NVMe Hot spare protection concepts

NVMe Hot spare devices are spare devices which reside on a system to replace failed NVMe devices.

The NVMe Hot spare protects NVMe devices which contain mirror-protected disk units. The NVMe Hot spare device automatically replaces a failed NVMe.

When a failure occurs on an NVMe device that contains mirror protected disk units, the system licensed internal code locates an NVMe Hot spare device which has the correct attributes:

• adequate storage capacity to clone the mirror protected disk units under the failing NVMe device
• the same blocksize as the failing NVMe device
• adequate namespace capacity to match the failing NVMe device
• identical drive-write-per-day characteristics as the failing NVMe device

The system licensed internal code automatically does the following procedure:

1. suspend every mirror protected disk unit under the failing NVMe device
2. stop the Hot spare designation on the NVMe Hot spare device
3. create namespaces (disk units) on the former NVMe Hot spare device. These new disk units exactly match the capacity of the suspended disk units that reside on the failing NVMe device
4. zero out the pages of the newly created disk units
5. synchronize the data in the disk units (copy the data from each Active member of the mirror set over to the replacement disk unit)