Cloud Service Model · 5 of 6

DBaaS — Database as a Service

A managed database — the cloud handles backups, patching, replication, failover, and (often) scaling. You point your app at a connection string and forget about everything else.

Managed DBAuto-backupsHA & FailoverRead ReplicasServerless DB
← Back to Cloud
Quick Facts

At a Glance

Basic Concepts

  • You get the database, not the box. No SSH, no apt install postgresql.
  • Backups, patches, point-in-time recovery are automatic.
  • HA via standby replicas in another availability zone — failover in seconds.
  • Scaling via vertical resize, read replicas, or serverless auto-scale.
  • You pay a managed-service premium over running it yourself — almost always worth it.
Landscape

The Major Options

TypeMajor DBaaS offerings
Relational (SQL)AWS RDS, Aurora · Azure SQL · Cloud SQL · Cloud Spanner · Oracle Autonomous
Postgres specialistsNeon, Supabase, Crunchy Bridge, Aiven, Timescale Cloud
MySQL specialistsPlanetScale (Vitess), MariaDB SkySQL
Document / NoSQLMongoDB Atlas, DynamoDB, Cosmos DB, Firestore
Key-valueDynamoDB, Redis Cloud, Upstash Redis, Memorystore
Wide-columnAstra DB (Cassandra), ScyllaDB Cloud, Bigtable
GraphNeptune, Neo4j AuraDB
SearchElastic Cloud, OpenSearch Service, Algolia, MeiliSearch Cloud
VectorPinecone, Weaviate Cloud, Qdrant Cloud, MongoDB Atlas Vector
Time seriesInfluxDB Cloud, Timescale Cloud, AWS Timestream
WarehousesSnowflake, BigQuery, Redshift, Databricks SQL
Mechanics

What "Managed" Really Means

What the Provider Handles
  • Provisioning — instance spins up in minutes.
  • Patching — minor version upgrades on a maintenance window.
  • Backups — daily snapshots + continuous WAL for point-in-time recovery.
  • Replication & failover — multi-AZ standby with automatic promotion.
  • Monitoring — CPU, memory, query latency, slow-query log.
  • Encryption at rest & in transit by default.
What's Still On You
  • Schema design & migrations (Liquibase, Flyway, Alembic, Prisma).
  • Indexing strategy & query optimization.
  • Connection pooling (PgBouncer, RDS Proxy, Hyperdrive).
  • Capacity planning — choose instance size, IOPS, storage type.
  • Data lifecycle — archival, retention, GDPR-style deletes.
Serverless & Scale-to-Zero DBs

A new class that auto-pauses when idle and bills per request:

  • Aurora Serverless v2 — auto-scale by ACUs.
  • Neon — Postgres with branching + scale-to-zero.
  • PlanetScale — MySQL with branching schemas.
  • DynamoDB on-demand — pay per request, no capacity planning.
  • D1 (Cloudflare), Turso (libSQL), Xata — edge-native serverless DBs.
Pricing Levers
  • Instance hours (vCPU + RAM).
  • Storage GB-month + IOPS / throughput.
  • Backup storage beyond included quota.
  • Read replicas = extra instances.
  • Egress — moving data out across regions / out of cloud.
  • Reserved capacity can cut steady-state cost 30–60%.
Trade-offs

Strengths & Weaknesses

Strengths
  • Eliminates the most thankless ops work.
  • HA & backups out of the box.
  • Easy to scale up or add replicas.
  • Security defaults are enterprise-grade.
Weaknesses
  • Pricier per CPU than self-managed.
  • Limited extensions / tunables on locked-down services.
  • Vendor lock-in for proprietary engines (Aurora, Spanner, Cosmos).
  • Major version upgrades still require care.
Continue

Other Service Models

IaaSPaaSSaaSFaaS / ServerlessContainers as a Service ↑ Back to Cloud