Add signoz

gateway/app/main.py

@@ -10,7 +10,7 @@ import resource
 import os
 from fastapi import FastAPI, Request, HTTPException, Depends, WebSocket, WebSocketDisconnect
 from fastapi.middleware.cors import CORSMiddleware
-from fastapi.responses import JSONResponse, StreamingResponse
+from fastapi.responses import JSONResponse, StreamingResponse, Response
 import httpx
 import time
 from shared.redis_utils import initialize_redis, close_redis, get_redis_client
@@ -27,7 +27,42 @@ from app.middleware.demo_middleware import DemoMiddleware
 from app.middleware.read_only_mode import ReadOnlyModeMiddleware
 from app.routes import auth, tenant, notification, nominatim, subscription, demo, pos, geocoding, poi_context
 from shared.monitoring.logging import setup_logging
-from shared.monitoring.metrics import MetricsCollector
+from shared.monitoring.metrics import MetricsCollector, add_metrics_middleware
+
+# OpenTelemetry imports
+from opentelemetry import trace
+from opentelemetry.sdk.trace import TracerProvider
+from opentelemetry.sdk.trace.export import BatchSpanProcessor
+from opentelemetry.exporter.otlp.proto.grpc.trace_exporter import OTLPSpanExporter
+from opentelemetry.instrumentation.fastapi import FastAPIInstrumentor
+from opentelemetry.instrumentation.httpx import HTTPXClientInstrumentor
+from opentelemetry.instrumentation.redis import RedisInstrumentor
+from opentelemetry.sdk.resources import Resource
+
+# Configure OpenTelemetry tracing
+def setup_tracing(service_name: str = "gateway"):
+    """Initialize OpenTelemetry tracing with OTLP exporter for Jaeger"""
+    # Create resource with service name
+    resource = Resource.create({"service.name": service_name})
+
+    # Configure OTLP exporter (sends to OpenTelemetry Collector)
+    otlp_exporter = OTLPSpanExporter(
+        endpoint=os.getenv("OTEL_EXPORTER_OTLP_ENDPOINT", "http://otel-collector.monitoring.svc.cluster.local:4317"),
+        insecure=True  # Use insecure connection for internal cluster communication
+    )
+
+    # Configure tracer provider
+    provider = TracerProvider(resource=resource)
+    processor = BatchSpanProcessor(otlp_exporter)
+    provider.add_span_processor(processor)
+
+    # Set global tracer provider
+    trace.set_tracer_provider(provider)
+
+    return provider
+
+# Initialize tracing
+tracer_provider = setup_tracing("gateway")
 
 # Setup logging
 setup_logging("gateway", settings.LOG_LEVEL)
@@ -75,9 +110,21 @@ app = FastAPI(
     redirect_slashes=False  # Disable automatic trailing slash redirects
 )
 
+# Instrument FastAPI with OpenTelemetry
+FastAPIInstrumentor.instrument_app(app)
+
+# Instrument httpx for outgoing requests
+HTTPXClientInstrumentor().instrument()
+
+# Instrument Redis (will be active once redis client is initialized)
+RedisInstrumentor().instrument()
+
 # Initialize metrics collector
 metrics_collector = MetricsCollector("gateway")
 
+# Add metrics middleware to track HTTP requests
+add_metrics_middleware(app, metrics_collector)
+
 # Redis client for SSE streaming
 redis_client = None
 
@@ -182,8 +229,11 @@ async def health_check():
 
 @app.get("/metrics")
 async def metrics():
-    """Metrics endpoint for monitoring"""
-    return {"metrics": "enabled"}
+    """Prometheus metrics endpoint"""
+    return Response(
+        content=metrics_collector.get_metrics(),
+        media_type="text/plain; version=0.0.4; charset=utf-8"
+    )
 
 # ================================================================
 # SERVER-SENT EVENTS (SSE) HELPER FUNCTIONS