GCP-ADP Data Preparation And Ingestion Practice Test 3 — 15 Questions

Question 1

You are using Cloud Data Fusion to build a pipeline that reads from Cloud Storage and writes to BigQuery. You need to ensure that the pipeline can handle schema evolution, where new columns are added to the source files over time. Which TWO approaches should you use? (Choose two.)

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Set the BigQuery sink to use the 'Update Schema' operation.. To handle schema evolution in Cloud Data Fusion when writing to BigQuery, enable 'Allow Schema Drift' on the BigQuery sink and set the operation to 'Update Schema'. These options allow the pipeline to automatically detect new columns and update the BigQuery table schema, ensuring that new data is ingested without manual intervention.

Answer

Use a Wrangler transform to manually add new columns to the schema.

Answer

Configure the source plugin to infer schema from the first file only.

Answer

Use a JavaScript transform to dynamically add new columns to each record.

Question 2

You are preparing data for analytics in BigQuery. You have a Cloud Storage bucket containing JSON files with inconsistent schemas and nested arrays. You need to load this data into BigQuery while preserving the nested structure and minimizing manual schema definition. Which TWO approaches should you use? (Choose two.)

Accepted Answer

Use the bq load command with the --autodetect flag and specify --source_format=NEWLINE_DELIMITED_JSON.. BigQuery can infer schemas for newline-delimited JSON, including nested and repeated fields, through autodetect. Using the bq load command with NEWLINE_DELIMITED_JSON and --autodetect applies that capability from the command line, so you can load JSON from Cloud Storage while preserving nested arrays and avoiding hand-written schema definitions. These two approaches are essentially the same autodetect capability accessed via the console or CLI, and both satisfy the requirements.

Answer

Create an external table over the JSON files and query them directly with SQL.

Answer

Define the full BigQuery schema manually with RECORD and REPEATED types for every nested field.

Answer

Use the Cloud Data Fusion Wrangler to transform the JSON into a flat CSV before loading.

Question 3

A logistics company ingests GPS pings from delivery vehicles into Pub/Sub. A Dataflow streaming pipeline writes the raw messages to BigQuery using Storage Write API with exactly-once semantics, and a separate pipeline aggregates them. The team notices duplicate rows in the raw BigQuery table during pipeline restarts. They confirm the pipeline uses the Storage Write API with exactly-once enabled. What is the most likely cause?

Accepted Answer

The pipeline is writing to BigQuery using streaming inserts instead of the Storage Write API exactly-once path.. The Storage Write API exactly-once feature prevents duplicate rows in BigQuery only when the sink is actually configured to use it. If the pipeline falls back to legacy streaming inserts or another write method, retries after a restart can append duplicates. Partitioning and worker machine type are unrelated to write deduplication.

Answer

The BigQuery table does not have a partitioning column defined, so rows are appended without deduplication.

Answer

The Dataflow workers are using the default machine type, which cannot maintain exactly-once state.

Answer

The Pub/Sub subscription is using at-least-once delivery, so duplicates are expected and should be deduplicated downstream.

Question 4

You are designing a data ingestion pipeline that must handle late-arriving data in a streaming pipeline. Events can arrive up to 2 hours after their event timestamp. You need to ensure that windows are triggered only after all data for a given window has arrived, but you also want to limit resource usage. What should you configure in Dataflow?

Accepted Answer

Set a fixed window of 1 hour with a allowed lateness of 2 hours and use the default trigger (watermark-based).. The default watermark-based trigger in Dataflow fires when the watermark passes the end of the window, which is delayed by the allowed lateness setting. By setting allowed lateness to 2 hours, the watermark is held back, so windows trigger only after all data up to 2 hours late has arrived. This ensures completeness and avoids unnecessary early firings, limiting resource usage. Other triggers or window types would either fire early or not handle late data correctly.

Answer

Set a session window with a gap duration of 2 hours and no allowed lateness.

Answer

Set a sliding window of 1 hour with a allowed lateness of 2 hours and trigger every 5 minutes.

Answer

Set a fixed window of 1 hour with a allowed lateness of 2 hours and trigger with a repeating interval of 1 minute.

Question 5

A financial analytics team needs to move several terabytes of on-premises Oracle tables into BigQuery each night. The source database cannot accept inbound connections from Google Cloud, and the team wants minimal operational overhead for the transfer. They plan to use the BigQuery Data Transfer Service. Which connector should they configure?

Accepted Answer

A BigQuery Data Transfer Service Oracle connector configured with an agent installed in the on-premises environment.. The source database cannot accept inbound connections, so the transfer must be initiated from inside the on-premises network. The BigQuery Data Transfer Service database connector uses an agent that makes outbound connections, satisfying the network constraint while providing a managed, scheduled load into BigQuery. Cloud Storage transfers do not parse Oracle data, federated queries need inbound reachability, and a Pub/Sub streaming design adds unnecessary components for a nightly batch job.

Answer

A federated query from BigQuery that uses an external connection to read the Oracle tables directly.

Answer

A Pub/Sub subscription that receives change events streamed from the Oracle database by a third-party tool.

Answer

A Cloud Storage transfer that reads database export files placed in a bucket by an on-premises script.

Question 6

You need to store sensitive data in GCS. Which feature ensures data is encrypted at rest?

