This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
Choose the best response for each of the following questions.
A team runs one compact language model for each instance of an internal API. The model must start, stop, and scale with the API. Low-latency local calls are also important. Which hosting approach best fits these requirements?
Deploy the model as a separate shared service with its own scaling policy.
Run the model in an App Service sidecar beside the main API container.
Add the model process to the main API container image.
A sidecar-enabled app contains an API and a model server. External requests must reach only the API, while the API calls the model server locally. How should the container roles be configured?
Set isMain: true on the API and isMain: false on the model server.
isMain: true
isMain: false
Set isMain: true on both containers and assign different target ports.
Set isMain: false on both containers and configure WEBSITES_PORT for the API.
WEBSITES_PORT
A production web app must pull private main and sidecar images from an Azure Container Registry that uses RBAC Registry Permissions. The app can't store registry passwords. Which configuration meets the requirement?
Enable the registry admin account and store its credentials in DOCKER_REGISTRY_SERVER_* app settings.
DOCKER_REGISTRY_SERVER_*
Make the repositories public and omit authentication from each site container definition.
Assign a managed identity to the app, grant it AcrPull, and reference the identity in each private site container definition.
AcrPull
A model sidecar produces artifacts that must remain durable independently of the web app lifecycle and serve as a system of record for other applications. Where should the application store the artifacts?
Store the artifacts under the web app's shared /home directory.
/home
Use Azure Blob Storage or Azure Files as the durable storage boundary.
Write the artifacts to each container's writable image layer.
The main API receives a connection-refused error when it calls http://localhost:11434. Which diagnostic step most directly tests the failing boundary?
http://localhost:11434
Compare the API endpoint port with the sidecar target port, then inspect whether the sidecar process listens on that port.
Scale the App Service plan before inspecting the container configuration.
Change the API endpoint from localhost to the sidecar container name.
localhost
You must answer all questions before checking your work.
Was this page helpful?
Need help with this topic?
Want to try using Ask Learn to clarify or guide you through this topic?