AZ-305 Practice Question 35: On Premises Data Ingestion
Microsoft Azure Solutions Architect Expert (AZ-305) · Test 4 · question 5 of 10
A retailer wants to land nightly extracts from an on-premises SQL Server instance and an on-premises SAP HANA system into Azure Data Lake Storage Gen2, where Synapse serverless SQL pools will query them. The network team will not open any inbound firewall port and will not fund an ExpressRoute circuit or a site-to-site VPN for this project. The orchestration itself must be a managed Azure service with scheduling, retries, and monitoring.
What should you include in the design?
Show answer and explanation
Answer:
The self-hosted integration runtime is the component that runs copy activities between a cloud data store and a data store in a private network. Microsoft's guidance is explicit: if you want to perform data integration securely in a private network environment that has no direct line of sight from the public cloud, install a self-hosted IR behind the firewall, and it makes only outbound HTTP-based connections to the internet, so no inbound port is opened. It is also the recommended runtime for sources that need a bring-your-own driver, which SAP HANA does. Orchestration, scheduling, retries, and monitoring stay in the managed Data Factory service.
Option A points the wrong runtime at the problem. An Azure integration runtime connects to data stores with publicly accessible endpoints, or over Private Link to Azure resources in a managed virtual network; a private endpoint terminates in Azure and cannot reach a server sitting in the retailer's datacenter without hybrid connectivity, which is out of scope here.
Option C is for lifting and shifting existing SSIS packages, not for building a new pipeline, and it inherits the same problem: on-premises data access for an Azure-SSIS IR is supported by joining it to a virtual network that is connected to the on-premises network, the connectivity the retailer declined to fund.
Option D can move the files but gives up everything else. There is no managed scheduling, no retry or lineage, no monitoring, and no connector for SAP HANA; the extracts, the failure handling, and the alerting all become scripts the retailer maintains.
Reference: Integration runtime in Azure Data Factory