Computer system with greater than fifteen drive fault tolerance
AssigneeHewlett-Packard Development Company, L.P.
InventorPurna C. Murthy
Life of this patent
Estimating the term…
Abstract
A computer system with an array of disk drives is disclosed. The drive array is capable of supporting greater than 15 drive fault tolerance accomplished by using coefficients for parity equations spanning greater than eight bits.
Claims22
A computer system comprising:
- a processor;
- a main memory coupled to the processor via a host bridge device;
- and
- an array controller coupled to said processor via a secondary expansion bus, wherein said array controller is adapted to perform greater than four bit encrypted resultant multiplication in multiplying parity equation coefficients and write data to determine error correction data for storage device fault recovery.
A method of operating a plurality of storage devices in an array having multiple device fault tolerance comprising:
- calculating error correction information using greater than four bit
- coefficients to facilitate said fault tolerant recovery capability.
An array controller that couples to a host computer system, comprising:
- a processor;
- a main memory coupled to said processor;
- a bridge device coupled to said processor by way of an array controller bus, said bridge device adapted to couple the array controller to a host computer system;
- an interface circuit coupled to said processor, said interface circuit couples to a plurality of storage devices;
- and
- said array controller adapted to calculate parity information using greater than four bit coefficients.
In a computer system having an array of storage devices coupled to an array controller capable of multiple device fault recovery, a method of operating the computer system comprising:
- writing data from said computer system to said array controller;
- dividing said data into a plurality of data subsets;
- calculating error correction information based on values of each of the data subsets;
- writing said data subsets and error correction information in a striped fashion across all the devices of the array of storage devices;
- and
- said calculating error correction information using coefficients whose values are represented by greater than four bits.
In an array controller having data to be stored on an array of storage devices, a method of operating said array controller comprising:
- dividing said data into a plurality of data subsets;
- calculating error correction information based on the value of each data subset;
- writing said data subsets and error correction information in a striped fashion across all the devices of the array of storage devices;
- and
- said calculating error correction information using coefficients whose values are represented by greater than four bits.
Dependencies are inferred from the claim text.