Not taking hardware RAID cards into account, what kind of theoretical performance improvements should I see when using 4 disks in a RAID-10 array vs 1 disk by itself?
For example, on RAID-0, I would expect to see something like
singleDrivePerformance * driveCount, as reads and writes are spread out pretty evenly across all physical disks in the array. On RAID-1, I'd expect to see
singleDrivePerformance / 2. Does this mean that with RAID-10, I should see double the performance of a single drive, albeit with parity?