Back-up and restoration of data between volatile and flash memory

• US 9,501,404 B2
• Filed: 07/29/2013
• Issued: 11/22/2016
• Est. Priority Date: 11/14/2012
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Palmer, Michael J., Smith, Peter M., Wong, Kelvin
• Primary Examiner(s): KUDIRKA, JOSEPH R
• Application Number:
  US14/439,673
• Publication Number:
  US 20150317210A1
• Time in Patent Office: 1,212 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1441 Resetting or repowering
• CPC Class Codes: G06F 12/0246 in block erasable memory, e.g. flash memory
• CPC Class Codes: G06F 2212/1032 Reliability improvement, data loss prevention, degraded operation etc
• CPC Class Codes: G06F 2212/7202 Allocation control and policies
• CPC Class Codes: G06F 2212/7207 management of metadata or control data
• CPC Class Codes: G06F 2212/7208 Multiple device management, e.g. distributing data over multiple flash devices
• CPC Class Codes: G11C 11/005 comprising combined but independently operative RAM-ROM, RAM-PROM, RAM-EPROM cells
• CPC Class Codes: G11C 14/00 Digital stores characterised by arrangements of cells having volatile and non-volatile storage properties for back-up when the power is down

Abstract

In one embodiment, a method for back-up of data to flash memory is provided. Backed up data is organized into one or more stripes, wherein a stripe comprises a set of pages across all available flash memory devices which have a same block and page address. Responsive to encountering an error in a block of flash memory during back-up of a particular stripe of data, the particular stripe of data is rewritten starting at a next available page address and excluding a page of flash memory for the block having the error. Subsequent stripes of data in the block having the error are written to pages excluding the page of flash memory for the block having the error.

Claims

1. A method for back-up of data to flash memory, comprising:

organizing backed-up data into one or more stripes, wherein a stripe comprises a set of pages across all available flash memory devices which have a same block and page address; and

responsive to encountering an error in a block of flash memory during back-up of a particular stripe of data, performing;

re-writing the particular stripe of data starting at a next available page address and excluding a page of flash memory for the block having the error;

writing subsequent stripes of data in the block having the error to pages excluding the page of flash memory for the block having the error; and

embedding metadata for the particular stripe of data, the metadata including an indication of a location of the block having the error.

2. The method of claim 1, wherein embedding metadata for the particular stripe of data comprises embedding the metadata in each page of the particular stripe of data, and the metadata comprises:

an identification number of the particular stripe of data;

a stripe length field of an amount of data saved in the particular stripe of data; and

an indication of known error blocks for the particular stripe of data.

3. The method of claim 1, wherein each of the available flash memory devices includes a channel, multiple dies, and multiple planes;

and wherein a specific channel, die, and plane location of the block having the error is excluded from being written to.

4. The method of claim 1, further comprising:

reading metadata from pages of a second stripe of data being backed-up;

determining locations of known error blocks in the second stripe of data; and

avoiding writing back-up data of the second stripe of data to the known error blocks in the second stripe of data by writing the back-up data of the second stripe of data to a next available page in the second stripe of data.

5. The method of claim 1, further comprising:

updating the embedded metadata for the particular stripe of data and metadata for each subsequent stripe of data to include;

an updated stripe size to indicate a capacity taking into account the block having the error; and

the location of block having the error.

6. The method of claim 1, further comprising:

writing an end marker stripe to indicate an end of backed up data, wherein the end marker stripe includes an indication of stripe length of zero.

7. The method of claim 1, further comprising:

pre-erasing sufficient blocks of flash memory in the available flash memory devices for backing-up data.

8. The method of claim 1, further comprising:

restoring backed-up data to target memory by reading all pages from a third stripe of backed-up data to target memory, reading metadata for the third stripe embedded in one or more pages of the third stripe, and using the embedded metadata of the third stripe to allow for bad blocks.

9. The method of claim 8, wherein using the embedded metadata to allow for bad blocks includes:

determining which stripe is expected next from a stripe number in the embedded metadata for the third stripe; and

responsive to determining that the stripe expected next is a stripe that has already been read;

reading data and metadata of the third stripe including a position of a block in the third stripe having an error;

re-reading data of the third stripe excluding a page for the block in the third stripe having the error;

restoring the re-read data of the third stripe to a location in target memory that is associated with the stripe number; and

adjusting a next expected stripe number.

