Databricks MCP Server

--- description: "Comprehensive debugging for development and deployment issues" --- # Debug Your Databricks App I'll help you debug issues with your Databricks app by running comprehensive diagnostics and providing solutions. ## What I'll do: 1. **System Health Check** - Verify all tools and dependencies 2. **Development Environment Debug** - Check local development setup 3. **Authentication Validation** - Test Databricks connectivity 4. **Local App Testing** - Run app locally to identify issues 5. **Deployment Status Analysis** - Check deployed app health 6. **Provide Solutions** - Give specific fixes for identified problems ## Debugging Workflow **Step 1: System Health Check** ```bash # Check required tools uv --version bun --version databricks --version python --version # Check project structure ls -la # Verify key files exist ls -la .env.local ls -la server/app.py ls -la client/package.json ``` **Step 2: Development Environment Debug** ```bash # Check development server status ps aux | grep databricks-app # Check logs tail -20 /tmp/databricks-app-watch.log # Check PID file cat /tmp/databricks-app-watch.pid ``` **Step 3: Authentication Debug** ```bash # Test authentication databricks current-user me # With profile if configured databricks current-user me --profile "$DATABRICKS_CONFIG_PROFILE" # Check .env.local configuration cat .env.local ``` **Step 4: Local App Testing** ```bash # Run app locally with debug mode ./run_app_local.sh --verbose ``` **Step 5: Deployment Status Analysis** ```bash # Check app status ./app_status.sh --verbose # Check workspace files databricks workspace list "$DBA_SOURCE_CODE_PATH" ``` ## Common Issues and Solutions ### Development Server Issues **Problem: Port already in use** ```bash # Kill processes on ports 3000/8000 pkill -f "uvicorn server.app:app" pkill -f "vite" pkill -f "node.*3000" ``` **Problem: Development server won't start** ```bash # Clean restart kill $(cat /tmp/databricks-app-watch.pid) || pkill -f watch.sh rm -f /tmp/databricks-app-watch.pid rm -f /tmp/databricks-app-watch.log nohup ./watch.sh > /tmp/databricks-app-watch.log 2>&1 & ``` ### Import and Build Issues **Problem: TypeScript client missing** ```bash # Regenerate TypeScript client uv run python scripts/make_fastapi_client.py ``` **Problem: `@/lib/utils` import error** ```bash # Check if utils.ts exists in correct location ls -la src/lib/utils.ts ls -la client/src/lib/utils.ts # Copy if missing mkdir -p src/lib cp client/src/lib/utils.ts src/lib/utils.ts ``` **Problem: Python import errors** ```bash # Check virtual environment uv run python -c "import sys; print(sys.path)" # Reinstall dependencies uv sync ``` ### Authentication Issues **Problem: Databricks authentication failed** ```bash # Test different auth methods databricks current-user me databricks current-user me --profile "$DATABRICKS_CONFIG_PROFILE" # Reconfigure authentication ./setup.sh ``` **Problem: Invalid credentials** ```bash # Check token validity (for PAT) curl -H "Authorization: Bearer $DATABRICKS_TOKEN" \ "$DATABRICKS_HOST/api/2.0/current-user" # Re-authenticate databricks auth login --host "$DATABRICKS_HOST" ``` ### Deployment Issues **Problem: App deployment failed** ```bash # Check app status ./app_status.sh --verbose # Test locally first ./run_app_local.sh --verbose # Check deployment logs # Visit app URL + /logz in browser ``` **Problem: App not running after deployment** ```bash # Get detailed app info databricks apps get "$DATABRICKS_APP_NAME" # Check workspace sync databricks workspace list "$DBA_SOURCE_CODE_PATH" # Redeploy with verbose output ./deploy.sh --verbose ``` ### Build and Compilation Issues **Problem: Frontend build fails** ```bash # Check frontend dependencies cd client bun install # Build manually bun run build # Check for TypeScript errors bun run type-check ``` **Problem: Backend startup fails** ```bash # Test backend manually uv run uvicorn server.app:app --reload # Check Python dependencies uv run python -c "import fastapi; print('FastAPI OK')" uv run python -c "import databricks; print('Databricks SDK OK')" ``` ## Advanced Debugging ### Log Analysis ```bash # Development logs tail -f /tmp/databricks-app-watch.log # Filter for errors grep -i error /tmp/databricks-app-watch.log # Check for specific issues grep -i "port.*use" /tmp/databricks-app-watch.log ``` ### Network Debugging ```bash # Check if ports are available netstat -an | grep :3000 netstat -an | grep :8000 # Test API connectivity curl http://localhost:8000/health curl http://localhost:8000/docs ``` ### File System Issues ```bash # Check permissions ls -la .env.local ls -la scripts/ # Check disk space df -h # Check file integrity file server/app.py file client/package.json ``` ## Nuclear Reset (Last Resort) If all else fails, here's a complete reset: ```bash # Stop everything pkill -f watch.sh pkill -f uvicorn pkill -f vite # Clean up files rm -f /tmp/databricks-app-watch.pid rm -f /tmp/databricks-app-watch.log rm -f /tmp/local-app-test.log # Reinstall dependencies uv sync cd client && bun install && cd .. # Reconfigure environment ./setup.sh # Test locally ./run_app_local.sh # Start fresh ./watch.sh ``` ## Getting Help If you're still having issues: 1. **Check the logs** - Always start with `/tmp/databricks-app-watch.log` 2. **Test locally** - Use `./run_app_local.sh --verbose` 3. **Verify authentication** - Test with `databricks current-user me` 4. **Check app status** - Use `./app_status.sh --verbose` 5. **Visit `/logz`** - Check deployment logs in browser ## Success Indicators Your app is working when: - ✅ Development servers start without errors - ✅ Frontend loads at http://localhost:3000 - ✅ Backend responds at http://localhost:8000 - ✅ API docs work at http://localhost:8000/docs - ✅ Authentication tests pass - ✅ App deploys successfully - ✅ Deployed app shows "RUNNING" status You've got this! 🐛➡️✅