DynamoDB read and write traffic is stable and can be forecast.
Provision capacity in advance and reduce the unit cost while still following moderate changes.
Which design is most appropriate?
Provisioned capacity with Auto Scaling fits predictable DynamoDB traffic and can adjust capacity within defined limits.
Detailed explanation
Changing database models without considering access requirements is not a cost-optimization method.
Changing database models without considering access requirements is not a cost-optimization method.
Provisioned capacity can be sized for expected traffic and Auto Scaling can follow changes.
Provisioned capacity can be sized for expected traffic and Auto Scaling can follow changes.
CloudFront caches content and does not configure DynamoDB write capacity.
CloudFront caches content and does not configure DynamoDB write capacity.
On-demand is useful for uncertain traffic, but predictable usage should be compared with provisioned pricing.
On-demand is useful for uncertain traffic, but predictable usage should be compared with provisioned pricing.
Check official documentation
Compare the following points with the official documentation.
DynamoDBのプロビジョンド・オンデマンドキャパシティの公式説明を確認する。
What to confirm
負荷の予測可能性からDynamoDBの容量モードを選べる。
Key points
- Provisioned capacity
- DynamoDB Auto Scaling
- Predictable traffic
Notes
- Environment: AWS公式DynamoDBドキュメントの確認
- Confirm current AWS pricing and service behavior in the official documentation.
Foundation review
Capacity mode
Provisioned capacity can be economical for a stable baseline, while Auto Scaling handles normal variation.
Model fit
Capacity optimization does not justify changing to a different database without an access-pattern reason.