feat: Add `start_space.sh` to automate DVC setup, model pulling, and launch FastAPI and Streamlit services.

scripts/start_space.sh

@@ -16,55 +16,63 @@ USER=${DAGSHUB_USERNAME:-$MLFLOW_TRACKING_USERNAME}
 PASS=${DAGSHUB_TOKEN:-$MLFLOW_TRACKING_PASSWORD}
 
 if [ -n "$USER" ] && [ -n "$PASS" ]; then
-    echo "Configuring DVC authentication for DagsHub..."
+    echo "$(date) - Configuring DVC authentication for DagsHub..."
     # Configure local config (not committed)
     dvc remote modify origin --local auth basic
     dvc remote modify origin --local user "$USER"
     dvc remote modify origin --local password "$PASS"
 else
-    echo "WARNING: No DagsHub credentials found. DVC pull might fail if the remote is private."
+    echo "$(date) - WARNING: No DagsHub credentials found. DVC pull might fail if the remote is private."
 fi
 
-echo "Pulling models from DVC..."
+echo "$(date) - Pulling models from DVC..."
 # Pull only the necessary files for inference
 dvc pull models/random_forest_tfidf_gridsearch.pkl.dvc \
          models/tfidf_vectorizer.pkl.dvc \
-         models/label_names.pkl.dvc
+         models/label_names.pkl.dvc || echo "DVC pull failed, but continuing..."
 
 # Create Nginx temp directories
 mkdir -p /tmp/client_temp /tmp/proxy_temp /tmp/fastcgi_temp /tmp/uwsgi_temp /tmp/scgi_temp
 
-echo "Checking models existence..."
+echo "$(date) - Checking models existence..."
 ls -la models/
 
-echo "Starting FastAPI application in background..."
-uvicorn hopcroft_skill_classification_tool_competition.main:app --host 127.0.0.1 --port 8000 >> /tmp/fastapi.log 2>&1 &
+echo "$(date) - Starting FastAPI application in background..."
+# Using 0.0.0.0 to be safe
+uvicorn hopcroft_skill_classification_tool_competition.main:app --host 0.0.0.0 --port 8000 >> /tmp/fastapi.log 2>&1 &
 
 # Wait for API to start
-echo "Waiting for API to start..."
-sleep 15
+echo "$(date) - Waiting for API to start (30s)..."
+for i in {1..30}; do
+    if curl -s http://127.0.0.1:8000/health > /dev/null; then
+        echo "$(date) - API is UP!"
+        break
+    fi
+    echo "$(date) - Waiting... ($i/30)"
+    sleep 2
+done
 
-echo "Starting Nginx reverse proxy..."
-# Check if nginx is in path
+echo "$(date) - Starting Nginx reverse proxy..."
 if ! command -v nginx &> /dev/null; then
-    echo "ERROR: nginx not found in PATH"
+    echo "$(date) - ERROR: nginx not found in PATH"
     exit 1
 fi
 nginx -c /app/nginx.conf -g "daemon off;" >> /tmp/nginx_startup.log 2>&1 &
 
-echo "Waiting for Nginx to initialize..."
+echo "$(date) - Waiting for Nginx to initialize..."
 sleep 5
 
 # Check if Nginx is running
 if ps aux | grep -v grep | grep -q "nginx"; then
-    echo "Nginx is running."
+    echo "$(date) - Nginx is running."
 else
-    echo "ERROR: Nginx failed to start. Logs:"
+    echo "$(date) - ERROR: Nginx failed to start. Logs:"
     cat /tmp/nginx_startup.log
-    # Try to start it without the config to see if it's a binary issue
-    # nginx -v
 fi
 
-echo "Starting Streamlit application..."
+echo "$(date) - Final backend check before starting Streamlit..."
+curl -v http://127.0.0.1:8000/health || echo "FastAPI health check failed!"
+
+echo "$(date) - Starting Streamlit application on 0.0.0.0:8501..."
 export API_BASE_URL="http://127.0.0.1:8000"
-streamlit run hopcroft_skill_classification_tool_competition/streamlit_app.py --server.port 8501 --server.address 127.0.0.1
+streamlit run hopcroft_skill_classification_tool_competition/streamlit_app.py --server.port 8501 --server.address 0.0.0.0