10. The method of claim 8, wherein using the embedded metadata to allow for bad blocks comprises:

determining any known error blocks in the third stripe from the metadata of the third stripe;

re-reading data of the third stripe excluding a page for the the block in the third stripe having the error;

reading subsequent stripes of data in the block in the third stripe excluding the page for the block in the third stripe having the error.

11. A system for back-up of data to flash memory, comprising:

one or more computer processors;

a stripe component for organizing backed-up data into one or more stripes, wherein a stripe comprises a set of pages across all available flash memory devices which have a same block and page address;

a grown bad block component for performing, with the one or more computer processors, responsive to encountering an error in a block of flash memory during back-up of a particular stripe of data;

re-writing, with the one or more computer processors, the particular stripe of data starting at a next available page address and excluding a page of flash memory for the block having the error; and

writing, with the one or more computer processors, subsequent stripes of data in the block having the error to pages excluding the page of flash memory for the block having the error; and

a metadata component for embedding, with the one or more computer processors, metadata for the particular stripe of data, the metadata including an indication of a location of the block having the error.

12. The system of claim 11, wherein the metadata component for embedding metadata for the particular stripe of data embeds the metadata in each page of the particular stripe of data, the metadata including:

an identification number of the particular stripe of data;

a stripe length field of an amount of data saved in the particular stripe of data; and

an indication of known error blocks for the particular stripe of data.

13. The system of claim 12, wherein the grown bad block component further comprises:

a metadata updating component for updating the embedded metadata for the particular stripe of data and metadata for each subsequent stripe of data to include;

an updated stripe size to indicate a capacity taking into account the block having the error;

the location of the block having the error.

14. The system of claim 11, further comprising a known bad block component for:

reading metadata from pages of a second stripe of data being backed up;

determining locations of known error blocks in the second stripe of data; and

avoiding writing back-up data of the second stripe of data to the known error blocks in the second stripe of data by writing the back-up data of the second stripe of data to a next available page in the second stripe of data.

15. The system of claim 11, further comprising:

a restoration component for restoring backed up data to target memory by reading all pages from a third stripe of backed up data to target memory, reading metadata for the third stripe embedded in one or more pages of the third stripe, and using the embedded metadata of the third stripe to allow for bad blocks.

16. The system of claim 15, wherein the restoration component includes a grown bad block component for:

determining which stripe is expected next from a stripe number in the embedded metadata for the third stripe; and

responsive to determining that the stripe expected next is a stripe that has already been read;

reading data and metadata of the third stripe including a position of a block in the third stripe having an error;

re-reading data of the third stripe excluding a page for the block in the third stripe having the error;

restoring the re-read data of the third stripe to a location in target memory that is associated with the stripe number; and

adjusting a next expected stripe number.

17. A computer program product for back-up of data to flash memory, the computer program product comprising:

a non-transitory computer readable storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method comprising;

organizing backed up data into one or more stripes, wherein a stripe comprises a set of pages across all available flash memory devices which have a same block and page address; and

responsive to encountering an error in a block of flash memory during back-up of a particular stripe of data, performing;

re-writing the particular stripe of data starting at a next available page address and excluding a page of flash memory for the block having the error;

writing subsequent stripes of data in the block having the error to pages excluding the page of flash memory for the block having the error; and

embedding metadata for the particular stripe of data, the metadata including an indication of a location of the block having the error.

18. The computer program product of claim 17, wherein embedding metadata for the particular stripe of data comprises:

embedding metadata in each page of the particular stripe of data, the metadata including;

an identification number of the particular stripe of data;

a stripe length field of an amount of data saved in the particular stripe of data; and

an indication of known error blocks for the particular stripe of data.

19. The computer program product of claim 17, wherein the method further comprises:

reading metadata from pages of a second stripe of data being backed up;

determining locations of known error blocks in the second stripe of data; and

avoiding writing back-up data of the second stripe of data to the known error blocks in the second stripe of data by writing the back-up data of the second stripe of data to a next available page in the second stripe of data.