2.1 Planning the Data Pipelines PROFESSIONAL-DATA-ENGINEER Practice Questions (Page 2)

Questions 6–10

1. 6foundation · easy

   A data engineer is evaluating a new data source for a pipeline. The source is an on-premises relational database that must be queried in real time to capture changes as they occur. Which type of data source is this?

   • AA database source
   • BA batch file source
   • CAn API source
   • DA streaming source
   Reveal Solution Discussion Select an answer first
2. 7foundation · easy

   A data pipeline runs on a VM in a VPC. To allow the VM to communicate with a database in the same VPC, what is required?

   • AA VPN connection to the database
   • BA public IP address on the VM
   • CA firewall rule allowing traffic between the VM and the database
   • DA NAT gateway
   Reveal Solution Discussion Select an answer first
4. 8foundation · easy

   A pipeline has two steps: Step A must complete successfully before Step B can start. What is this relationship called?

   • AA schedule
   • BA retry policy
   • CAn error handler
   • DA dependency
   Reveal Solution Discussion Select an answer first
5. 9expert · hard

   A data engineering team is building a pipeline to process IoT sensor data from millions of devices. The data is ingested via Pub/Sub and processed in Dataflow. The pipeline must (1) aggregate temperature readings per device per minute, (2) detect anomalies where the temperature changes by more than 10 degrees within a minute, and (3) store both the raw and aggregated data. The team is debating whether to use a single Dataflow pipeline or two separate pipelines. Which consideration is most important in this decision?

   • ATwo separate pipelines are required because Dataflow cannot perform both aggregation and anomaly detection in a single job.
   • BA single pipeline is always more cost-effective because it avoids duplicate reads from Pub/Sub.
   • CA single pipeline can share state between the aggregation and anomaly detection, reducing the need for a separate state store.
   • DTwo separate pipelines allow independent scaling and failure isolation, but require a second read from Pub/Sub and duplicate processing.
   Reveal Solution Discussion Select an answer first
6. 10expert · hard

   A company is migrating its data pipeline to Google Cloud. The pipeline currently runs on-premises and reads from a database in a private subnet. The company has set up a Cloud VPN tunnel to Google Cloud. They plan to run Dataflow jobs in a VPC that is peered with the VPN. The Dataflow workers need to access the on-premises database. Which configuration is required to allow Dataflow workers to reach the on-premises database?

   • AUse a Cloud Interconnect instead of a VPN to ensure lower latency for the Dataflow workers.
   • BConfigure the Dataflow workers to use the VPC's private IP range and ensure the on-premises firewall allows traffic from that range.
   • CSet up a Cloud NAT gateway in the VPC to allow Dataflow workers to reach the on-premises database.
   • DEnsure the Dataflow workers are assigned external IP addresses so they can route through the VPN.
   Reveal Solution Discussion Select an answer first