Databases: AWS to Azure service map
34 AWS → Azure mappings: 10 direct, 17 partial, 7 with no equivalent.
Managed relational engines, cloud-native and serverless SQL, high availability and replication, NoSQL (key-value, document, wide-column, graph), time series, in-memory caches, and database migration. AWS mapped to Azure, verified against official docs.
Managed relational engines(7)
Managed PostgreSQL Direct match
Amazon RDS for PostgreSQL maps to Azure Database for PostgreSQL (flexible server). Both run community PostgreSQL with automated backups, point-in-time restore, standby-based HA, read replicas, and a large set of extensions.
Managed MySQL Direct match
Amazon RDS for MySQL maps to Azure Database for MySQL (flexible server). Both run MySQL Community Edition with automated backups, zone-redundant HA, and read replicas.
Managed MariaDB No equivalent
Azure no longer has a managed MariaDB service. Azure Database for MariaDB retired on Sep 19, 2025, and Microsoft's migration target is Azure Database for MySQL (flexible server). Amazon RDS for MariaDB is still available.
Managed Microsoft SQL Server Partial match
Amazon RDS for SQL Server maps to Azure SQL Managed Instance (instance-level compatibility, closest to RDS) or Azure SQL Database (database-scoped PaaS). Azure is the native home of SQL Server, so it has more options and deeper engine features than RDS.
SQL Server (or Oracle) with OS-level control Partial match
Amazon RDS Custom maps most closely to SQL Server on Azure Virtual Machines with the SQL IaaS Agent extension, which adds automated backup, patching, and license management to a VM you fully control.
Managed Oracle Database Partial match
Amazon RDS for Oracle has no Azure-native equivalent. The Azure option is Oracle AI Database@Azure, where Oracle runs Exadata, Autonomous, and Base Database services in Azure data centers. Its direct AWS counterpart is Oracle Database@AWS.
Managed IBM Db2 No equivalent
Azure has no managed Db2 service. On Azure, Db2 runs on virtual machines you manage, or through IBM or partner offerings in Azure Marketplace. Amazon RDS for Db2 offers Standard and Advanced editions as a managed service.
Cloud-native, serverless & distributed SQL(6)
Cloud-native PostgreSQL (separate compute and storage) Partial match
Microsoft maps Amazon Aurora PostgreSQL to Azure HorizonDB, a PostgreSQL service with disaggregated compute and storage. HorizonDB was in public preview when verified, so for production today the usual Azure target is Azure Database for PostgreSQL (flexible server).
Cloud-native MySQL-compatible database Partial match
Azure has no Aurora MySQL-style engine. The closest option is Azure Database for MySQL (flexible server), using the Business Critical tier with zone-redundant HA and read replicas.
Serverless relational database (autoscale and pause) Partial match
Microsoft maps Amazon Aurora Serverless v2 to Azure SQL Database serverless (and SQL database in Microsoft Fabric). Both scale compute between limits you set and can pause when idle, but they run different engines (Aurora is MySQL or PostgreSQL; Azure SQL is SQL Server).
Serverless, active-active distributed SQL No equivalent
Azure has no equivalent of Amazon Aurora DSQL. For multi-region active-active on Azure, Cosmos DB (NoSQL) offers multi-region writes; for relational workloads, Azure SQL failover groups and PostgreSQL cross-region replicas are active-passive.
Horizontally sharded PostgreSQL Partial match
Aurora PostgreSQL Limitless Database maps to elastic clusters in Azure Database for PostgreSQL, which is managed Citus. Azure Cosmos DB for PostgreSQL, the older Citus service, is on a retirement path.
Run SQL Server apps on PostgreSQL No equivalent
Azure has no equivalent of Babelfish for Aurora PostgreSQL. On Azure, SQL Server apps usually move to Azure SQL itself, which avoids an engine change.
Availability, scaling & operations(6)
In-region high availability with automatic failover Direct match
RDS Multi-AZ deployments map to zone-redundant high availability in Azure Database for PostgreSQL and MySQL, and to the zone-redundant configurations of Azure SQL Database and Managed Instance.
Read replicas (in-region and cross-region) Direct match
RDS read replicas map to read replicas in Azure Database for PostgreSQL and MySQL. Both use asynchronous native replication, can sit in another region, and can be promoted.
Cross-region relational disaster recovery Partial match
Aurora Global Database maps to Azure SQL active geo-replication and failover groups (for Azure SQL), and to cross-region read replicas for Azure Database for PostgreSQL and MySQL.
Database connection pooling Partial match
Amazon RDS Proxy maps to the built-in PgBouncer in Azure Database for PostgreSQL. Azure has no managed proxy for MySQL or Azure SQL (Azure SQL handles many connections natively).
