ENHANCING THE LIFETIME AND PERFORMANCE OF FLASH-BASED STORAGE

• US 20130262746A1
• Filed: 04/02/2012
• Published: 10/03/2013
• Est. Priority Date: 04/02/2012
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Srinivasan, KY
• Granted Patent: US 8,918,581 B2
• Time in Patent Office: Days
• US Class Current: 711/103
• CPC Class Codes: G06F 12/0246 in block erasable memory, e.g. flash memory
• CPC Class Codes: G06F 12/0804 with main memory updating G06F12/0806 takes precedence
• CPC Class Codes: G06F 12/0866 for peripheral storage systems, e.g. disk cache
• CPC Class Codes: G06F 12/0871 Allocation or management of cache space
• CPC Class Codes: G06F 2212/1024 Latency reduction
• CPC Class Codes: G06F 2212/1036 Life time enhancement
• CPC Class Codes: G06F 2212/466 Metadata, control data

Abstract

A storage management system decouples application write requests from write requests to a flash-based storage device. By placing a layer of software intelligence between application requests to write data and the storage device, the system can make more effective decisions about when and where to write data that reduce wear and increase performance of the storage device. An application has a set of performance characteristics and writes data with a frequency that is appropriate for the application, but not necessarily efficient for the hardware. By analyzing how data is being used by an application, the system can strategically place data in the storage device or even avoid using the storage device altogether for some operations to minimize wear. One technique for doing this is to create an in-memory cache that acts as a buffer between the application requests and the storage hardware.

Claims

1. A computer-implemented method to handle a write request from a software application directed to a flash-based storage device, the method comprising:

receiving from the application a request to write data to the flash-based storage device;

determining a storage device to which the received request is directed;

upon determining that the storage device is flash-based, identifying 240 an in-memory cache for collecting data requested by applications to be written to the storage device;

writing the data received with the request to the identified in-memory cache without writing the data to the flash-based storage device;

returning a write result to the requesting application indicating success or failure of the write operation to the in-memory cache; and

collecting at least one additional write operation in the in-memory cache before sending the data from the first write request and the additional write operation to the flash-based storage device,

wherein the preceding steps are performed by at least one processor.

2. The method of claim 1 wherein receiving the request comprises receiving the request in response to a user action in a software application, wherein the software application invokes an operating system application programming interface (API) for storing data on the flash-based storage device.

3. The method of claim 1 wherein receiving the request comprises receiving information related to the data including an indication of an expected frequency of access of the data.

4. The method of claim 1 wherein determining the storage device comprises determining whether the target storage device is flash-based and will benefit from in-memory caching of write requests.

5. The method of claim 1 wherein determining the storage device comprises reading information provides by a basic input/output system (BIOS) or extended firmware interface (EFI) to determine physical characteristics of the device.

6. The method of claim 1 wherein identifying an in-memory cache comprises determining that the cache does not yet exist and allocating an in-memory cache to handle the present and subsequent data write requests.

7. The method of claim 1 wherein identifying an in-memory cache comprises identifying an area of random access memory (RAM) to which the system can write data without incurring the wear and other penalties of immediately writing the data to the flash-based storage device.

8. The method of claim 1 wherein the cache collects multiple write requests and services any received read requests prior to flushing the data from the cache to the flash-based storage device.

9. The method of claim 1 wherein writing the data to the cache comprises updating metadata that indicates which data associated with the flash-based storage is currently held in the cache.

10. The method of claim 1 further comprising flushing the data from the in-memory cache to the flash-based storage device, or determining that writing the data is unnecessary because the data has been deleted.

11. The method of claim 1 wherein writing data to the cache and later flushing the data combines multiple write requests to a particular block of flash-based memory, thereby avoiding excessive wear on the flash-based device and improving device performance.

12. The method of claim 1 wherein returning a write result comprises providing the result to the application in less time than writing the data directly to the flash-based storage because of the lower latency of the in-memory cache.

13. A computer system for enhancing the lifetime and performance of flash-based storage, the system comprising:

a flash-based storage device that includes at least some flash-based non-volatile memory;

a device interface component that provides an interface between the other components of the system and the flash-based storage device;

a data cache component that provides an in-memory cache of data requested to be written to the flash-based storage device by an application;

a cache write component that receives requests to write data to the flash-based storage device and writes the data to the data cache component instead;

a cache read component that receives requests to read data from the flash-based storage device and determines whether the requested data is available from the data cache component without accessing the flash-based storage device;

a cache flush component that copies data stored in the data cache component to the flash-based storage device 110 at a time decoupled from application requests to write data;

an application interface component that provides an interface between the other components of the system and one or more applications that request to read from or write to the flash-based storage device.

14. The system of claim 13 wherein the flash-based storage device performs block-based erasing, and has a finite quantity of non-overlapping writes that can be performed per block before the data can no longer be reliably written to a block.

15. The system of claim 13 wherein the device interface component receives requests directed to one or more operating system application-programming interfaces (APIs) for accessing storage devices.

16. The system of claim 13 wherein the data cache component provides an intermediate location in which data requested to be stored by applications can be held without incurring the penalties of writing the data to flash-based storage, and while the system acts to reduce those penalties by reducing the number of writes/erasures of the flash-based storage.

17. The system of claim 13 wherein the cache write component determines whether a particular block of the flash-based storage device is already stored in the cache and stores information in a data structure indicating that new data for a particular block of the device is available in the cache.

18. The system of claim 13 wherein the cache flush component flushes data upon determining that the cache is full and contains no non-dirty locations for writing new data.

19. The system of claim 13 wherein the application interface component cooperates with an operating system as a kernel or user-mode device driver to intercept requests from applications and redirect requests to the flash-based storage device to the data cache component.

20. A computer-readable storage medium comprising instructions for controlling a computer system to flush cached data to a flash-based storage device, wherein the instructions, upon execution, cause a processor to perform actions comprising:

receiving from one or more applications multiple requests to store data on a flash-based storage device and caching the data in an intermediate cache without storing the data on the flash-based storage device at the time of servicing the request;

identifying the cached data in the intermediate cache and determining whether the cached data will be written to the flash-based storage device;

upon determining that one or more criteria for flushing the cache have been met,

identifying the flash-based storage device to which the received requests were directed, wherein the flash-based storage device acts as non-volatile storage for the data in the intermediate cache; and

writing at least some of the data stored in the intermediate cache to the identified flash-based storage device such that the writing of data to the flash-based storage device occurs after the original write operation that stored the data to the intermediate cache has already completed, resulting in a decoupling of application writes from device writes that reduces a quantity of writes and resulting wear on the device.