LangChain Agent MCP Server

# Google Cloud Run Deployment Guide Complete guide for deploying the LangChain Agent MCP Server to Google Cloud Run. ## Prerequisites 1. **Google Cloud Account** with billing enabled 2. **Google Cloud SDK (gcloud)** installed and configured 3. **Docker** installed and running 4. **OpenAI API Key** ready to configure ## Quick Start ### Option 1: Using Deployment Scripts (Recommended) **Windows PowerShell (Recommended for Windows users):** ```powershell .\deploy-cloud-run.ps1 -ProjectId "your-project-id" -Region "us-central1" ``` 📖 **For detailed Windows instructions, see [DEPLOY_CLOUD_RUN_WINDOWS.md](DEPLOY_CLOUD_RUN_WINDOWS.md)** **Linux/Mac:** ```bash chmod +x deploy-cloud-run.sh ./deploy-cloud-run.sh your-project-id us-central1 ``` ### Option 2: Manual Deployment Follow the step-by-step instructions below. ## Step-by-Step Deployment ### 1. Install and Configure Google Cloud SDK ```bash # Install gcloud CLI (if not already installed) # See: https://cloud.google.com/sdk/docs/install # Authenticate gcloud auth login # Set your project gcloud config set project YOUR_PROJECT_ID ``` ### 2. Enable Required APIs ```bash gcloud services enable cloudbuild.googleapis.com gcloud services enable run.googleapis.com gcloud services enable containerregistry.googleapis.com ``` ### 3. Build and Push Docker Image ```bash # Set variables PROJECT_ID="your-project-id" SERVICE_NAME="langchain-agent-mcp-server" IMAGE_NAME="gcr.io/${PROJECT_ID}/${SERVICE_NAME}" # Build the image docker build -t $IMAGE_NAME . # Push to Container Registry docker push $IMAGE_NAME ``` ### 4. Deploy to Cloud Run ```bash gcloud run deploy $SERVICE_NAME \ --image $IMAGE_NAME \ --platform managed \ --region us-central1 \ --allow-unauthenticated \ --memory 2Gi \ --cpu 2 \ --timeout 300 \ --max-instances 10 \ --min-instances 0 \ --set-env-vars "OPENAI_MODEL=gpt-4o-mini,MAX_ITERATIONS=10,VERBOSE=false" \ --port 8000 ``` ### 5. Configure OpenAI API Key **Option A: Using Environment Variable (Quick but less secure)** ```bash gcloud run services update $SERVICE_NAME \ --set-env-vars OPENAI_API_KEY=your-key-here \ --region us-central1 ``` **Option B: Using Secret Manager (Recommended for production)** 1. Create a secret: ```bash echo -n "your-openai-api-key" | gcloud secrets create openai-api-key \ --data-file=- \ --replication-policy="automatic" ``` 2. Grant Cloud Run access to the secret: ```bash PROJECT_NUMBER=$(gcloud projects describe $PROJECT_ID --format="value(projectNumber)") gcloud secrets add-iam-policy-binding openai-api-key \ --member="serviceAccount:${PROJECT_NUMBER}-compute@developer.gserviceaccount.com" \ --role="roles/secretmanager.secretAccessor" ``` 3. Update the service to use the secret: ```bash gcloud run services update $SERVICE_NAME \ --update-secrets=OPENAI_API_KEY=openai-api-key:latest \ --region us-central1 ``` ### 6. Verify Deployment ```bash # Get the service URL SERVICE_URL=$(gcloud run services describe $SERVICE_NAME \ --platform managed \ --region us-central1 \ --format 'value(status.url)') # Test health endpoint curl $SERVICE_URL/health # Test manifest endpoint curl $SERVICE_URL/mcp/manifest ``` ## Configuration Options ### Resource Allocation Adjust based on your needs: ```bash # For higher traffic or complex queries --memory 4Gi \ --cpu 4 \ --max-instances 20 # For cost optimization --memory 1Gi \ --cpu 1 \ --max-instances 5 \ --min-instances 0 # Scale to zero when not in use ``` ### Timeout Settings Cloud Run has a maximum timeout of 300 seconds (5 minutes). For longer-running agent tasks: ```bash --timeout 300 # Maximum allowed ``` ### Environment Variables Set additional environment variables: ```bash gcloud run services update $SERVICE_NAME \ --set-env-vars "OPENAI_MODEL=gpt-4,MAX_ITERATIONS=15,VERBOSE=true,API_KEY=your-api-key" \ --region us-central1 ``` ### CORS Configuration If you need to allow specific origins: ```bash gcloud run services update $SERVICE_NAME \ --set-env-vars "CORS_ORIGINS=https://yourdomain.com,https://app.yourdomain.com" \ --region us-central1 ``` ## Using Cloud Build (CI/CD) ### 1. Create cloudbuild.yaml The `cloudbuild.yaml` file is already included in the repository. It: - Builds the Docker image - Pushes to Container Registry - Deploys to Cloud Run ### 2. Set up Cloud Build Trigger ```bash # Create a trigger for GitHub gcloud builds triggers create github \ --name="deploy-langchain-mcp" \ --repo-name="LangchainMCP" \ --repo-owner="mcpmessenger" \ --branch-pattern="^main$" \ --build-config="cloudbuild.yaml" ``` ### 3. Set Secrets in Cloud Build ```bash # Store OpenAI API key as a secret echo -n "your-openai-api-key" | gcloud secrets create openai-api-key \ --data-file=- # Grant Cloud Build access PROJECT_NUMBER=$(gcloud projects describe $PROJECT_ID --format="value(projectNumber)") gcloud secrets add-iam-policy-binding openai-api-key \ --member="serviceAccount:${PROJECT_NUMBER}@cloudbuild.gserviceaccount.com" \ --role="roles/secretmanager.secretAccessor" ``` ### 4. Update cloudbuild.yaml to use secrets Add this step before the deploy step: ```yaml - name: 'gcr.io/google.com/cloudsdktool/cloud-sdk' entrypoint: 'bash' args: - '-c' - | gcloud run services update langchain-agent-mcp-server \ --update-secrets=OPENAI_API_KEY=openai-api-key:latest \ --region us-central1 ``` ## Monitoring and Logging ### View Logs ```bash # View recent logs gcloud run services logs read $SERVICE_NAME \ --platform managed \ --region us-central1 # Follow logs in real-time gcloud run services logs tail $SERVICE_NAME \ --platform managed \ --region us-central1 ``` ### Set up Monitoring 1. Go to Cloud Console → Cloud Run → Your Service 2. Click on "Monitoring" tab 3. Set up alerts for: - Request latency - Error rate - Memory usage - CPU usage ## Cost Optimization ### 1. Scale to Zero ```bash --min-instances 0 # Scales down when not in use ``` ### 2. Use Appropriate Resources Start with minimal resources and scale up as needed: - **Development**: 1 CPU, 1Gi memory - **Production**: 2 CPU, 2Gi memory - **High Traffic**: 4 CPU, 4Gi memory ### 3. Set Request Limits ```bash --max-instances 10 # Limit concurrent instances ``` ### 4. Use Cloud Run Pricing Calculator Estimate costs: https://cloud.google.com/run/pricing ## Troubleshooting ### Service Won't Start 1. Check logs: ```bash gcloud run services logs read $SERVICE_NAME --region us-central1 ``` 2. Verify environment variables: ```bash gcloud run services describe $SERVICE_NAME --region us-central1 ``` 3. Test locally with Docker: ```bash docker run -p 8000:8000 -e OPENAI_API_KEY=your-key gcr.io/$PROJECT_ID/$SERVICE_NAME ``` ### High Latency 1. Increase memory/CPU: ```bash gcloud run services update $SERVICE_NAME \ --memory 4Gi --cpu 4 --region us-central1 ``` 2. Check agent iteration limits: ```bash gcloud run services update $SERVICE_NAME \ --set-env-vars MAX_ITERATIONS=5 --region us-central1 ``` ### Authentication Issues If you need to restrict access: ```bash # Remove --allow-unauthenticated and use IAM gcloud run services update $SERVICE_NAME \ --no-allow-unauthenticated \ --region us-central1 # Grant access to specific users gcloud run services add-iam-policy-binding $SERVICE_NAME \ --member="user:email@example.com" \ --role="roles/run.invoker" \ --region us-central1 ``` ## Security Best Practices 1. **Use Secret Manager** for API keys (not environment variables) 2. **Enable VPC** if accessing private resources 3. **Set up IAM** policies for service access 4. **Enable Cloud Armor** for DDoS protection 5. **Use HTTPS only** (enabled by default) 6. **Set up API key authentication** in the application ## Next Steps - Set up custom domain: https://cloud.google.com/run/docs/mapping-custom-domains - Configure CDN: Use Cloud CDN with Cloud Run - Set up monitoring: Configure alerts in Cloud Monitoring - Enable tracing: Use Cloud Trace for request tracing ## Support For issues or questions: - Check Cloud Run logs - Review [Cloud Run documentation](https://cloud.google.com/run/docs) - Check application logs in Cloud Console