Tutorial: Performance Testing

Learn how to perform load and stress testing with Performance. This class allows you to simulate multiple concurrent users to evaluate your API's performance.

What is Performance Testing?

RapidTest's Performance module allows you to: - Simulate multiple concurrent users - Measure response times - Calculate success rate - Evaluate throughput (requests per second)

Initial Setup

from rapidtest import Performance

# Create instance
perf = Performance(
    base_url="http://localhost:8000",
    users=10,        # Concurrent users
    duration=10,    # Duration in seconds
    timeout=10,     # Request timeout
    delay=0.1       # Delay between requests
)

Constructor Parameters

Parameter	Type	Default	Description
base_url	str	None	Base URL of the API
users	int	10	Number of concurrent users
duration	int	10	Test duration in seconds
timeout	int	10	Maximum request timeout
delay	float	0.1	Delay between requests

Adding Test Tasks

Use the add_task() method to define endpoints to test:

# Simple GET request
perf.add_task(endpoint="/health", method="GET")

# GET with parameters
perf.add_task(
    endpoint="/users",
    method="GET",
    params={"page": 1}
)

# POST request
perf.add_task(
    endpoint="/login",
    method="POST",
    json={"username": "test", "password": "test123"}
)

# PUT request
perf.add_task(
    endpoint="/users/1",
    method="PUT",
    json={"name": "Updated"}
)

# DELETE request
perf.add_task(endpoint="/users/1", method="DELETE")

# With headers
perf.add_task(
    endpoint="/protected",
    method="GET",
    headers={"Authorization": "Bearer token"}
)

Run the Test

results = perf.run()
print(results)

Returned Metrics

The run() method returns a dictionary with:

{
    'total_requests': 1500,        # Total requests made
    'successful_requests': 1485,  # Successful requests
    'failed_requests': 15,        # Failed requests
    'success_rate': 99.0,         # Success percentage
    'avg_response_time': 45.2,    # Average time in ms
    'min_response_time': 12.1,    # Minimum time in ms
    'max_response_time': 89.7,    # Maximum time in ms
    'requests_per_second': 150.0, # Requests per second
    'duration': 10,               # Actual duration in seconds
    'users': 10                  # Concurrent users
}

Complete Example: Load Test

from rapidtest import Performance

# Configure load test
perf = Performance(
    base_url="http://localhost:8000",
    users=50,          # 50 concurrent users
    duration=30,       # 30 seconds
    timeout=15,        # 15 seconds timeout
    delay=0.2          # 200ms between requests
)

# Add tasks
print("Setting up load test...")

# Health endpoint test
perf.add_task(endpoint="/health", method="GET")

# User list test
perf.add_task(endpoint="/users", method="GET")

# User creation test
perf.add_task(
    endpoint="/users",
    method="POST",
    json={"name": "Load Test User", "email": "load@test.com"}
)

# Run test
print("\n🚀 Starting load test...\n")
results = perf.run()

# Analyze results
print("\n📊 RESULTS SUMMARY:")
print(f"   Success: {results['success_rate']}%")
print(f"   Throughput: {results['requests_per_second']} req/s")
print(f"   Average time: {results['avg_response_time']}ms")

Example: Result Validation

from rapidtest import Performance

def run_and_validate():
    perf = Performance(
        base_url="http://localhost:8000",
        users=100,
        duration=60,
        timeout=10
    )

    perf.add_task(endpoint="/api/health", method="GET")
    results = perf.run()

    # Validations
    assert results['success_rate'] >= 95, f"Success rate too low: {results['success_rate']}%"
    assert results['avg_response_time'] <= 500, f"Average time too high: {results['avg_response_time']}ms"
    assert results['requests_per_second'] >= 10, f"Throughput too low: {results['requests_per_second']}"

    print(f"✅ All validations passed!")
    return results

run_and_validate()

Example: Multiple Scenarios

from rapidtest import Performance

def test_api_endpoints():
    scenarios = [
        {
            "name": "Health Check",
            "endpoint": "/health",
            "method": "GET",
            "expected_success_rate": 99,
            "max_response_time": 200
        },
        {
            "name": "User List",
            "endpoint": "/users",
            "method": "GET",
            "expected_success_rate": 95,
            "max_response_time": 500
        },
        {
            "name": "User Creation",
            "endpoint": "/users",
            "method": "POST",
            "json": {"name": "Test", "email": "test@example.com"},
            "expected_success_rate": 90,
            "max_response_time": 1000
        }
    ]

    for scenario in scenarios:
        print(f"\n🧪 Testing: {scenario['name']}")

        perf = Performance(
            base_url="http://localhost:8000",
            users=20,
            duration=10
        )

        perf.add_task(
            endpoint=scenario["endpoint"],
            method=scenario["method"],
            json=scenario.get("json")
        )

        results = perf.run()

        # Validate against expectations
        if results['success_rate'] < scenario['expected_success_rate']:
            print(f"   ⚠️  Warning: success rate {results['success_rate']}% < {scenario['expected_success_rate']}%")

        if results['avg_response_time'] > scenario['max_response_time']:
            print(f"   ⚠️  Warning: time {results['avg_response_time']}ms > {scenario['max_response_time']}ms")

test_api_endpoints()

Results Interpretation

Indicator	Ideal Value	Meaning
success_rate	≥ 95%	API is stable
avg_response_time	≤ 500ms	Good response times
requests_per_second	High	Higher throughput
max_response_time	≤ 2x average	No significant outliers

Tips for Effective Testing

1. Start gradually: Begin with few users and increase
2. Measure baseline: First establish a baseline
3. Isolate endpoints: Test one endpoint at a time for clear results
4. Consider delay: Very low delay can saturate the server
5. Monitor the server: Watch CPU, memory, and API connections

Next Step

Mastered the basics? Explore more in: - API Reference