Databases: Cloud SQL, Firestore, and Bigtable

Databases: Cloud SQL, Firestore, and Bigtable

Cloud SQL manages supported relational engines. Firestore is a serverless document database with application SDKs and realtime capabilities. Bigtable is a distributed wide-column database for very large, low-latency operational and analytical workloads.

Core ideas

• Cloud SQL high availability uses a regional setup for failover across zones.
• Firestore organizes documents in collections and indexes queries; document paths and query patterns shape design.
• Bigtable row keys control ordering and distribution, so monotonically increasing prefixes can create hotspots.
• No one database is universally best; consistency, transactions, query shape, scale, latency, and operations drive selection.

Design and operating model

Write the access patterns and transaction boundaries before selecting a product. Use Cloud SQL for relational constraints and familiar SQL. Use Firestore for document-centric mobile, web, or server applications. Use Bigtable for huge sparse datasets with key-range access. Test realistic data size and skew.

Example

gcloud sql instances create training-sql \
  --database-version=POSTGRES_17 \
  --region=asia-south1 \
  --availability-type=REGIONAL \
  --cpu=2 \
  --memory=7680MiB

gcloud firestore databases list
gcloud bigtable instances list

Run examples in a disposable training project. Replace project IDs, Regions, identities, resource names, sizes, and policies with reviewed values, and verify current Google Cloud CLI and service requirements before production use.

Production guidance

• Use private connectivity and IAM or database authentication appropriate to the engine.
• Enable automated backups and point-in-time capabilities, then rehearse restore into isolation.
• Monitor query latency, connections, replication, capacity, index growth, and hot partitions.

The chapter quiz follows the lesson and checks both service vocabulary and architecture decisions.