Your Drives
RAID IS NOT A BACKUP
RAID keeps a server running when a drive dies. It does not protect against deletion, ransomware, fire, theft or controller faults — those all replicate straight onto the array. Every RAID volume still needs its own backup, ideally off-site or immutable. If the array matters to the business, talk to Triset about backup as well.
Your Setup
| Array | |
| Drives |
Result
| Usable capacity | |
| Raw capacity | |
| Storage efficiency | |
| Survives simultaneous failures of |
Comparison for the Same Drives
| Level | Usable | Efficiency | Can lose |
|---|
HOW THE NUMBERS WORK
- RAID 0 — speed, zero safety. All drives stripe together: 4 × 4 TB = 16 TB usable, full efficiency. One drive fails and everything is gone. Only for throwaway cache and scratch disks.
- RAID 1 — the honest mirror. Exactly two drives, each a full copy of the other: 2 × 4 TB = 4 TB usable, 50% efficiency. Either drive can fail. Simple, fast to rebuild, the right answer for small servers and NVRs.
- RAID 5 — the small-business default. One drive's worth of parity spread across the set: (n−1) × size usable. Four 4 TB drives give 12 TB usable at 75% efficiency, and any single drive can fail without data loss.
- RAID 6 — protection for big drives. Two parity drives: (n−2) × size usable, any two drives can fail. With 10 TB+ drives, a rebuild can take a day or more — long enough that a second failure during rebuild is a genuine risk, which is why big arrays should be RAID 6.
- RAID 10 — speed with mirrors. Drives paired into mirrors, striped across pairs: (n÷2) × size usable, needs an even count (4+). Survives one failure per pair and rebuilds quickly because only one small mirror copies at a time. The pick for busy databases and virtualisation hosts.
- And always: RAID is availability, not backup. The comparison table on the plan makes that point wherever it's printed, because it's the one that gets forgotten.