While preparing for the GCP Associate Cloud Engineer (ACE) exam, many candidates get confused by database service selection. In the real world, this is fundamentally a decision about balancing strong consistency and transactional integrity against ease of migration and operational overhead. Let’s drill into a simulated scenario.
The Scenario #
FinCentric, a fast-growing global fintech startup, manages a critical internal application for processing transactional orders. This app is exclusively accessed by users at the company’s New York headquarters. The app requires strict ACID compliance and strong consistency across multiple tables, ensuring all financial transactions are flawless and real-time. The initial implementation is on PostgreSQL.
Key Requirements #
Migrate this transactional order management system to the cloud with minimal code changes, preserving transactional properties and query speed.
The Options #
- A. BigQuery
- B. Cloud SQL
- C. Cloud Spanner
- D. Cloud Datastore
Correct Answer #
B. Cloud SQL
The Architect’s Analysis #
Correct Answer #
Option B: Cloud SQL.
Step-by-Step Winning Logic #
Cloud SQL provides a fully managed PostgreSQL-compatible relational database that delivers ACID compliance and strong consistency out of the box. It supports multi-table transactions without code changes and offers a familiar environment for developers already using PostgreSQL. Cloud SQL handles backups, patching, and failover with minimal operational effort, aligning with the principle of “Managed Services first” in SRE. Given the app’s tight consistency and transactional requirements in a single location, Cloud SQL fits perfectly without the complexity or overhead of distributed multi-region databases.
The Traps (Distractor Analysis) #
-
Why not A) BigQuery?
BigQuery is designed for large-scale analytics and OLAP queries, not transactional workloads with multi-row ACID guarantees. It lacks strong consistency and fine-grained multi-table transactions needed here. -
Why not C) Cloud Spanner?
While Cloud Spanner offers global strongly consistent transactions, it introduces unnecessary complexity and cost for a single-physical-location app with limited scale. Also, some PostgreSQL features may require code changes. -
Why not D) Cloud Datastore?
Cloud Datastore is a NoSQL document database optimized for scalability and eventual consistency, which violates the strict ACID and strong consistency needs.
The Architect Blueprint #
Mermaid Diagram illustrating Cloud SQL replacing on-premise PostgreSQL with managed operations:
Diagram Note:
This shows users at a single site interacting with the cloud-hosted managed PostgreSQL instance, preserving transactional integrity with minimal app changes.
Real-World Practitioner Insight #
Exam Rule #
For the ACE exam, always pick Cloud SQL when you see strong ACID transactional requirements closely tied to PostgreSQL workloads in a single region or physical site.
Real World #
In production, if the business later requires horizontal scaling across multiple regions or global strong consistency, Cloud Spanner might be considered. But at early stages, Cloud SQL balances cost, operational simplicity, and developer productivity. Avoid premature optimization into distributed systems when your workload fits a managed relational database.