perf-test-reference

Separating write and read latency metrics in performance testing isolates the distinct behavior of different operations. This structured approach applies a single-variable change per round, ensuring fair comparisons and accurate identification of operation-specific bottlenecks.

The ramping-arrival-rate executor in k6 precisely controls the number of requests per second, allowing you to deterministically simulate traffic scenarios. It provides the structured runtime template needed to execute repeatable performance tests for web services.

You can use this performance testing framework for large-scale traffic scenarios by configuring the k6 ramping-arrival-rate scenarios. It provides deterministic execution protocols designed to reproduce high traffic volumes and identify operational performance bottlenecks.

To collect latency metrics separately for read and write operations, configure your k6 scenarios to track distinct metrics per request type. The framework's structured result reporting supports separate latency measurements for accurate analysis.