Find usable RAID capacity from drive count, drive size, hot-spare planning, and mixed-drive assumptions, then compare RAID 0, 1, 5, 6, and 10 by efficiency.
Last updated
Common storage presets
RAID level
Selected RAID result
12 TB
RAID 5 with 4 installed drives gives 12 TB usable storage from 16 TB raw capacity.
The active array is using 4 drives at an effective size of 4 TB with no parked hot spare.
16 TB
Total raw capacity
4 TB
Parity or mirror overhead
1
Drives that can fail
75%
Storage efficiency
10.9 TiB
Usable capacity in binary units
0 GB
Capacity parked as hot spare
Selected-level breakdown
Measure
Value
RAID level
RAID 5
Minimum drives
3
Installed raw capacity
16 TB (14.6 TiB)
Active array raw capacity
16 TB
Usable capacity
12 TB (10.9 TiB)
Reserved for parity or mirroring
4 TB
Reserved in binary units
3.64 TiB
Hot-spare capacity
0 GB
Effective active drive size
4 TB
Mixed-drive penalty
0 GB
Efficiency
75%
Capacity overhead
25%
Parity / mirror drives
1
Fault tolerance
Any 1 drive
Best fit
Best for balanced NAS capacity with one-drive redundancy.
Rebuild note
Rebuilds stress the remaining drives because the missing data must be reconstructed from parity.
Planner note
RAID 5 gives the best usable-space ratio in small arrays, but the rebuild window is still a single-disk-parity exposure.
Planning note RAID improves availability and fault tolerance, but it is still not a backup. Pair the array design with a separate backup plan and a rebuild-risk check for the drive count you expect to run. RAID 5 gives the best usable-space ratio in small arrays, but the rebuild window is still a single-disk-parity exposure.
Selected RAID planning scenarios
Scenario
Usable
Installed efficiency
Planning note
Current plan
12 TB
75%
The current hot-spare and active-drive assumptions used in the main result.
Add 1 hot spare
8 TB
50%
Shows the usable-capacity drop if one installed disk is parked as a ready-to-rebuild spare.
Same drives across RAID levels
Level
Usable
Efficiency
Fault tolerance
Planning note
RAID 0
16 TB
100%
No fault tolerance
RAID 0 is capacity-first and the hot-spare setting does not add protection because striping has no redundancy.
RAID 1
4 TB
25%
Up to 3 active drives in the mirror
RAID 1 keeps the planning model simple: one mirrored copy stays available while the other members cover failures.
RAID 5
12 TB
75%
Any 1 drive
RAID 5 gives the best usable-space ratio in small arrays, but the rebuild window is still a single-disk-parity exposure.
RAID 6
8 TB
50%
Any 2 drives
RAID 6 trades another drive worth of capacity for a second parity layer, which matters most in large-capacity NAS rebuilds.
RAID 10
8 TB
50%
1 drive per mirror pair
RAID 10 usually gives the cleanest rebuild behaviour because only the failed mirror pair has to be recopied.
RAID 0, 1, 5, 6, and 10 capacity, fault tolerance, and efficiency
A RAID calculator shows how much usable storage a disk array provides, how much parity overhead you are paying, and how many drives can fail before data is lost.
How each RAID level uses drive capacity
RAID stands for Redundant Array of Independent Disks. Each level trades raw capacity for a different mix of performance, redundancy, and minimum drive count. The usable capacity depends on how many drives are dedicated to parity or mirroring versus actual data storage.
RAID 0 — striping across all drives. Usable = all drives. No fault tolerance: any single drive failure loses all data.
RAID 1 — full mirror. Usable = one drive size regardless of drive count. All but one drive may fail.
RAID 5 — distributed parity. Usable = (n − 1) drives. One drive may fail. Minimum 3 drives.
RAID 6 — double parity. Usable = (n − 2) drives. Two drives may fail simultaneously. Minimum 4 drives.
RAID 10 — mirrored stripes. Usable = half the raw capacity. One drive per mirror pair may fail. Requires even drive count.
Choosing the right RAID level
RAID 5 is the most common choice for general file storage where capacity efficiency and single-drive redundancy are both important. RAID 6 adds a second layer of protection at the cost of one more drive, which matters most during the rebuild window after a drive failure — when a second drive is at elevated risk.
RAID 10 has higher rebuild speed and better write performance than parity-based arrays, making it a preferred choice for databases and high-write workloads. RAID 0 provides maximum usable capacity with no overhead but should only be used where the data is either expendable or backed up elsewhere.
RAID is not a backup. A hardware failure, ransomware attack, or accidental deletion affects all drives in the array simultaneously. Regular backups stored separately are still essential regardless of RAID level.
The highest usable-capacity answer is not always the best operational answer. A RAID calculator is more useful when it compares the array you can build against the one you can realistically maintain: how much usable space you lose by choosing RAID 6 over RAID 5, whether RAID 10 buys you a simpler rebuild path, and whether a hot spare is worth parking for the workload you are trying to protect.
RAID planning for NAS rebuilds and mixed drive sizes
Most people searching for a RAID calculator are really trying to answer a planning question: should this NAS use RAID 5, RAID 6, or RAID 10? The right answer depends on whether you care most about usable storage, rebuild safety, or write performance. RAID 5 usually maximises usable space from a small number of drives, RAID 6 is a better fit when rebuild risk matters, and RAID 10 is often the safer choice for write-heavy workloads.
Mixed drive sizes can work, but they usually waste capacity because the array behaves like the smallest drive sets the baseline. If your storage plan relies on reusing older disks, the calculator is most useful when you compare each RAID level side by side before committing to a configuration.
