Remap raid to maintain raid level

1. A device comprising a processor, the processor comprising:

a detection unit to detect a failure of at least one of a plurality of drives of a redundant array of independent disks (RAID);

a remap unit to remap a remainder of the plurality of drives, if one of the plurality of drives fails; and

a mapping table, wherein

the remap unit is to maintain a RAID level of the RAID and to reduce a size of the RAID, if the remap unit remaps the remainder of the plurality of drives,

the remap unit is to shift logical block addresses (LBAs) of the RAID based on which of the plurality of drives failed,

the mapping table is to map the LBAs of the RAID to physical blocks of data stored at a main memory, and

the RAID is to be a cache for the main memory.

2. The processor of claim 1 , wherein the remap unit is to reduce a number of the LBAs while maintaining striping if the RAID level includes striping and at least one of the drives fails.

3. The processor of claim 1 , wherein the remap unit is to reduce a number of the LBAs while maintaining mirroring for the RAID if the RAID level includes mirroring and at least one of the drives fails.

4. The processor of claim 1 , wherein the remap unit is to reduce a number of the LBAs while maintaining parity if the RAID level includes parity and at least one of the drives fails.

5. The processor of claim 4 , wherein,

the remap unit is to remap a data block of the remainder of the plurality of drives to a parity block, if the failed drive included a parity block, and

the remap unit is to mark data located in a mapping table and associated with the remapped block as invalid.

6. The processor of claim 5 , wherein the remap unit is to calculate and store parity information to the remapped block based on an XOR calculation of data blocks having a same stripe as the remapped block.

7. The processor of claim 4 , wherein the remap unit is to update a parity block of at least one of the stripes based on a reduced RAID stripe size, if at least one of the drives fails.

8. The device of claim 1 , wherein the mapping table is to change the map between the LBAs and the physical blocks without shifting the physical blocks of data at the remainder of the plurality of drives, if one of the plurality of drives fails.

9. A method, comprising:

determining that at least one of a plurality of drives of a redundant array of independent disks (RAID) has failed; and

remapping a remainder of the plurality of drives to maintain a same RAID level and to reduce a number of logical blocks addresses (LBAs), wherein

the LBAs are to be mapped to the remainder of the plurality of drives and to be continuous,

the remapping is to shift the LBAs of the RAID based on which of the plurality of drives failed,

the remapping is to map, at a mapping table, the LBAs of the RAID to physical blocks of data stored at a main memory, and

the RAID is to be a cache for the main memory.

10. The method of claim 9 , wherein the remapping is to maintain at least one of mirroring, striping and parity for the RAID when the same RAID level is maintained.

11. The method of claim 9 , wherein the remapping is to calculate Previously presented parity information for the remainder of the drives based on which of the drives has failed.

12. A non-transitory computer-readable storage medium storing instructions that, if executed by a processor of a device, cause the processor to:

determine which of a plurality of drives of a redundant array of independent disks (RAID) has failed; and

remap a remainder of the drives of the RAID based on a RAID level of the RAID before the drive failed, wherein

the remainder of the drives are remapped to maintain the RAID level,

logical block addresses (LBAs) of the RAID are shifted based on which of the plurality of drives failed when the remainder of the drives are remapped,

the LBAs of the RAID are mapped at a mapping table to physical blocks of data stored at a main memory, and

the RAID is to be a cache for the main memory.

13. The non-transitory computer-readable storage medium of claim 12 , wherein the LBAs corresponding to the plurality of the drives are reduced when the RAID is remapped.