Performance Test Scripts for Critical Endpoints
Write load test scripts with realistic user behavior.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
Write load test scripts with realistic user behavior. Tool: [LOAD_TEST_TOOL]. Endpoints and user journeys: [ENDPOINTS_AND_JOURNEYS]. Expected traffic (average and peak requests per second, concurrent users): [TRAFFIC]. Service level goals: [SLO]. Test environment: [ENVIRONMENT]. Model your answer on this example structure for a login and search journey: - Setup: create test users in advance; authenticate once per virtual user and reuse the token. - Scenario: 70 percent browse search results (think time 3 to 8 seconds), 20 percent view an item, 10 percent add to cart. - Load shape: ramp up for 5 minutes, hold at the target for 15 minutes, spike to 2 times for 2 minutes, ramp down. - Checks: status codes, response contents, and a response time threshold per endpoint. - Thresholds: 95th percentile under 500 ms, error rate under 1 percent. For my case, write the full script with a realistic data feeder (varied ids and search terms), correlation of dynamic values, think times, and thresholds. Include a smoke test (1 user), a load test, a stress test to find the limit, and a soak test outline. Explain how to run the test safely (non-production environment, with data isolation, notifying the team), what to monitor on the server side (CPU, memory, database connections, queue depth), and how to interpret and report results. Warn against testing third-party services without permission.
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into ChatGPT, Gemini, Claude or another assistant.