Performance and Scaling

Guidelines for scaling Pixashot deployments, including CPU and memory optimization, worker configuration, and performance tuning.

This guide covers scaling considerations and performance optimization for Pixashot deployments.

Resource Requirements

Base Requirements Per Instance

  • CPU: 1 core minimum
  • RAM: 2GB minimum
  • Storage: 1GB base + temporary storage
  • Network: Reliable outbound connection

Scaling Formula

Workers = MIN(CPU_CORES * 2, AVAILABLE_MEMORY_GB)
Concurrent_Requests = Workers * 10

Memory Scaling

Memory Per Request Type

Standard Screenshot:   ~50MB
Full Page:            ~100MB
PDF Generation:       ~150MB
Complex Interactions: ~200MB

Instance Sizing

# Light load (<10 req/min)
--memory=2g --workers=2

# Medium load (10-50 req/min)
--memory=4g --workers=4

# Heavy load (50+ req/min)
--memory=8g --workers=8

Worker Configuration

Environment Variables

# Light usage
export WORKERS=2
export MAX_REQUESTS=500
export KEEP_ALIVE=300

# Medium usage
export WORKERS=4
export MAX_REQUESTS=1000
export KEEP_ALIVE=300

# Heavy usage
export WORKERS=8
export MAX_REQUESTS=2000
export KEEP_ALIVE=300

Performance Optimization

Browser Context

{
    "block_media": true,            // Reduce memory usage
    "wait_for_network": "mostly_idle",  // Faster than "idle"
    "pixel_density": 1.0,          // Lower for large viewports
    "format": "jpeg",              // More efficient than PNG
    "image_quality": 85            // Balance quality vs size
}

Caching

# Enable response caching
export CACHE_MAX_SIZE=1000        # Adjust based on memory
export RATE_LIMIT_ENABLED=true    # Prevent overload
export RATE_LIMIT_CAPTURE="5 per second"

Cloud Run Auto-scaling

Configuration

# Optimal scaling settings
gcloud run deploy pixashot \
  --min-instances=1 \
  --max-instances=10 \
  --concurrency=80 \
  --cpu=2 \
  --memory=4Gi \
  --timeout=300

Load Balancing

  • Set appropriate min-instances for steady traffic
  • Configure max-instances based on peak load
  • Use concurrency to control requests per instance
  • Monitor cold start performance

Monitoring Metrics

Key metrics to track:

  • Memory usage per request
  • CPU utilization
  • Response times
  • Error rates
  • Cold start frequency
  • Cache hit rates

Health Endpoints

# Monitor these endpoints
/health          # Overall health
/health/live     # Liveness check
/health/ready    # Readiness check

Scaling Checklist

  1. Resource Monitoring

    • Configure CPU alerts
    • Set memory usage alerts
    • Monitor network latency
    • Track error rates

  2. Performance Tuning

    • Optimize worker count
    • Configure request limits
    • Enable caching
    • Set appropriate timeouts

  3. Load Testing

    • Test concurrent requests
    • Measure memory usage
    • Monitor response times
    • Verify error handling

Best Practices

  1. Resource Management

    • Start with conservative limits
    • Monitor actual usage
    • Scale gradually
    • Implement proper cleanup

  2. Error Handling

    • Set appropriate timeouts
    • Implement retry logic
    • Monitor failure rates
    • Log errors comprehensively

  3. Performance

    • Use efficient formats
    • Optimize wait strategies
    • Enable caching
    • Monitor response times

For detailed monitoring configuration, see the Monitoring Guide.

Get the Latest Updates