That is also why the calculator separates the nominal drive size from the smallest active drive in the set. The nominal figure tells you what you intended to buy or install; the smallest-drive figure tells you what the active array will actually behave like once one undersized disk participates in parity or mirroring.
Further reading
TrueNAS ZFS storage primer — Official TrueNAS reference explaining practical storage redundancy tradeoffs and why parity versus mirrored layouts behave differently during rebuilds.
Hot spares, rebuild windows, and why parked capacity can still be worth it
A hot spare is an installed disk that is not contributing to current usable capacity. Instead, it waits for an active member to fail so the array can begin rebuilding sooner. A RAID calculator with hot-spare planning is useful because it shows the real tradeoff: less usable storage today in exchange for a smaller operational gap between a drive failure and the rebuild starting.
That tradeoff matters most in parity arrays. RAID 5 can survive one failed disk, but after that first failure it is exposed until rebuild completes. RAID 6 keeps one more layer of parity during rebuild, which is why many large-capacity NAS designs choose it even though the headline usable TB is lower. RAID 10 often gives up more raw capacity than parity layouts, but rebuilds are usually simpler because one mirror pair is being recopied instead of reconstructing parity across the whole set.
Further reading
IBM — RAID types — Reference overview of RAID layout behaviour and the tradeoffs between parity, mirroring, and striping.
Decimal TB versus binary TiB in RAID capacity planning
Drive vendors usually label disks in decimal GB or TB, where 1 TB means 1,000 GB. Operating systems often display binary GiB or TiB, where 1 TiB is 1,024 GiB. Both units are valid; they are just describing the same bytes with different measurement systems.
That means a RAID capacity calculator can look inconsistent if it only shows one unit system. The decimal total is useful when you are planning purchases from drive labels. The binary total is useful when you want a closer match to what many NAS interfaces and operating systems display after the array is built.
Worked example: 4 × 4 TB drives
With four 4 TB drives, raw capacity is 16 TB before redundancy. RAID 0 uses all 16 TB but has no fault tolerance, RAID 1 gives only 4 TB because every block is mirrored, RAID 5 gives 12 TB with one-drive protection, and RAID 6 gives 8 TB with two-drive protection.
That same hardware also makes RAID 10 easy to compare: it delivers 8 TB usable because half the raw capacity is dedicated to mirroring. A RAID calculator is most useful when you compare the same drive set across multiple levels instead of looking at only one mode in isolation.
If the same four-bay system parks one drive as a hot spare, the planning question changes. RAID 5 with three active 4 TB drives gives 8 TB usable instead of 12 TB, but the spare is ready for faster rebuild response. If one of the active drives were only 3 TB, the effective baseline for the active set becomes 3 TB and usable capacity falls again. That is why the calculator exposes both the smallest active drive and the hot-spare count as separate planning controls.
Frequently asked questions
Does RAID replace backups?
No. RAID protects against drive hardware failure, not against data loss from accidental deletion, ransomware, fire, theft, or controller failure. The 3-2-1 backup rule — three copies, two media types, one offsite — applies even when RAID is in use.
Can I mix different drive sizes in a RAID array?
Most RAID controllers allow mixed drive sizes, but usable capacity is based on the smallest drive in the array. All drives behave as if they are the same size as the smallest one. Using identically-sized drives maximises usable capacity.
What happens during a RAID rebuild?
When a failed drive is replaced, the controller reconstructs the missing data from the remaining drives and writes it to the new drive. During this rebuild the array is in a degraded state with no redundancy. RAID 5 rebuilds are slower and put stress on remaining drives; RAID 6 keeps one layer of protection during a rebuild.
Why does RAID 10 show less usable storage than RAID 5?
RAID 10 mirrors half the drives, so only half the raw capacity is usable. RAID 5 uses one drive worth of parity instead, which usually gives more usable space from the same hardware. The tradeoff is that RAID 10 typically rebuilds faster and handles write-heavy workloads better.
When should I choose RAID 6 over RAID 5?
Choose RAID 6 when rebuild safety matters more than the extra drive worth of usable capacity. It adds a second parity layer, which makes the array more resilient if another drive fails during rebuild.
What does a hot spare do in a RAID calculator?
A hot spare reserves one installed disk outside the active array so rebuild can start sooner after a failure. It reduces current usable capacity because that disk is not storing live data, but it can reduce the time the array spends waiting for a replacement drive.
Why does the smallest active drive change the result?
Most RAID layouts normalise participating members to the smallest active drive. If one active disk is smaller, every parity or mirror calculation has to fit within that smaller baseline, which strands some of the capacity on the larger disks.
Is RAID 5 still safe for large disks?
RAID 5 is still common, but larger disks increase the rebuild window and the amount of data that must be read after a failure. That is why many capacity-heavy NAS plans compare RAID 5 against RAID 6 or RAID 10 rather than choosing the highest-usable-space answer automatically.
Why does my NAS show TiB instead of TB?
Drive labels are usually decimal TB, while many operating systems and NAS interfaces report binary TiB or GiB. The total bytes are the same; the unit system changes the displayed number.
Does a hot spare increase fault tolerance immediately?
Not in the same way parity or mirroring does. A hot spare does not add another live copy of your data while the array is healthy. Its main benefit is that it can shorten the time between a failure and a rebuild starting.
Should I use RAID 10 instead of RAID 6 for a write-heavy server?
Often yes, if write performance and rebuild simplicity matter more than raw storage efficiency. RAID 10 usually gives up more usable capacity than RAID 6, but it avoids parity calculations and typically rebuilds by copying one mirror pair instead of reconstructing the whole array.
Why is RAID 1 so space-inefficient with many drives?
In an n-way mirror every active drive holds the same data, so usable capacity stays equal to one drive size no matter how many members are in the mirror. The extra drives improve redundancy, not usable space.
