Storage system and method of preventing deterioration of write performance in storage system

• US 7,761,655 B2
• Filed: 01/02/2008
• Issued: 07/20/2010
• Est. Priority Date: 03/15/2007
• Status: Expired due to Fees

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Nakamura, Shuji, Mizushima, Nagamasa
• Primary Examiner(s): Nguyen; Hiep T
• Application Number:
  US11/968,218
• Publication Number:
  US 20080229003A1
• Time in Patent Office: 930 Days
• Field of Search: None
• 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 2212/1032 Reliability improvement, data loss prevention, degraded operation etc
• CPC Class Codes: G06F 2212/2022 Flash memory
• CPC Class Codes: G06F 2212/7205 Cleaning, compaction, garbage collection, erase control

Abstract

Provided is a storage system capable of inhibiting the deterioration of its write performance. This storage system includes a flash memory, a cache memory, and a controller for controlling the reading, writing and deletion of data of the flash memory and the reading and writing of data of the cache memory, and detecting the generation of a defective block in the flash memory. When the controller detects the generation of a defective block in the flash memory, it migrates prescribed data stored in the flash memory to the cache memory and, even upon receiving from the host computer a command for updating the migrated data, disables the writing of data in the flash memory based on the command.

Claims

1. A storage system, comprising:

a flash memory for writing data in page units and erasing said data in units of a block configured from a plurality of said pages, and having a plurality of said blocks and requiring the deletion of said blocks containing said pages in order to update said data;

a cache memory for writing and temporarily storing data to be written in said flash memory at a speed that is faster than said flash memory;

a controller for controlling the reading, writing and deletion of data of said flash memory and the reading and writing of data of said cache memory, and detecting the generation of a defective block in said flash memory; and

a host computer for issuing a command requesting the write processing of said data,

wherein said controller migrates prescribed data stored in said flash memory to said cache memory upon detecting the generation of said defective block in said flash memory and, even upon receiving from said host computer a command for updating the migrated data, disables the writing of data in said flash memory based on said command.

2. The storage system according to claim 1, wherein data migrated to said cache memory is updated based on a command for updating said migrated data.

3. The storage system according to claim 1, further comprising:

a cache management table for managing said data to be stored in said cache memory,

wherein said cache management table retains disable information for disabling the writing of said data in said flash memory.

4. The storage system according to claim 1, further comprising:

an address translation table for managing the correspondence of a logical page address and a physical page address of data to be stored in said flash memory,

wherein said address translation table retains address nonexistent information representing that a physical page address corresponding to a logical page address of said data does not exist.

5. The storage system according to claim 1, wherein said controller calculates a hit rate showing the probability of data of a logical page address that is the same as the logical page address of said data existing in said cache memory upon writing said data in said cache memory based on a command from said host computer;

and

upon detecting the generation of said defective block, calculates a first write performance of said storage system when another block in said flash memory is applied as a substitute destination of said defective block and a second write performance of said storage system when a part of said data stored in said flash memory is migrated to said cache memory based on said hit rate calculated up to said detection of said defective block, and migrates said data stored in said flash memory to said cache memory based on said calculation result when said second write performance is determined to be superior to said first write performance.

6. The storage system according to claim 1, wherein said cache memory is a nonvolatile random access memory.

7. The storage system according to claim 6, wherein said nonvolatile random access memory is a phase-change Random Access Memory (RAM).

8. The storage system according to claim 1, wherein said data migrated to said cache memory is retained redundantly by said cache memory.

9. The storage system according to claim 8, further comprising:

a cache memory that is different from said cache memory, and

wherein said redundant retention is performed by copying said data, which was migrated to said cache memory, to said different cache memory.

10. The storage system according to claim 8, further comprising:

a plurality of cache memories that are different from said cache memory, and

wherein said redundant retention is performed by configuring said data migrated to said cache memory as a Redundant Arrays of Inexpensive Disks (RAID) group with said plurality of different cache memories.

11. A method for preventing deterioration of write performance in a storage system comprising:

a flash memory for writing data in page units and erasing said data in units of a block configured from a plurality of said pages, and having a plurality of said blocks and requiring the deletion of said blocks containing said pages in order to update said data;

a cache memory for writing and temporarily storing data to be written in said flash memory at a speed that is faster than said flash memory;

a controller for controlling the reading, writing and deletion of data of said flash memory and the reading and writing of data of said cache memory, and detecting the generation of a defective block in said flash memory; and

a host computer for issuing a command requesting the write processing of said data;

wherein said controller performs the steps of;

detecting the generation of said defective block in said flash memory;

migrating prescribed data stored in said flash memory to said cache memory; and

even upon receiving from said host computer a command for updating the migrated data, disabling the writing of data in said flash memory based on said command.

12. The method for preventing deterioration of write performance in a storage system according to claim 11, wherein said data is disabled from being written in said flash memory based on disable information retained in a cache management table for storing said data to be stored in said cache memory.

13. The method for preventing deterioration of write performance in a storage system according to claim 11, wherein said controller calculates a hit rate showing the probability of data of a logical page address that is the same as the logical page address of said data existing in said cache memory upon writing said data in said cache memory based on a command from said host computer;

and

upon detecting the generation of said defective block, calculates a first write performance of said storage system when another block in said flash memory is applied as a substitute destination of said defective block and a second write performance of said storage system when a part of said data stored in said flash memory is migrated to said cache memory based on said hit rate calculated up to said detection of said defective block, and migrates said data stored in said flash memory to said cache memory based on said calculation result when said second write performance is determined to be superior to said first write performance.

14. The method for preventing deterioration of write performance in a storage system according to claim 11, wherein said cache memory is a nonvolatile random access memory.

15. The method for preventing deterioration of write performance in a storage system according to claim 11, wherein said data migrated to said cache memory is retained redundantly by said cache memory.

16. The method for preventing deterioration of write performance in a storage system according to claim 15, wherein said redundant retention is performed by copying said data, which was migrated to said cache memory, to a cache memory that is different from said different cache memory.

17. The method for preventing deterioration of write performance in a storage system according to claim 15, wherein said storage system further comprising:

a plurality of cache memories that are different from said cache memory, and

wherein said redundant retention is performed by configuring said data migrated to said cache memory as a RAID group with a plurality of cache memories that are different from said cache memory.