Electronic system with declustered data protection by parity based on reliability and method of operation thereof

• US 10,365,836 B1
• Filed: 01/27/2015
• Issued: 07/30/2019
• Est. Priority Date: 01/27/2015
• Status: Active Grant

• Current Assignee: JP Morgan Chase Bank, N.A. (JP Morgan Chase & Co)
• Original Assignee: Western Digital Technologies Incorporated (Western Digital Corporation)
• Inventors: Xu, Jun, Xi, Wei, Mackey, Grant, Chu, Sanghoon, Yu, Jie
• Primary Examiner(s): Warren, Tracy A
• Application Number:
  US14/607,055
• Time in Patent Office: 1,645 Days
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 3/0619 in relation to data integrity, e.g. data losses, bit errors
• CPC Class Codes: G06F 3/0644 Management of space entities, e.g. partitions, extents, pools
• CPC Class Codes: G06F 3/0647 Migration mechanisms
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD

Abstract

An apparatus includes: an adaptive declustered RAID array configured of data storage devices (DSDs), the DSDs comprise data chunks allocated as data, a local parity, or a global parity; and a controller configured to generate a reliability indicator reflective of a reliability status of at least a portion of the adaptive declustered RAID array for reallocating the data chunks by dynamically increasing or decreasing the data chunks allocated as the local parity, the global parity, or a combination thereof.

Claims

1. An apparatus, comprising:

an adaptive declustered RAID array comprising data storage devices (DSDs), the DSDs comprising data chunks, each of the data chunks allocated as a corresponding one of data, a local parity, and a global parity, the local parity to protect the data, the global parity to protect the local parity; and

processing circuitry configured to;

generate, at a first time according to a sampling frequency, a first reliability indicator, the first reliability indicator indicating a first reliability status of at least a portion of the adaptive declustered RAID array;

generate, at a second time after the first time according to the sampling frequency, a second reliability indicator, the second reliability indicator indicating a second reliability status of the at least the portion of the adaptive declustered RAID array;

compare the first reliability indicator and the second reliability indicator;

determine whether a difference between the first reliability indicator and the second reliability indicator is greater than a predetermined threshold; and

in response to determining that the difference is greater than the predetermined threshold;

evaluate upgrade and downgrade configuration rules to determine whether to upgrade or to downgrade the adaptive declustered RAID array; and

reallocate the data chunks, by dynamically increasing or decreasing the data chunks allocated as the local parity, the global parity, or a combination thereof, according to the evaluation of the upgrade and downgrade configuration rules.

2. The apparatus of claim 1, further comprising:

DSD reliability managing circuitry in one of the DSDs; and

a reliability communication interface connected between the processing circuitry and the DSD reliability managing circuitry to provide a dedicated interface for managing reliability of the adaptive declustered RAID array.

3. The apparatus of claim 1, wherein reallocating the data chunks according to the evaluation of the upgrade and downgrade configuration rules comprises reallocating, as a spare data chunk, at least one of the data chunks allocated as one of the local parity and the global parity based on the second reliability indicator.

4. The apparatus of claim 1, wherein reallocating the data chunks according to the evaluation of the upgrade and downgrade configuration rules comprises reallocating, as the global parity, one of the data chunks allocated as the local parity, based on the second reliability indicator.

5. The apparatus of claim 1, wherein reallocating the data chunks according to the evaluation of the upgrade and downgrade configuration rules comprises reallocating, as the local parity, one of the data chunks allocated as the global parity, based on the second reliability indicator.

6. The apparatus of claim 1, further comprising central reliability monitoring circuitry configured to dynamically evaluate the second reliability indicator.

7. The apparatus of claim 1, wherein reallocating the data chunks according to the evaluation of the upgrade and downgrade configuration rules comprises reallocating the data chunks based on a neural network reliability warning scheme of reliability metrics.

8. The apparatus of claim 1, wherein reallocating the data chunks according to the evaluation of the upgrade and downgrade configuration rules comprises reallocating the data chunks based on Bayesian analytical approaches.

9. The apparatus of claim 1, wherein the data chunks further comprise a spare data chunk.

10. The apparatus of claim 9, wherein reallocating the data chunks according to the evaluation of the upgrade and downgrade configuration rules comprises reallocating the spare data chunk as the local parity based on the second reliability indicator.

11. The apparatus of claim 9, wherein reallocating the data chunks according to the evaluation of the upgrade and downgrade configuration rules comprises real locating the spare data chunk as the global parity based on the second reliability indicator.

12. The apparatus of claim 9, wherein:

reallocating the data chunks according to the evaluation of the upgrade and downgrade configuration rules comprises reallocating the spare data chunk for data migration of user data from one of the data chunks to the spare data chunk, based on the second reliability indicator.

13. The apparatus of claim 1, wherein the processing circuitry is further configured to generate the second reliability indicator based on early life failure, random failure, and wear out failure life cycles of the DSDs.

