Design a service that retrieves and updates user profiles by user ID. Each request targets one item; the workload does not require joins or complex aggregations.
Traffic is unpredictable, and the service needs millisecond latency with automatic horizontal scaling.
Which database is most appropriate?
Choose Amazon DynamoDB for key-value access with a partition key, low latency, and demand-based scaling.
Detailed explanation
Redshift is an analytical data warehouse, not the primary store for frequent single-item OLTP lookups with millisecond latency.
Redshift is an analytical data warehouse, not the primary store for frequent single-item OLTP lookups with millisecond latency.
S3 is object storage and does not provide database-style frequent single-item updates and low-latency key lookups.
S3 is object storage and does not provide database-style frequent single-item updates and low-latency key lookups.
Aurora supports relational workloads, but a fixed-size deployment does not automatically match an unpredictable key-value workload's scaling needs.
Aurora supports relational workloads, but a fixed-size deployment does not automatically match an unpredictable key-value workload's scaling needs.
DynamoDB is designed for key-value and document access and can scale single-item partition-key requests at low latency.
DynamoDB is designed for key-value and document access and can scale single-item partition-key requests at low latency.
Check official documentation
Compare the following points with the official documentation.
Amazon DynamoDBのキー・バリューアクセス、パーティションキー、オンデマンドまたは自動スケールの公式説明を確認する。
Official references
What to confirm
データモデル、アクセスパターン、レイテンシー、負荷の予測可能性からデータベースエンジンを選択できる。
Key points
- Amazon DynamoDB
- Partition key
- Key-value access
Notes
- Environment:
- Confirm current behavior in the official SAA-C03 guide and AWS service documentation.
Foundation review
Data model and queries
When joins and complex aggregations are unnecessary and items are read by key, a key-value model can fit better than relational tables.
Predictability of load
For unpredictable traffic, compare demand-based scaling with fixed-instance capacity planning.