Here is the reference table of the most commonly used RAID levels: principle, usable capacity, fault tolerance, minimum number of drives, and recommended usage. For n drives of identical capacity C.
RAID Levels Table
| Level | Principle | Usable Capacity | Fault Tolerance | Min. Drives | Usage |
|---|---|---|---|---|---|
| RAID 0 | Striping | 100% (n × C) | None | 2 | Performance, non-critical data |
| RAID 1 | Mirroring | 50% (C) | 1 drive | 2 | System, small server |
| RAID 5 | Distributed parity | (n − 1) × C | 1 drive | 3 | Good compromise |
| RAID 6 | Double parity | (n − 2) × C | 2 drives | 4 | Large arrays |
| RAID 10 | Mirroring + striping | 50% (n / 2 × C) | 1 per mirror | 4 | Databases, virtualization |
How to Read This Table
The usable capacity is the actual available space after redundancy. The fault tolerance indicates how many drives can fail without data loss. RAID 5 and RAID 6 offer the best capacity/security balance; RAID 10 prioritizes performance. For a precise calculation based on your drives, use our RAID calculator.
FAQ: RAID Levels
Which level offers the best compromise?
RAID 5 (high capacity, one failure tolerated) or RAID 6 (two failures) for large arrays.
Does RAID 10 lose half the capacity?
Yes, like RAID 1: it’s the cost of performance and write robustness.
Does RAID replace backups?
No. It protects against drive failure, not deletion or disasters.
To learn more, read our RAID levels guide, and browse our hard drives & SSDs, our servers, and our storage arrays.