Accepted Answer

Default Encryption. Google Cloud Storage automatically encrypts all data at rest using AES-256 encryption before it is written to disk, with no configuration required. This default encryption ensures that sensitive data is protected without any additional action from the user. While CMEK and Cloud DLP offer additional control or inspection, the fundamental guarantee of encryption at rest is provided by default encryption.

Answer

Identity and Access Management

Answer

Cloud DLP

Answer

Customer-Managed Encryption Keys (CMEK)

Question 7

Which GCS storage class is most cost-effective for data accessed only once per year?

Accepted Answer

Archive. Archive is the most cost-effective GCS storage class for data accessed only once per year because it is designed for long-term storage of data that is accessed less than once a year. Archive storage has the lowest storage cost among GCS classes, but it also has the highest retrieval cost and a minimum storage duration of 365 days. For data accessed only once per year, Archive provides the best cost savings.

Answer

Standard

Answer

Coldline

Answer

Nearline

Question 8

You are designing a streaming pipeline in Dataflow. You need to ensure that data is processed in the order it was generated, even if it arrives late. What do you use?

Accepted Answer

Watermarks and Allowed Lateness. Watermarks track the progress of event time in a streaming pipeline, and Allowed Lateness defines how long after the watermark passes a window will still accept late-arriving elements. Together they let Dataflow emit correct, event-time-ordered results while tolerating out-of-order arrivals. Without allowed lateness, late data is dropped once the watermark advances past the window.

Answer

Processing time triggers

Answer

Dataflow shuffle service

Answer

Fixed windows only

Question 9

You need to perform data cleansing on a large dataset in Dataprep. Which execution engine should you choose for scalability?

Accepted Answer

Dataflow. Dataflow is the correct execution engine for Dataprep because Dataprep is built on top of Cloud Dataflow, which provides distributed, autoscaling processing for large datasets. Dataflow handles the parallel execution of Dataprep recipes across many workers, so cleansing operations scale horizontally without manual cluster management. Choosing Dataflow ensures the job can handle terabyte-scale data efficiently.

Answer

Local Desktop

Answer

BigQuery

Answer

Cloud Run

Question 10

You need to ingest data from an external API that requires an OAuth token. Which tool is best for executing this periodically?

Accepted Answer

Cloud Functions. Cloud Functions is the best fit because it can run scheduled (via Cloud Scheduler), execute custom code to perform the OAuth token exchange, call the external API, and write results to a destination like BigQuery or Cloud Storage. It supports secrets via Secret Manager and scales automatically. This makes it ideal for periodic, code-driven API ingestion requiring authentication.

Answer

Pub/Sub

Answer

BigQuery Data Transfer Service

Answer

Storage Transfer Service

Question 11

You are processing streaming data in Dataflow and notice 'stuck' elements causing pipeline latency. Which feature helps debug this?

Accepted Answer

Dataflow Monitoring UI. The Dataflow Monitoring UI provides detailed visibility into pipeline execution, including per-stage metrics, watermarks, and the ability to identify stuck or slow elements causing latency. It surfaces system lag, data freshness, and hot keys that indicate bottlenecks. This makes it the right tool for debugging streaming pipeline latency.

Answer

Dataprep Jobs view

Answer

Cloud Logging

Answer

Cloud Trace

Question 12

You need to set up an alert when a GCS bucket exceeds a certain size. Which tool do you use?

Accepted Answer

Cloud Monitoring. Cloud Monitoring is the GCP service responsible for collecting metrics, creating alerting policies, and notifying on threshold breaches. GCS automatically exports storage metrics (such as storage.googleapis.com/storage/total_bytes) to Cloud Monitoring, so you can build an alerting policy on bucket size and route it to email, Pub/Sub, or PagerDuty. This is the canonical way to alert on resource metrics in GCP.

Answer

Cloud Build

Answer

IAM

Answer

Cloud Logging

Question 13

You need to transfer files from an Amazon S3 bucket to GCS. What is the most efficient method?

Accepted Answer

Storage Transfer Service. Storage Transfer Service is a fully managed Google Cloud service designed for high-performance, scalable transfers from Amazon S3 to Cloud Storage. It handles scheduling, retries, and data integrity checks, making it the most efficient method for this use case.

Answer

Dataflow

Answer

gsutil rsync

Answer

Write a Python script

Question 14

You are migrating a high-throughput on-premises database to BigQuery. You need to ensure zero downtime. What is your best strategy?

Accepted Answer

Datastream. Datastream is a serverless change data capture (CDC) service that replicates data from on-premises databases to BigQuery with minimal downtime. It continuously streams changes, ensuring zero downtime during migration.

Answer

Cloud Pub/Sub

Answer

Dataflow batch jobs

Answer

Batch Export/Import

Question 15

You need to provide temporary access to a GCS file for a third party. What should you use?

Accepted Answer

Signed URL. A signed URL provides temporary, time-limited access to a specific Cloud Storage object, granting the bearer permission to perform a specific action (e.g., GET) for a limited duration. This is ideal for sharing a file with a third party without granting them a Google account or IAM role. Signed URLs are generated using a service account key or by a user with sufficient permissions, and they include authentication information in the query string.

Answer

Public access

Answer

Bucket Policy

Answer

IAM Role