Store structured and semi-structured raw data for a long period and read it with Athena or EMR only when needed.
There are few row updates and storage must grow without managing servers.
Which data store is most appropriate?
Use Amazon S3 as durable, scalable object storage for the raw data lake layer.
Detailed explanation
A row-oriented transactional database is not the economical primary store for unbounded raw files.
A row-oriented transactional database is not the economical primary store for unbounded raw files.
A messaging queue delivers messages and is not a long-term analytics file store.
A messaging queue delivers messages and is not a long-term analytics file store.
An in-memory cache is not intended for durable, long-term raw data.
An in-memory cache is not intended for durable, long-term raw data.
S3 stores large objects durably and lets analytics services read them on demand.
S3 stores large objects durably and lets analytics services read them on demand.
Check official documentation
Compare the following points with the official documentation.
Amazon S3の耐久性、ストレージクラス、データレイク利用の公式説明を確認する。
What to confirm
オブジェクトストレージ・トランザクションDB・キャッシュ・キューの役割を区別できる。
Key points
- Amazon S3
- Data lake
- Durability
- Object storage
Notes
- Environment: AWS公式DEA-C01試験ガイドと各サービスの公式ドキュメントの確認
- Confirm current AWS service behavior in the official DEA-C01 documentation.
Foundation review
Storage and compute separation
Keep raw data in object storage and use an analytics engine only when processing is required.
Access tier
Frequency, retrieval time, and retention determine the suitable S3 storage class.