Quiz: RAID

← Back to quiz index

7 questions

L1 (4 questions)

1. RAID 5 is degraded and another disk has a predictive failure. What do you do?

Show answer

Do NOT wait. Replace the predictive failure disk first (if hot spare isn't active). During rebuild, a second failure kills the array. Consider RAID 6 or RAID 10 for critical data.

2. What is the difference between RAID 1, RAID 5, and RAID 10?

Show answer

RAID 1: mirroring, 50% usable capacity, survives 1 disk loss. RAID 5: striping with parity, (N-1) usable disks, survives 1 disk loss, write penalty from parity calc. RAID 10: striped mirrors, 50% capacity, survives 1 disk per mirror pair, best write performance.

3. RAID is not a backup. Explain why.

Show answer

RAID protects against disk failure, not data loss. Accidental deletion, corruption, ransomware, and controller failure affect all RAID members. Backups are separate copies stored independently, ideally offsite.

4. What is the practical difference between RAID 5 and RAID 6 for fault tolerance?

Show answer

RAID 5 tolerates 1 disk failure (single parity). RAID 6 tolerates 2 simultaneous failures (double parity). With modern large disks (4TB+), RAID 5 rebuilds take hours — during which a second failure destroys the array. RAID 6 is strongly recommended for drives >1TB. Trade-off: RAID 6 loses 2 disks of capacity vs 1 for RAID 5.

L2 (3 questions)

1. What is the RAID write penalty and why does it matter for RAID 5?

Show answer

RAID 5 write penalty is 4 (each write = 2 reads + 2 writes for parity). RAID 10 penalty is 2 (1 write to each mirror). For write-heavy workloads (databases), RAID 5 throughput is significantly worse.

2. Why is RAID 5 risky with large modern disks?

Show answer

Rebuild time on large disks (4TB+) can take hours to days. During rebuild, a second disk failure or an unrecoverable read error (URE) destroys the array. RAID 6 (dual parity) or RAID 10 is safer for large disks.

3. A RAID 5 array is rebuilding after a disk replacement. What is the risk and how do you mitigate it?

Show answer

During rebuild, every sector of every surviving disk is read. Unrecoverable Read Errors (URE) on large disks (spec: 1 per 10^14 bits ~= 1 per 12TB read) can cause a second failure, destroying the array. Mitigations:
1. Use RAID 6 or RAID
10.
2. Use enterprise disks with lower URE rates.
3. Patrol reads to detect bad sectors proactively.
4. Maintain hot spares for immediate rebuild.