obs: Add Prometheus metrics endpoint for request counting and latency

quantara/pyproject.toml

@@ -34,6 +34,7 @@ faker = "^30.8.1"
 notebook = "^7.2.2"
 sentry-sdk = {extras = ["fastapi"], version = "^2.18.0"}
 slowapi = "^0.1.9"
+prometheus-client = "0.21.1"
 structlog = ">=24.1.0"
 [tool.poetry.group.dev.dependencies]
 black = "24.8.0"

quantara/web_app/api/main.py

@@ -31,6 +31,7 @@
 from web_app.api.leaderboard import router as leaderboard_router
 from web_app.api.referal import router as referal_router
 from web_app.api.wallet_auth import router as auth_router
+from web_app.api.metrics import router as metrics_router, PrometheusMiddleware
 from web_app.config_validator import assert_valid_config
 from web_app.api.middleware import MaxBodySizeMiddleware
 from web_app.db.database import init_db
@@ -148,6 +149,7 @@ async def global_exception_handler(request: Request, exc: Exception):
 # full middleware stack and can reject requests before they reach routers.
 app.add_middleware(SlowAPIMiddleware)
 app.add_middleware(MaxBodySizeMiddleware, max_body_size=1024*1024)
+app.add_middleware(PrometheusMiddleware)
 
 
 @app.middleware("http")
@@ -207,3 +209,4 @@ async def health_check(response: Response, db: Session = Depends(get_database)):
 app.include_router(leaderboard_router)
 app.include_router(referal_router)
 app.include_router(auth_router)
+app.include_router(metrics_router)

quantara/web_app/api/metrics.py

@@ -0,0 +1,67 @@
+"""
+Prometheus metrics endpoint and middleware for the QUANTARA API.
+
+Exposes /metrics with:
+- http_requests_total: counter by method, endpoint, status
+- http_request_duration_seconds: histogram of request latency
+- http_requests_in_flight: gauge of concurrent requests
+"""
+
+import time
+
+from fastapi import APIRouter
+from prometheus_client import (
+    Counter,
+    Gauge,
+    Histogram,
+    generate_latest,
+    CONTENT_TYPE_LATEST,
+)
+from starlette.middleware.base import BaseHTTPMiddleware
+from starlette.requests import Request
+from starlette.responses import Response
+
+router = APIRouter()
+
+REQUEST_COUNT = Counter(
+    "http_requests_total",
+    "Total HTTP request count",
+    ["method", "endpoint", "status"],
+)
+REQUEST_LATENCY = Histogram(
+    "http_request_duration_seconds",
+    "HTTP request latency in seconds",
+    ["method", "endpoint"],
+)
+REQUESTS_IN_FLIGHT = Gauge(
+    "http_requests_in_flight",
+    "Number of HTTP requests currently being processed",
+)
+
+
+class PrometheusMiddleware(BaseHTTPMiddleware):
+    """Middleware that records request count, latency, and in-flight gauge."""
+
+    async def dispatch(self, request: Request, call_next) -> Response:
+        endpoint = request.url.path
+        method = request.method
+        REQUESTS_IN_FLIGHT.inc()
+        start = time.perf_counter()
+        try:
+            response = await call_next(request)
+            status = str(response.status_code)
+        except Exception:
+            status = "500"
+            raise
+        finally:
+            duration = time.perf_counter() - start
+            REQUEST_COUNT.labels(method=method, endpoint=endpoint, status=status).inc()
+            REQUEST_LATENCY.labels(method=method, endpoint=endpoint).observe(duration)
+            REQUESTS_IN_FLIGHT.dec()
+        return response
+
+
+@router.get("/metrics", tags=["Observability"], summary="Prometheus metrics endpoint")
+async def metrics() -> Response:
+    """Expose Prometheus metrics in text format."""
+    return Response(content=generate_latest(), media_type=CONTENT_TYPE_LATEST)