RAIDZ Capacity Calculator
Estimate raw, usable, and practical capacity for a single RAIDZ vdev using equal-size disks. This tool is ideal for ZFS planning before you buy drives.
Assumptions: single vdev, identical disks, no special allocation classes, and decimal TB marketing sizes converted to TiB for binary reporting.
What this RAIDZ calculator is for
Most people underestimate the gap between “raw” storage and real usable storage. ZFS RAIDZ gives you redundancy, data integrity, and reliability, but parity and best-practice free space policies reduce available capacity. This calculator helps you plan correctly by showing all major layers:
- Raw pool capacity from all active disks.
- Parity overhead based on RAIDZ1, RAIDZ2, or RAIDZ3.
- Usable capacity after parity.
- Practical capacity after estimated overhead and recommended headroom.
How the math works
Core formula
For a single RAIDZ vdev with equal-size drives, a simplified capacity model is:
Usable = (Active Disks − Parity Disks) × Disk Size
Where Active Disks is total disks minus hot spares. Hot spares improve resilience but do not contribute capacity until activated after failure.
Practical vs theoretical space
Even with perfect parity math, your filesystem still needs room for metadata, copy-on-write behavior, snapshots, and fragmentation tolerance. ZFS pools generally perform best when not near 100% full. That is why this page includes:
- Filesystem overhead % for real-world loss.
- Recommended free space % so your pool stays healthy and fast.
Quick example configurations
| Layout | Disks | Disk Size | Parity | Usable (before extra overhead) |
|---|---|---|---|---|
| RAIDZ1 | 6 | 12 TB | 1 disk | 60 TB |
| RAIDZ2 | 8 | 16 TB | 2 disks | 96 TB |
| RAIDZ3 | 12 | 20 TB | 3 disks | 180 TB |
Choosing RAIDZ1, RAIDZ2, or RAIDZ3
RAIDZ1
Single parity. Best capacity efficiency, but least fault tolerance. A second disk failure during resilver can mean data loss. Usually better for small arrays and less critical data.
RAIDZ2
Double parity. A strong default for many home lab and SMB systems. It tolerates two concurrent disk failures in a vdev and gives better safety margin during rebuild events.
RAIDZ3
Triple parity. Usually selected for large-capacity arrays, long rebuild windows, or mission-critical datasets where extra parity is worth the usable-capacity tradeoff.
Important planning tips
- RAID is not backup. You still need off-site or offline backups.
- Wider vdevs are not always better. Capacity improves, but resilver risk and operational complexity may rise.
- Leave free space. ZFS performance and stability often degrade as pools get very full.
- Use similar disks. Mixed drive sizes are effectively limited by the smallest drive in a vdev.
- Test restores. Backup that has never been restored is only a theory.
Common mistakes this calculator helps avoid
- Buying disks based on raw TB numbers without accounting for parity.
- Ignoring binary conversion differences (TB vs TiB).
- Forgetting to reserve headroom for snapshots and growth.
- Assuming hot spares add active capacity.
Frequently asked questions
Does this model support multiple vdevs?
This page models one RAIDZ vdev for clarity. If you have multiple vdevs in one pool, calculate each vdev and sum the results.
Why does my observed capacity still differ?
Real systems vary due to ashift, sector alignment, special metadata behavior, snapshots, compression, small-block workloads, and pool history. The result here is a planning estimate, not an exact byte-for-byte simulator.
What free-space target should I use?
Many operators target around 20% free. Heavy random-write workloads may benefit from even more headroom.
Use the calculator above to experiment with combinations before purchasing hardware. A few minutes of planning can prevent expensive upgrades, painful migrations, and unexpected “out of space” events later.