Databricks MCP Server

--- description: "Deploy your Databricks app to production" --- # Deploy to Databricks Apps I'll deploy your Databricks app to production with comprehensive validation and monitoring. ## What I'll do: 1. **Validate environment** - Check authentication and configuration 2. **Test locally first** - Run app locally to catch issues before deployment 3. **Check app status** - Verify if app exists or needs creation 4. **Deploy to Databricks** - Build, sync, and deploy using proper workflow 5. **Monitor deployment** - Verify successful deployment and provide URL 6. **Provide next steps** - Give you monitoring and debugging information ## Deployment Workflow **Step 1: Environment Validation** ```bash # Check if .env.local exists and is configured cat .env.local # Test Databricks authentication databricks current-user me ``` **Step 2: Local Testing (Critical)** ```bash # Test app locally first to catch issues ./run_app_local.sh ``` **Step 3: App Status Check** ```bash # Check if app exists ./app_status.sh # Get app details databricks apps get "$DATABRICKS_APP_NAME" ``` **Step 4: Deployment Decision** Based on app status: - **If app exists**: Deploy with `./deploy.sh` - **If app doesn't exist**: Ask if you want to create it - **If you want to create**: Use `./deploy.sh --create` **Step 5: Deploy** ```bash # Deploy (with creation if needed) ./deploy.sh --create --verbose ``` **Step 6: Deployment Verification** ```bash # Check final status ./app_status.sh # Verify app is running databricks apps get "$DATABRICKS_APP_NAME" ``` ## Deployment Options **Standard Deployment:** ```bash ./deploy.sh ``` **Create New App:** ```bash ./deploy.sh --create ``` **Verbose Deployment:** ```bash ./deploy.sh --verbose ``` ## What Happens During Deployment 1. **Authentication** - Validates Databricks credentials 2. **App Creation** - Creates app if using `--create` and doesn't exist 3. **Frontend Build** - Builds React app for production 4. **Requirements Generation** - Creates requirements.txt from pyproject.toml 5. **Workspace Sync** - Uploads source code to Databricks workspace 6. **App Deployment** - Deploys via Databricks CLI 7. **Verification** - Confirms successful deployment ## Monitoring Your Deployment **Check App Status:** ```bash ./app_status.sh ``` **View Deployment Logs:** - Visit your app URL + `/logz` in browser - Requires OAuth authentication - Cannot be accessed via curl **Debug Deployment Issues:** ```bash # Get verbose status ./app_status.sh --verbose # Check workspace files databricks workspace list "$DBA_SOURCE_CODE_PATH" ``` ## Common Deployment Issues **Authentication Problems:** - Check `.env.local` configuration - Test with `databricks current-user me` - Reconfigure with `./setup.sh` **App Creation Issues:** - Verify you have app creation permissions - Check if app name is available - Use `./deploy.sh --create` explicitly **Build/Import Errors:** - Test locally first with `./run_app_local.sh` - Check TypeScript compilation - Verify all dependencies are installed **Deployment Failures:** - Check app logs at URL + `/logz` - Use `./app_status.sh --verbose` for details - Verify workspace file sync ## Success Criteria Deployment is successful when: - ✅ App status shows "RUNNING" - ✅ App URL returns 200 OK - ✅ No errors in `/logz` endpoint - ✅ App functionality works as expected ## Next Steps After Deployment 1. **Test your app** at the provided URL 2. **Monitor logs** via `/logz` endpoint 3. **Use `/status`** to check health regularly 4. **Use `/debug`** if issues arise 5. **Iterate and deploy** as needed Your app is now live! 🚀