AWS Database Services

1. Relational

Amazon RDS (Relational Database Service)

• Fully managed relational database service
• Supports: MySQL, PostgreSQL, MariaDB, SQL Server, Oracle
• Handles provisioning, patching, backups, and recovery automatically
• Offers Single DB Instance, Multi-AZ DB Instance, and Multi-AZ DB Cluster
• Gives a connection endpoint (read / read-write)
• Storage types: io2, io1, gp3, gp2
• Storage autoscaling supported
• Security group controls inbound access
• Public endpoint available (not recommended for production)

Amazon Aurora

• AWS cloud-native relational database
• Compatible with MySQL and PostgreSQL
• Up to 5× faster than standard MySQL, 3× faster than standard PostgreSQL
• Automatically replicates data across 3 AZs (6 copies)
• Auto-scaling storage up to 128 TB
• Variants:
  • Aurora Serverless v2 — scales instantly based on load
  • Aurora Global Database — cross-region replication with < 1s lag
  • Aurora DSQL — distributed SQL (see below)

2. In-Memory

Amazon ElastiCache

• Managed in-memory caching service
• Engines:
  • Redis — supports persistence, pub/sub, sorted sets, Lua scripting
  • Memcached — simple, multi-threaded, pure caching
• Sub-millisecond latency
• Use cases: session caching, leaderboards, real-time analytics
• Reduces database load by caching frequent queries

Amazon MemoryDB

• Redis-compatible primary database (not just a cache)
• Fully durable — data persisted to a Multi-AZ transaction log
• Sub-millisecond read and single-digit ms write latency
• Suitable when you need Redis data structures with full durability
• Use cases: financial applications, gaming leaderboards, user session stores

3. NoSQL

Amazon DynamoDB

• Fully managed, serverless key-value and document database
• Single-digit millisecond performance at any scale
• No servers to manage — pay per request or provisioned capacity
• Features:
  • DynamoDB Streams — capture item-level changes
  • Global Tables — active-active multi-region replication
  • TTL — automatically expire items
  • DAX — in-memory cache for DynamoDB (microsecond latency)
• Use cases: e-commerce carts, gaming, IoT, user profiles

Amazon DocumentDB

• Fully managed MongoDB-compatible document database
• Stores, queries, and indexes JSON documents natively
• Scales storage automatically up to 64 TB
• Replicates 6 copies of data across 3 AZs
• Use cases: content management, catalogs, user profiles, mobile apps

4. Wide-Column

Amazon Keyspaces

• Fully managed Apache Cassandra-compatible database
• Serverless — no cluster management needed
• Scales automatically for high-throughput workloads
• Supports Cassandra Query Language (CQL)
• Use cases: IoT device data, time-series logs, large-scale industrial apps

5. Graph

Amazon Neptune

• Fully managed graph database
• Supports two graph models:
  • Property Graph — queried with Gremlin or openCypher
  • RDF Graph — queried with SPARQL
• Stores billions of relationships, sub-second query latency
• Replicates across 3 AZs, up to 15 read replicas
• Use cases: social networks, fraud detection, recommendation engines, knowledge graphs

6. Time-Series

Amazon Timestream

• Fully managed time-series database
• Purpose-built for IoT and operational data
• Automatically scales ingestion, storage, and queries
• Built-in time-series analytics functions (smoothing, interpolation, approximation)
• Stores recent data in memory, historical data in magnetic store automatically
• Use cases: IoT sensor data, application metrics, DevOps monitoring, clickstream data

7. Distributed SQL

Aurora DSQL

• AWS’s newest distributed SQL database (serverless)
• Active-active multi-region writes with strong consistency
• PostgreSQL-compatible
• Zero infrastructure management — fully serverless
• Designed for globally distributed applications needing relational semantics
• Use cases: global SaaS apps, financial systems, multi-region transactional workloads

8. Managed Oracle

Oracle Database@AWS

• Run Oracle Database on dedicated AWS infrastructure
• Full Oracle compatibility — existing licenses, tools, and workloads work as-is
• Integrated with AWS services (VPC, IAM, CloudWatch)
• Managed by AWS for infrastructure; Oracle manages the database engine
• Use cases: lift-and-shift of on-premise Oracle workloads to AWS

Quick Comparison Table

Service	Type	Engine / Protocol	Key Strength
RDS	Relational	MySQL, PG, MSSQL, Oracle	Broad engine support, managed
Aurora	Relational	MySQL, PostgreSQL	High performance, cloud-native
ElastiCache	In-Memory Cache	Redis, Memcached	Sub-ms latency, reduce DB load
MemoryDB	In-Memory DB	Redis	Durable Redis as primary DB
DynamoDB	Key-Value / Doc	Proprietary	Serverless, infinite scale
DocumentDB	Document	MongoDB-compatible	JSON docs, managed MongoDB
Keyspaces	Wide-Column	Cassandra (CQL)	High-write throughput, serverless
Neptune	Graph	Gremlin, SPARQL, openCypher	Relationship-heavy queries
Timestream	Time-Series	SQL-like	IoT & metrics, auto-tiered storage
Aurora DSQL	Distributed SQL	PostgreSQL-compatible	Multi-region active-active SQL
Oracle@AWS	Relational	Oracle DB	Oracle compatibility on AWS infra

Choose Database

• Structured data with relationships → RDS or Aurora
• Massive scale, flexible schema → DynamoDB
• JSON documents → DocumentDB
• Speed up reads (caching) → ElastiCache
• Redis as a durable primary DB → MemoryDB
• Graph relationships → Neptune
• Time-stamped metrics / IoT → Timestream
• Cassandra workloads → Keyspaces
• Global SQL with multi-region writes → Aurora DSQL
• Existing Oracle workloads → Oracle Database@AWS