add pyhton logger code

Author: Lanting687

backend/app/logging_config.py

@@ -0,0 +1,17 @@
+import logging
+import sys
+
+
+def setup_logging() -> None:
+    handler = logging.StreamHandler(sys.stdout)
+    handler.setFormatter(logging.Formatter(
+        "%(asctime)s %(levelname)s [%(name)s] %(message)s",
+        datefmt="%Y-%m-%d %H:%M:%S",
+    ))
+
+    root = logging.getLogger()
+    root.setLevel(logging.INFO)
+    root.handlers = [handler]
+
+    # Quiet down noisy third-party loggers
+    logging.getLogger("httpx").setLevel(logging.WARNING)

backend/app/main.py

@@ -2,10 +2,13 @@
 from fastapi import FastAPI
 from fastapi.middleware.cors import CORSMiddleware
 from app.config import settings
+from app.logging_config import setup_logging
 from app.utils.qdrant_service import get_qdrant_service
 from app.routers.chat import router as chat_router
 from app.routers.documents import router as documents_router
 
+setup_logging()
+
 
 @asynccontextmanager
 async def lifespan(app: FastAPI):

backend/app/routers/chat.py

@@ -1,7 +1,9 @@
+import logging
 from fastapi import APIRouter, HTTPException
 from app.schemas import ChatRequest, ChatResponse, DocumentRead
 from app.utils.rag import retrieve_relevant_documents, ask_gemini, ask_gemini_with_grounding, is_audit_related
 
+logger = logging.getLogger(__name__)
 router = APIRouter()
 
 CONFIDENCE_THRESHOLD = 0.65
@@ -17,19 +19,28 @@ async def chat(request: ChatRequest):
     if not request.question.strip():
         raise HTTPException(status_code=400, detail="Question cannot be empty.")
 
+    logger.info("Question received: %r", request.question)
+
     if not await is_audit_related(request.question):
+        logger.info("Rejected as non-audit-related")
         return ChatResponse(answer=NON_AUDIT_RESPONSE, sources=[])
 
     try:
         relevant = await retrieve_relevant_documents(request.question, top_k=8)
     except Exception as exc:
+        logger.error("Vector search failed: %s", exc)
         raise HTTPException(status_code=502, detail=f"Vector search failed: {exc}")
 
+    top_score = max((doc.score for doc in relevant), default=0.0)
+    logger.info("Retrieved %d docs, top score=%.3f", len(relevant), top_score)
+
     if _is_confident(relevant):
+        logger.info("Confidence threshold met — answering from Qdrant context")
         try:
             snippets = [f"Title: {doc.title}\nContent: {doc.content}" for doc in relevant]
             answer = await ask_gemini(request.question, snippets)
         except RuntimeError as exc:
+            logger.error("Gemini chat failed: %s", exc)
             raise HTTPException(status_code=502, detail=f"LLM request failed: {exc}")
 
         return ChatResponse(
@@ -40,9 +51,11 @@ async def chat(request: ChatRequest):
             ],
         )
 
+    logger.info("Confidence threshold not met — falling back to Gemini web grounding")
     try:
         answer = await ask_gemini_with_grounding(request.question)
     except RuntimeError as exc:
+        logger.error("Gemini grounding failed: %s", exc)
         raise HTTPException(status_code=502, detail=f"LLM request failed: {exc}")
 
     return ChatResponse(answer=answer, sources=[])

backend/app/utils/rag.py

@@ -1,10 +1,13 @@
 import asyncio
 import json
+import logging
 from typing import List
 import httpx
 from app.config import settings
 from app.utils.qdrant_service import get_qdrant_service
 
+logger = logging.getLogger(__name__)
+
 
 def _build_headers(endpoint: str) -> dict:
     headers = {"Content-Type": "application/json"}
@@ -127,7 +130,8 @@ async def _do_request(c: httpx.AsyncClient) -> bool:
             return await _do_request(client)
         async with httpx.AsyncClient(timeout=30.0) as owned_client:
             return await _do_request(owned_client)
-    except Exception:
+    except Exception as exc:
+        logger.warning("is_audit_related classification failed, failing open (allowing question): %s", exc)
         return True  # fail open — don't block users if classification itself fails
 
 

ec2-logs.sh

@@ -0,0 +1,28 @@
+#!/bin/bash
+# Quick access to EC2 application logs.
+#
+# Usage:
+#   ./ec2-logs.sh                 # last 50 lines, all services
+#   ./ec2-logs.sh backend         # last 50 lines, backend only
+#   ./ec2-logs.sh backend -f      # follow backend logs live
+#   ./ec2-logs.sh -f              # follow all services live
+
+KEY_PATH="$(dirname "$0")/RAG_KEY_PAIR.pem"
+EC2_HOST="[email protected]"
+
+SERVICE=""
+FOLLOW_FLAG=""
+
+for arg in "$@"; do
+  if [ "$arg" = "-f" ] || [ "$arg" = "--follow" ]; then
+    FOLLOW_FLAG="-f"
+  else
+    SERVICE="$arg"
+  fi
+done
+
+if [ -n "$FOLLOW_FLAG" ]; then
+  ssh -i "$KEY_PATH" "$EC2_HOST" "cd ~/RAG && docker compose logs -f $SERVICE"
+else
+  ssh -i "$KEY_PATH" "$EC2_HOST" "cd ~/RAG && docker compose logs --tail=50 $SERVICE"
+fi