Load Testing
Testing that evaluates how a system performs under expected and peak traffic conditions by simulating concurrent users or requests.
Load testing applies controlled levels of traffic to a system to measure response times, throughput, error rates, and resource consumption. Variants include stress testing (pushing past expected limits to find the breaking point), spike testing (sudden traffic surges), and soak testing (sustained load over a long period to detect memory leaks or degradation).
Popular load testing tools include k6, Apache JMeter, Locust, and Gatling. Cloud-based services can generate load from multiple geographic regions to simulate real-world conditions.
Why It Matters for QA Teams
A website that performs well for 50 users may collapse under 5,000. Load testing reveals performance bottlenecks before a product launch, marketing campaign, or seasonal traffic spike exposes them to real customers.
Example
Before a Black Friday sale, the team uses k6 to simulate 10,000 concurrent users browsing products and 2,000 simultaneously checking out. The test reveals that the payment API begins returning timeouts above 1,500 concurrent checkout requests. The infrastructure team scales the service and re-tests to confirm the fix.