Blue/green database upgrades No equivalent
Azure has no managed blue/green deployment for databases like Amazon RDS Blue/Green Deployments. Azure PostgreSQL offers in-place major version upgrades, and you can build a blue/green pattern yourself with read replicas or logical replication.
Managed databases on your own hardware Partial match
Amazon RDS on Outposts maps most closely to SQL Managed Instance enabled by Azure Arc. RDS on Outposts runs on AWS-owned racks, while Arc runs SQL Managed Instance on Kubernetes on your own infrastructure.
NoSQL (key-value, document, wide-column, graph)(8)
Serverless key-value and document database Direct match
Amazon DynamoDB maps to Azure Cosmos DB for NoSQL. Both partition data by key, scale throughput elastically, offer serverless and provisioned capacity, and provide change streams and multi-region replication. Their data models, consistency, indexing, and pricing differ, so plan data-model changes rather than a straight port.
Multi-region, multi-active NoSQL replication Direct match
DynamoDB global tables map to Cosmos DB global distribution with multi-region writes. Both replicate every write to all regions and let any region accept writes.
NoSQL change data capture stream Direct match
DynamoDB Streams maps to the Azure Cosmos DB change feed. Both expose an ordered log of item changes that Lambda or Azure Functions can trigger on.
Built-in cache for a NoSQL database Partial match
DynamoDB Accelerator (DAX) maps to the Azure Cosmos DB integrated cache, which runs on a dedicated gateway. Both are read-through caches tied to the database's own API.
Low-cost key-attribute table storage Partial match
Microsoft also maps DynamoDB to Azure Table Storage (simple and very cheap, in a storage account) and Azure Cosmos DB for Table (the same API with Cosmos DB scale and global distribution). Table Storage is a good fit for simple DynamoDB workloads that don't need its advanced features.
MongoDB-compatible document database Direct match
Amazon DocumentDB maps to Azure DocumentDB (vCore-based, open-source DocumentDB engine) or Azure Cosmos DB for MongoDB (RU-based). Both clouds implement the MongoDB wire protocol on their own engines rather than running MongoDB itself.
Apache Cassandra-compatible wide-column database Direct match
Amazon Keyspaces maps to Azure Cosmos DB for Apache Cassandra (serverless-style, RU-based) or Azure Managed Instance for Apache Cassandra (managed open-source Cassandra clusters).
Graph database Partial match
Amazon Neptune maps to Azure Cosmos DB for Apache Gremlin for property graphs. Azure has no managed RDF/SPARQL or openCypher service, so those Neptune workloads need another design on Azure.
Time series, in-memory & ledger(5)
Serverless time-series database Partial match
Amazon Timestream for LiveAnalytics maps to Azure Data Explorer (or Eventhouse in Microsoft Fabric), which runs on the same engine. Timestream for LiveAnalytics closed to new customers on June 20, 2025, and AWS now steers new workloads to Timestream for InfluxDB.
Managed InfluxDB No equivalent
Azure has no first-party managed InfluxDB service like Amazon Timestream for InfluxDB. On Azure, the options are InfluxData's own cloud offering, self-managed InfluxDB on VMs or AKS, or a move to Azure Data Explorer.
Managed in-memory cache (Redis-compatible) Direct match
Amazon ElastiCache maps to Azure Managed Redis. Azure Cache for Redis, the older service, is retiring (Enterprise tiers on March 31, 2027, and Basic, Standard, and Premium tiers on September 30, 2028), so new designs should use Azure Managed Redis.
Durable in-memory primary database No equivalent
Microsoft states that Azure has no direct equivalent of Amazon MemoryDB. Azure Managed Redis supports persistence (RDB or AOF) and high availability, but it's designed as a cache, not a durable primary database with a multi-AZ transaction log.
Tamper-evident ledger data Partial match
Amazon QLDB was retired on July 31, 2025, and AWS recommended Aurora PostgreSQL for migrations. On Azure, the ledger feature in Azure SQL Database, SQL Managed Instance, and SQL Server provides tamper-evident tables inside a normal relational database.
Migration & analytics integration(2)
Database migration and continuous replication Partial match
AWS Database Migration Service maps to Azure Database Migration Service. AWS DMS is a general replication engine with many sources and targets, plus Schema Conversion for engine changes. Azure DMS focuses on migrations to Azure SQL targets and Azure Database for MySQL, while other engines use their own tools (such as the PostgreSQL migration service in flexible server).
Zero-ETL replication into analytics Partial match
Aurora and RDS zero-ETL integrations (to Amazon Redshift or a SageMaker lakehouse) map to Microsoft Fabric mirroring, which continuously replicates Azure SQL, Cosmos DB, PostgreSQL, and other sources into OneLake as Delta tables.