Google Cloud Professional Cloud Database Engineer tests whether you can design, run, migrate and scale databases on Google Cloud: Cloud SQL, AlloyDB, Spanner, Bigtable, Firestore and Memorystore, plus Database Migration Service. The current exam guide also includes vector data and databases that support generative AI and LLM use cases. It is aimed at experienced DBAs: Google recommends 5+ years of database and IT experience, including 2 years on Google Cloud databases.
This guide follows Google's official exam guide: four sections with weights, how to choose between the database services, a 6-week plan and 15 original scenario questions.
Quick facts (Google Cloud, November 2026)
| Item | Detail |
|---|---|
| Length | 2 hours |
| Questions | 50–60, multiple choice and multiple select |
| Fee | USD 200 (plus tax where applicable) |
| Language | English |
| Delivery | Online-proctored or at a test centre |
| Prerequisites | None; 5+ years of database and IT experience, including 2 years hands-on with Google Cloud databases, recommended |
Read the official Professional Cloud Database Engineer page and exam guide before booking. Google also publishes free sample questions.
Who should take it?
- Oracle, SQL Server, MySQL and PostgreSQL DBAs moving to Google Cloud.
- Data and backend engineers who own production databases.
- Engineers planning database migrations. Compare with the Professional Data Engineer, which focuses on pipelines and analytics.
Sections and weights
| Section | Weight |
|---|---|
| 1. Design innovative, scalable and highly available cloud database solutions | ~32% |
| 2. Manage a solution that can span multiple database technologies | ~25% |
| 3. Migrate data solutions | ~23% |
| 4. Deploy scalable and highly available databases in Google Cloud | ~20% |
Choosing the right database
| Need | Usual choice |
|---|---|
| Lift-and-shift MySQL, PostgreSQL or SQL Server | Cloud SQL |
| PostgreSQL-compatible with high performance and mixed transactional and analytical work | AlloyDB |
| Global relational database with strong consistency and horizontal scale | Spanner |
| Very high-throughput time series, IoT or ad-tech data | Bigtable |
| Mobile and web app documents with offline sync | Firestore |
| In-memory cache or session store | Memorystore (Redis or Valkey) |
| Oracle workloads that must stay on Oracle | Bare Metal Solution or partner offerings |
| Vector search for RAG | pgvector on Cloud SQL or AlloyDB, or vector search in Spanner |
Section 1: Design (~32%)
- Capacity planning: sizing from current workload metrics and future growth; trade-offs between machine and storage types.
- HA and DR: multi-regional vs regional vs zonal deployments; maintenance windows and notifications.
- Connectivity: networking, key management and encryption; session poolers (connection pooling); audit policies.
- Choosing services: managed vs self-managed, bare metal and partner databases; SQL vs NoSQL (structured, semi-structured, unstructured, vector); cost comparison; dependencies; compliance and organisation policies; multi-database solutions (federation, exports, hybrid); databases for generative AI and LLM use cases.
Section 2: Manage (~25%)
- Access: IAM and policies for database connectivity; database users, authentication and access.
- Monitoring: slow queries, locking and missing indexes; CPU, RAM, storage, I/O and audit logs; quotas; resource contention; alerts.
- Backup and recovery: automated backups; export and import; RTO, RPO and point-in-time recovery; retention.
- Cost and performance: scaling up vs out; replication strategies; continuous cost review; query optimisation.
- Automation: index rebuilds and exports; scheduled exports; managed upgrades; monitoring SLAs and SLOs.
-- Find slow queries in Cloud SQL for PostgreSQL (pg_stat_statements enabled)
SELECT query, calls, round(mean_exec_time::numeric, 1) AS avg_ms, rows
FROM pg_stat_statements
ORDER BY mean_exec_time DESC
LIMIT 10;
-- Check whether a query uses an index
EXPLAIN (ANALYZE, BUFFERS)
SELECT * FROM orders WHERE customer_id = 4471 AND created_at > now() - interval '30 days';Section 3: Migrate (~23%)
- Migration strategies: zero or near-zero downtime, extended outage, and fallback plans.
- Reverse replication from Google Cloud back to the source, so you can fall back.
- DDL and DML conversion for heterogeneous migrations (for example Oracle to PostgreSQL).
- Choosing tools: Database Migration Service, native replication, Datastream, export and import, or partner tools.
Section 4: Deploy (~20%)
- Provisioning HA databases and testing HA and DR.
- Multi-regional replication; deploying and scaling read replicas.
- Automating provisioning (Terraform, gcloud) and monitoring HA databases.
