RAID & Synology SHR Calculator: Real Usable Capacity

Mixed drive sizes: SHR, RAID 5 and RAID 6

Conventional RAID accepts drives of different sizes. Classic RAID on Synology sizes every member to the smallest drive in the pool; space above it is unused. (Synology, checked 2026-10-07)

Synology states pool capacity as: RAID 5 = (drives minus 1) x smallest drive; RAID 6 = (drives minus 2) x smallest; RAID 10 = (drives / 2) x smallest. (Synology, checked 2026-10-07)

Synology Hybrid RAID (SHR) splits the drives into size tiers and protects each tier on its own, so larger drives add space when enough of them reach a tier. That makes it more capacity-efficient than conventional RAID with mixed sizes, but not every combination uses every terabyte: the calculator shows what is left unused. Synology's own RAID Calculator counts SHR-2 capacity only where at least four drives reach a size tier; capacity above the fourth largest drive is counted as unused. (Synology RAID Calculator, checked 2026-10-07)

Worked examples: the same drives in four layouts

Layout capacity before platform reserves, decimal TB. SHR-1 and RAID 5 survive one failed drive; SHR-2 and RAID 6 survive two.

Upgrading 4 x 4 TB to 8 TB drives, one at a time

• On an SHR pool of equal sized drives, capacity grows only after at least two drives (SHR-1) or four drives (SHR-2) have been replaced with larger ones, one at a time. (Synology, checked 2026-10-07)
• On an SHR pool of mixed sizes, Synology says to replace the smallest drives with drives at least as large as the largest drive already in the pool. The planner refuses a smaller replacement on a mixed SHR pool rather than guess what DSM would do. (Synology, checked 2026-10-07)
• On classic RAID (not SHR) with equal sized drives, every drive must be replaced with a larger one before capacity grows. (Synology, checked 2026-10-07)
• On classic RAID with mixed sizes, replacing a drive other than the smallest does not change capacity; the smallest must be replaced first. (Synology, checked 2026-10-07)
• On QNAP the Expand Capacity step becomes available only after every disk in the RAID group has been replaced and rebuilt. (QNAP, checked 2026-10-07)
• Synology Storage Manager replaces one drive at a time. Each replacement has to finish rebuilding before the next one starts. (Synology, checked 2026-10-07)

Single or dual parity?

There is no drive size where single parity stops being reasonable. What matters is how long your array would run degraded and what a loss would cost you. These push towards dual parity (SHR-2, RAID 6, RAIDZ2):

• a long rebuild, which grows with drive size and with how busy the NAS is while it rebuilds;
• a long wait for a replacement drive, because the array runs without redundancy until the rebuild finishes;
• more drives sharing the risk during that window, and drives of the same age;
• no tested backup of the data.

The cost is one more drive of capacity, which this calculator shows. To put numbers on the window for your own array, use the RAID Rebuild Time and Risk Estimator and the Drive Failure Risk Estimator.

How we calculate

1. Raw capacity is the sum of the drive sizes you enter, in decimal TB (1 TB = 1,000 GB). Spare drives are set aside first. A spare must be able to stand in for the drives it protects: on SHR it must be at least as large as your largest drive; on classic RAID and ZFS at least as large as the smallest drive left in the pool. The calculator picks the smallest drives that qualify.
2. Layout capacity follows the RAID type. RAID 5 is (drives minus 1) x the smallest drive, RAID 6 (drives minus 2) x the smallest, RAID 10 half the drives x the smallest. RAID 1 is one drive's worth when every drive holds a copy, or half the drives when they are mirrored in pairs. JBOD and RAID 0 use every drive in full, the way Synology's RAID Calculator counts RAID 0. What a layout cannot use is shown as unused, not hidden.
3. SHR is counted tier by tier: each slice between two drive sizes adds (drives that reach it minus 1) x the slice for SHR-1, or minus 2 for SHR-2, when enough drives reach it. A tier too few drives reach is unused.
4. Unraid: your largest drive (two for dual parity) is parity; usable space is the sum of the other drives.
5. ZFS (one vdev): nominal capacity is (drives minus parity) x the smallest drive. The result shows the space ZFS reports, which counts RAIDZ space as if every block were 128K, minus the slop space ZFS holds back.
6. Platform reserves come from the vendor's own figures, each on its own line: Synology system partitions per drive and file system metadata; QNAP storage pool, RAID group and ext4 metadata reserves (a range); ZFS slop space. A snapshot reserve counts only if you enter one. Other platforms show the layout maths only.
7. Available for your data is layout capacity minus those reserves and any other space you enter.
8. Plan on storing applies your free-space percentage once, to what is available. It is your choice, not a vendor reserve.
9. Expansion applies each replacement or added drive in turn with the platform's own rule (Synology, QNAP, Unraid, ZFS) and recalculates the layout after each step.

Assumptions NTK made

• RAID 0 counts every drive in full on every platform, as Synology's RAID Calculator does. Some systems stripe at the smallest drive instead, which would give less on mixed sizes.
• QNAP's 1% RAID group reserve is applied to the raw capacity of the group, at most 150 GB: QNAP's sentence does not say raw or usable, and raw is the larger figure.
• Unraid has no hot spares, so on an Unraid array the spare drives you enter are your smallest drives, left out of the array; the largest drives become parity.
• Rebuild speed ranges in the planner are NTK planning assumptions shared with the RAID Rebuild Time and Risk Estimator.
• QNAP is modelled as QTS with ext4. QuTS hero (ZFS) is not modelled. ZFS is one vdev; for several vdevs, record sizes and sector size use the ZFS vdev calculator.
• ZFS assumes 4 KiB sectors (ashift 12), no special, log or cache devices, no compression and no dataset quotas or reservations. The result is the free space ZFS reports. How much data actually fits depends on record size: on RAIDZ, small records carry proportionally more parity than the 128K reporting assumes, so a pool of small files holds less. RAIDZ expansion (adding a disk to an existing RAIDZ vdev) needs OpenZFS 2.3.0 or later.

Sources this calculator uses

Vendor figures and rules checked at source: 2026-10-06 to 2026-10-14.

Worked example: Synology SHR, 4 x 4 TB, Btrfs, 10% free space kept

Raw capacity 16 TB. SHR-1 uses one drive's worth for redundancy: layout capacity 12 TB. The system partitions take 0.04 TB and Btrfs metadata 0.48 TB, leaving 11.48 TB available. Keeping 10% free, plan on storing about 10.33 TB. In binary units the available space is about 10.44 TiB.

AU NAS drive market prices

NAS-grade drives sold by Australian retailers, with the current market price for each. Compare individual retailers on Staticice.

Market prices, last changed 23 October 2026. Each figure is the typical going rate across the Australian retailers we track, checked daily against their listings. A drive stocked by fewer than 3 retailers shows no price. Confirm the price with the retailer before you buy.

Warranty: a drive's warranty term is the manufacturer's own, so check its datasheet. Your first step for a faulty product is usually the retailer that sold it. The ACCC explains your rights under the Australian Consumer Law at accc.gov.au. The array runs with less redundancy until a replacement arrives and rebuilds, so a spare drive on the shelf shortens that wait.

Frequently Asked Questions