Method for removing duplicate data from a storage array

• US 8,930,307 B2
• Filed: 09/30/2011
• Issued: 01/06/2015
• Est. Priority Date: 09/30/2011
• Status: Active Grant

• Current Assignee: Pure Storage, Inc.
• Original Assignee: Pure Storage, Inc.
• Inventors: Colgrove, John, Hayes, John, Miller, Ethan, Hasbani, Joseph S., Sandvig, Cary
• Primary Examiner(s): Chen, Susan
• Application Number:
  US13/250,570
• Publication Number:
  US 20130086006A1
• Time in Patent Office: 1,194 Days
• Field of Search: 707/650, 707/692, 707/802, 707/522, 707/784, 707/791, 711/103, 711/114, 711/137, 711/154, 711/216, 711/135, 711/170, 715/205, 726/24
• US Class Current: 707/610
• CPC Class Codes: G06F 16/137 Hash-based content-based indexing of textual data G06F16/31
• CPC Class Codes: G06F 16/1752 based on file chunks
• 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 3/0688 Non-volatile semiconductor memory arrays

Abstract

A system and method for efficiently removing duplicate data blocks at a fine-granularity from a storage array. A data storage subsystem supports multiple deduplication tables. Table entries in one deduplication table have the highest associated probability of being deduplicated. Table entries may move from one deduplication table to another as the probabilities change. Additionally, a table entry may be evicted from all deduplication tables if a corresponding estimated probability falls below a given threshold. The probabilities are based on attributes associated with a data component and attributes associated with a virtual address corresponding to a received storage access request. A strategy for searches of the multiple deduplication tables may also be determined by the attributes associated with a given storage access request.

Claims

1. A computer system comprising:

a non-transitory data storage medium;

a first fingerprint table comprising a first plurality of entries and a second fingerprint table comprising a second plurality of entries, wherein each entry of the first and the second plurality of entries is configured to store a fingerprint corresponding to a data component already stored in the system, wherein the first fingerprint table has fewer entries than the second fingerprint table and wherein the second fingerprint table comprises a fingerprint for at least one deduplicated data components not included in the first fingerprint table; and

a data storage controller comprising hardware;

wherein in response to receiving a write request, the data storage controller is configured to;

search the first fingerprint table during inline deduplication prior to the second fingerprint table based on a first fingerprint corresponding to the write request;

in response to detecting a hit on a matching entry in the first fingerprint table during said search;

write a reference to the data corresponding to the matching entry in the first table; and

in response to detecting a miss in the first fingerprint table during said search;

postpone further deduplication to offline deduplication; and

write data corresponding to the write request in the data storage medium.

2. The system as recited in claim 1, wherein in response to detecting a miss in the first fingerprint table during the search, the data component is stored in the data storage medium prior to any search of the second deduplication table with the first hash.

3. The system as recited in claim 2, wherein the first fingerprint table and the second fingerprint table are stored across a hierarchy of different storage media.

4. The system as recited in claim 3, wherein the data storage controller is further configured to restrict a search of the first fingerprint table or the second fingerprint table to a subset of the hierarchy of different storage media and a number of accesses to a given level of the hierarchy based on attributes of a corresponding data component and characteristics of the storage media.

5. The system as recited in claim 1, wherein in response to detecting a match between a first fingerprint and a second fingerprint, the data storage controller is configured to compare a first data component corresponding to the first fingerprint with a second data component corresponding to the second fingerprint.

6. The system as recited in claim 5, wherein in response to detecting that the content of the first data component and the content of the second data component are the same, the data storage controller is further configured to:

store a virtual-to-physical address mapping corresponding to the write in a first mapping table; and

store a reverse physical-to-virtual address mapping corresponding to the write in a second mapping table.

7. The computer system as recited in claim 1, wherein in response to initiation of a deduplication operation corresponding to given data stored in the data storage medium, the data storage controller is configured to:

determine a likelihood of the given data being deduplicated;

if said likelihood exceeds a given threshold, store an entry corresponding to the given data in the first fingerprint table rather than the second fingerprint table; and

if said likelihood does not exceed the given threshold, store an entry corresponding to the given data in the second fingerprint table rather than the first fingerprint table.