14. The apparatus of claim 1, further comprising a user interface connected to the processing circuitry for sending reliability directives to the processing circuitry for dynamically increasing or decreasing a number of the data chunks.

15. The apparatus of claim 1, wherein the second reliability indicator is indicative of mean time to data loss or mean time to failure of the at least the portion of the adaptive declustered RAID array at the second time.

25. The method of claim 7, further comprising reallocating the spare data chunk as the local parity based on the second reliability indicator.

16. A method of operating an apparatus, the method comprising:

configuring an adaptive declustered RAID) array comprising data storage devices (DSDs), the DSDs comprising data chunks, each of the data chunks allocated as a corresponding one of data, a local parity, and a global parity, the local parity to protect the data, the global parity to protect the local parity;

generating, by processing circuitry, at a first time according to a sampling frequency, a first reliability indicator, the first reliability indicator indicating a first reliability status of at least a portion of the adaptive declustered RAID array;

generating, by the processing circuitry, at a second time after the first time according to the sampling frequency, a second reliability indicator, the second reliability indicator indicating a second reliability status of the at least the portion of the adaptive declustered RAID array;

determining, by the processing circuitry, a difference between the first reliability indicator and the second reliability indicator; and

reallocating, by the processing circuitry, the data chunks, by dynamically increasing or decreasing the data chunks allocated as the local parity, the global parity, or a combination thereof, according to the difference.

17. The method of claim 16, further comprising:

connecting a reliability communication interface between a control circuit and DSD reliability managing circuitry in one of the DSDs to provide a dedicated interface for managing reliability of the adaptive declustered RAID array.

18. The method of claim 16, wherein reallocating the data chunks includes reallocating at least one of the data chunks, which is allocated as one of the local parity and the global parity, to be a spare data chunk, based on the second reliability indicator.

19. The method of claim 16, wherein reallocating the data chunks includes reallocating a data chunk which is allocated as the local parity, to be the global parity, based on the second reliability indicator.

20. The method of claim 16, wherein reallocating the data chunks includes reallocating a data chunk, which is allocated as the global parity to be the local parity, based on the second reliability indicator.

21. The method of claim 16, further comprising dynamically evaluating the second reliability indicator by central reliability monitoring circuitry.

22. The method of claim 16, wherein reallocating the data chunks is based on a neural network reliability warning scheme of reliability metrics.

23. The method of claim 16, wherein reallocating the data chunks is based on Bayesian analytical approaches.

24. The method of claim 16, wherein configuring the adaptive declustered RAID array includes configuring a spare data chunk.

26. The method of claim 24, further comprising reallocating the spare data chunk as the global parity based on the second reliability indicator.

27. The method of claim 24, further comprising:

allocating the spare data chunk for data migration of user data from one of the data chunks to the spare data chunk, based on the second reliability indicator.

28. The method of claim 16, wherein reallocating the data chunks includes reallocating the data chunks based on early life failure, random failure, and wear out failure life cycles of the DSDs.

29. The method of claim 16, further comprising connecting a user interface to the processing circuitry for sending reliability directives to a controller for dynamically increasing or decreasing a number of the data chunks.

30. A non-transitory computer readable medium including stored thereon instructions to be executed by processing circuitry, the instructions when executed by the processing circuitry cause the processing circuitry to:

configure an adaptive declustered RAID array comprising data storage devices (DSDs), the DSDs comprising data chunks, each of the data chunks allocated as a corresponding one of data, a local parity, and a global parity, the local parity to protect the data, the global parity to protect the local parity;

generate, at a first time according to a sampling frequency, a first reliability indicator, the first reliability indicator indicating a first reliability status of at least a portion of the adaptive declustered RAID array;

generate, at a second time after the first time according to the sampling frequency, a second reliability indicator, the second reliability indicator indicating a second reliability status of the at least the portion of the adaptive declustered RAID array;

determine a difference between the first reliability indicator and the second reliability indicator; and

reallocate the data chunks, by dynamically increasing or decreasing the data chunks allocated as the local parity, the global parity, or a combination thereof, according to the difference.

31. A controller, comprising:

means for configuring an adaptive declustered RAID array comprising data storage devices (DSDs), the DSDs comprising data chunks, each of the data chunks allocated as a corresponding one of data, a local parity, and a global parity, the local parity to protect the data, the global parity to protect the local parity;

means for generating a first reliability indicator, the first reliability indicator indicating a first reliability status of at least a portion of the adaptive declustered RAID array;

means for generating a second reliability indicator, the second reliability indicator indicating a second reliability status of the at least the portion of the adaptive declustered RAID array;

means for determining a difference between the first reliability indicator and the second reliability indicator; and

means for reallocating the data chunks, by dynamically increasing or decreasing the data chunks allocated as the local parity, the global parity, or a combination thereof according to the difference.