Study Google Cloud ACE Storage and Database Choices: key concepts, common traps, and exam decision cues.
This lesson covers one of the most common ACE judgment calls: which storage or database service matches the workload? Google Cloud expects you to choose based on access pattern, scale, structure, and retention behavior instead of memorizing product names blindly.
Cold storage class: Lower-cost storage tier intended for infrequently accessed objects with different retrieval trade-offs.
Operational database: Data service optimized for live application reads and writes rather than warehouse-style analytics.
Google Cloud wants you to distinguish:
| Need | Strongest first lane | Why it fits |
|---|---|---|
| Large-scale analytical SQL over managed datasets | BigQuery | Analytics warehouse path |
| Relational application data with familiar SQL engine behavior | Cloud SQL | Managed relational database |
| Globally scaled relational consistency and high-end transactional design | Spanner | Strongly consistent globally distributed relational system |
| Document-shaped app data with flexible schema and app-facing reads | Firestore | Document database path |
| Massive key-value or wide-column throughput | Bigtable | High-scale operational data pattern |
| Durable files, objects, backups, or archives | Cloud Storage | Object storage, not database query engine |
| Boot or attached block storage for VMs | Persistent Disk | VM storage, not shared object or database service |
| If the question says | Think first about |
|---|---|
| files, media, backups, archives, static objects | Cloud Storage |
| SQL application, transactions, joins, familiar relational admin model | Cloud SQL |
| app records shaped like documents, flexible schema, mobile or web app backing store | Firestore |
| huge analytical datasets and reporting queries | BigQuery |
| extreme scale key-value or wide-column access | Bigtable |
ACE likes to mix two different decisions:
| Decision type | Example |
|---|---|
| Product choice | BigQuery versus Cloud SQL |
| Storage class choice | Standard versus colder Cloud Storage classes |
That means a prompt about rarely accessed archived objects is not mainly a database-engine question. It is first a Cloud Storage plus storage-class question.
| Trap | Better reading |
|---|---|
| “Any data question means BigQuery.” | BigQuery is for analytics, not every operational workload. |
| “Cloud Storage can replace a relational database because both store data.” | Object storage and relational query engines solve different problems. |
| “Firestore and Bigtable are interchangeable app databases.” | Firestore is document-oriented; Bigtable is wide-column and scale-heavy. |
| “Cold storage class means different database engine.” | Storage class is a tier choice inside Cloud Storage, not a database selection. |
A team needs three data services:
The strongest first fit is:
Correct answer: 1. The workload types are different, so the strongest answer uses object storage, relational SQL, and analytics separately.