Dic Service Benchmark Bundle Laravel Package

Getting Started

Minimal Steps

1. Installation:

   composer require dawen/dic-service-benchmark-bundle
   

   No kernel registration is required—this is a zero-configuration package.

2. First Run: Execute the bundled test directly via PHPUnit (disable coverage for faster results):

   ./vendor/bin/phpunit -c app vendor/dawen/dic-service-benchmark-bundle/Tests/DicServiceBenchmarkTest.php
   

   Outputs a list of registered services with their instantiation times.

3. Threshold Check: The test fails if any service exceeds 50ms by default. Useful for performance regression detection.

Implementation Patterns

Workflows

1. Performance Profiling:

   • Integrate into CI/CD pipelines to catch slow service instantiation early.
   • Example: Run post-commit or pre-deploy:

     ./vendor/bin/phpunit -c app --filter=DicServiceBenchmarkTest
     

2. Ignoring Services:

   • Exclude non-critical services (e.g., debug tools) via phpunit.xml:

     <php>
         <var name="dic-service-benchmark-ignore" value="web_profiler.controller.router:my_ignored_service"/>
     </php>
     

   • Format: Colon-separated list of service IDs (supports wildcards? Not documented—verify).

3. Custom Thresholds:

   • Override the default 50ms threshold by extending the test class:

     // tests/DicServiceBenchmarkCustomTest.php
     use Dawen\Bundle\DicServiceBenchmarkBundle\Tests\DicServiceBenchmarkTest;
     class CustomBenchmarkTest extends DicServiceBenchmarkTest {
         protected $maxAllowedTime = 100; // ms
     }
     

4. Integration with Laravel:

   • Manual Trigger: Call the benchmark logic programmatically in a custom command:

     use Dawen\Bundle\DicServiceBenchmarkBundle\Service\BenchmarkService;
     $benchmark = new BenchmarkService($container);
     $results = $benchmark->run();
     

   • Laravel Service Provider: Bind the BenchmarkService to the container for reuse.

Gotchas and Tips

Pitfalls

1. False Positives:

   • Heavy services (e.g., database connections, HTTP clients) may legitimately exceed 50ms.
   • Fix: Ignore them via dic-service-benchmark-ignore or adjust the threshold.

2. Environment Dependencies:

   • Instantiation times vary across environments (dev/staging/prod).
   • Tip: Run benchmarks in a staging-like environment for realistic results.

3. Lazy Services:

   • Services with lazy initialization (e.g., App\Services\LazyService::class) may show inconsistent times.
   • Workaround: Force eager loading in tests or exclude them.

4. No Laravel-Specific Docs:

   • The package targets Symfony’s DIC. Laravel users must:
     • Use Illuminate\Container\Container as the DIC.
     • Manually bind the BenchmarkService if not auto-discovered.

5. Test Isolation:

   • The bundled test assumes a clean DIC. Avoid running it mid-request (e.g., in a controller).

Debugging Tips

• Verbose Output: Check the test’s run() method for debug logs (not exposed in README).
• Service-Specific Issues:
  • Slow services often indicate:
    • Missing @Lazy annotations (Symfony).
    • Heavy constructor logic (e.g., loading config files).
    • Circular dependencies.
  • Action: Profile individual services with microtime(true) or Xdebug.

Extension Points

1. Custom Benchmark Logic:

   • Extend Dawen\Bundle\DicServiceBenchmarkBundle\Service\BenchmarkService to:
     • Add pre/post-hooks for each service.
     • Log results to a database or file.
   • Example:

     class ExtendedBenchmarkService extends BenchmarkService {
         public function run() {
             $results = parent::run();
             $this->logResults($results); // Custom logic
             return $results;
         }
     }
     

2. Visualization:

   • Parse output into a dashboard (e.g., Grafana) for trend analysis.
   • Example CLI output format:

     Service: app.service.foo | Time: 12.34ms | Status: OK
     Service: app.service.bar | Time: 67.89ms | Status: FAIL (threshold: 50ms)
     

3. Parallel Testing:

   • Use PHPUnit’s --parallel flag to speed up benchmarking across multiple services.