zfs

List all datasets

$ zfs list

$ sudo zfs create [pool/dataset]

$ sudo zfs snapshot [pool/dataset@snapshot_name]

$ zfs list -t snapshot

$ sudo zfs rollback [pool/dataset@snapshot_name]

$ sudo zfs set [compression=lz4] [pool/dataset]

$ zfs get [compression] [pool/dataset]

$ sudo zfs destroy [pool/dataset]

zfs command [options] [arguments]

zfs manages ZFS filesystems, snapshots, clones, and volumes. ZFS is a combined filesystem and volume manager with advanced features like copy-on-write, snapshots, checksums, and built-in compression.
Key concepts:
- Datasets are filesystems or volumes within a pool
- Snapshots are read-only point-in-time copies
- Clones are writable copies created from snapshots
- Properties control behavior (compression, quota, mountpoint)
ZFS uses hierarchical datasets: pool/parent/child inherits properties from parent. Properties can be set locally to override inheritance.
Common properties include compression, quota, reservation, recordsize, atime, and mountpoint.

list [-t type] [-o properties] [dataset]

List datasets and properties

Create filesystem or volume (-p creates parents)

Destroy dataset (-r recursive)

Create snapshot

Rollback to snapshot

Create clone from snapshot

Generate stream for replication

Receive stream into dataset

Set property

Get property value

Mount filesystem

Unmount filesystem

Most operations require root privileges.
zfs destroy is irreversible. Snapshots protect against accidental deletion but must be created proactively.
ZFS can consume significant memory for caching (ARC). Tune zfs_arc_max on memory-constrained systems.
Snapshots are cheap but not free. Many snapshots with high data churn can consume significant space.

zpool(8), zdb(8), zfs-send(8), zfs-receive(8)