Turning business requirements into a concrete storage design

Mapping business requirements to storage means matching data shape (structured, semi-structured, unstructured), access pattern (OLTP vs OLAP vs cache), scale, and consistency needs to the right managed service. The exam tests your ability to read a scenario and quickly rule out mismatched services based on throughput, latency, transactionality, and schema flexibility clues.

Must-know

• Cloud SQL suits traditional relational OLTP workloads that fit on a single instance (with read replicas) and need standard SQL with ACID transactions at moderate scale.
• Cloud Spanner is the choice when you need horizontal scalability, global distribution, strong consistency, and SQL with ACID transactions simultaneously, something Cloud SQL cannot do.
• Bigtable is built for high-throughput, low-latency NoSQL workloads (IoT, time-series, ad-tech) with petabyte-scale data, but only becomes cost-effective at large scale (generally recommended above ~1TB) and lacks multi-row transactions or joins.
• Firestore (Datastore mode/Native mode) targets semi-structured document data for web/mobile apps needing real-time sync, offline support, and flexible schemas rather than heavy analytical queries.
• BigQuery is the default for OLAP/analytical workloads on structured or semi-structured data at any scale, offering serverless SQL analytics but is not designed as a transactional, row-level update system despite supporting DML.
• Cloud Storage is the answer whenever requirements mention unstructured data, data lakes, backups, or staging before ETL/ELT, and it is priced/tiered (Standard, Nearline, Coldline, Archive) based on access frequency and retention needs.

Coverage checked against the published exam guide on Aug 11, 2026.

These are independent practice questions, written against this certification's published exam guide. They are not the certification vendor's own questions, and not the real exam.