Efficient distributed hot sparing scheme in a parity declustered RAID organization

A system, comprising:

• a processor;
• a plurality of physical drives;
• and
• a memory coupled with the processor, the memory storing instructions which, when executed by the processor, cause the system to perform a process comprising;
• configuring a RAID group with a plurality of logical drives, wherein the RAID group provides data redundancy for the system;
• distributing the RAID group to the plurality of physical drives, wherein the plurality of logical drives are distributed in a parity-declustered layout on the plurality of physical drives;
• identifying a value of a maximum number of logical drives that are allowed to fail as indicated by a redundancy level associated with the configured RAID group;
• preallocating one or more units of spare space on each of the plurality of physical drives for use in failure recovery, wherein the number of units of spare space on each of the plurality of physical drives is dynamically allocated as a function of the identified value of the maximum number of logical drives as indicated by the redundancy level;
• and
• upon failure of a first physical drive in the plurality of physical drives leaving remaining of the plurality of physical drives as surviving physical drives, reconstructing data stored in the failed first physical drive to the one or more units of spare space of the surviving physical drives, wherein after reconstruction, the plurality of logical drives remain in the parity-declustered layout on the surviving physical drives.