ZFS RAIDZ1

ZFS single parity. Same risk profile as RAID 5 but with checksums, snapshots, and used-block resilver.

Min. drives
3
Usable capacity
(N-1) × smallest
Fault tolerance
1
Performance
Fast reads, slow writes

How it works

Functionally similar to RAID 5: data and parity stripes across drives. ZFS adds per-block checksums (detects bit rot), copy-on-write (no write hole), snapshots, and used-block resilver — rebuilds only re-create actually-used data, not entire drives. The URE-during-rebuild concern remains: a single uncorrectable read on a surviving drive can leave data lost.

Formula: (N − 1) × min(drives)

ZFS RAIDZ1 — D = data, P = parityDrive 1D1D4PD9Drive 2D2PD7D10Drive 3PD6D8D11Drive 4D3D5D8P
Layout diagram

Pros / Cons

Pros

  • Block-level checksums detect bit rot
  • Used-block resilver — only rebuilds actual data
  • Native snapshots, replication, compression
  • Strong data integrity guarantees
  • Free / open source / battle-tested

Cons

  • Same URE-during-rebuild concern as RAID 5
  • Avoid with drives ≥ 12 TB
  • ZFS learning curve
  • RAM-hungry (ECC recommended)
  • vdev expansion is awkward

When to use

ZFS pools with 3-4 small drives where capacity matters. Test labs, single-vdev secondary pools, backup targets.

When NOT to use

Drives ≥ 12 TB — RAIDZ2 is strongly preferred. Large pools (8+ drives) — too much URE exposure.

Rebuild math example

4 × 8 TB RAIDZ1 at 80% fill: resilver reads (4-1) × 8 × 0.8 = ~19 TB at 70 MB/s ≈ 25 hours. URE risk during resilver remains as in RAID 5; ZFS does NOT magically protect against URE during rebuild.

Check rebuild risk →

Try it now

Related

Bottom line: RAIDZ1 is fine for small ZFS pools with small drives, but RAIDZ2 is the new default for any pool that matters.