Method and apparatus for rapid data restoration including on-demand output of sorted logged changes

Claim 3

The method of claim 1 , the changes being logged into the changed log according to chronological order of receipt, the method further including sorting the logged changes, where the application of the logged changes to the backup dataset comprises updating the backup dataset according to the sorted logged changes.

Claim 4

The method of claim 2 , the sorting of the logged changes comprising sorting the logged changes by subpart and then according to the chronological order of receipt of the changes by the storage system.

Claim 5

The method of claim 1 , the changes being logged into the change log by storage in multiple change logs, the method further including combining the multiple change logs, where the application of the logged changes to the backup dataset comprises updating the backup dataset according to the combined change log.

Claim 6

The method of claim 1 , the changes being logged into the change log according to chronological order of receipt in multiple change logs, the method further including combining and sorting the multiple change logs, where the application of the logged changes to the backup dataset comprises updating the backup dataset according to the combined and sorted change log.

Claim 7

The method of claim 5 , the combining and sorting of the multiple change logs further including storing the combined and sorted multiple change such that each constituent logged change is randomly accessible.

Claim 8

The method of claim 6 , the combined and sorted multiple change logs being stored in key sequence dataset (KSDS) format.

Claim 9

The method of claim 1 , the method further comprising: before applying the logged changes to the backup dataset, performing a second backup operation using the backup dataset to create a second backup dataset.

Claim 10

The method of claim 1 , the operation of providing an output of the subpart from the log comprising applying the changes not yet applied to the subpart of the backup dataset to the backup dataset, and then providing an output of the subpart from the backup dataset.

Claim 11

A signal-bearing medium tangibly embodying a program of machine-readable instructions executable by a digital processing apparatus to perform operations for restoring a backup dataset in a storage system, where contents of the backup dataset comprise contents of a corresponding original dataset upon creation of the backup dataset, and where changes to the original dataset received by the system after creating the backup dataset are logged by the system in a change log, the operations comprising: in response to unavailability of the original dataset, commencing application of the logged changes to the backup dataset; and during the application of the logged changes to the backup dataset, receiving one or more requests to access one or more subparts of the original dataset, and responsive to each request, for each subpart: determining whether the change log contains any changes not yet applied to the subpart of the backup dataset; and satisfying the request by: if the log contains one or more changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the log, the output comprising a most recent one of the logged changes not yet applied; if the log does not contains change any not yet applied to the subpart of the backup dataset, providing an output of the subpart from the backup dataset.

Claim 12

A signal-bearing medium tangibly embodying a program of machine-readable instructions executable by a digital processing apparatus to perform operations for restoring a backup dataset in a storage system, where contents of the backup dataset comprise contents of a corresponding original dataset upon creation of the backup dataset, and where changes to the original dataset received by the system after creating the backup dataset are logged by the system in a change log, the operations comprising: in response to unavailability of the original dataset, commencing application of the logged changes to the backup dataset; and during the application of the logged changes to the backup dataset, receiving one or more requests to access one or more subparts of the original dataset, and responsive to each request, for each subpart: determining whether the change log contains any changes not yet applied to the subpart of the backup dataset; and satisfying the request by: if the log contains one or more changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the log, the output comprising a most recent one of the logged changes not yet applied; if the log does not contains change any not yet applied to the subpart of the backup dataset, providing an output of the subpart from the backup dataset.

Claim 13

The medium of claim 10 , the changes being logged into the change log according to chronological order of receipt, the operations further including sorting the logged changes, where the application of the logged changes to the backup dataset comprises updating the backup dataset according to the sorted logged changes.

Claim 14

The medium of claim 11 , the sorting of the logged changes comprising sorting the logged changes by subpart and then according to the chronological order of receipt of the changes by the storage system.

Claim 15

The medium of claim 10 , the changes being logged into the change log by storage in multiple change logs, the operations further including combining the multiple change logs, where the application of the logged changes to the backup dataset comprises updating the backup dataset according to the combined change log.

Claim 16

The medium of claim 10 , the changes being logged into the change log according to chronological order of receipt by storage in multiple change logs, the operations further including combining and sorting the multiple change logs, where the application of the logged changes to the backup dataset comprises updating the backup dataset according to the combined and sorted change log.

Claim 17

The medium of claim 14 , the combining and sorting of the multiple change logs further including storing the combined and sorted multiple change such that each constituent logged change is randomly accessible.

