The attributes behind your decision
- Decision input
- Measured workload
- Primary variables
- Capacity, control, operations
- Validation
- Representative testing
- Next step
- dedicated servers
RAID changes availability, not backup history
RAID can keep storage available after selected device failures. It does not protect against deletion, corruption, compromise or loss of the entire server. Maintain separate, tested backups.
Compare mirrors and parity layouts
RAID 1 mirrors data and offers simple recovery characteristics. RAID 5 and 6 use parity with different capacity and failure tolerance. RAID 10 combines mirrors and striping and requires additional devices.
Include rebuild time and workload pressure
Large drives and busy arrays may rebuild slowly. Performance and exposure during rebuild matter as much as nominal failure tolerance, especially for write-heavy applications.
Calculate usable capacity before ordering
Use device count and size to estimate raw and usable capacity, then reserve room for filesystem overhead, snapshots, logs and growth. Validate controller and operating-system support.
Worked example: four equal disks
Four 2 TB disks provide 8 TB of raw decimal capacity. RAID 10 provides about 4 TB before filesystem overhead; RAID 5 about 6 TB; RAID 6 about 4 TB. These are capacity calculations, not performance or availability guarantees, and assume equal-sized disks with no hot spare.
Choose the layout alongside the workload’s write pattern, controller or software support and rebuild procedure. Test monitoring for a degraded array and document replacement responsibilities. RAID can help tolerate some disk failures; it does not undo accidental deletion, corruption or a compromised administrator account. Keep separately recoverable backups.