Skip to content
Draft: this topic has not been reviewed yet. Check facts against the linked AWS documentation before relying on them.

AWS database services compared

Cloud Practitioner · Domain 3 · Cloud Technology and ServicesTask 3.4: Identify AWS database servicesLevel 120 minComparisonDraft
Difficulty
Confidence

Almost every application needs a database. AWS offers many, because different data fits different shapes. The good news for the exam: you mostly need to match the kind of data to the kind of database, and know that AWS can run the database for you.

First choice: run it yourself, or let AWS run it?

You can install a database on an EC2 instance yourself, or use a managed database service.

Database on Amazon EC2 AWS managed database
You manage Operating system, installation, patching, backups, replication, scaling, failover Your data, schema, queries and who can access them
AWS manages The hardware Hardware, OS, database software patching, automated backups, failover and much of scaling
Choose when You need a database engine or setting that no managed service offers, or full OS access Almost always: it removes routine work

Managed databases are a good example of the shared responsibility model shifting: on RDS, AWS patches the database software; on EC2, you do.

Relational databases

Relational databases store data in tables with rows and columns, linked by keys, and are queried with SQL. They suit structured data with clear relationships, such as customers, orders and invoices.

Amazon RDS

Amazon Relational Database Service (Amazon RDS) runs popular database engines for you: MySQL, PostgreSQL, MariaDB, Oracle, Microsoft SQL Server and Db2.

  • Automated backups and point-in-time restore.
  • Multi-AZ deployments: a standby copy in another Availability Zone takes over automatically if the main one fails, for high availability.
  • Read replicas: extra read-only copies to handle more read traffic.

Amazon Aurora

Amazon Aurora is AWS’s own relational database, compatible with MySQL and PostgreSQL, so most applications work without changes.

  • Built for high performance and availability.
  • Stores six copies of your data across three Availability Zones.
  • Offers a serverless option that scales capacity automatically.

NoSQL databases

NoSQL databases do not use fixed tables. They trade some relational features for flexible data and huge scale.

Amazon DynamoDB

DynamoDB is a key-value and document database.

  • Serverless: no servers or instances to manage; it scales automatically.
  • Single-digit millisecond response times at any scale, even millions of requests per second.
  • Pay for the reads, writes and storage you use (or reserve capacity).

Typical uses: shopping carts, user profiles, gaming leaderboards, session data, and anything needing fast lookups by key at huge scale.

In-memory: Amazon ElastiCache

ElastiCache keeps data in memory for microsecond reads. It is usually put in front of another database as a cache, so popular data is served without querying the database every time.

It supports Valkey, Memcached and Redis OSS. Typical uses: caching database results, user sessions, real-time leaderboards.

Other purpose-built databases

Service Type Use it for
Amazon DocumentDB Document (MongoDB-compatible) JSON documents, using MongoDB tools and drivers
Amazon Neptune Graph Highly connected data: social networks, recommendation engines, fraud detection
Amazon Redshift Data warehouse Analytics on large amounts of data, such as sales reports across years; not for everyday transactions

Migrating databases to AWS

Tool What it does
AWS Database Migration Service (AWS DMS) Moves data from a source database to AWS while the source keeps running, and keeps replicating changes until you switch over
AWS Schema Conversion Tool (AWS SCT) Converts the database schema and code when the target engine differs, such as Oracle to Aurora PostgreSQL
  • Same engine (for example MySQL to RDS for MySQL): DMS alone.
  • Different engine (for example SQL Server to Aurora MySQL): SCT to convert the schema, then DMS to move the data.

Putting it together: an online store

Data Database Why
Orders, customers, invoices Amazon Aurora Structured, relational data needing transactions
Shopping carts and sessions Amazon DynamoDB Fast key lookups at any scale
Popular product pages Amazon ElastiCache Microsecond reads in front of Aurora
“Customers who bought this also bought” Amazon Neptune Relationships between products and customers
Five years of sales analysis Amazon Redshift Large analytic queries

Common mix-ups

  • RDS vs Aurora. Aurora is one of the engines you can run through RDS, built by AWS for performance and availability; RDS also runs standard MySQL, PostgreSQL and others.
  • DynamoDB vs RDS. DynamoDB is NoSQL key-value; RDS is relational SQL.
  • Redshift is for analytics, not for running a website’s day-to-day transactions.
  • Multi-AZ vs read replicas. Multi-AZ is for availability (failover); read replicas are for read performance.

Check yourself

Which database is serverless, NoSQL and returns results in single-digit milliseconds at any scale?

Amazon DynamoDB.

An application reads the same product data thousands of times per second. What reduces load on the database?

Amazon ElastiCache as an in-memory cache in front of the database.

A company wants a managed MySQL-compatible database with high performance and six copies of data across three AZs. Which service?

Amazon Aurora.

A team is moving from Oracle to Amazon Aurora PostgreSQL. Which tools does it need?

AWS SCT to convert the schema and code, and AWS DMS to move and replicate the data.

Key takeaways

  • Managed databases move patching, backups and failover to AWS.
  • Relational: RDS and Aurora. NoSQL: DynamoDB. In-memory: ElastiCache. Document: DocumentDB. Graph: Neptune. Warehouse: Redshift.
  • DMS migrates data (with ongoing replication); SCT converts schemas between engines.

Cloud Study Lab is an independent study site. It is not affiliated with or endorsed by Amazon Web Services. AWS and the names of AWS certifications are trademarks of Amazon.com, Inc. or its affiliates.