Exam DP-100 topic 6 question 3 discussion

B is the correct

The MLClient class is the primary entry point for SDK operations, allowing you to: 1. Connect to the Azure ML workspace. 2. Access deployment details, including logs. 3. Use methods like online_deployments.get_logs() to retrieve inference server logs, which include scoring script outputs. SSH/Docker tools (A/C): Not required for SDK-based log retrieval. OnlineDeploymentOperations (D): Accessed via ml_client.online_deployments, not instantiated directly.

https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-online-endpoints?view=azureml-api-2&tabs=python#get-container-logs https://learn.microsoft.com/en-us/python/api/azure-ai-ml/azure.ai.ml.operations.onlinedeploymentoperations?view=azure-python You should not instantiate this class directly. Instead, you should create an MLClient instance that instantiates it for you and attaches it as an attribute.

you create a OnlineDeploymentOperation by referencing 'ml_client.online_deployments'. OnlineDeploymentOperation has then the method 'get_logs(...)' References: https://learn.microsoft.com/en-us/python/api/azure-ai-ml/azure.ai.ml.mlclient?view=azure-python https://learn.microsoft.com/en-us/python/api/azure-ai-ml/azure.ai.ml.operations.onlinedeploymentoperations?view=azure-python#azure-ai-ml-operations-onlinedeploymentoperations-be

I think the answer should be B because you need an instance of ml_client class to call get_logs on online_deployment instance.

To review the container logs from an online endpoint by using the Azure ML Python SDK v2, you need to do the following steps: Create an instance of the OnlineDeploymentOperations class by passing the workspace object and the name of the online endpoint. Call the get_logs method on the OnlineDeploymentOperations object by passing the name of the deployment. This method returns a dictionary with the logs for the deployment, including the console log from the inference server and the print/log statements from the model’s scoring script.

chatgpt and bard say it should be D

Correct. ml_client is the instance for MLCLient class, and online_deployment is the instance for either ManagedOnlineDeployment class or KubernetesOnlineDeployment class. https://github.com/MicrosoftDocs/azure-docs/blob/main/articles/machine-learning/how-to-troubleshoot-online-endpoints.md#get-container-logs