Edit

Deploy a C# Azure AI Search app to Azure Container Apps

Note

Azure AI Search is available through the Azure portal, REST APIs, and Azure SDKs. It also underpins Foundry IQ, the managed knowledge layer that transforms enterprise content into reusable, permission-aware knowledge bases for agents in the Microsoft Foundry portal.

Deploy the search-enabled website to Azure Container Apps by using the Azure Developer CLI (azd). This deployment includes the React client, the .NET API, Azure AI Search, managed identity, and supporting Azure resources.

The deployment uses managed identity as the default authentication method for Azure AI Search. You can opt in to key-based authentication by setting USE_KEYLESS_AUTH to false before deployment.

Sign in to Azure

  1. In Visual Studio Code, open a terminal at the repository root, such as azure-search-static-web-app.

  2. Sign in to Azure by using azd.

    azd auth login

  3. If prompted by your browser, complete the sign-in flow by using the Azure account that you're using for this tutorial.

Create an azd environment

  • Create an azd environment. Replace YOUR-ENVIRONMENT-NAME with a short name that identifies this tutorial deployment.

    azd env new YOUR-ENVIRONMENT-NAME

(Optional) Configure key-based authentication

The default authentication method is managed identity, so most environments can skip this section. Complete this section only if your environment requires API keys.

Set USE_KEYLESS_AUTH to false.

azd env set USE_KEYLESS_AUTH false

If you don't set USE_KEYLESS_AUTH, the deployment uses managed identity. The Bicep infrastructure grants the container app identity access to the Azure AI Search data plane.

Deploy with azd up

  1. From the repository root, run azd up.

    azd up

  2. When prompted, select the Azure subscription and location for the resources.

  3. Wait for azd to provision the infrastructure, build the containers, push the images, and deploy the services to Azure Container Apps.

  4. Review the output after deployment. The output includes the client and server fully qualified domain names (FQDNs) for the Azure Container Apps deployment.

    AZURE_CLIENT_URL: <client-container-app-fqdn>
AZURE_SERVER_URL: <server-container-app-fqdn>
SEARCH_SERVICE_NAME: <search-service-name>

    Use the AZURE_CLIENT_URL value to open the website. The client app calls the server app through the endpoint configured during deployment.

    The deployment also created the Azure AI Search service and loaded the good-books index automatically through the postprovision hook in the sample repository.

Use search in your Container Apps website

  1. Open the client FQDN from the azd up output in a browser.

  2. In the website search bar, enter a search query, such as code. The autocomplete feature suggests matching book titles.

  3. Select a suggestion or continue entering your own query. Select Enter when you finish your search query.

  4. Review the results, and then select a book to see more details.

Troubleshooting

If the web app doesn't deploy or work, use the following list to determine and fix the issue:

  • Did azd up complete?

    Review the azd up output for the first failed provisioning, build, or deploy step. If deployment stops during provisioning, check your subscription permissions and the selected region. If deployment stops during the container build or deploy steps, rerun azd up after you fix the reported issue.

  • Can you open the client endpoint?

    Open the AZURE_CLIENT_URL value from the azd up output. If the page doesn't load, go to the Azure portal, open the client container app and review its revision status and logs.

  • Can the client call the server endpoint?

    Open your browser developer tools and review the network calls from the client app. If calls to the API fail, confirm that the server container app is running and that the client app received the server FQDN during deployment.

  • Can the server query Azure AI Search?

    If searches return errors, review the server container app logs. For managed identity, confirm that the identity has Azure AI Search data-plane access. For key-based authentication, confirm that you set USE_KEYLESS_AUTH to false before deployment and that the generated container environment contains the search service configuration.

Clean up resources

Use azd down --purge to delete the Azure resources that this tutorial created. The --purge flag permanently removes resources that support soft-delete, instead of leaving them in a recoverable state for their retention period. Purging avoids naming conflicts if you redeploy this tutorial later.

  1. From the repository root, run the clean-up command.

    azd down --purge

  2. Review the resources that azd lists.

  3. Confirm the deletion when prompted.

If you created an Azure AI Search service outside the azd deployment, go to your search service in the Azure portal and select Delete at the top of the page.

Next step

Explore the search integration code

  • Last updated on