RAID 0
Striping without redundancy
At least 2 drives. Data is spread across the drives, so all raw capacity is usable. There is no fault tolerance: a single drive failure can cause loss of the array.
Storage, made clear
Calculate usable RAID capacity, storage efficiency and redundancy overhead.
≈ 21.83 TiB
4 × 8 TB drives in RAID 5
Multiple drive failures may be tolerated if they occur in different mirror pairs.
Estimated RAID-level capacity. Formatting, filesystem metadata and system reservations can reduce the space available for files.
Correct the input to see an estimated usable capacity.
You’ll need 4 × 8 TB drives.
| RAID | Usable capacity | Efficiency | Protection |
|---|
Mirror-based protection depends on which drives fail. RAID 10 tolerates at least one drive failure.
Start with the number of drives and the capacity of each drive. Multiplying them gives the raw capacity: the total storage before RAID redundancy. The RAID level determines how much of that total can hold your data.
Mirroring stores copies of data. Parity stores information that can help reconstruct data after a drive fails. The capacity used for either is the RAID overhead.
For example, four 8 TB drives in RAID 5 provide 32 TB raw capacity, 24 TB usable capacity and 8 TB overhead: 75% storage efficiency.
RAID 0
At least 2 drives. Data is spread across the drives, so all raw capacity is usable. There is no fault tolerance: a single drive failure can cause loss of the array.
RAID 1
Exactly 2 drives in this calculator. Both drives hold the same data, leaving one drive’s capacity usable. The array can tolerate 1 drive failure.
RAID 5
At least 3 drives. The equivalent of one drive’s capacity is used for distributed parity. Usable capacity is (N − 1) × drive size, with tolerance for 1 drive failure.
Use the RAID 5 calculator →RAID 6
At least 4 drives. The equivalent of two drives’ capacity is used for distributed parity. Usable capacity is (N − 2) × drive size, with tolerance for 2 drive failures.
Use the RAID 6 calculator →RAID 10
At least 4 drives and an even number of drives. Half of the raw capacity is usable. At least 1 drive failure is tolerated. Multiple drive failures may be tolerated if they occur in different mirror pairs.
Use the RAID 10 calculator →Drive manufacturers use decimal units: 1 TB is 10¹² bytes and 1 GB is 10⁹ bytes. Binary units use powers of two: 1 TiB is 2⁴⁰ bytes. As a result, 24 TB is approximately 21.83 TiB.
Some operating systems show binary capacity, sometimes with a TB label. That smaller-looking number does not mean bytes have disappeared. Filesystem formatting and system reservations can separately reduce the space available for files.
Multiply drive capacity by the number of drives available for data. RAID 0 uses all drives; RAID 1 uses one of two; RAID 5 reserves the equivalent of one drive; RAID 6 reserves two; RAID 10 uses half.
Mirroring keeps duplicate data and parity stores recovery information. Both use space that cannot also hold unique data. RAID 0 has no redundancy overhead.
A TB is 1,000,000,000,000 bytes. A TiB is 1,099,511,627,776 bytes. The same storage therefore has a smaller numerical value in TiB.
RAID 5 requires at least 3 drives. This calculator assumes all drives have the same capacity.
No. RAID 5 tolerates one drive failure. A second drive failure before redundancy is restored can cause data loss.
RAID 10 can offer better write performance because it does not need parity calculations. Actual performance depends on the drives, controller and workload; this calculator estimates capacity only.
No. RAID redundancy helps with certain drive failures. Keep separate backups to protect against deletion, malware, theft, fire and other causes of data loss.