Week 03 · lesson
Lesson 3: RAID Is Not Backup
Core path: 42 minutes
RAID can improve performance, fault tolerance, or both.
It does not automatically protect your data from everything.
That distinction matters enough to make it the title.
What RAID does
RAID combines multiple physical drives into an array with a defined data layout.
Common levels include:
- RAID 0 — striping for performance, no fault tolerance;
- RAID 1 — mirroring, usually survives one drive failure in a two-drive mirror;
- RAID 5 — striping with single parity, minimum three drives;
- RAID 6 — striping with double parity, minimum four drives;
- RAID 10 — mirrored pairs combined with striping, minimum four drives.
What RAID does not do
A mirrored array can reproduce accidental deletion perfectly.
Ransomware can encrypt every mounted copy.
Fire, theft, electrical damage, or controller problems can affect the whole system.
So:
redundancy is not the same thing as backup.
A common backup principle is 3-2-1:
- 3 copies of important data;
- 2 different media or storage systems;
- 1 copy kept off-site or otherwise isolated from the primary failure domain.
Compare the tradeoffs
| RAID | Main idea | Minimum drives | Typical fault tolerance |
|---|---|---|---|
| 0 | striping | 2 | none |
| 1 | mirroring | 2 | one drive in a two-drive mirror |
| 5 | single parity | 3 | one drive |
| 6 | double parity | 4 | two drives |
| 10 | striped mirrors | 4 | depends on which drives fail |
Do not memorize the table without reasoning about what happens when a drive disappears.
Array failure evidence
A RAID problem can surface through:
- controller/drive LED status indicators;
- a degraded or failed-array status;
- a missing array;
- an audible alarm from supported controllers/enclosures;
- one or more missing drives;
- degraded read/write performance during failure/rebuild.
A loud alarm is not a diagnosis by itself. Record the controller/array state before changing disks.
Guided case
A RAID 5 array remains online but reports one failed member and an audible alert.
A weak response is:
It still works, so ignore it.
A stronger response is:
The array is operating in a degraded state with reduced fault tolerance. Preserve/verify backup status, identify the failed member from controller evidence, follow the approved replacement/rebuild procedure, then verify the array returns to a healthy state.
Student action
Choose one scenario:
- esports lab file server;
- media archive workstation;
- small office shared storage;
- personal gaming PC.
Recommend one RAID level or recommend no RAID at all.
Your answer must include:
- what problem you are solving;
- capacity cost;
- fault-tolerance behavior;
- what happens after a drive failure;
- which status evidence you would inspect;
- why a separate backup is still required.
Evidence checkpoint
If your explanation contains the phrase "RAID backs up the files," revise it.
RAID may keep a system operating through some drive failures. Backup is a separate recovery strategy.
Read it. Prove it.