Storage device array architecture with copyback cache

Claim 2

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units for storing information as stripes of blocks, including data blocks and associated error-correction blocks; (b) at least one copyback cache storage unit, divided into logical areas, for temporarily storing data blocks; (c) at least two storage unit controllers, coupled to the plurality of failure independent storage units and to the at least one copyback cache storage unit, each controller having a corresponding unique logical area in each copyback cache storage unit and including a controller buffer, each storage unit controller for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into at least one associated logical area of the copyback cache storage unit; (iii) during idle time of at least some of the plurality of failure independent storage units: (1) reading at least one pending data block from the controller buffer; (2) accessing at least one of the plurality of failure independent storage units and reading information corresponding to each read pending data block; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and associated error-correction block to a stripe of the idle storage units.

Claim 3

The fault tolerant storage device array of claim 1, wherein: (a) the storage unit controller acknowledges that each data block has been written to the array immediately after each pending data block is written to the copyback cache unit.

Claim 4

The fault-tolerant storage device array of claim 1, wherein each storage unit controller and the corresponding unique logical areas in each of the copyback cache storage units are associated with a unique logical array of the failure independent storage units.

Claim 5

The fault-tolerant storage device array of claim 3, wherein: (a) each storage unit controller is associated with a unique copyback cache storage unit; (b) each storage unit controller writes pending data blocks to a corresponding copyback cache storage unit if the corresponding copyback cache storage unit is available; if the corresponding copyback cache storage unit is unavailable, each storage unit controller writes pending data blocks to the corresponding unique logical area within another available copyback cache storage unit.

Claim 6

The fault-tolerant storage device array of claim 3, wherein: (a) at least one stripe of each logical array of the failure independent storage units is dedicated for storing pending data blocks, each logical array having a corresponding array controller selected from the at least two storage unit controllers; and (b) pending data blocks are written by each array controller to one such dedicated stripe if no copyback cache storage units are available.

Claim 7

The fault-tolerant storage device array of claim 4, wherein at least one copyback cache storage unit is used as a hot spare.

Claim 8

The fault-tolerant storage device array of claim 1, wherein: (a) the copyback cache storage unit includes a track buffer for receiving data blocks to be written to the copyback cache storage unit; and (b) the storage unit controller is further for acknowledging that each data block has been written to the array immediately upon completely writing a pending data block to a track buffer within a copyback cache storage unit.

Claim 9

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units logically configured as redundant logical arrays for storing information as stripes of blocks, including data blocks and associated error-correction blocks; (b) at least one copyback cache storage unit, divided into logical areas, for temporarily storing data blocks; (c) at least two storage unit controllers, coupled to the plurality of failure independent storage units and to the at least one copyback cache storage unit, each storage unit controller associated with a unique one of the logical areas and including a controller buffer and control means, the control means being adapted for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into at least one associated logical area within the at least one copyback cache storage unit, thereby establishing read pending data blocks, each read pending data block having corresponding information; (iii) during idle time of at least some of the plurality of failure independent storage units: (1) reading at least one read pending data block from the associated logical area of the at least one copyback cache storage unit, the associated logical area being associated with a reading storage unit controller; (2) accessing the failure independent storage units and reading the corresponding information of the read pending data blocks; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and the associated error-correction block to a stripe of the idle storage units.

Claim 10

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units logically configured as redundant logical arrays for storing information as stripes of blocks, including data blocks and associated error-correction blocks; (b) at least one copyback cache storage unit, divided into logical areas, for temporarily storing data blocks; (c) at least two storage unit controllers, coupled to the plurality of failure independent storage units and to the at least one copyback cache storage unit, each storage unit controller associated with a unique one of the logical areas and including a controller buffer and control means, the control means being adapted for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into at least one associated logical area within the at least one copyback cache storage unit, thereby establishing read pending data blocks, each read pending data block having corresponding information; (iii) during idle time of at least some of the plurality of failure independent storage units: (1) reading at least one read pending data block from the associated logical area of the at least one copyback cache storage unit, the associated logical area being associated with a reading storage unit controller; (2) accessing the failure independent storage units and reading the corresponding information of the read pending data blocks; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and the associated error-correction block to a stripe of the idle storage units.

Claim 11

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units logically configured as redundant logical arrays for storing information as stripes of blocks, including at least one stripe comprising data blocks and associated error-correction blocks and at least one stripe dedicated to temporarily storing pending data blocks; (b) at least one copyback cache storage unit, divided into logical areas, for temporarily storing data blocks; (c) at least two storage unit controllers, coupled to the plurality of failure independent storage units and to the at least one copyback cache storage unit, each storage unit controller associated with a unique one of the logical areas and including a controller buffer and control means, the control means being adapted for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into at least one associated logical area within the at least one copyback cache storage unit if the copyback cache storage unit is available; if the copyback cache storage unit is unavailable, writing received data blocks into a dedicated stripe within the plurality of storage units; (iii) during idle time of at least some of the plurality of storage units: (1) reading at least one pending data block from the associated logical area of the copyback cache storage unit if the copyback cache storage unit is available, the logical area being associated with a reading storage unit controller; if the copyback cache storage unit is unavailable, reading at least one pending data block from the dedicated stripe; (2) accessing the storage units and reading information corresponding to each read pending data block; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and associated error-correction block to a corresponding stripe of the idle storage units.

