1 · Choose a RAID level
Block striping with dual distributed parity. Survives two failures.
2 · Configure drives
3 · Drive class
3.5" nearline SAS/SATA capacity HDD — indicative figures.
Advanced — read/write mix, URE rate
Calculated for planning. We don't publish prices — a 24-year UK reseller, Servnet confirms the exact drives, array and pricing on quote. IOPS, throughput & rebuild are indicative estimates.
Calculate a specific RAID level
Each level has its own calculator and a plain-English guide — capacity formula, pros and cons, what it is best for, and a worked example.
RAID 0 →
Striping — maximum capacity and speed, zero redundancy.
RAID 1 →
Mirroring — a full copy of your data for simple, fast protection.
RAID 5 →
Single distributed parity — capacity-efficient, one-drive protection.
RAID 6 →
Dual distributed parity — survives two failures, the large-array default.
RAID 10 →
Striped mirrors — top write performance and fast rebuilds.
RAID 50 →
Striped RAID 5 groups — more performance and safer rebuilds than RAID 5.
RAID 60 →
Striped RAID 6 groups — dual parity per group for very large pools.
RAIDZ1 →
Single-parity ZFS — capacity-efficient, with real ZFS overhead modelled.
RAIDZ2 →
Double-parity ZFS — the resilient default for wide vdevs.
RAIDZ3 →
Triple-parity ZFS — survives three failures per vdev.
RAID 1E →
Striped mirroring — RAID 10-style protection across an odd drive count.
RAID 4 →
Block striping with a dedicated parity disk — RAID 5 with parity on one drive.
RAID 5E →
RAID 5 with an integrated hot spare — faster failover, one less usable drive.
RAID-DP →
NetApp double parity on WAFL — RAID 6 resilience without the ×6 write penalty.
RAID-TEC →
NetApp triple parity (triple erasure coding) — three drives of protection on WAFL.
Capacity, resilience and speed — the trade-off
Every RAID level trades three things: how much of your raw capacity is usable, how many drives you can lose, and how fast writes are. RAID 0 is all capacity and speed but zero safety; mirrors and RAID 10 are fast and safe but only 50% efficient; parity levels (RAID 5/6, RAIDZ) keep more capacity but pay a write penalty.
Our calculator goes further than most: it models the real ZFS overheads (slop reservation and (parity+1) padding), applies the correct RAIDZ per-vdev IOPS rule, treats RAID-DP/TEC as WAFL (not a ×6 penalty), and reports the dual-parity URE nuance correctly — the places generic calculators get it wrong.
Guides, ZFS & storage tools
RAID calculator — common questions
What does this RAID calculator work out?
For any RAID level — RAID 0, 1, 5, 6, 10, 50, 60, RAID-DP/TEC and ZFS RAIDZ1/2/3 — it calculates usable capacity (shown in both decimal TB and binary TiB), raw capacity, space efficiency, how many drives you can lose, the write penalty, front-end read/write/mixed IOPS, sequential throughput, an estimated rebuild time and the probability of an unrecoverable read error (URE) during a single-drive rebuild. You can also compare two levels side by side.
How is RAID usable capacity calculated?
It depends on the level. RAID 0 gives n × drive size; RAID 5 gives (n−1) × size; RAID 6 and RAID-DP give (n−2) × size; RAID 10 gives n/2 × size; RAID 50/60 multiply per group; and RAIDZ removes parity per vdev then accounts for ZFS slop and allocation padding. The calculator applies the correct formula automatically and shows the parity overhead.
Why does it show both TB and TiB?
Drives are sold in decimal terabytes (1 TB = 10¹² bytes), but operating systems and ZFS report binary tebibytes (1 TiB = 2⁴⁰ bytes ≈ 1.1 TB). A 16 TB drive shows as 14.55 TiB. Showing both removes the single most common storage-sizing confusion — the “missing” capacity is just unit conversion.
How accurate is it, and where do the figures come from?
Capacity, efficiency and fault-tolerance are exact, industry-standard formulas — our ZFS capacity output matches the OpenZFS model. IOPS, throughput and rebuild time are clearly-labelled estimates based on conservative, user-overridable drive-class figures, because they depend on your exact drive model, controller and workload. Anything that is an estimate is badged as one.
What is the URE-on-rebuild risk?
When a drive fails, the array reads the surviving drives to rebuild it. An unrecoverable read error during that read can fail the rebuild. For single-parity levels (RAID 5, RAIDZ1) there is no redundancy left during rebuild, so a URE means data loss — the classic reason RAID 5 is risky on large drives. For dual/triple parity (RAID 6, RAIDZ2/3) a URE during a single-drive rebuild is recoverable, and the calculator says so.
Does it show pricing?
No — we do not publish prices. The calculator gives you the full technical picture for planning; for pricing, enter your email and we send itemised pricing for the exact drives and array within one business day. As a 24-year UK reseller, Servnet confirms the final specification on quote.