What is RAID 10?

• A. Mirrored stripes offering high performance and redundancy at the cost of capacity efficiency.
• B. Striping without mirroring, no redundancy.
• C. A parity-based RAID level emphasizing data integrity.
• D. A software-only RAID mode.

Answer: (A)

RAID 10 combines mirroring with striping: data is written to mirrored pairs, and those mirrors are striped across to other pairs. This gives the speed benefits of striping (through parallel I/O) while preserving data through mirroring (each piece of data has a duplicate). Because mirroring duplicates data, you lose usable capacity equal to the number of mirrors, so total usable space is about half of the raw capacity. You also gain redundancy: you can survive a single drive failure in each mirrored pair, and still stay online as long as not both drives in the same mirror fail.

That description matches the idea of mirrored stripes with high performance and redundancy, at the cost of capacity efficiency. The other options describe different RAID concepts—striping without redundancy (no protection), parity-based RAID (uses parity rather than mirroring), or a mode that implies software-only implementation (RAID 10 can be hardware or software, but isn’t defined as software-only).