Claim 12

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units logically configured as redundant logical arrays for storing information as stripes of blocks, including at least one stripe comprising data blocks and associated error-correction blocks and at least one stripe dedicated to temporarily storing pending data blocks; (b) at least one copyback cache storage unit, divided into logical areas, for temporarily storing data blocks; (c) at least two storage unit controllers, coupled to the plurality of failure independent storage units and to the at least one copyback cache storage unit, each storage unit controller associated with a unique one of the logical areas and including a controller buffer and control means, the control means being adapted for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into at least one associated logical area within the at least one copyback cache storage unit if the copyback cache storage unit is available; if the copyback cache storage unit is unavailable, writing received data blocks into a dedicated stripe within the plurality of storage units; (iii) during idle time of at least some of the plurality of storage units: (1) reading at least one pending data block from the associated logical area of the copyback cache storage unit if the copyback cache storage unit is available, the logical area being associated with a reading storage unit controller; if the copyback cache storage unit is unavailable, reading at least one pending data block from the dedicated stripe; (2) accessing the storage units and reading information corresponding to each read pending data block; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and associated error-correction block to a corresponding stripe of the idle storage units.

Claim 13

The fault-tolerant storage device array of claim 9, wherein: (a) the at least one copyback cache storage unit includes a track buffer for receiving data blocks to be written to the at least one copyback cache storage unit; and (b) the at least two storage unit controllers are further for acknowledging that each data block has been written to the fault-tolerant storage device array immediately upon completely writing each pending data block to a track buffer within a copyback cache storage unit.

Claim 14

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units for storing information as stripes of blocks, including at least one stripe comprising data blocks and associated error-correction blocks and at least one stripe dedicated for temporarily storing pending data blocks; (b) a storage unit controller, coupled to the plurality of storage units, including a controller buffer for receiving pending data blocks and control means, coupled to the controller buffer, for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into the at least one dedicated stripe, thereby establishing read pending data blocks, each read pending data block having corresponding information; (iii) during idle time of at least some of the plurality of failure independent storage units: (1) reading at least one read pending data block from at least one dedicated stripe; (2) accessing the failure independent storage units and reading the corresponding information of the read pending data blocks; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and the associated error-correction block to a stripe of the idle failure independent storage units.

Claim 15

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units for storing information as stripes of blocks, including at least one stripe comprising data blocks and associated error-correction blocks and at least one stripe dedicated for temporarily storing pending data blocks; (b) a storage unit controller, coupled to the plurality of storage units, including a controller buffer for receiving pending data blocks and control means, coupled to the controller buffer, for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into the at least one dedicated stripe, thereby establishing read pending data blocks, each read pending data block having corresponding information; (iii) during idle time of at least some of the plurality of failure independent storage units: (1) reading at least one read pending data block from at least one dedicated stripe; (2) accessing the failure independent storage units and reading the corresponding information of the read pending data blocks; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and the associated error-correction block to a stripe of the idle failure independent storage units.

Claim 16

The fault-tolerant storage device array of claim 11, wherein: (b) at least one of the plurality of storage units includes a track buffer for receiving data blocks to be written to the at least one storage unit; and (c) the storage unit controller is further for acknowledging that each data block has been written to the fault-tolerant storage device array immediately upon completely writing each pending data block to a track buffer within a storage unit.

Claim 17

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units for storing information as stripes of blocks, including at least one stripe comprising data blocks and associated error-correction blocks and at least one stripe dedicated to temporarily storing pending data blocks; (b) a storage unit controller, coupled to the plurality of storage units, including a controller buffer for receiving pending data blocks and control means, coupled to the controller buffer, for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into at least one dedicated stripe, thereby establishing read pending data blocks, each read pending data block having corresponding information; (iii) during idle time of at least some of the plurality of the failure independent storage units: (1) reading at least one read pending data block from the controller buffer; (2) accessing the failure independent storage units and reading the corresponding information of the read pending data blocks; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and the associated error-correction block to a corresponding stripe of the idle failure independent storage units.

