You migrated a data backend for an application that serves 10 PB of historical product data for analytics. Only the last known state for a product, which is about 10 GB of data, needs to be served…

Question

Options

• A
  1. Store the historical data in BigQuery for analytics.
• B
  1. Store the products as a collection in Firestore with each product having a set of historical changes.
• C
  1. Store the historical data in Cloud SQL for analytics.
• D
  1. Store the historical data in BigQuery for analytics.

• A
  79% (11)
• C
  14% (2)
• D
  7% (1)

Explanation

Option A correctly pairs BigQuery for the 10 PB analytics workload because BigQuery is purpose-built for petabyte-scale analytics at low cost using serverless columnar storage - it's the go-to GCP service when "large-scale analytics" appears in a question. The second part of Option A likely pairs BigQuery with Bigtable (or Cloud Spanner) for the 10 GB API layer, which handles 1,000+ QPS at sub-second latency efficiently.

Option B (Firestore) is wrong because Firestore is a document database optimized for transactional, real-time app data - not petabyte-scale analytical queries. Storing 10 PB of historical changes as Firestore documents would be prohibitively expensive and analytically impractical.

Option C (Cloud SQL) is wrong because Cloud SQL is a managed relational database suited for gigabyte-to-terabyte OLTP workloads - scaling it to 10 PB would be extremely costly and technically unsupported.

Option D is likely a distractor that pairs BigQuery with a wrong second component (e.g., Cloud SQL or Firestore for the API layer) rather than a low-latency, high-throughput NoSQL store like Bigtable.

Memory tip: Use the "two-bucket" pattern - BigQuery for big analytical cold data, Bigtable/Spanner for hot serving data requiring high QPS and low latency. When you see "PB-scale analytics + high-QPS API," think BigQuery + Bigtable.

No community discussion yet for this question.