8. A method comprising:

maintaining a first fingerprint table comprising a first plurality of entries and a second fingerprint table comprising a second plurality of entries in a computer system, wherein each entry of the first and the second plurality of entries is configured to store a fingerprint corresponding to a data component already stored in the system, wherein the first fingerprint table has fewer entries that the second fingerprint table and wherein the second fingerprint table comprises a fingerprint for at least one deduplicated data components not included in the first fingerprint table;

in response to receiving a storage access request;

searching the first fingerprint table during inline deduplication prior to the second fingerprint table based on a first fingerprint corresponding to the write request;

in response to detecting a hit on a matching entry in the first fingerprint table during said search;

write a reference to the data corresponding to the matching entry in the first table; and

in response to detecting a miss in the first fingerprint table during said search;

postpone further deduplication to offline deduplication; and

write data corresponding to the write request in the data storage medium.

9. The method as recited in claim 8, wherein in response to detecting a miss in the first fingerprint table during the search, the data component is stored in the data storage medium prior to any search of the second fingerprint table with the first hash.

10. The method as recited in claim 9, wherein the first fingerprint table and the second fingerprint table are stored across a hierarchy of different storage media.

11. The method as recited in claim 10, further comprising restricting a search of the first fingerprint table or the second fingerprint table to a subset of the hierarchy of different storage media and a number of accesses to a given level of the hierarchy based on attributes of a corresponding data component and the characteristics of the storage media.

12. The method as recited in claim 8, wherein in response to detecting a match between a first fingerprint and a second fingerprint, the method further comprises comparing a first data component corresponding to the first fingerprint with a second data component corresponding to the second fingerprint.

13. The method as recited in claim 12, wherein in response to detecting that the content of the first data component and the content of the second data component are the same, the method further comprises:

storing a virtual-to-physical address mapping corresponding to the write in a first mapping table; and

storing a reverse physical-to-virtual address mapping corresponding to the write in a second mapping table.

14. A non-transitory computer readable storage medium comprising program instructions, wherein said program instructions are executable to:

maintain a first fingerprint table comprising a first plurality of entries and a second fingerprint table comprising a second plurality of entries in a computer system, wherein each entry of the first and the second plurality of entries is configured to store a fingerprint corresponding to a data component already stored in the system, wherein the first fingerprint table has fewer entries than the second fingerprint table and wherein the second fingerprint table comprises a fingerprint for at least one deduplicated data components not included in the first fingerprint table;

in response to receiving a storage access request;

search the first fingerprint table prior to the second fingerprint table based on a first fingerprint corresponding to the write request;

in response to detecting a hit on a matching entry in the first fingerprint table during said search;

write a reference to the data corresponding to the matching entry in the first table; and

in response to detecting a miss in the first fingerprint table during said search;

postpone further deduplication to offline deduplication; and

write data corresponding to the write request in the data storage medium.

15. The storage medium as recited in claim 14, wherein in response to detecting a miss in the first fingerprint table during the search, the data component is stored in the data storage medium prior to any search of the second fingerprint table with the first hash.

16. The storage medium as recited in claim 15, wherein the first fingerprint table and the second fingerprint table are stored across a hierarchy of different storage media.

17. The storage medium as recited in claim 16, wherein the program instructions are further executable to restrict a search of the first fingerprint table or the second fingerprint table to a subset of the hierarchy of different storage media and a number of accesses to a given level of the hierarchy based on attributes of a corresponding data component and characteristics of the storage media.

18. The storage medium as recited in claim 14, wherein in response to detecting a match between a first fingerprint and a second fingerprint, the program instructions are further executable to compare a first data component corresponding to the first fingerprint with a second data component corresponding to the second fingerprint.

19. The storage medium as recited in claim 18, wherein in response to detecting that the content of the first data component and the content of the second data component are the same, the program instructions are further executable to:

store a virtual-to-physical address mapping corresponding to the write in a first mapping table; and

store a reverse physical-to-virtual address mapping corresponding to the write in a second mapping table.