Policy Based Data Retrieval Performance for Deduplicated Data

• US 20110289281A1
• Filed: 05/24/2010
• Published: 11/24/2011
• Est. Priority Date: 05/24/2010
• Status: Active Grant

• Current Assignee: Quantum Corporation (Chi Ko Investment Co., Ltd.)
• Original Assignee: Quantum Corporation (Chi Ko Investment Co., Ltd.)
• Inventors: Spackman, Stephen P.
• Granted Patent: US 8,244,992 B2
• Time in Patent Office: Days
• US Class Current: 711/154
• CPC Class Codes: G06F 12/0223 User address space allocation, e.g. contiguous or non contiguous base addressing

Abstract

A method that includes, by one or more computer systems, determining a data retrieval rate policy based on at least one data retrieval rate parameter. The method also includes determining at least one storage subsystem performance parameter. The method further includes determining a fragmentation value based on the data retrieval rate policy and the at least one storage subsystem performance parameter. The method additionally includes determining a storage subsystem fragmentation of a first data object. The storage subsystem fragmentation includes fragmenting the first data object into a plurality of first data object fragments. The method also includes deduplicating the first data object based on the fragmentation value and the storage subsystem fragmentation.

Claims

1. A method comprising, by one or more computer systems:

determining a data retrieval rate policy based on at least one data retrieval rate parameter;

determining at least one storage subsystem performance parameter;

determining a fragmentation value based on the data retrieval rate policy and the at least one storage subsystem performance parameter;

determining a storage subsystem fragmentation of a first data object, the storage subsystem fragmentation fragmenting the first data object into a plurality of first data object fragments; and

deduplicating the first data object based on the fragmentation value and the storage subsystem fragmentation.

2. The method of claim 1, further comprising receiving a user specified data retrieval rate parameter.

3. The method of claim 1, wherein determining at least one storage subsystem performance parameter comprises determining at least one of a data transfer rate associated with the storage subsystem and a data access latency time associated with the storage subsystem.

4. The method of claim 1, wherein:

determining a fragmentation value comprises determining an excess read capacity budget based on the data retrieval rate policy and the at least one storage subsystem performance parameter; and

deduplicating the first data object comprises;

reducing the excess read capacity budget for by a first cost for each fragment of the plurality of first data object fragments; and

determining a second cost associated with a deduplication of the first data object, wherein the first data object is deduplicated upon the second cost being less than or equal to the reduced excess read capacity budget.

5. The method of claim 4, wherein the first cost is based in part on an actual seek time for access between two of the plurality of first data object fragments as stored by the storage subsystem.

6. The method of claim 1, wherein deduplicating the first data object is further based on a priority associated with a plurality of potential deduplication sub-blocks.

7. The method of claim 1, wherein each fragmentation has a cost associated therewith, wherein:

upon a fragmentation for a deduplication of the first data object occurring within a consecutive block of data, the cost comprises a seek time to the deduplicated data and a seek time to return to the consecutive block of data; and

upon the fragmentation occurring at the end of one of the plurality of first data object fragments, the cost comprises a seek time to the next first data object fragment.

8. One or more computer-readable storage media embodying instructions that are operable when executed by one or more computer systems to:

determine a data retrieval rate policy based on at least one data retrieval rate parameter;

determine at least one storage subsystem performance parameter;

determine a fragmentation value based on the data retrieval rate policy and the at least one storage subsystem performance parameter;

determine a storage subsystem fragmentation of a first data object, the storage subsystem fragmentation fragmenting the first data object into a plurality of first data object fragments; and

deduplicate the first data object based on the fragmentation value and the storage subsystem fragmentation.

9. The computer-readable storage media of claim 8, wherein the instructions are further operable to receive a user specified data retrieval rate parameter.

10. The computer-readable storage media of claim 8, wherein the instructions operable to determine at least one storage subsystem performance parameter are further operable to determine at least one of a data transfer rate associated with the storage subsystem and a data access latency time associated with the storage subsystem.

11. The computer-readable storage media of claim 8, wherein the instructions operable to:

determine a fragmentation value are further operable to determine an excess read capacity budget based on the data retrieval rate policy and the at least one storage subsystem performance parameter; and

deduplicate the first data object are further operable to;

reduce the excess read capacity budget for by a first cost for each fragment of the plurality of first data object fragments; and

determine a second cost associated with a deduplication of the first data object, wherein the first data object is deduplicated upon the second cost being less than or equal to the reduced excess read capacity budget.

12. The computer-readable storage media of claim 11, wherein the first cost is based in part on an actual seek time for access between two of the plurality of first data object fragments as stored by the storage subsystem.

13. The computer-readable storage media of claim 8, wherein the instructions operable to deduplicate the first data object are further operable to deduplicate the first data object based on a priority associated with a plurality of potential deduplication sub-blocks.

14. The computer-readable storage media of claim 8, wherein each fragmentation has a cost associated therewith, wherein:

upon a fragmentation for a deduplication of the first data object occurring within a consecutive block of data, the cost comprises a seek time to the deduplicated data and a seek time to return to the consecutive block of data; and

upon the fragmentation occurring at the end of one of the plurality of first data object fragments, the cost comprises a seek time to the next first data object fragment.

15. One or more computer-readable storage media embodying data that was stored on the media at least in part by:

determining a data retrieval rate policy based on at least one data retrieval rate parameter;

determining at least one storage subsystem performance parameter;

determining a fragmentation value based on the data retrieval rate policy and the at least one storage subsystem performance parameter;

determining a storage subsystem fragmentation of a first data object, the storage subsystem fragmentation fragmenting the first data object into a plurality of first data object fragments; and

deduplicating the first data object based on the fragmentation value and the storage subsystem fragmentation.

16. The media of claim 15, wherein the data was stored on the media at least further in part by receiving a user specified data retrieval rate parameter.

17. The media of claim 15, wherein determining at least one storage subsystem performance parameter comprised determining at least one of a data transfer rate associated with the storage subsystem and a data access latency time associated with the storage subsystem.

18. The media of claim 15, wherein:

determining a fragmentation value comprised determining an excess read capacity budget based on the data retrieval rate policy and the at least one storage subsystem performance parameter; and

deduplicating the first data object comprised;

reducing the excess read capacity budget for by a first cost for each fragment of the plurality of first data object fragments; and

determining a second cost associated with a deduplication of the first data object, wherein the first data object is deduplicated upon the second cost being less than or equal to the reduced excess read capacity budget.

19. The media of claim 18, wherein the first cost is based in part on an actual seek time for access between two of the plurality of first data object fragments as stored by the storage subsystem.

20. The media of claim 15, wherein deduplicating the first data object was further based on a priority associated with a plurality of potential deduplication sub-blocks.

21. The media of claim 15, wherein each fragmentation had a cost associated therewith, wherein:

upon a fragmentation for a deduplication of the first data object occurring within a consecutive block of data, the cost comprised a seek time to the deduplicated data and a seek time to return to the consecutive block of data; and

upon the fragmentation occurring at the end of one of the plurality of first data object fragments, the cost comprised a seek time to the next first data object fragment.