Архитектура

Руководство по интеграции

Интеграция с quic-test и CloudBridge Relay

AI Routing Lab - Integration Guide

How to integrate AI Routing Lab with quic-test and CloudBridge Relay


Integration Overview

AI Routing Lab integrates with two main components:

  1. quic-test - For data collection and model validation
  2. CloudBridge Relay - For production route optimization

Integration with quic-test

Step 1: Setup quic-test with Prometheus Export

# Start quic-test server with Prometheus metrics
cd cloudbridge/quic-test
./bin/quic-server --prometheus-port 9090 --prometheus-path /metrics

Step 2: Configure Prometheus

# prometheus.yml
scrape_configs:
  - job_name: 'quic-test'
    scrape_interval: 15s
    static_configs:
      - targets: ['localhost:9090']

Step 3: Collect Data

from data.collectors.quic_test_collector import PrometheusCollector

collector = PrometheusCollector(prometheus_url="http://localhost:9090")
metrics = collector.collect_all_metrics()

Step 4: Validate Predictions

# Run quic-test with ML validation mode
./bin/quic-client \
  --ml-validation-mode \
  --ml-predictions-file predictions.json \
  --output results.json

Integration with CloudBridge Relay

Step 1: Deploy ML Prediction API

# Start inference service
python inference/predictor.py --model models/latency_predictor.pkl --port 8080

Step 2: Configure CloudBridge Relay

# relay-config.yaml
ml_routing:
  enabled: true
  prediction_api_url: "http://ai-routing-lab:8080/api/v1/predict"
  fallback_to_baseline: true
  prediction_timeout: 100ms

Step 3: Route Selection

CloudBridge Relay will query ML API for route predictions and select optimal route based on predictions.


Data Collection Workflow

  1. quic-test generates traffic → Exports metrics to Prometheus
  2. AI Routing Lab collects metrics → Processes and stores
  3. ML models train → Generate predictions
  4. Predictions validated → Against quic-test results
  5. Production deployment → CloudBridge Relay uses predictions

Validation Workflow

  1. Generate predictions for test routes
  2. Run quic-test with same routes
  3. Compare predictions vs actual latency/jitter
  4. Calculate accuracy metrics
  5. Report results and iterate

For detailed integration code, see: integration/quic_test_client.py and integration/relay_integration.py