Claim 18

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units for storing information as stripes of blocks, including at least one stripe comprising data blocks and associated error-correction blocks and at least one stripe dedicated to temporarily storing pending data blocks; (b) a storage unit controller, coupled to the plurality of storage units, including a controller buffer for receiving pending data blocks and control means, coupled to the controller buffer, for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into at least one dedicated stripe, thereby establishing read pending data blocks, each read pending data block having corresponding information; (iii) during idle time of at least some of the plurality of the failure independent storage units: (1) reading at least one read pending data block from the controller buffer; (2) accessing the failure independent storage units and reading the corresponding information of the read pending data blocks; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and the associated error-correction block to a corresponding stripe of the idle failure independent storage units.

Claim 19

The fault-tolerant storage device array of claim 13, wherein: (a) at least one of the plurality of storage units includes a track buffer for receiving data blocks to be written to the at least one storage unit; and (b) the storage unit controller is further for acknowledging that each data block has been written to the array immediately upon completely writing each pending data block to a track buffer within a storage unit.

Claim 20

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units for storing information as stripes of blocks, including at least data blocks and associated error-correction blocks; (b) at least one copyback cache storage unit for temporarily storing data blocks, including a track buffer for receiving data to be stored in the copyback cache storage unit; (c) a storage unit controller, coupled to the plurality of failure independent storage units and to the at least one copyback cache storage unit, including a controller buffer and control means, the control means being adapted for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into the track buffer of at least one copyback cache storage unit, thereby establishing read pending data blocks, each read pending data block having corresponding information; (iii) immediately acknowledging that a data block has been written to the fault-tolerant storage device array each time a pending data block is completely received in the track buffer of the at least one copyback cache storage unit; (iv) during idle time of at least some of the plurality of failure independent storage units: (1) reading at least one pending data block from the at least one copyback cache storage unit; (2) accessing the plurality of failure independent storage units and reading the corresponding information of the read pending data blocks; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and the associated error-correction block to a stripe of the idle storage units.

Claim 21

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units for storing information as stripes of blocks, including at least data blocks and associated error-correction blocks; (b) at least one copyback cache storage unit for temporarily storing data blocks, including a track buffer for receiving data to be stored in the copyback cache storage unit; (c) a storage unit controller, coupled to the plurality of failure independent storage units and to the at least one copyback cache storage unit, including a controller buffer and control means, the control means being adapted for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into the track buffer of at least one copyback cache storage unit, thereby establishing read pending data blocks, each read pending data block having corresponding information; (iii) immediately acknowledging that a data block has been written to the fault-tolerant storage device array each time a pending data block is completely received in the track buffer of the at least one copyback cache storage unit; (iv) during idle time of at least some of the plurality of failure independent storage units: (1) reading at least one pending data block from the at least one copyback cache storage unit; (2) accessing the plurality of failure independent storage units and reading the corresponding information of the read pending data blocks; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and the associated error-correction block to a stripe of the idle storage units.

Claim 22

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units for storing information as stripes of blocks, including at least data blocks and associated error-correction blocks; (b) at least one copyback cache storage unit for temporarily storing data blocks, including a track buffer for receiving data to be stored in the copyback cache storage unit; (c) a storage unit controller, coupled to the plurality of failure independent storage units and to the at least one copyback cache storage unit, including a controller buffer and control means, the control means being adapted for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into the track buffer of at least one copyback cache storage unit, thereby establishing read pending data blocks, each read pending data block having corresponding information; (iii) immediately acknowledging that a data block has been written to the fault-tolerant storage device array each time a pending data block is completely received in the track buffer of the at least one copyback cache storage unit; (iv) during idle time of at least some of the plurality of failure independent storage units: (1) reading at least one pending data block from the controller buffer; (2) accessing the plurality of failure independent storage units and reading the corresponding information of the read pending data blocks; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and the associated error-correction block to a corresponding stripe of the idle storage units.

Claim 23

A fault-tolerant storage device array including: (a) a plurality of failure independent storage units for storing information as stripes of blocks, including at least data blocks and associated error-correction blocks; (b) at least one copyback cache storage unit for temporarily storing data blocks, including a track buffer for receiving data to be stored in the copyback cache storage unit; (c) a storage unit controller, coupled to the plurality of failure independent storage units and to the at least one copyback cache storage unit, including a controller buffer and control means, the control means being adapted for: (i) receiving data blocks into the controller buffer as pending data blocks; (ii) writing each pending data block into the track buffer of at least one copyback cache storage unit, thereby establishing read pending data blocks, each read pending data block having corresponding information; (iii) immediately acknowledging that a data block has been written to the fault-tolerant storage device array each time a pending data block is completely received in the track buffer of the at least one copyback cache storage unit; (iv) during idle time of at least some of the plurality of failure independent storage units: (1) reading at least one pending data block from the controller buffer; (2) accessing the plurality of failure independent storage units and reading the corresponding information of the read pending data blocks; (3) generating an associated error-correction block from the read information and each read pending data block; (4) writing each such read pending data block and the associated error-correction block to a corresponding stripe of the idle storage units.