You are configuring a new database server that will use eight identical 1 TB 12 Gb/s SAS drives in the same backplane. Management wants maximum transaction performance, fast rebuilds, and protection from disk failures, so you create a hardware RAID 10 array with all eight drives. Which outcome should you expect after the volume finishes building?
About 6 TB of usable capacity because two disks are reserved for distributed parity.
About 4 TB of usable capacity, but rebuilding a failed disk requires reading every surviving drive, making recovery slower than RAID 6.
About 4 TB of usable capacity and the array stays online as long as no more than one drive fails in each mirrored pair.
About 8 TB of usable capacity and protection against any two simultaneous drive failures, regardless of which disks fail.
RAID 10 stripes data across several mirrored pairs. Because every block is written twice (once to each disk in a pair), only half of the aggregate raw capacity is available for data, so eight 1 TB disks yield about 4 TB of usable space. Fault-tolerance depends on the mirror sets: the array can survive multiple simultaneous failures provided no two failed disks are in the same pair-effectively one disk per mirror. If both members of a pair fail, that stripe is lost and the whole array goes offline. Parity-based levels such as RAID 5/6 would offer more usable capacity but slower writes and slower, full-disk rebuilds because every surviving drive must be read to regenerate parity; RAID 10 rebuilds require only the surviving partner in the affected mirror. Therefore, the expected result is 50 % usable capacity (≈4 TB) with tolerance of one disk failure per mirrored pair.