# Cloud SQL for PostgreSQL: HA primary plus a read replica (asia-south1 = Mumbai)
gcloud sql instances create orders-db --database-version=POSTGRES_16 \
--region=asia-south1 --availability-type=REGIONAL --tier=db-custom-4-16384 \
--enable-point-in-time-recovery --backup-start-time=20:30
gcloud sql instances create orders-db-replica --master-instance-name=orders-db \
--region=asia-south26-week study plan
| Week | Focus | Lab |
|---|---|---|
| 1 | Service selection: Cloud SQL, AlloyDB, Spanner, Bigtable, Firestore, Memorystore | Write a one-page choice for five sample apps |
| 2 | Cloud SQL: HA, backups, PITR, read replicas, private IP, Auth Proxy, IAM authentication | HA instance with a cross-region replica; restore to a point in time |
| 3 | AlloyDB and Spanner: architecture, scaling, multi-region, schema design | Spanner instance with interleaved tables |
| 4 | Monitoring: Query Insights, slow queries, locks, indexes, alerts, quotas | Find and fix a missing index |
| 5 | Migration: Database Migration Service, Datastream, reverse replication, Oracle to PostgreSQL | Migrate a MySQL VM to Cloud SQL with minimal downtime |
| 6 | Bigtable, Firestore, vector search for GenAI, cost optimisation; sample questions | pgvector similarity query on AlloyDB; book the exam |
15 scenario questions with answers
These are original scenarios written for this guide, not exam questions.
1. A banking app needs a relational database with strong consistency across India, Europe and the US. Which service?
Spanner with a multi-region configuration.
2. An on-premises PostgreSQL database must move to Google Cloud with minimal changes. Which service?
Cloud SQL for PostgreSQL (or AlloyDB if it needs more performance).
3. Millions of sensor readings per second must be stored and read by key and time. Which service?
Bigtable.
4. A Cloud SQL database must survive a zone failure automatically. What do you configure?
High availability (regional) configuration with a standby in another zone.
5. Reports slow down the primary database. What do you add?
Read replicas, and point reporting at them.
6. The business can lose at most 5 minutes of data. What feature supports this?
Point-in-time recovery with transaction log retention (it defines your RPO).
7. Thousands of short-lived serverless connections overwhelm the database. What helps?
A connection pooler, such as PgBouncer or Cloud SQL managed connection pooling where available.
8. Apps must connect to Cloud SQL securely without managing SSL certificates or allowlisting IPs. What do you use?
The Cloud SQL Auth Proxy (or connectors) with IAM.
9. Which tool finds the slowest queries in Cloud SQL or AlloyDB?
Query Insights.
10. A MySQL database must move to Cloud SQL with near-zero downtime. Which service?
Database Migration Service with continuous replication, then a short cutover.
11. After cutover, the team wants a fallback path to the old database. What do you set up?
Reverse replication from Google Cloud to the source.
12. Oracle stored procedures must run on PostgreSQL. What part of the migration is this?
DDL and DML (schema and code) conversion in a heterogeneous migration.
13. A RAG chatbot needs similarity search next to transactional data in PostgreSQL. What do you use?
The pgvector extension on AlloyDB or Cloud SQL for PostgreSQL.
14. Database costs are high but CPU is under 10% most of the day. What do you do?
Right-size the instance, and review replicas and storage type.
15. Upgrades must not happen during business hours. What do you configure?
A maintenance window and maintenance notifications.
After Professional Cloud Database Engineer
- Practise SQL interview questions and vector database interview questions.
- Read the Professional Cloud Architect guide and the Professional Data Engineer guide.
- See live database and cloud jobs.
Cloud Soft Solutions in Ameerpet, Hyderabad teaches SQL, cloud databases, data engineering and Google Cloud with real-time projects. See the APEX program and all courses.
📞 Call +91 96660 19191 · WhatsApp · Flat No. 513, 5th Floor, Nilagiri Block, Aditya Enclave, Ameerpet, Hyderabad
Frequently Asked Questions
How long is the Professional Cloud Database Engineer exam?
2 hours, with 50 to 60 multiple-choice and multiple-select questions.
How much does the exam cost?
USD 200, plus tax where applicable.
Which languages is the exam offered in?
English only.
Which section has the most weight?
Designing innovative, scalable and highly available cloud database solutions, at about 32%.
How much experience is recommended?
5+ years of overall database and IT experience, including 2 years of hands-on experience with Google Cloud database solutions.
Does the exam cover generative AI?
Yes. The design section includes using database technologies to support generative AI and LLM use cases, and vector data.
