ZFS, Actually Storage you can finally reason about.

Snapshots, backups, and actually sleeping at night

chain

6 steps, in order — start anywhere

  1. Copy-on-Write ZFS never writes over existing data—every change lands on fresh space first, then takes effect in one clean switch. The habit everything else stands on.
  2. Snapshots A frozen moment of a dataset that copies nothing—it just keeps the past from being thrown away. Free to take; the cost arrives slowly, as the present drifts from the moment you kept.
  3. Snapshots and Disappearing Space You deleted the files and the free space barely moved—because snapshots are still holding those files. The space changed owners; it didn't come back.
  4. Snapshots Are Not Backups A snapshot lives on the same disks, in the same pool, behind the same mistakes as the data. It's the versioning half of a backup strategy; replication is the other half.
  5. Send and Receive How a frozen moment travels to another machine—an exact copy, and after the first trip, only what changed. Not a file-comparing sync tool: it already knows what's different.
  6. Sanoid and Syncoid Not one "backup system"—two small helpers: sanoid takes and prunes snapshots on a schedule; syncoid ships them to another machine. Both just automate things you can do by hand.