Cheshire Cat Load Testing

This project contains scripts and configuration for load testing the Cheshire Cat API using Locust.

Overview

The load testing suite is designed to simulate various levels of user traffic to the Cheshire Cat API, allowing for performance analysis and bottleneck identification. It uses Locust, a popular open-source load testing tool, to generate traffic and collect metrics.

Files

• run_load_test.sh: Main script to execute the load tests
• config.sh: Configuration file for test parameters
• locustfile.py: Locust test scenarios definition

Prerequisites

• Bash shell
• Python 3.7+
• Locust (pip install -r requirements.txt)
• Docker (for running the Cheshire Cat API)

Configuration

Edit config.sh to modify the following parameters:

• LOCUST_FILE: Path to the Locust file (default: "locustfile.py")
• RESULTS_DIR: Directory to store test results (default: "results")
• USER_COUNTS: Array of user counts for each test iteration (default: 10, 50, 100, 500)
• TEST_DURATION: Duration of each test run (default: "3m")
• PAUSE_TIME: Pause time between test iterations in seconds (default: 60)

Usage

1. Ensure the Cheshire Cat API is running (defaults on http://localhost:1865)

2. Run the load test script:

   ./run_load_test.sh
   

3. The script will execute multiple test iterations based on the USER_COUNTS defined in config.sh

4. Results for each iteration will be saved in the RESULTS_DIR with timestamps

Test Scenarios

The locustfile.py defines two main tasks:

1. get_status: Simulates checking the API status (weight: 1)
2. send_message: Simulates sending a message to the Cheshire Cat using the HTTP message endpoint (weight: 2)

Results

After each test iteration, you'll find the following in the results directory:

• HTML report (report.html)
• CSV files with detailed metrics

The current load tests are configured to run against a Cheshire Cat instance with the following resource constraints:

deploy:
  resources:
    limits:
      cpus: '2'
      memory: 4G
    reservations:
      cpus: '2'
      memory: 4G

Customization

To add or modify test scenarios, edit the locustfile.py file. Refer to the Locust documentation for more information on creating custom user behaviors and tasks.