RAIDZ3 Calculator
ZFS RAIDZ3 is triple-parity RAID — the most fault-tolerant standard RAID level available anywhere, surviving three simultaneous disk failures.
The RAIDZ3 formula
Usable Capacity = (N − 3) × Disk Size
Three disks' worth of capacity go to triple parity. No traditional hardware RAID level offers 3-disk fault tolerance at all — RAIDZ3 is a ZFS-specific capability.
Worked example
8 disks × 4 TB = 20 TB usable out of 32 TB raw, 62.5% storage efficiency, tolerates 3 simultaneous disk failures. On the same 8 disks, RAIDZ2 would give 24 TB (75% efficiency) but only survive 2 failures.
When RAIDZ3 is actually worth the extra overhead
Very large vdevs (12+ disks) where resilver times stretch long enough that a second failure during rebuild becomes a realistic risk even with RAIDZ2, or archival/cold-storage pools where drives sit mostly idle and failures may go unnoticed for longer than in an actively monitored system, extending the effective exposure window. For typical vdevs under 10-12 disks, RAIDZ2 usually provides adequate protection at a better capacity ratio — RAIDZ3 is a deliberate choice for higher-risk or higher-stakes configurations, not a default.
What's the minimum number of disks for RAIDZ3?
5 disks. At that size the overhead is 60%, which only makes sense if triple-failure tolerance is specifically the goal — otherwise RAIDZ2 or RAIDZ1 gives better capacity efficiency.
When would I actually need RAIDZ3 over RAIDZ2?
Large vdevs (12+ disks) with long resilver windows, archival pools where failed drives might not be noticed quickly, or environments with a documented history of correlated multi-disk failures (same batch, same enclosure, shared power/cooling faults) where a single or double failure realistically could become a triple failure.
Need a different RAID level? Use the full RAID Calculator covering RAID 0, 1, 5, 6, 10, 50, 60 and ZFS RAIDZ1/2/3, or read the RAID levels comparison guide.