A new date=2026-08-20 partition arrives in S3.
Athena must recognize the partition without waiting for a full data reload.
Which approach is most appropriate?
Run a crawler or register the location with ALTER TABLE ADD PARTITION so catalog metadata points to the new data.
Detailed explanation
Object versions do not automatically create Athena table partitions.
Object versions do not automatically create Athena table partitions.
Permissions do not create partition metadata in the catalog.
Permissions do not create partition metadata in the catalog.
VACUUM reorganizes Redshift tables and does not register S3 partitions.
VACUUM reorganizes Redshift tables and does not register S3 partitions.
The new partition's location and value become catalog metadata that Athena can query.
The new partition's location and value become catalog metadata that Athena can query.
Check official documentation
Compare the following points with the official documentation.
Glue Catalog partition、Athena ALTER TABLE ADD PARTITION、MSCK REPAIR TABLEの公式説明を確認する。
What to confirm
物理パスの作成と、Catalogメタデータ登録の違いを説明できる。
Key points
- Partition registration
- ALTER TABLE
- MSCK REPAIR
- Catalog metadata
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
Physical and logical metadata
Files can exist in S3 before the catalog knows their partition values.
Refresh strategy
Balance crawler latency, explicit registration, and query freshness requirements.