Key points

• Partition-key design and access distribution are fundamental; a hot key can limit performance regardless of total capacity.
• On-demand and provisioned capacity fit different demand predictability and planning needs.
• Global tables, backups, streams, indexes, and consistency choices should be added for explicit requirements because they affect behavior and cost.

Best-known use cases

• Store sessions, carts, metadata, or high-scale entities accessed by key.
• Build multi-Region key-value applications with compatible conflict and consistency requirements.

What candidates often confuse it with

• DynamoDB favors key-oriented horizontal scale; RDS/Aurora favors relational queries, joins, and constraints.
• DynamoDB is the system of record; ElastiCache is commonly a cache for reusable low-latency data.

Key takeaway

Choose DynamoDB when access patterns are key-oriented and the application can design for partitioned serverless scale.

Related services

• Amazon ElastiCache
• Amazon RDS
• Amazon Aurora
• AWS Lambda

Relevant exam tasks

• D2.1
• D3.3
• D4.3

Learn more

• Amazon DynamoDB