LangChain Agent MCP Server

# Google Cloud Run Deployment Guide - Windows Complete Windows-focused 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 - Download: https://cloud.google.com/sdk/docs/install - Run the installer (it will add gcloud to PATH) - Restart PowerShell after installation 3. **Docker Desktop** installed and running - Download: https://www.docker.com/products/docker-desktop - Make sure Docker Desktop is running (whale icon in system tray) 4. **OpenAI API Key** ready to configure 5. **PowerShell** (comes with Windows) > **Note:** If you don't have `winget`, use the manual download links above. ## Quick Start (Windows) ### Step 1: Install Prerequisites **Install Google Cloud SDK:** ```powershell # Using winget (Windows 11/10 with App Installer) winget install Google.CloudSDK # Or download from: https://cloud.google.com/sdk/docs/install ``` **Install Docker Desktop:** ```powershell winget install Docker.DockerDesktop ``` **Verify installations:** ```powershell gcloud --version docker --version ``` ### Step 2: Authenticate with Google Cloud ```powershell # Login to Google Cloud gcloud auth login # Set your project gcloud config set project YOUR_PROJECT_ID # Verify gcloud config list ``` ### Step 3: Enable Required APIs ```powershell gcloud services enable cloudbuild.googleapis.com gcloud services enable run.googleapis.com gcloud services enable containerregistry.googleapis.com gcloud services enable secretmanager.googleapis.com ``` ### Step 4: Deploy Using PowerShell Script ```powershell # Navigate to project directory cd "C:\Users\senti\OneDrive\Desktop\Langchain MCP\restful-data-gateway-main" # Run deployment script .\deploy-cloud-run.ps1 -ProjectId "your-project-id" -Region "us-central1" ``` The script will: - ✅ Build the Docker image - ✅ Push to Google Container Registry - ✅ Deploy to Cloud Run - ✅ Show you the service URL ### Step 5: Configure OpenAI API Key **Option A: Using Secret Manager (Recommended)** ```powershell # Create the secret $apiKey = "your-openai-api-key-here" $apiKey | gcloud secrets create openai-api-key --data-file=- # Get project number $projectNumber = (gcloud projects describe $env:GOOGLE_CLOUD_PROJECT --format="value(projectNumber)") # Grant Cloud Run access gcloud secrets add-iam-policy-binding openai-api-key ` --member="serviceAccount:$projectNumber-compute@developer.gserviceaccount.com" ` --role="roles/secretmanager.secretAccessor" # Update the service to use the secret gcloud run services update langchain-agent-mcp-server ` --update-secrets=OPENAI_API_KEY=openai-api-key:latest ` --region us-central1 ``` **Option B: Using Environment Variable (Quick but less secure)** ```powershell gcloud run services update langchain-agent-mcp-server ` --set-env-vars OPENAI_API_KEY=your-key-here ` --region us-central1 ``` ### Step 6: Test Your Deployment ```powershell # Get the service URL $serviceUrl = gcloud run services describe langchain-agent-mcp-server ` --platform managed ` --region us-central1 ` --format "value(status.url)" # Test health endpoint Invoke-WebRequest -Uri "$serviceUrl/health" | Select-Object -ExpandProperty Content # Test manifest endpoint Invoke-WebRequest -Uri "$serviceUrl/mcp/manifest" | Select-Object -ExpandProperty Content ``` ## Manual Deployment (Step-by-Step) If you prefer to run commands manually: ### 1. Set Environment Variables ```powershell $env:GOOGLE_CLOUD_PROJECT = "your-project-id" $env:REGION = "us-central1" $env:SERVICE_NAME = "langchain-agent-mcp-server" $env:IMAGE_NAME = "gcr.io/$env:GOOGLE_CLOUD_PROJECT/$env:SERVICE_NAME" ``` ### 2. Build Docker Image ```powershell docker build -t $env:IMAGE_NAME . ``` ### 3. Push to Container Registry ```powershell docker push $env:IMAGE_NAME ``` ### 4. Deploy to Cloud Run ```powershell gcloud run deploy $env:SERVICE_NAME ` --image $env:IMAGE_NAME ` --platform managed ` --region $env:REGION ` --allow-unauthenticated ` --memory 2Gi ` --cpu 2 ` --timeout 300 ` --max-instances 10 ` --min-instances 0 ` --set-env-vars "OPENAI_MODEL=gpt-4o-mini,MAX_ITERATIONS=10,VERBOSE=false" ` --port 8000 ``` ## Windows-Specific Tips ### Using PowerShell vs Command Prompt **PowerShell (Recommended):** - Better error handling - More modern syntax - Better integration with gcloud **Command Prompt:** - Use `set` instead of `$env:` - Use `^` for line continuation instead of backtick ### Docker Desktop on Windows Make sure Docker Desktop is running before building: ```powershell # Check if Docker is running docker ps # If not running, start Docker Desktop from Start Menu ``` ### Path Issues If you get "command not found" errors: ```powershell # Add gcloud to PATH (usually done automatically, but verify) $env:PATH += ";C:\Program Files (x86)\Google\Cloud SDK\google-cloud-sdk\bin" # Or restart PowerShell after installing gcloud ``` ### Line Continuation in PowerShell Use backtick (`) for line continuation: ```powershell gcloud run deploy langchain-agent-mcp-server ` --image gcr.io/project/image ` --region us-central1 ``` ## Troubleshooting (Windows) ### Docker Build Fails **Issue:** "Cannot connect to Docker daemon" ```powershell # Solution: Make sure Docker Desktop is running # Check Docker status docker info ``` **Issue:** "Build context is too large" ```powershell # Solution: Check .dockerignore file # Make sure large files are excluded ``` ### gcloud Command Not Found **Issue:** PowerShell can't find gcloud ```powershell # Solution 1: Restart PowerShell after installation # Solution 2: Add to PATH manually $env:PATH += ";C:\Program Files (x86)\Google\Cloud SDK\google-cloud-sdk\bin" # Solution 3: Use full path & "C:\Program Files (x86)\Google\Cloud SDK\google-cloud-sdk\bin\gcloud.cmd" --version ``` ### Authentication Issues **Issue:** "You do not currently have an active account selected" ```powershell # Solution: Re-authenticate gcloud auth login gcloud auth application-default login ``` ### Permission Denied Errors **Issue:** "Permission denied" when pushing images ```powershell # Solution: Configure Docker to use gcloud as credential helper gcloud auth configure-docker ``` ### PowerShell Execution Policy **Issue:** "Cannot run script because execution policy" ```powershell # Solution: Allow script execution (run as Administrator) Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser # Or run script with bypass powershell -ExecutionPolicy Bypass -File .\deploy-cloud-run.ps1 ``` ## Viewing Logs (Windows) ```powershell # View recent logs gcloud run services logs read langchain-agent-mcp-server ` --platform managed ` --region us-central1 # Follow logs in real-time gcloud run services logs tail langchain-agent-mcp-server ` --platform managed ` --region us-central1 ``` ## Updating the Service ```powershell # Update environment variables gcloud run services update langchain-agent-mcp-server ` --set-env-vars "OPENAI_MODEL=gpt-4,MAX_ITERATIONS=15" ` --region us-central1 # Update resources gcloud run services update langchain-agent-mcp-server ` --memory 4Gi ` --cpu 4 ` --region us-central1 # Update image (after rebuilding) gcloud run services update langchain-agent-mcp-server ` --image gcr.io/$env:GOOGLE_CLOUD_PROJECT/langchain-agent-mcp-server:latest ` --region us-central1 ``` ## Cost Optimization ### Scale to Zero (Save Money When Not in Use) ```powershell gcloud run services update langchain-agent-mcp-server ` --min-instances 0 ` --region us-central1 ``` ### Set Maximum Instances ```powershell gcloud run services update langchain-agent-mcp-server ` --max-instances 5 ` --region us-central1 ``` ## Quick Reference Commands ```powershell # Get service URL gcloud run services describe langchain-agent-mcp-server --region us-central1 --format "value(status.url)" # List all services gcloud run services list --region us-central1 # Delete service gcloud run services delete langchain-agent-mcp-server --region us-central1 # View service details gcloud run services describe langchain-agent-mcp-server --region us-central1 ``` ## Next Steps 1. ✅ Set up monitoring in Cloud Console 2. ✅ Configure custom domain (optional) 3. ✅ Set up CI/CD with Cloud Build (optional) 4. ✅ Review cost estimates in Cloud Console ## Getting Help - **Cloud Run Docs**: https://cloud.google.com/run/docs - **gcloud CLI Reference**: https://cloud.google.com/sdk/gcloud/reference - **View logs in Console**: https://console.cloud.google.com/run --- **Ready to deploy?** Run: `.\deploy-cloud-run.ps1 -ProjectId "your-project-id"`