Scenarios and Tests

A scenario is the definition of a performance test: which endpoints are called, with which HTTP methods, and under what parameters. This page explains scenario creation, key parameters, and the run lifecycle step by step.

Creating a scenario

In LoadEng, every test is defined with a scenario. A scenario can include multiple tasks; each task maps to one target URL and one HTTP method (GET, POST, PUT, PATCH, DELETE). This allows you to send load to different endpoints with different methods in the same run.

  • Tasks: Each task contains a URL and HTTP method. For example, use GET for homepage, POST for login, and PUT for API updates. Tasks are executed according to the scenario's distribution rules.
  • Parameters: Configure concurrent users (or virtual users), total test duration, and ramp-up time. Start low and increase gradually to identify system breaking points.
  • Pre-check: Before running the scenario, use the Check tool on Uptime Check to verify target URLs are reachable and responding to the correct method.

Running tests

After saving the scenario, start the test. The system distributes the scenario to registered traffic generators, which send calls according to task definitions. Test duration and concurrency are applied from your configured scenario values.

  • Real-time monitoring: RPS (requests per second), average/percentile latency, and error rate are tracked live. You can stop the test early if needed.
  • Traffic distribution: You can view how many traffic generators are active and how traffic is distributed. For distributed setups, see Workers and Deployment.

Reports and results

When the test completes, summary and detailed reports are generated. Reports include total request count, success/failure distribution, latency statistics (min, max, average, p95, p99), and time-series data. You can compare historical runs for capacity planning or post-optimization verification.

Important notes

  • For targets you test for the first time, start with low concurrency and short durations; then scale gradually.
  • Run performance tests only against systems you own or explicitly have authorization to test.

Next steps

Go back to Getting Started for core concepts. Continue with Workers and Deployment for worker setup and distributed execution. Use API Reference to manage scenarios and runs via API.