Claim 18

The medium of claim 15 , the combined and sorted multiple change logs being stored in a key sequence dataset (KSDS) format.

Claim 19

The medium of claim 10 , the operations further comprising: before applying the logged changes to the backup dataset, performing a second backup operation using the backup dataset to create a second backup dataset.

Claim 20

The medium of claim 10 , the operation of providing an output of the subpart from the log comprising applying the changes not yet applied to the subpart of the backup dataset to the backup dataset, and then providing an output of the subpart from the backup dataset.

Claim 21

A computer-driven backup system, comprising: a data storage containing a backup dataset having contents that comprise contents of a corresponding original dataset upon creation of the backup dataset, the storage maintaining a change log containing changes to the original dataset received by the system after creating the backup dataset; and a storage controller coupled to the storage and programmed to restore the backup dataset by performing operations comprising: in response to unavailability of the original dataset, commencing application of the logged changes to the backup dataset; and during the application of the logged changes to the backup dataset, receiving one or more requests to access one or more subparts of the original dataset, and responsive to each request, for each subpart: determining whether the change log contains any changes not yet applied to the subpart of the backup dataset; and satisfying the request by: if the log contains one or more changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the log, the output comprising a most recent one of the logged changes not yet applied; if the log does not contains any changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the backup dataset.

Claim 22

A computer-driven backup system, comprising: a data storage containing a backup dataset having contents that comprise contents of a corresponding original dataset upon creation of the backup dataset, the storage maintaining a change log containing changes to the original dataset received by the system after creating the backup dataset; and a storage controller coupled to the storage and programmed to restore the backup dataset by performing operations comprising: in response to unavailability of the original dataset, commencing application of the logged changes to the backup dataset; and during the application of the logged changes to the backup dataset, receiving one or more requests to access one or more subparts of the original dataset, and responsive to each request, for each subpart: determining whether the change log contains any changes not yet applied to the subpart of the backup dataset; and satisfying the request by: if the log contains one or more changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the log, the output comprising a most recent one of the logged changes not yet applied; if the log does not contains any changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the backup dataset.

Claim 23

The system of claim 19 , the changes being logged into the change log according to chronological order of receipt, the operations further including sorting the logged changes, where the application of the logged changes to the backup dataset comprises updating the backup dataset according to the sorted logged changes.

Claim 24

The system of claim 20 , the sorting of the logged changes comprising sorting the logged changes by subpart and then according to the chronological order of receipt of the changes by the storage system.

Claim 25

The system of claim 19 , the changes being logged into the change log by storage in multiple change logs, the operations further including combining the multiple change logs, where the application of the logged changes to the backup dataset comprises updating the backup dataset according to the combined change log.

Claim 26

The system of claim 19 , the changes being logged into the change log according to chronological order of receipt by storage in multiple change logs, the operations further including combining and sorting the multiple change logs, where the application of the logged changes to the backup dataset comprises updating the backup dataset according to the combined and sorted change log.

Claim 27

The system of claim 23 , the combining and sorting of the multiple change logs further including storing the combined and sorted multiple change such that each constituent logged change is randomly accessible.

Claim 28

The system of claim 24 , the combined and sorted multiple change logs being stored in key sequence dataset (KSDS) format.

Claim 29

The system of claim 19 , the operations further comprising: before applying the logged changes to the backup dataset, performing a second backup operation using the backup dataset to create a second backup dataset.

Claim 30

The system of claim 19 , the storage controller being programmed such that the operation of providing an output of the subpart from the log comprises applying the changes not yet applied to the subpart of the backup dataset to the backup dataset, and then providing an output of the subpart from the backup dataset.

Claim 33

A signal-bearing medium tangibly embodying a program of machine-readable instructions executable by a digital processing apparatus to perform operations for restoring a backup dataset in a storage system, where contents of the backup dataset comprise contents of a corresponding original dataset upon creation of the backup dataset, and where changes to the original dataset received by the system after creating the backup dataset are logged by the system in a change log, the operations comprising: in response to unavailability of the original dataset, commencing application of the logged changes to the backup dataset; and during the application of the logged changes to the backup dataset, receiving one or more requests to access one or more subparts of the original dataset, and responsive to each request, for each subpart: determining whether the change log contains any changes not yet applied to the subpart of the backup dataset; and satisfying the request by: if the log contains one or more changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the log, the output comprising a most recent one of the logged changes not yet applied; if the log does not contains change any not yet applied to the subpart of the backup dataset, providing an output of the subpart from the backup dataset; responsive to receipt of each change to the original dataset, representing the change in an update map; upon commencing application of the logged changes to the backup dataset, setting an indicator; and responsive to application of a logged change to the backup dataset, removing representation of the logged change from the update map; where the operation of determining of whether the change log contains any changes not yet applied to the subpart of the backup dataset comprises determining whether the indicator is set, and if the indicator is set, determining whether the update map contains a representation of the subpart.

