# Performance Testing Suite

This directory contains performance tests for the Aniworld API and download system.

## Test Categories

### API Load Testing (`test_api_load.py`)

Tests API endpoints under concurrent load to ensure acceptable performance:

-   **Load Testing**: Concurrent requests to endpoints
-   **Sustained Load**: Long-running load scenarios
-   **Concurrency Limits**: Maximum connection handling
-   **Response Times**: Performance benchmarks

**Key Metrics:**

-   Requests per second (RPS)
-   Average response time
-   Success rate under load
-   Graceful degradation behavior

### Download Stress Testing (`test_download_stress.py`)

Tests the download queue and management system under stress:

-   **Queue Operations**: Concurrent add/remove operations
-   **Capacity Testing**: Queue behavior at limits
-   **Memory Usage**: Memory leak detection
-   **Concurrency**: Multiple simultaneous downloads
-   **Error Handling**: Recovery from failures

**Key Metrics:**

-   Queue operation success rate
-   Concurrent download capacity
-   Memory stability
-   Error recovery time

## Running Performance Tests

### Run all performance tests:

```bash
conda run -n AniWorld python -m pytest tests/performance/ -v -m performance
```

### Run specific test file:

```bash
conda run -n AniWorld python -m pytest tests/performance/test_api_load.py -v
```

### Run with detailed output:

```bash
conda run -n AniWorld python -m pytest tests/performance/ -vv -s
```

### Run specific test class:

```bash
conda run -n AniWorld python -m pytest \
    tests/performance/test_api_load.py::TestAPILoadTesting -v
```

## Performance Benchmarks

### Expected Results

**Health Endpoint:**

-   RPS: ≥ 50 requests/second
-   Avg Response Time: < 0.1s
-   Success Rate: ≥ 95%

**Anime List Endpoint:**

-   Avg Response Time: < 1.0s
-   Success Rate: ≥ 90%

**Search Endpoint:**

-   Avg Response Time: < 2.0s
-   Success Rate: ≥ 85%

**Download Queue:**

-   Concurrent Additions: Handle 100+ simultaneous adds
-   Queue Capacity: Support 1000+ queued items
-   Operation Success Rate: ≥ 90%

## Adding New Performance Tests

When adding new performance tests:

1. Mark tests with `@pytest.mark.performance` decorator
2. Use `@pytest.mark.asyncio` for async tests
3. Include clear performance expectations in assertions
4. Document expected metrics in docstrings
5. Use fixtures for setup/teardown

Example:

```python
@pytest.mark.performance
class TestMyFeature:
    @pytest.mark.asyncio
    async def test_under_load(self, client):
        \"\"\"Test feature under load.\"\"\"
        # Your test implementation
        metrics = await measure_performance(...)
        assert metrics["success_rate"] >= 95.0
```

## Continuous Performance Monitoring

These tests should be run:

-   Before each release
-   After significant changes to API or download system
-   As part of CI/CD pipeline (if resources permit)
-   Weekly as part of regression testing

## Troubleshooting

**Tests timeout:**

-   Increase timeout in pytest.ini
-   Check system resources (CPU, memory)
-   Verify no other heavy processes running

**Low success rates:**

-   Check application logs for errors
-   Verify database connectivity
-   Ensure sufficient system resources
-   Check for rate limiting issues

**Inconsistent results:**

-   Run tests multiple times
-   Check for background processes
-   Verify stable network connection
-   Consider running on dedicated test hardware

## Performance Optimization Tips

Based on test results, consider:

1. **Caching**: Add caching for frequently accessed data
2. **Connection Pooling**: Optimize database connections
3. **Async Processing**: Use async/await for I/O operations
4. **Load Balancing**: Distribute load across multiple workers
5. **Rate Limiting**: Implement rate limiting to prevent overload
6. **Query Optimization**: Optimize database queries
7. **Resource Limits**: Set appropriate resource limits

## Integration with CI/CD

To include in CI/CD pipeline:

```yaml
# Example GitHub Actions workflow
- name: Run Performance Tests
  run: |
      conda run -n AniWorld python -m pytest \
        tests/performance/ \
        -v \
        -m performance \
        --tb=short
```

## References

-   [Pytest Documentation](https://docs.pytest.org/)
-   [HTTPX Async Client](https://www.python-httpx.org/async/)
-   [Performance Testing Best Practices](https://docs.python.org/3/library/profile.html)