Data protection in a random access disk array

1. A disk array memory system comprising

a plurality of disks in a disk array for storage of content data and parity data in stripes, content data in a same stripe sharing parity bits of said parity data, each stripe having a spare stripe capacity including at least some of a predefined array spare capacity, said array spare capacity providing a dynamic space reserve over said array to permit data recovery following a disk failure event;

a cache for caching content data for aggregation prior to writing to said disk array;

a controller configured to select a stripe currently having a largest spare stripe capacity, for a current write operation of aggregated data from said cache, thereby to write all said aggregated data of said current write operation on a same stripe, thereby to maximize sharing of parity bits per write operation and minimize separate parity write operations.

2. The disk array memory system of claim 1 , further comprising a table for mapping of physical locations of said content data as defined by respective write operations with corresponding logical locations of said data.

3. The disk array of claim 2 , wherein said data is written on said disks in blocks and wherein a granularity of said table is a size of said blocks.

4. The disk array memory system of claim 2 , wherein said controller is configured to use said table to balance write operations between respective disks of said array.

5. The disk array memory system of claim 2 , wherein said controller is configured to use said table to divert a current write operation from a disk which is temporarily unavailable.

6. The disk array memory system of claim 1 , wherein said disk array has a stripe size, said stripe size being differentially variable.

7. The disk array memory system of claim 6 , wherein said disks in said array have respectively different capacities, said stripe size being such as to provide for even distribution of stripes over said disks, in proportion to said respective capacity at each disk.

8. The disk array memory system of claim 7 , said stripes having parity columns for said parity bits, said parity columns being distributed evenly over said disks.

9. The disk array memory system of claim 1 , wherein said controller is configured to use said parity data to recover content data lost in said disk failure event and to write said recovered content data into said dynamic space reserve.

10. The disk array memory system of claim 1 , wherein said disk array is arranged into blocks, and wherein each block is assigned one of three states, in use, not in use and not in parity, said not in parity state allowing for ignoring physical data and treating it as zeroed out.

11. The disk array memory system of claim 1 , configured as a RAID 6 array using two parity columns, said two columns being distributed in respectively different directions.

12. The disk array memory system of claim 1 , wherein different amounts of spare stripe capacities are defined for respective sections, thereby to provide variable levels of performance over said array.

13. A disk array memory system comprising

a plurality of disks in a disk array for storage of content data and parity data in stripes, content data in a same stripe sharing parity bits of said parity data,

a cache for caching content data prior to writing to said disk array;

a controller configured to select a stripe for a current write operation of data from said cache, thereby to write all said content data of said current write operation and corresponding parity data on a same stripe at a same physical location, and irrespective of logical relationships of parts of said content data; and the controller utilizing a mapping table to map between physical locations and logical relationships of said content data.

14. A disk array memory method comprising

providing a plurality of disks in a disk array,

storing content data and parity data in stripes within said disk array, wherein content data in a same stripe shares parity bits of said parity data,

for said array defining an array spare capacity, said array spare capacity providing a dynamic space reserve over said array to permit data recovery following a disk failure event;

caching content data to aggregate data from multiple write operations into a single aggregated write operation prior to writing to said disk array;

selecting a stripe having a largest spare stripe capacity;

writing said aggregated cached data in a single write operation to said selected stripe, thereby to maximize sharing of parity bits per write operation and minimize separate parity write operations.

15. The disk array memory method of claim 14 , comprising mapping, using a table, of physical locations of said content data as defined by respective write operations with corresponding logical locations of said data.

16. The disk array method of claim 15 , wherein said table has a granularity of a data block.

17. The disk array memory method of claim 15 , comprising using said table to balance write operations between respective disks of said array.

18. The disk array memory method of claim 15 , comprising using said table to divert a current write operation from a disk which is temporarily unavailable.

19. The disk array memory method of claim 14 , wherein said disk array has a stripe size, said stripe size being differentially variable.

20. The disk array memory method of claim 19 , wherein said disks in said array have respectively different capacities, said stripe size being such as to provide for even distribution of stripes over said disks, in proportion to said respective capacity at each disk.

21. The disk array memory method of claim 20 , said stripes having parity columns for said parity bits, said parity columns being distributed evenly over said disks.

22. The disk array memory method of claim 14 , comprising using said parity data to recover content data lost in said disk failure event and to write said recovered content data into said dynamic space reserve.

23. The disk array memory method of claim 14 , wherein said disk array is arranged into blocks, and wherein each block is assigned one of three states, in use, not in use and not in parity, said not in parity state allowing for ignoring physical data and treating it as zeroed out.

24. The disk array memory method of claim 14 , configured as a RAID 6 array using two parity columns, said two columns being distributed in respectively different directions.

25. The disk array memory method of claim 14 , wherein different amounts of spare stripe capacities are defined for respective sections, thereby to provide variable levels of performance over said array.

26. A disk array memory method comprising

providing a plurality of disks in a disk array for storage of content data and parity data in stripes, content data in a same stripe sharing parity bits of said parity data,

caching content data prior to writing to said disk array;

selecting a stripe for a current write operation of data from said cache, thereby to write all said content data of said current write operation and corresponding parity data on a same stripe at a same physical location, and irrespective of logical relationships of parts of said content data; and

utilizing a mapping table to map between physical locations and logical relationships of said content data.