k6-load-testing

Validate API performance against SLOs using k6 by executing staged load profiles to measure latency, throughput, and error rates. Threshold-gated CI integration automatically fails pipelines when defined performance targets are breached.

Smoke, stress, and spike load testing scenarios represent different staged testing profiles that measure system capacity under normal, elevated, and sudden burst workloads. These open-model arrival-rate executors ensure accurate latency measurements.

Automate load testing in CI by configuring threshold-gated k6 pipelines that automatically fail deployments when latency or error rate thresholds are breached. This ensures performance regressions are caught before deployment.

k6 load testing explicitly handles rate-limited back-pressure differently from actual system failures by providing a rigorous, threshold-gated pipeline. This distinction prevents designed rate-limiting from causing false performance test failures.

Simulate open-model workloads with k6 arrival-rate executors to ensure accurate capacity and latency measurements. This approach supports staged load profiles including smoke, average, stress, and spike testing scenarios.

Database isolation is required for local API performance testing using a throwaway environment setup. This ensures accurate threshold-gated CI integration and prevents data pollution during load simulation.