STORAGE DEVICES AND METHODS FOR OPTIMIZING USE OF STORAGE DEVICES BASED ON STORAGE DEVICE PARSING OF FILE SYSTEM METADATA IN HOST WRITE OPERATIONS

• US 20160054931A1
• Filed: 11/26/2014
• Published: 02/25/2016
• Est. Priority Date: 08/20/2014
• Status: Active Grant

• Current Assignee: SanDisk Technologies LLC (Western Digital Corporation)
• Original Assignee: Sandisk Technologies Incorporated (Western Digital Corporation)
• Inventors: Hahn, Judah Gamliel, Tuers, Daniel Edward, Romanovsky, Leonid, Meza, Joseph Robert
• Granted Patent: US 10,228,854 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 12/0246 in block erasable memory, e.g. flash memory
• CPC Class Codes: G06F 3/0608 Saving storage space on storage systems
• CPC Class Codes: G06F 3/0643 Management of files
• CPC Class Codes: G06F 3/0659 Command handling arrangements, e.g. command buffers, queues, command scheduling
• CPC Class Codes: G06F 3/0679 Non-volatile semiconductor memory device, e.g. flash memory, one time programmable memory [OTP]

Abstract

The subject matter described herein includes processing file system metadata in host write requests to determine information about future host write operations. The information regarding future host write operations can be used by a device controller to prepare the non-volatile memory for the future host write operations. For example, the device controller may prepare the non-volatile storage device for future sequential host write access patterns or random host write access patterns depending on the content of the file system metadata. The file system metadata may also be usable to determine when it is optimal to perform memory management operations.

Claims

1. A storage device comprising:

nonvolatile memory;

a metadata identifier for determining whether an incoming write request received from a host system includes file system metadata;

a descriptor block (DB) parser for, in response to a determination that the incoming write request includes file system metadata, locating a descriptor block in the write request and parsing file system metadata following the descriptor block to determine information regarding future host write operations; and

a device controller for preparing the nonvolatile memory for the future host write operations based on the information regarding the future host write operations determined from the file system metadata.

2. The storage device of claim 1 wherein the nonvolatile memory comprises flash memory.

3. The storage device of claim 2 wherein the flash memory comprises NAND flash memory or NOR flash memory having a two-dimensional or three-dimensional configuration.

4. The storage device of claim 1 wherein the metadata identifier is configured to identify the presence of file system metadata in the write request based on memory address information determined from the write request.

5. The storage device of claim 4 wherein the metadata identifier is configured to identify the presence of file system metadata in the write request based on the presence of memory address information that corresponds to the location of a file system journal in the nonvolatile memory.

6. The storage device of claim 1 wherein the DB parser is configured to determine logical block addresses (LBAs) of memory blocks to be written in the nonvolatile memory from inodes following the descriptor block.

7. The storage device of claim 1 comprising a commit block (CB) identifier for identifying a CB in the write request, wherein the device controller is configured to perform memory management operations on the nonvolatile memory after the CB is identified.

8. The storage device of claim 7 wherein the memory management operations include activating or deactivating garbage collection.

9. The storage device of claim 1 wherein the DB parser is configured to determine, from the file system metadata, that future sequential host write operations will occur and wherein, in response to the determination that the future sequential host write operations will occur, the device controller is configured to prepare contiguous blocks of the nonvolatile memory for the future sequential host write operations.

10. The storage device of claim 1 wherein the DB parser is configured to determine, from the file system metadata, that future random host write access patterns will occur.

11. The storage device of claim 1 wherein the device controller is configured to map logical addresses for the future host write operations to physical addresses associated with a particular memory die or dies to increase the efficiency of utilization of the memory die or dies.

12. A storage device comprising:

non-volatile storage;

a hint derivation and memory utilization optimization module for deriving hints regarding accesses to the non-volatile storage and from file system metadata and for utilizing the hints to optimize utilization of the non-volatile storage; and

a hint table maintained in the non-volatile storage for storing the hints.

13. A storage device comprising:

non-volatile storage;

means for determining whether an incoming write request received from a host system includes file system metadata;

means for, in response to a determination that the incoming write request includes file system metadata, locating a descriptor block in the write request and parsing file system metadata following the descriptor block to determine information regarding future host write operations; and

means for preparing the nonvolatile memory for the future host write operations based on the information regarding the future host write operations determined from the file system metadata.

14. A method for using file system metadata to enhance utilization of nonvolatile memory, the method comprising:

in a nonvolatile memory device having nonvolatile memory;

determining whether an incoming write request received from a host system includes file system metadata;

in response to a determination that the incoming write request includes file system metadata;

locating a descriptor block in the write request;

parsing file system metadata following the descriptor block;

identifying, from the file system metadata, memory addresses of future host write operations; and

preparing the nonvolatile memory for the future host write operations based on the file system metadata.

15. The method of claim 14 wherein the nonvolatile memory comprises flash memory.

16. The method of claim 15 wherein the flash memory comprises NAND flash memory or NOR flash memory having a two-dimensional or three-dimensional configuration.

17. The method of claim 14 wherein identifying the presence of file system metadata in the write request comprises identifying the presence of file system metadata based on memory address information determined from the write request.

18. The method of claim 17 wherein identifying the presence of file system metadata in the write request includes identifying memory address information that corresponds to the location of a file system journal in the nonvolatile memory.

19. The method of claim 14 wherein identifying memory addresses of the future write operations includes identifying logical block addresses (LBAs) of memory blocks to be written in the nonvolatile memory from inodes following the descriptor block.

20. The method of claim 14 comprising identifying a commit block (CB) in the write request, and performing memory management operations on the nonvolatile memory after the CB is identified.

21. The method of claim 20 wherein the memory management operations include activating or deactivating garbage collection.

22. The method of claim 14 comprising determining, from the file system metadata, that future sequential host write operations will occur and, in response to determining that the future sequential host write operations will occur, preparing contiguous blocks of the nonvolatile memory for the future sequential host write operations.

23. The method of claim 14 comprising determining, from the file system metadata, that future random host write access patterns will occur.

24. The method of claim 14 comprising mapping logical addresses for the future host write operations to physical addresses associated with a particular memory die or dies to increase the efficiency of utilization of the memory die or dies.