- Same-region/cross-region replication (possible)
- 99.99999999999% durability
- Data-lake
- Analytics
- Data-archiving
- Static website-hosting
- ...
- Athena - Queries S3
Query Functionality on top of S3.
- SelectObjectContents - SDK
- InputType
- SQL
- InputType
- DataFormats
- CSV
- JSON
- Parquet
- Compression
- GZIP
- BZIP2 for CSV/JSON
- GZIP Columnar for Parquet
- Encoding
- UTF-8
- Standard - Frequently accessed data
- Standard IA - Long-lived, infrequently accessed (but still ms response when needed)
- Intelligent Tiering - Changing or unknown pattern
- One-Zone IA - Non-critical, infrequently accessed data (usually that can be recreated)
- Glacier - Long-term data-archiving - minutes to hours of retrieval time
- Glacier Deep Archive - Rarely accessed data - 12h+ retrieval time
- RRS - Frequently accessed non-critical data (usually that can be recreated)
- Lifecycle rules to graduate files to colder storage classes e.g.: from Standard to Glacier
- Transition always goes one-way only, from Standard -> to colder classes
- Encrypted objects remain encrypted
- Archived objects can't use life-cycle policy
- Copies objects asynchronously across S3 buckets
- CRR - Cross Region Replication
- SRR - Same Region Replication