Data reliability schemes for data storage systems

• US 9,021,339 B2
• Filed: 11/29/2012
• Issued: 04/28/2015
• Est. Priority Date: 11/29/2012
• Status: Expired due to Fees

• Current Assignee: Western Digital Technologies Incorporated (Western Digital Corporation)
• Original Assignee: Skyera, Inc. (Western Digital Corporation), Western Digital Technologies Incorporated (Western Digital Corporation)
• Inventors: Lu, Guangming, Ho, Leader, Danilak, Radoslav, Mullendore, Rodney N., Jones, Justin, Tomlin, Andrew J.
• Primary Examiner(s): Rizk, Sam
• Application Number:
  US13/689,673
• Publication Number:
  US 20140149826A1
• Time in Patent Office: 880 Days
• Field of Search: 714/770, 714/773, 711/114
• US Class Current: 714/773
• CPC Class Codes: G06F 11/108 Parity data distribution in semiconductor storages, e.g. in SSD

Abstract

A data storage system configured to implement a data reliability scheme is disclosed. In one embodiment, a data storage system controller detects uncorrectable errors using intra page parity when data units are read from a set of pages. When an uncorrectable error is detected, the data storage system controller attempts to recover user data using inter page parity without using all data from each page of the set of pages. Recovery of user data can thereby be performed without reading all data from each page. As a result, the amount of time needed to read data can be reduced in some cases and overall data storage system performance can be increased.

Claims

1. A data storage system, comprising:

a non-volatile memory array comprising a plurality of data pages; and

a controller configured to;

store a plurality of data units and intra page parity units in each data page of a first set of data pages of the plurality of data pages and a plurality of inter page parity units in each data page of a second set of data pages of the plurality of data pages;

in response to detecting an error using the intra page parity units of a first data page of the first set when a first data unit is read from the first data page, perform a recovery for data of the first data unit using corresponding data units from the data pages of the first set and corresponding inter page parity units from the data pages of the second set; and

in response to determining that the recovery did not result in successfully determining the data of the first data unit because a number of data pages from the first set for which data read errors are detected exceeds an error correction capability for inter data page error correction;

read one or more data units of the first set other than the first data unit using a modified read parameter or decode the one or more data units of the first set other than the first data unit using a modified decoding parameter, and

in response to determining that the number of data pages from the first set for which the data read errors are detected no longer exceeds the error correction capability for inter data page error correction, successfully determine the data of the first data unit using the corresponding data units from the data pages of the first set and the corresponding inter page parity units from the data pages of the second set,

wherein the controller is configured to perform the recovery without using all data units from the data pages of the first set and all inter page parity units from the data pages of the second set.

2. The data storage system of claim 1, wherein each data unit of the first set comprises a common number of octets of data after padding with a pre-defined data set, and each inter page parity unit of the second set comprises the common number of octets of data and corresponds to a parity based at least in part on the corresponding data units from the data pages of the first set.

3. The data storage system of claim 1, wherein each data page of the plurality of data pages comprises a flash page (F-page), and each data unit comprises an error-correcting code page (E-page).

4. The data storage system of claim 3, wherein the first set and the second set together form a super-page (S-page) comprising one F-page from each array die of a plurality of array dies.

5. The data storage system of claim 1, wherein the plurality of intra page parity units comprise low-density parity-check (LDPC) parity units, and the plurality of inter page parity units comprise Reed-Solomon (RS) parity units.

6. The data storage system of claim 1, wherein the controller is further configured to read the first data unit using the modified read parameter in response to detecting the error using the intra page parity units of the first data page when the first data unit is read from the first data page.

7. The data storage system of claim 6, wherein the modified read parameter comprises at least one of:

a single-read/multiple-read and a voltage threshold adjustment.

8. The solid-state storage system of claim 1, wherein the controller is configured to read the one or more data units of the first set other than the first data unit using the modified read parameter in response to determining that the recovery did not result in successfully determining the data of the first data unit.

9. The data storage system of claim 1, wherein the controller is further configured to decode the first data unit using the modified decoding parameter in response to detecting the error using the intra page parity units of the first data page when the first data unit is read from the first data page.

10. The data storage system of claim 1, wherein the controller is configured to decode the one or more data units of the first set other than the first data unit using the modified decoding parameter in response to determining that the recovery did not result in successfully determining the data of the first data unit.

11. The data storage system of claim 1, wherein the controller is further configured to:

assign individual data pages of the plurality of data pages to either the first set or the second set based on quality data about the individual data pages, and

determine each inter page parity unit of the second set based at least in part on the corresponding data units from the data pages of the first set.

12. In a data storage system comprising a controller and a non-volatile memory array including a plurality of data pages, a method of performing data recovery, the method comprising:

storing a plurality of data units and intra page parity units in each data page of a first set of data pages of the plurality of data pages and a plurality of inter page parity units in each data page of a second set of data pages of the plurality of data pages;

in response to detecting an error using the intra page parity units of a first data page of the first set when a first data unit is read from the first data page, performing a recovery for data of the first data unit using corresponding data units from the data pages of the first set and a-corresponding inter page parity units from the data pages of the second set; and

in response to determining that the recovery did not result in successfully determining the data of the first data unit because a number of data pages from the first set for which data read errors are detected exceeds an error correction capability for inter data page error correction;

reading one or more data units of the first set other than the first data unit using a modified read parameter or decoding the one or more data units of the first set other than the first data unit using a modified decoding parameter, and

in response to determining that the number of data pages from the first set for which the data read errors are detected no longer exceeds the error correction capability for inter data page error correction, successfully determining the data of the first data unit using the corresponding data units from the data pages of the first set and the corresponding inter page parity units from the data pages of the second set,

wherein the recovery is performed using less than all of the data units from the data pages of the first set and less than all inter page parity units from the data pages of the second set.

13. The method of claim 12, wherein each data unit of the first set comprises a common number of octets of data after padding with a pre-defined data set, and each inter page parity unit of the second set comprises the common number of octets of data and corresponds to a parity based at least in part on the corresponding data units from the data pages of the first set.

14. The method of claim 12, wherein each data page of the plurality of data pages comprises a flash page (F-page), and each data unit comprises an error-correcting code page (E-page).

15. The method of claim 14, wherein the first set and the second set together form a super-page (S-page) comprising one F-page from each array die of a plurality of array dies.

16. The method of claim 12, wherein plurality of the intra page parity units comprise low-density parity-check (LDPC) parity units, and plurality of the inter page parity units comprise Reed-Solomon (RS) parity units.

17. The method of claim 12, further comprising reading the first data unit using the modified read parameter in response to detecting the error using the intra page parity units of the first data page when the first data unit is read from the first data page.

18. The method of claim 17, wherein the modified read parameter comprises at least one of:

a single-read/multiple-read and a voltage threshold adjustment.

19. The method of claim 12, further comprising reading the one or more data units of the first set other than the first data unit using the modified read parameter in response to determining that the recovery did not result in successfully determining the data of the first data unit.

20. The method of claim 12, further comprising decoding the first data unit using the modified decoding parameter in response to detecting the error using the intra page parity units of the first data page when the first data unit is read from the first data page.

21. The method of claim 12, further comprising decoding the one or more data units of the first set other than the first data unit using the modified decoding parameter in response to determining that the recovery did not result in successfully determining the data of the first data unit.

22. The method of claim 12, further comprising:

assigning individual data pages of the plurality of data pages to either the first set or the second set based on quality data about the individual data pages, and

determining each inter page parity unit of the second set based at least in part on the corresponding data units from the data pages of the first set.