You use BigQuery as your centralized analytics platform. New data is loaded every day, and an ETL pipeline modifies the original data and prepares it for the final users. This ETL pipeline is…

Question

Options

• AOrganize your data in a single table, export, and compress and store the BigQuery data in Cloud Storage.
• BOrganize your data in separate tables for each month, and export, compress, and store the data in Cloud Storage.
• COrganize your data in separate tables for each month, and duplicate your data on a separate dataset in BigQuery.
• DOrganize your data in separate tables for each month, and use snapshot decorators to restore the table to a time prior to the corruption.

• A
  9% (4)
• B
  77% (36)
• C
  4% (2)
• D
  11% (5)

Explanation

Organizing data in monthly tables limits recovery scope - when an error is detected after 2 weeks, you only need to reload and reprocess the affected month rather than an entire monolithic table. Exporting, compressing, and storing those monthly snapshots in Cloud Storage is the optimal cost strategy because Cloud Storage is significantly cheaper than keeping duplicate data in BigQuery itself, and the backups persist indefinitely regardless of when the error is discovered.

Why the distractors fail:

• A uses a single table, so recovery requires restoring and reprocessing all data rather than just the affected period - costly and risky.
• C duplicates monthly tables within BigQuery, which provides granular recovery but at full BigQuery storage prices, making it far more expensive than Cloud Storage - fails the "optimized for storage costs" requirement.
• D is disqualified by timing: BigQuery's time travel (snapshot decorators) has a maximum lookback window of 7 days, but the question states errors may not surface until 2 weeks (14 days) later - the snapshot will already be gone.

Memory tip: When you see "2 weeks" + "storage cost" in a BigQuery backup question, immediately eliminate any answer involving BigQuery-native snapshots (7-day limit) or BigQuery duplication (expensive). The winning pattern is always partition by time + export compressed to Cloud Storage.

No community discussion yet for this question.