Azure provides a comprehensive portfolio of managed database services. Picking the right one depends on your data model, consistency requirements, expected scale, and team expertise.
Azure SQL Database
Azure SQL Database is a fully managed, cloud-native relational database built on Microsoft SQL Server. It supports multiple deployment models:
- Single Database: Isolated database with dedicated resources. Best for new cloud applications.
- Elastic Pool: Multiple databases sharing a resource pool. Cost-effective for SaaS apps with variable per-tenant load.
- Managed Instance: Near-100% SQL Server compatibility. Ideal for lift-and-shift migrations that rely on SQL Server-specific features.
Key features: automatic backups, point-in-time restore up to 35 days, built-in high availability (99.99% SLA), intelligent query processing, and Microsoft Defender for SQL.
Azure Database for PostgreSQL and MySQL
Fully managed open-source relational databases — same familiar engines, without the operational overhead.
- Azure Database for PostgreSQL — supports Flexible Server (recommended), with zone-redundant HA and read replicas
- Azure Database for MySQL — supports Flexible Server with similar HA options
Best for: Teams already using PostgreSQL or MySQL on-premises and wanting a managed cloud equivalent with minimal migration effort.
Azure Cosmos DB
Cosmos DB is Azure's globally distributed, multi-model NoSQL database. It's designed for applications that need low latency at global scale.
Key capabilities:
- Multi-model APIs: NoSQL (document), MongoDB, Cassandra, Table, Gremlin (graph) — same underlying engine
- Global distribution: Replicate data to any number of Azure regions with a single click
- Five consistency levels: Strong, Bounded Staleness, Session (default), Consistent Prefix, Eventual
- Guaranteed SLAs: 99.999% availability, <10ms reads, <15ms writes at p99
- Serverless mode: Pay per request — ideal for variable or spiky workloads
Best for: Global web and mobile apps, IoT telemetry, real-time personalisation, e-commerce product catalogs — wherever you need low latency at scale and flexible schema.
Azure Cache for Redis
A fully managed, in-memory data store based on Redis. Used for:
- Session caching — store user session state at sub-millisecond speed
- Database query caching — reduce database load for frequently read data
- Message brokering — pub/sub and stream processing
- Leaderboards and real-time counters
Choosing the Right Database
| Need | Azure Service |
|---|---|
| SQL Server-compatible relational | Azure SQL Database / Managed Instance |
| Open-source PostgreSQL | Azure Database for PostgreSQL |
| Open-source MySQL / MariaDB | Azure Database for MySQL |
| Global NoSQL documents | Cosmos DB (NoSQL API) |
| MongoDB-compatible NoSQL | Cosmos DB for MongoDB |
| Cassandra-compatible wide-column | Cosmos DB for Cassandra |
| In-memory caching | Azure Cache for Redis |
| Data warehouse analytics | Azure Synapse Analytics |