Measuring DB limits with pre-event load testing — using k6 to find the point where the DB breaks first

If you’re told to “run a load test” before an event, use k6 to ramp up requests per second step by step while monitoring database metrics—CPU, active threads, connection pool, lock waits, buffer pool, and replication lag—all on one screen. Identify what breaks first, then set numerical thresholds for system limits and when to stop. Reproduce everything locally with Docker.