GCP Data Analysis & Reporting 5

Question 1

Which BigQuery feature allows you to share datasets with other Google Cloud projects without copying data?

Accepted Answer

Analytics Hub

Answer

Analytics Hub enables organizations to publish and subscribe to shared datasets (data exchanges) across projects without duplicating the underlying data.

Question 2

In a Dataflow streaming pipeline, what does 'windowing' accomplish?

Accepted Answer

It groups unbounded streaming data into finite chunks for aggregation

Answer

Windowing divides an unbounded data stream into finite time-based or count-based buckets so aggregations (sums, averages) can be computed over streaming data.

Question 3

You notice a BigQuery query scanning 500GB but only returning 100 rows. What is the most effective optimization?

Accepted Answer

Add a WHERE clause filtering on a partitioned or clustered column

Answer

Filtering on a partition or cluster column prunes the data scanned at the storage layer, dramatically reducing bytes processed and query cost.

Question 4

Which BigQuery ML model type would you use to predict a continuous numerical value, such as next month's sales revenue?

Accepted Answer

Linear regression

Answer

Linear regression in BigQuery ML predicts a continuous output variable based on input features, making it appropriate for numerical forecasting tasks.

Question 5

What is the purpose of the INFORMATION_SCHEMA views in BigQuery?

Accepted Answer

To provide metadata about datasets, tables, jobs, and schema definitions

Answer

INFORMATION_SCHEMA is a set of system views that expose metadata about BigQuery resources including table schemas, column details, job history, and storage statistics.

GCP Test Practice Test

GCP Test Practice Test

GCP Data Analysis & Reporting 5