Microsoft Azure Architect Design (AZ-304) Practice Test

The self-hosted integration runtime in Azure is the correct choice for transforming data files on a virtual machine before loading them into Azure Data Lake Storage. This runtime allows you to connect on-premises data sources with cloud-based services, enabling you to move and transform data securely.

When you need to perform data transformations on files locally before they are sent to Azure, the self-hosted integration runtime can execute the necessary data workflows and transformations directly on the virtual machine. It operates in your network environment, which provides the flexibility needed to handle any custom transformations or pre-processing of the data files.

In contrast, while the Azure Pipelines agent can manage and deploy tasks in a CI/CD pipeline, it is primarily focused on continuous integration and delivery workflows rather than data transformation tasks. The Azure File Sync agent is designed for syncing files between on-premises and Azure, but it does not typically handle data transformations. The On-premises data gateway allows for secure data transfer between on-premises data sources and Azure services but is not designed specifically for file transformations.

This is why the self-hosted integration runtime is the most suitable option for the scenario described.