coordinating timed failures for integration tests.
- **Failure Injector Interface**: `src/__Tests/__Libraries/StellaOps.Testing.Chaos/IFailureInjector.cs` -- contract for failure injection strategies (network partition, service crash, resource exhaustion).
- **Convergence Tracker Interface**: `src/__Tests/__Libraries/StellaOps.Testing.Chaos/IConvergenceTracker.cs` -- tracks system convergence after failure injection, verifying the system recovers to a consistent state within a timeout.
- **Chaos Models**: `src/__Tests/__Libraries/StellaOps.Testing.Chaos/Models.cs` -- data models for failure scenarios, injection points, and convergence results.
- **Tests**: `src/__Tests/__Libraries/StellaOps.Testing.Chaos.Tests/FailureChoreographerTests.cs`, `FailureInjectorTests.cs`, `ConvergenceTrackerTests.cs`
## E2E Test Plan
- [ ] Configure a `FailureChoreographer` with a network partition scenario and verify it injects the failure at the specified time and restores connectivity after the duration
- [ ] Inject a service crash via `IFailureInjector` and verify `IConvergenceTracker` detects the system has not converged within the expected timeout
- [ ] Inject a failure and verify the system eventually converges to a consistent state after the failure is removed
- [ ] Run a choreographed sequence of 3 failures (network delay, service restart, resource exhaustion) and verify each failure is applied in order with correct timing
- [ ] Verify the chaos tests are isolated and do not affect other test suites running in parallel