AWS has launched S3 Files, connecting general-purpose S3 buckets to a file system that can be mounted by compute workloads.
The service uses Amazon EFS underneath and presents objects as files and directories. Changes made through the file system are exported back to S3, while bucket changes become visible through the mounted view after synchronization.
Why it matters
- Existing file-oriented tools gain a route to data in S3.
- Multiple compute resources can share a mounted data set.
- File-system capacity, synchronization and access permissions become part of the design.
This is an additional access layer with its own behavior and costs. It should be budgeted separately from ordinary S3 API traffic; the storage calculator’s S3 Standard estimate does not include S3 Files or EFS charges.