Databases for AWSs
AWS databases are managed data services so you do not run database servers yourself. RDS and Aurora offer relational engines with backups and failover, DynamoDB is a key-value and document store that scales by partition key, and ElastiCache provides Redis or Memcached caches that reduce load and latency.
Stage: Data · Part of the AWS roadmap · Last reviewed 19 November 2026
What to learn in Databases
- Launch an RDS instance in private subnets with Multi-AZ
- Explain what read replicas are and when to use them
- Design a DynamoDB table around its partition key and access patterns
- Add a cache layer in front of a slow query
- Choose between relational and NoSQL for a given use case
Databases interview questions
What is the difference between Multi-AZ and a read replica in RDS?
Multi-AZ keeps a standby copy in another zone with synchronous replication for failover, improving availability, not read capacity. A read replica uses asynchronous replication and serves read traffic to offload the primary, but promoting it to primary is a manual step for recovery.
Why does DynamoDB need careful key design?
DynamoDB distributes data by partition key and is fast only when queries use the key. A poorly chosen key can create hot partitions, and queries not matching the key need scans, which are slow and costly. You model tables from access patterns, not from entities alone.
Where Databases fits in the AWS roadmap
Databases belongs to the Data stage (intermediate level), which takes about 25 hours. Store and query.