Volume snapshot deletion + physical purge semantics

, , ,
1

Doing a data-deletion review. The snapshot docs cover retention (1–60 days) and that snapshots survive volume deletion within retention — we’ve verified that behavior. Three things the docs don’t cover:

  1. Can a specific snapshot be deleted on demand (CLI/API)? fly volumes snapshots only has create/list. If not, is retention expiry the only mechanism?
  2. When a snapshot expires, what’s the maximum latency before its data is physically purged, including the AWS backup storage listed in your sub-processors?
  3. After a volume is deleted and its snapshots expire, does any copy of the data persist anywhere (backups/replication/DR), and for how long?
3

As I mentioned as a reply to your other post, you’ll get the best answer to your questions from compliance@fly.io, but here’s the gist: volume snapshots cannot be deleted on demand; they are automatically deleted after expiry on the next snapshot run (every 24h). Volume data is only ever stored on the host (deleted 24h after the volume is destroyed) or in S3 backups (deleted at most 24h after the snapshot expires), there is no other copy.