Cloud Pro: Google Cloud Databases: Choose, Compare and Launch

What's inside

1. Level 1: Every Google Cloud Database on One PageFree
   Learn the full Google Cloud database lineup grouped by type: relational, distributed SQL, key-value and document, wide-column, in-memory, graph and time series, and migration. Each service is listed with its current and former names and its lifecycle status.
2. Level 2: The Standard Workload ProfileFree
   Build the one-page workload profile that every later choice is measured against. It covers the data model, read/write mix, QPS and latency, transactions and consistency, size and growth, scaling, availability, RPO and RTO, multi-region needs, engine compatibility and licensing, OLTP versus analytics, team skills and the cost model.
3. Level 3: Cloud SQL: Engines, Editions and Limits
   Understand what Cloud SQL is for MySQL, PostgreSQL and SQL Server: where the Enterprise and Enterprise Plus editions differ, which workloads it fits and which it does not, and which documented limits bite in production.
4. Level 4: Cloud SQL: Costs, Operations and Real Uses
   Price a Cloud SQL deployment across compute, storage, backups, HA, replicas, SQL Server licences and egress. Then draw the line between what you operate and what Google operates, using a pros and cons table and real use cases.
5. Level 5: Cloud SQL Quick Start and Decision
   Launch the smallest working Cloud SQL instance with gcloud and with Terraform, connect securely through the Cloud SQL Auth Proxy, run a query and tear everything down. Then record when to choose Cloud SQL and when to avoid it.
6. Level 6: AlloyDB for PostgreSQL: Fit, Limits and Cost
   Learn how AlloyDB's split of compute and storage, its read pools and its columnar engine change what a PostgreSQL workload can do. Compare its limits, pricing and operating model with Cloud SQL for PostgreSQL.
7. Level 7: AlloyDB Quick Start and Decision
   Stand up a minimal AlloyDB cluster and primary instance with gcloud and Terraform, including the private networking it needs. Connect, run queries, clean up, then record when AlloyDB is the right call.
8. Level 8: AlloyDB Omni: AlloyDB Anywhere
   Learn what AlloyDB Omni is: the downloadable AlloyDB engine you run yourself on containers or Kubernetes, on premises or on other clouds. Cover its licensing and support model, what you take on operating it, and a local quick start.
9. Level 9: Spanner: Editions, Model and Limits
   Understand Spanner as a globally distributed, strongly consistent SQL database. Cover the Standard, Enterprise and Enterprise Plus editions, compute measured in nodes and processing units, the limits that bite, and how schema and key design avoid hotspots.
10. Level 10: Spanner Quick Start and PostgreSQL Interface
    Create the smallest Spanner instance with gcloud and Terraform, using both the GoogleSQL and PostgreSQL dialects. Connect through PGAdapter, query, clean up, then decide when Spanner is worth its cost.
11. Level 11: Bigtable: Wide-Column Fit, Limits and Cost
    Learn Bigtable's wide-column data model and its row-key design. Cover the workloads it fits (time series, IoT, personalisation, high-throughput key lookups), its limits, and pricing across nodes, storage type, replication and backups.
12. Level 12: Bigtable Quick Start and Decision
    Create a minimal Bigtable instance with gcloud and Terraform, write and read rows with the cbt CLI and a client library, delete the instance, and record when to choose Bigtable.
13. Level 13: Firestore Native Mode: Documents at Scale
    Understand Firestore's document model in Native and Datastore modes: indexes, transactions, real-time listeners, the documented limits that bite, and pricing driven by reads, writes, deletes and storage.
14. Level 14: Firestore with MongoDB Compatibility
    Learn what Firestore with MongoDB compatibility supports and what it leaves out. Launch a database with gcloud and Terraform, connect with a standard MongoDB driver or mongosh, clean up, and decide between it and other document stores.
15. Level 15: Memorystore for Valkey and Redis Cluster
    Learn the recommended in-memory options: Memorystore for Valkey and Memorystore for Redis Cluster. Cover shards, replicas, persistence, limits, node-based pricing, and a quick start using Private Service Connect.
16. Level 16: Memorystore for Redis and Memcached: Legacy Paths
    Learn where Memorystore for Redis (standalone) and Memorystore for Memcached stand today, including Memcached's deprecation timeline, and how to plan a move to Memorystore for Valkey.
17. Level 17: Oracle Database@Google Cloud
    Understand Oracle Database@Google Cloud: Exadata and Autonomous Database run by Oracle inside Google Cloud regions. Cover regional availability, Marketplace purchasing, licensing (BYOL versus license-included), networking, and who operates what.
18. Level 18: Bare Metal Solution: Allowlist-Only Legacy
    Learn what Bare Metal Solution offered for Oracle and other specialised workloads, why it is now allowlist-only, and how existing users move to Oracle Database@Google Cloud or other targets.
19. Level 19: BigQuery as the Analytics Alternative
    Learn when a workload is really analytics and belongs in BigQuery rather than an operational database. Compare on-demand and capacity (editions) pricing, and connect operational stores through federated queries, Datastream and change streams.
20. Level 20: Database Migration Service in Practice
    Learn what Database Migration Service supports: homogeneous migrations into Cloud SQL and AlloyDB, and heterogeneous migrations with conversion workspaces. Cover its limits and costs, and run a minimal continuous migration job end to end.
21. Level 21: High Availability Across Every Service
    Learners compare how each Google Cloud database stays up: zonal and regional failover, read replicas, multi-region configurations and published SLAs. Each design is tied to the availability target in the workload profile.
22. Level 22: Backup, Point-in-Time Recovery and Disaster Recovery
    Learners design backup, point-in-time recovery and cross-region disaster recovery for each service, sized to the RPO and RTO in the workload profile, and prove it works with a timed restore.
23. Level 23: Security, Encryption and Access Control
    Learners lock down every database service in layers: private networking, IAM and database-level authentication, encryption at rest and in transit with Google-managed or customer-managed keys, secrets handling and audit logging.
24. Level 24: Performance Tuning and Fair Benchmarking
    Learners find bottlenecks with each service's monitoring and query-insight tools, apply the tuning levers that matter most (schema, indexes, keys, connection pooling, caching, sizing) and run benchmarks that compare services fairly.
25. Level 25: A Worked Total Cost of Ownership
    Learners build a dated, line-by-line monthly and three-year cost for one workload on two candidate services. The cost covers compute, storage, I/O, backups, replicas, licences, egress, committed-use discounts and staff time.
26. Level 26: Migration Paths: Homogeneous and Heterogeneous
    Learners plan a migration from start to finish by choosing between same-engine and cross-engine paths. They pick the tool (Database Migration Service, native replication, dump and restore, or third-party tools) and plan validation, cutover and rollback.
27. Level 27: Building a Weighted Selection Scorecard
    Learners turn the Level 2 workload profile into a weighted scorecard with hard pass/fail constraints and weighted criteria, then score every candidate Google Cloud database consistently and with evidence.
28. Level 28: Scoring Google Cloud Against Another Cloud
    Learners apply the same scorecard to equivalent services on another cloud. They gather evidence dated against the current docs and pricing pages, and keep the comparison neutral towards both providers.
29. Level 29: Final Challenge: Decision Workshop and Selection Record Exam
    Learners use the whole course on a realistic case: profile the workload, shortlist and score services, compare against another cloud, cost the choice, then design HA, DR, security and migration. The result is a one-page selection record.

Access

The first 2 levels are free with a free account. Every level, the podcast edition and the AI tutor come with All Access at £4.99/month or any Creator plan — see pricing.

Start free →