Managing data deduplication in storage systems based on I/O activities

• US 9,460,102 B1
• Filed: 12/26/2013
• Issued: 10/04/2016
• Est. Priority Date: 12/26/2013
• Status: Active Grant

• Current Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Original Assignee: EMC Corporation (Dell Technologies Inc.)
• Inventors: Ku, Mark K., Bono, Jean-Pierre, Duprey, Dennis T., Chen, Xiangping, Armangau, Philippe, Chaudhary, Monica
• Primary Examiner(s): AL HASHEMI, SANA A
• Application Number:
  US14/141,221
• Time in Patent Office: 1,013 Days
• Field of Search: 707/637, 707/687, 707/692, 707/694, 707/685, 707/758
• US Class Current: 1/1
• CPC Class Codes: G06F 16/1748 De-duplication implemented within the file system, e.g. based on file segments de-duplication techniques in storage systems for the management of data blocks G06F3/0641
• CPC Class Codes: G06F 3/0608 Saving storage space on storage systems
• CPC Class Codes: G06F 3/0641 De-duplication techniques
• CPC Class Codes: G06F 3/0683 Plurality of storage devices

Abstract

A method is used in managing data deduplication in storage systems based on I/O activities. I/O activity of first and second data objects is evaluated. The first and second data objects are selected for applying a deduplicating technique. Based on the evaluation, the deduplicating technique is applied to the first and second data objects.

Claims

1. A method for use in managing data deduplication in storage systems based on I/O activities, the method comprising:

evaluating input/output (I/O) operations activity of first and second data objects based on I/O access patterns for the first and second data objects, wherein the first and second data objects are selected for applying a deduplicating technique; and

based on the evaluation, applying the deduplicating technique to the first and second data objects, wherein applying the deduplicating technique includes updating mapping information of the first and second data objects, wherein the second data object is deduplicated to the first data object upon determining that I/O operations activity for the first data object is higher than I/O operations activity for the second data object, wherein the first data object is deduplicated to the second data object upon determining that I/O operations activity for the second data object is higher than I/O operations activity for the first data object.

2. The method of claim 1, wherein the first and second data objects are selected from the group consisting of a deduplication domain, a storage extent, a Logical Unit Number (LUN), a file and a data block, wherein the data block is a fixed size chunk of a physical disk storage.

3. The method of claim 2, wherein a deduplication domain comprises a set of storage extents, wherein each storage extent of the set of storage extents comprises a set of LUNs, each LUN of the set of LUNs is a logical representation of a subset of physical disk storage.

4. The method of claim 1, wherein a mapping pointer is associated with a data block, wherein the mapping pointer includes a distributed weight indicating whether the data block has been shared.

5. The method of claim 1, wherein a mapping pointer for a data block points to a Virtual Block Mapping object, wherein the Virtual Block Mapping object includes another mapping pointer to refer to a data block and a total distributed weight indicating whether the Virtual Block Mapping object has been shared.

6. The method of claim 1, further comprising:

based on the evaluation of I/O activity of the first and second data objects, determining whether to keep the first data object as a master deduplicated data object;

based on the determination, updating a Virtual Block Mapping object referred to by a mapping pointer of the second data object by changing the mapping pointer included in the Virtual Block Mapping object to point to a Virtual Block Mapping object referred to by a mapping pointer of the first data object; and

freeing the second data object.

7. The method of claim 6, further comprising:

updating a total distributed weight of the Virtual Block Mapping object referred to by the mapping pointer of the first data object.

8. The method of claim 1, wherein a master deduplicated copy is selected from the group consisting of the first and second data objects based on evaluation of the I/O activity of the first and second data objects.

9. The method of claim 1, wherein the first data object is selected as a master deduplicated copy upon determining that I/O activity of the first data object is more than I/O activity of the second data object.

10. A system for use in managing data deduplication in storage systems based on I/O activities, the system comprising:

first logic evaluating input/output (I/O) operations activity of first and second data objects based on I/O access patterns for the first and second data objects, wherein the first and second data objects are selected for applying a deduplicating technique; and

second logic applying, based on the evaluation, the deduplicating technique to the first and second data objects, wherein applying the deduplicating technique includes updating mapping information of the first and second data objects, wherein the second data object is deduplicated to the first data object upon determining that I/O operations activity for the first data object is higher than I/O operations activity for the second data object, wherein the first data object is deduplicated to the second data object upon determining that I/O operations activity for the second data object is higher than I/O operations activity for the first data object.

11. The system of claim 10, wherein the first and second data objects are selected from the group consisting of a deduplication domain, a storage extent, a Logical Unit Number (LUN), a file and a data block, wherein the data block is a fixed size chunk of a physical disk storage.

12. The system of claim 11, wherein a deduplication domain comprises a set of storage extents, wherein each storage extent of the set of storage extents comprises a set of LUNs, each LUN of the set of LUNs is a logical representation of a subset of physical disk storage.

13. The system of claim 10, wherein a mapping pointer is associated with a data block, wherein the mapping pointer includes a distributed weight indicating whether the data block has been shared.

14. The system of claim 10, wherein a mapping pointer for a data block points to a Virtual Block Mapping object, wherein the Virtual Block Mapping object includes another mapping pointer to refer to a data block and a total distributed weight indicating whether the Virtual Block Mapping object has been shared.

15. The system of claim 10, further comprising:

third logic determining, based on the evaluation of I/O activity of the first and second data objects, whether to keep the first data object as a master deduplicated data object;

fourth logic updating, based on the determination, a Virtual Block Mapping object referred to by a mapping pointer of the second data object by changing the mapping pointer included in the Virtual Block Mapping object to point to a Virtual Block Mapping object referred to by a mapping pointer of the first data object; and

fifth logic freeing the second data object.

16. The system of claim 15, further comprising:

sixth logic updating a total distributed weight of the Virtual Block Mapping object referred to by the mapping pointer of the first data object.

17. The system of claim 10, wherein a master deduplicated copy is selected from the group consisting of the first and second data objects based on evaluation of the I/O activity of the first and second data objects.

18. The system of claim 10, wherein the first data object is selected as a master deduplicated copy upon determining that I/O activity of the first data object is more than I/O activity of the second data object.