Predict, then measure: queueing model versus open-loop load test

Predict a service’s maximum throughput and latency knee from measured service times, then measure latency percentiles under open-loop load and explain every gap.

The exercise happens on Self-directed project (open-loop load generator). Canonical link: https://github.com/giltene/wrk2. Link check: link responded (5 Oct 2026). Opened from this app 0 times; opening is not counted as an attempt.

Supports: Quantitative performance reasoning for systems research and performance-sensitive roles.