Skip to content

Step 11: Monitor Cloud Orchestration with DTS Dashboard

Now that the application is deployed to Azure, you can monitor the multi-agent orchestration running in the cloud using the Durable Task Scheduler (DTS) dashboard.

  1. Generate the DTS Dashboard URL:

    The repository includes helper scripts to generate the DTS dashboard URL. In your VS Code terminal (ensure you're in the repository root directory), run:

    ./scripts/get-dts-dashboard-url.sh

    The script will output a URL similar to:

    https://dashboard.durabletask.io/subscriptions/{subscription-id}/schedulers/{dts-name}/taskhubs/{taskhub-name}?endpoint={encoded-endpoint}&tenantId={tenant-id}

  2. Copy and open the dashboard URL:

    • Copy the entire URL from the terminal output
    • Open it in your browser (you may need to sign in with your Azure credentials)
    • You should see the Durable Task Scheduler dashboard for your deployed application

  3. Trigger a Cloud Orchestration:

    Back in VS Code, open GitHub Copilot Chat and run the multi-agent orchestration against your deployed function. Use this exact prompt to ensure it calls the comprehensive documentation tool:

    #remote-snippy use the generate_comprehensive_documentation tool to create documentation and save it in a new file called cloud-documentation.md
    • Select Allow when Copilot asks to use the generate_comprehensive_documentation tool
    • This time the orchestration will run on the Azure Durable Task Scheduler service in the cloud

  4. Monitor the Orchestration in the Dashboard:

    • Switch to the browser tab with the DTS dashboard
    • You should see a new orchestration instance appear
    • Watch as it progresses through the same stages you saw locally:
      • DeepWikiAgent initial documentation generation
      • DeepWikiAgent refinement pass
      • CodeStyleAgent style guide generation
    • Select the orchestration instance to view detailed execution history:
      • Timeline of agent calls and durations
      • Agent outputs and intermediate results
      • Vector search tool invocations
      • State transitions throughout the workflow

  5. Review the Generated Documentation:

    Once the orchestration completes (1-2 minutes), check your workspace:

    • Open cloud-documentation.md in the project root
    • Compare it with the comprehensive-documentation.md generated during local testing
    • Both files demonstrate the same multi-agent workflow, but one ran locally and one in Azure

What You've Demonstrated:

You've now run the same multi-agent orchestration in two environments:

  • Local: Using Docker-based Durable Task emulator for development
  • Cloud: Using Azure Durable Task Scheduler for production

The DTS dashboard provides:

  • Real-time monitoring of orchestration progress
  • Detailed execution history and debugging information
  • Visibility into agent calls, tool invocations, and state management
  • The ability to track and troubleshoot complex multi-agent workflows

This demonstrates how Durable Functions provides a consistent development experience from local to cloud, with enterprise-grade observability for AI agent orchestrations.