Claim 34

A signal-bearing medium tangibly embodying a program of machine-readable instructions executable by a digital processing apparatus to perform operations for restoring a backup dataset in a storage system, where contents of the backup dataset comprise contents of a corresponding original dataset upon creation of the backup dataset, and where changes to the original dataset received by the system after creating the backup dataset are logged by the system in a change log, the operations comprising: in response to unavailability of the original dataset, commencing application of the logged changes to the backup dataset; and during the application of the logged changes to the backup dataset, receiving one or more requests to access one or more subparts of the original dataset, and responsive to each request, for each subpart: determining whether the change log contains any changes not yet applied to the subpart of the backup dataset; and satisfying the request by: if the log contains one or more changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the log, the output comprising a most recent one of the logged changes not yet applied; if the log does not contains change any not yet applied to the subpart of the backup dataset, providing an output of the subpart from the backup dataset; responsive to receipt of each change to the original dataset, representing the change in an update map; upon commencing application of the logged changes to the backup dataset, setting an indicator; and responsive to application of a logged change to the backup dataset, removing representation of the logged change from the update map; where the operation of determining of whether the change log contains any changes not yet applied to the subpart of the backup dataset comprises determining whether the indicator is set, and if the indicator is set, determining whether the update map contains a representation of the subpart.

Claim 35

A computer-driven backup system, comprising: a data storage containing a backup dataset having contents that comprise contents of a corresponding original dataset upon creation of the backup dataset, the storage maintaining a change log containing changes to the original dataset received by the system after creating the backup dataset; and a digital data processing apparatus coupled to the storage and programmed to restore the backup dataset by performing operations comprising: in response to unavailability of the original dataset, commencing application of the logged changes to the backup dataset; and during the application of the logged changes to the backup dataset, receiving one or more requests to access one or more subparts of the original dataset, and responsive to each request, for each subpart: determining whether the change log contains any changes not yet applied to the subpart of the backup dataset; and satisfying the request by: if the log contains one or more changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the log, the output comprising a most recent one of the logged changes not yet applied; if the log does not contains any changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the backup dataset; responsive to receipt of each change to the original dataset, representing the change in an update map; upon commencing application of the logged changes to the backup dataset, setting an indicator; and responsive to application of a logged change to the backup dataset, removing representation of the logged change from the update map; and where the operation of determining of whether the change log contains any changes not yet applied to the subpart of the backup dataset comprises determining whether the indicator is set, and if the indicator is set, determining whether the update map contains a representation of the subpart.

Claim 36

A computer-driven backup system, comprising: a data storage containing a backup dataset having contents that comprise contents of a corresponding original dataset upon creation of the backup dataset, the storage maintaining a change log containing changes to the original dataset received by the system after creating the backup dataset; and a digital data processing apparatus coupled to the storage and programmed to restore the backup dataset by performing operations comprising: in response to unavailability of the original dataset, commencing application of the logged changes to the backup dataset; and during the application of the logged changes to the backup dataset, receiving one or more requests to access one or more subparts of the original dataset, and responsive to each request, for each subpart: determining whether the change log contains any changes not yet applied to the subpart of the backup dataset; and satisfying the request by: if the log contains one or more changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the log, the output comprising a most recent one of the logged changes not yet applied; if the log does not contains any changes not yet applied to the subpart of the backup dataset, providing an output of the subpart from the backup dataset; responsive to receipt of each change to the original dataset, representing the change in an update map; upon commencing application of the logged changes to the backup dataset, setting an indicator; and responsive to application of a logged change to the backup dataset, removing representation of the logged change from the update map; and where the operation of determining of whether the change log contains any changes not yet applied to the subpart of the backup dataset comprises determining whether the indicator is set, and if the indicator is set, determining whether the update map contains a representation of the subpart.