run_locust
Execute load tests on web applications by configuring virtual users, spawn rates, and test duration to measure performance under simulated traffic.
Instructions
Run Locust with the given configuration.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| test_file | Yes | ||
| host | No | http://localhost:8089 | |
| users | No | ||
| spawn_rate | No | ||
| runtime | No | 30s | |
| headless | No |
Implementation Reference
- locust_server.py:14-52 (handler)The run_locust tool handler: an async function decorated with @mcp.tool that runs the Locust load testing tool via subprocess.run, constructing the command based on parameters and returning success/error status with output.
@mcp.tool(name="run_locust", description="Run Locust with the given configuration.") async def run_locust(test_file: str, host: str = os.getenv("LOCUST_HOST", "http://localhost:8089"), users: int = int(os.getenv("LOCUST_USERS", "100")), spawn_rate: int = int(os.getenv("LOCUST_SPAWN_RATE", "10")), runtime: str = os.getenv("LOCUST_RUNTIME", "30s"), headless: bool = os.getenv("LOCUST_HEADLESS", "true").lower() == "true") -> Any: """ Run Locust with the given configuration. Args: test_file: Path to the Locust test file host: Target host URL to load test users: Number of concurrent users to simulate spawn_rate: Rate at which users are spawned per second runtime: Duration of the test (e.g., "30s", "1m", "5m") headless: Whether to run in headless mode (no web UI) """ locust_bin = os.getenv("LOCUST_BIN", "locust") cmd = [locust_bin, "-f", test_file, "--host", host] if headless: cmd.extend(["--headless"]) cmd.extend(["-u", str(users)]) cmd.extend(["-r", str(spawn_rate)]) cmd.extend(["-t", runtime]) try: result = subprocess.run(cmd, capture_output=True, text=True, check=True) return { "status": "success", "output": result.stdout, "error": result.stderr } except subprocess.CalledProcessError as e: return { "status": "error", "output": e.stdout, "error": e.stderr }