You have an Azure App Service API that enables users to upload documents to the cloud with a mobile device. A mobile app connects to the service by using REST API calls. When a new document is uploaded to the service, the service extracts the document metadata. Usage statistics for the app show significant increases in app usage.

The extraction process is CPU$-$intensive. You have to modify the API to use a queue. You need to ensure that the solution scales, handles request spikes, and reduces costs between request spikes.

What should you do$?$

Configure a series of CPU$-$optimized virtual machine instances and install extraction logic to process a queue.

Move the extraction logic into an Azure Function. Create a queue$-$triggered function to process the queue.

Configure a CPU$-$optimized virtual machine and install the web app service on the new instance.

Configure an Azure Container Service to retrieve items from a queue and run across a pool of virtual machine nodes using the extraction logic.