hesela.devRAID 01
RAID 01 mirrors complete RAID 0 stripe sets, placing the redundancy layer above the striping layer.
RAID 0+1 reverses the nesting order of conventional RAID 1+0, which stripes over mirror groups. [1] In a two-side layered implementation, one failed member makes its stripe set unavailable as a complete component; the other side supplies the volume. [2]
With stripe sets A/C and B/D, losing A and C leaves B/D intact, but losing A and D leaves neither complete side. This example assumes initially synchronized copies, readable current survivors and no rebuild. It describes component availability, not permanent destruction of every block. [4]
An interface can display mirrored stripe sets while software recognizes and mirrors corresponding components individually. Check the implementation and member map before inferring failure tolerance from the label. [3]
Relations
Limitations
- A capacity or nesting label does not establish repair duration, controller availability or a measured failure probability. [4]
Citations
- SNIA Swordfish 1.2.4a: RAIDType placement definitions
- Oracle Linux 10: Software RAID Levels
- Solaris Volume Manager: Providing RAID-1+0 and RAID-0+1
- Hesela: RAID 0+1 and RAID 1+0 component availability
Sources & caveats
- Source synthesis and an explicitly scoped four-member calculation, not a hardware experiment. The Solaris example is implementation-specific.