Retry Lab

Retry Lab exposes deterministic endpoints that intentionally misbehave. Use it to validate exponential backoff, circuit breakers, and telemetry pipelines.

Trigger a scenario

1. Open Retry Lab in the UI.
2. Select a scenario such as slow-start, timeout, truncate, or flip-flop.
3. Copy the generated endpoint and hit it from your client under test.

Each scenario tracks request counts, success ratios, and latency buckets.

Monitor the progress bars and success ratios while your client exercises the selected retry scenario.

Observe metrics

• The dashboard updates in realtime via the retry.stats channel.
• GET /api/retrylab/stats returns the full aggregate payload for automation.
• Progress bars reset manually or when you clear the state via the UI.

Automate checks

Create a heartbeat job (for example, via GitHub Actions or a custom runner) that calls target scenarios on a schedule and alerts you if retry success falls below SLA thresholds. Record baselines in your observability platform.

Clean slate

Use Clear metrics to reset counters between test runs. This action only affects aggregates; individual HTTP responses are not stored beyond in-memory statistics.