Exam AZ-204 topic 2 question 55 discussion

ACI - Because the GPU usage. Kubernetes can manage ACIs but is not a 'compute target' and it will increment the cost. "The container instances in the group can access one or more NVIDIA Tesla GPUs while running container workloads such as CUDA and deep learning applications." https://learn.microsoft.com/en-us/azure/container-instances/container-instances-gpu Restart Policy - To stop the container after the execution (In fact is avoiding to restart it after a succeeded execution) "Set an appropriate restart policy for the container instance, depending on whether the command-line specifies a long-running task or a run-once task. For example, a restart policy of Never or OnFailure is recommended for a run-once task." https://learn.microsoft.com/en-us/azure/container-instances/container-instances-start-command#command-line-guidelines

Got it in exam on 13.03.2024. Score: 910. Answer ACI, restart policy

Got this in the exam today! May 15, 2025. Went most voted. 940/1000. 50 of 51 from examtopics. Cases Munson’s Pickles and Preserves Farm (7) and Fourth Coffee (3)

Azure Container Instance - Since scaling is not an requirement, the right choice would be to use ACI instead of ACA or Kubernetes.

To deploy a Python application for image rendering with the specified requirements on Azure, you can use Azure Container Instances (ACI). ACI provides an easy and cost-effective way to run containers in the cloud. Here are the options you can choose: Compute Value: Choose the ACI option for running your Linux container. Container Termination: Since you want the container to be stopped when the image rendering is complete to minimize cost, you can use the "Single" container group mode in Azure Container Instances. This mode is suitable for scenarios where you want to run a single container task and stop it when it's done. This command creates a single container instance (--restart-policy Never) and stops it when the image rendering is complete, minimizing cost. Adjust the resource limits (--cpu and --memory) based on your application's requirements.

Will this question be on the exam?

As suggested ACI is preferred for one time or demand script and jobs execution whereas as ACA is preferred for deploying container continually running full lifecycle apps, I am leaning towards ACI.

If I choose C# in the exam, will this question still appear?

given that AKS is not in the exam curriculum , only ACA and ACI are left. out of which, only ACI supports GPU (though still in preview). As for the container termination-> restart policy set to never should be true