lastprøving

Lastprøving, commonly translated as load testing, is a software engineering practice that evaluates how a system performs under expected and peak load. The aim is to verify that an application meets performance, reliability, and scalability requirements under realistic conditions, and to identify bottlenecks before deployment.

In practice, lastprøving involves defining a workload model that mirrors real usage, such as concurrent users, transactions per second, or API calls. Test scenarios are executed with automated tooling that simulates this load. Popular tools include JMeter, Gatling, Locust, and k6, which can generate traffic from one or more test environments and allow monitoring of system behavior in real time.

Key objectives include measuring response times, throughput, error rates, and resource utilization (CPU, memory, disk I/O, and network). Acceptable thresholds are usually tied to service level agreements or user experience targets. Results help validate capacity, reveal performance regressions, and guide tuning of code, databases, caching, and infrastructure. Findings may lead to architectural adjustments, such as scaling strategies or database optimizations, and inform capacity planning.

Best practices emphasize testing in production-like environments, using realistic data, and maintaining reproducible test scripts. It is often complemented by other testing types, such as functional testing and stress testing (which pushes beyond normal limits to find breaking points). Overall, lastprøving is a core component of performance engineering and capacity planning.