# RAID Redundancy

Record: `arch:raid-redundancy`
Kind: technique
Layer: [Correctness Core](https://banes-lab.com/records/layer/correctness-core.md)
Severity: mandatory for managed infrastructure
Scope: storage, redundancy, infrastructure
Canonical: https://banes-lab.com/ontology#arch-raid-redundancy

## Repair

- Refactored by: Place data on a mirrored or parity RAID array (RAID 1/5/10)
- Detected by: total data loss when one drive fails
- Violated by: durable data written to a single disk with no physical redundancy
- Measured by: tolerated simultaneous disk failures
- Enforced by: storage architecture review

## requires

- [Multiple Physical Disks](https://banes-lab.com/records/lex/multiple-physical-disks.md)

## reinforces

- [Redundancy](https://banes-lab.com/records/arch/redundancy.md)
- [Fault Tolerance](https://banes-lab.com/records/arch/fault-tolerance.md)

## enables

- [Disk-Failure Survival](https://banes-lab.com/records/lex/disk-failure-survival.md)
- [Parity-Based Recovery](https://banes-lab.com/records/lex/parity-based-recovery.md)

## conflicts-with

- [Single-Disk Point of Failure](https://banes-lab.com/records/lex/single-disk-point-of-failure.md)

## tensions-with

- [Write Amplification](https://banes-lab.com/records/lex/write-amplification.md)

## tensions

- [RAID Redundancy Write Amplification](https://banes-lab.com/records/tension/raid-redundancy-write-amplification.md)