ZFS – Introduction and Key Concepts

Use at your own risk. All guides and scripts are provided for educational purposes only. Always review and understand any code before running it — especially with administrative privileges. Test in a safe environment before using in production. Your system, your responsibility.

ZFS is a combined filesystem and volume manager originally developed by Sun Microsystems. It has become one of the most trusted storage solutions for servers, NAS devices, and home labs — and for good reason.

This guide explains what ZFS is, how it works, and why you might want to use it.

What Makes ZFS Different

Traditional filesystems (ext4, NTFS, APFS) manage files. ZFS manages everything — the physical disks, the volumes, and the filesystem — in one unified system.

Key features:

Copy-on-Write (CoW) ZFS never overwrites existing data. When you modify a file, ZFS writes the new data to a new location and updates the pointer. Old data stays intact until the new write is confirmed. This means ZFS can never be in an inconsistent state — no need for fsck after a crash.

Data Integrity with Checksums Every block of data has a checksum. When ZFS reads data, it verifies the checksum. If corruption is detected, ZFS automatically repairs it using a mirror or RAIDZ copy. Silent data corruption — bit rot — is caught and fixed automatically.

Snapshots A ZFS snapshot is an instant, point-in-time copy of a dataset. Snapshots are nearly free to create (they use no extra space until data changes) and can be used for backups, rollbacks, or cloning.

Storage Pools (zpools) Instead of partitions, ZFS uses storage pools. You add disks to a pool and ZFS manages the space. Adding more disks to a pool expands it automatically.

Built-in RAID (RAIDZ) ZFS has its own RAID implementation — RAIDZ — that avoids the RAID 5 write hole problem that affects hardware RAID controllers.

ZFS Terminology

Term	What it means
zpool	A storage pool made up of one or more disks
vdev	A virtual device — a disk or group of disks in a pool
dataset	A filesystem or volume inside a pool
snapshot	A read-only point-in-time copy of a dataset
clone	A writable copy of a snapshot
scrub	A background scan that verifies and repairs data integrity
ARC	Adaptive Replacement Cache — ZFS RAM cache
L2ARC	Secondary cache on a fast SSD
SLOG/ZIL	Separate Intent Log — write cache on a fast SSD

ZFS RAID Levels

Type	Disks needed	Fault tolerance	Notes
Single disk	1	None	No redundancy
Mirror	2+	1 disk failure	Best read performance
RAIDZ1	3+	1 disk failure	Like RAID 5 but safer
RAIDZ2	4+	2 disk failures	Like RAID 6
RAIDZ3	5+	3 disk failures	Maximum redundancy

For home labs: a mirror (2 disks) is the simplest and most performant option. RAIDZ1 or RAIDZ2 is better for larger setups with 4+ disks.

What ZFS is Not

ZFS is not a backup — RAIDZ protects against disk failure, not against accidental deletion, ransomware, or fire. Always maintain separate backups.
ZFS does not like being starved of RAM — it uses RAM aggressively for caching (ARC). 8 GB+ is recommended, more is better.
ZFS pools should not run above 80% capacity — performance degrades significantly when a pool is nearly full.

Where ZFS is Used

TrueNAS — built entirely around ZFS
Proxmox VE — native ZFS support for VM storage
Ubuntu/Debian — ZFS available as a native filesystem option
FreeBSD — ZFS is the default filesystem
Home NAS builds — ZFS is the go-to choice for DIY NAS setups