Rapid recovery from loss of storage device cache

• US 9,063,908 B2
• Filed: 05/31/2012
• Issued: 06/23/2015
• Est. Priority Date: 05/31/2012
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Loaiza, Juan R., Umamageswaran, Kothanda, Erlikhman, Boris, Friedman, David, Shi, Jia, Tao, Zuoyu, Tsukerman, Alex
• Primary Examiner(s): Rojas, Midys
• Application Number:
  US13/485,557
• Publication Number:
  US 20130326152A1
• Time in Patent Office: 1,118 Days
• Field of Search: 711/141, 711/100, 711/113, 711/143, 711/154, 711/138, 711/E12.02
• US Class Current: 1/1
• CPC Class Codes: G06F 11/2082 Data synchronisation
• CPC Class Codes: G06F 12/00 Accessing, addressing or allocating within memory systems or architectures digital input from, or digital output to record carriers, e.g. to disk storage units, G06F3/06
• CPC Class Codes: G06F 12/08 in hierarchically structured memory systems, e.g. virtual memory systems
• CPC Class Codes: G06F 12/0802 Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches
• CPC Class Codes: G06F 3/0629 Configuration or reconfiguration of storage systems

Abstract

Dirty data in a storage device is made current through rapid re-silvering, which uses a mirrored and up-to-date version of the dirty data from another storage device to recover the data. Because under rapid re-silvering cache metadata in volatile memory survives the failure of the cache, the cache metadata is used to determine which subset of data from the other storage device needs to be copied to the storage device being re-silvered. During re-silvering, cache metadata is used to determine which I/O requests from clients are requests for data that is not stale.

Claims

1. A method, comprising steps of:

a storage device detecting a failure of a storage device cache that said storage device uses to cache data stored in a primary storage of said storage device;

in response to detecting a failure;

said storage device making a determination of a set of data blocks in primary storage that are stale;

based on the determination of the set of data blocks in primary storage that are stale, said storage device generating a re-silvering list that identifies data blocks in said primary storage to re-synchronize; and

based on the re-silvering list, performing re-silvering of the storage device, wherein re-silvering comprises re-synchronizing data blocks identified by the re-silvering list by at least overwriting a subset of said data blocks with up-to-date versions of said subset of said data blocks from another storage device.

2. The method of claim 1, the steps further including:

while re-silvering the storage device;

the storage device receiving a request for a data block from a storage device client; and

in response to said request, the storage device returning said data block to said storage device client.

3. The method of claim 1, wherein the steps further include:

while re-silvering the storage device;

the storage device receiving a request for a data block from a storage device client;

in response to receiving the request for a data block, the storage device determining that the re-silvering list does not identify the data block as stale; and

in response to determining that the re-silvering list does not identify the data block as stale, said storage device returning said data block to said storage device client.

4. The method of claim 1, further including the steps of:

maintaining in a volatile memory of said storage device, cache metadata that identifies data blocks in said cache that are dirty;

wherein the determination of a set of data blocks in primary storage that are stale is based on the cache metadata.

5. The method of claim 1, wherein the steps further include:

the storage device sending said re-silvering list to a storage device client of said storage device; and

receiving from said storage device client write requests to overwrite particular data blocks in said re-silvering list with up-to-date versions of said particular data blocks from one or more mirror buddies of said storage device.

6. The method of claim 5, further including receiving from said storage device client updates to the re-silvering list, wherein receiving from said storage device client updates to the re-silvering list includes receiving an update that indicates that a certain data block has been synchronized, wherein a request to overwrite said certain data block to re-synchronize said certain data block has not been received by said storage device before receiving said update.

7. The method of claim 1, the steps further including:

a storage device client receiving said re-silvering list from said storage device;

based on an examination of the re-silvering list, said storage device client requesting up-to-date versions of particular data blocks from one or more mirror buddies of said storage device;

issuing to said storage device write requests to overwrite the particular data blocks;

updating a copy of said re-silvering list to reflect completion of said write requests; and

sending to said storage device said client updates to the re-silvering list that reflect completion of said write requests.

8. The method of claim 7, wherein sending to said storage device said client updates to the re-silvering list includes sending an update to the re-silvering list to identify a particular data block as synchronized without having sent a write request to overwrite said particular data block with an up-to-date version of said particular data block.

9. The method of claim 1, wherein a storage device client of said storage device is database server.

10. The method of claim 1, wherein said storage device cache is non-volatile memory.

11. A computer-readable non-transitory storage medium that stores instructions which, when executed by one or more processors, cause the one of more processors to perform:

a storage device detecting a failure of a storage device cache that said storage device uses to cache data stored in a primary storage of said storage device;

in response to detecting a failure;

said storage device making a determination of a set of data blocks in primary storage that are stale;

based on the determination of the set of data blocks in primary storage that are stale, said storage device generating a re-silvering list that identifies data blocks in said primary storage to re-synchronize; and

based on the re-silvering list, performing re-silvering of the storage device, wherein re-silvering comprises re-synchronizing data blocks identified by the re-silvering list by at least overwriting a subset of said data blocks with up-to-date versions of said subset of said data blocks from another storage device.

12. The computer-readable non-transitory storage medium of claim 11, wherein the instructions, when executed by said one or more processors, further cause:

while re-silvering the storage device;

the storage device receiving a request for a data block from a storage device client; and

in response to said request, the storage device returning said data block to said storage device client.

13. The computer-readable non-transitory storage medium of claim 11, wherein the instructions, when executed by said one or more processors, further cause:

while re-silvering the storage device;

the storage device receiving a request for a data block from a storage device client;

in response to receiving the request for a data block, the storage device determining that the re-silvering list does not identify the data block as stale; and

in response to determining that the re-silvering list does not identify the data block as stale, said storage device returning said data block to said storage device client.

14. The computer-readable non-transitory storage medium of claim 11, wherein the instructions, when executed by said one or more processors, further cause:

maintaining in a volatile memory of said storage device cache metadata that identifies data blocks in said cache that are dirty;

wherein the determination of a set of data blocks in primary storage that are stale is based on the cache metadata.

15. The computer-readable non-transitory storage medium of claim 11, wherein the instructions, when executed by said one or more processors, further cause:

while re-silvering the storage device;

the storage device sending said re-silvering list to a storage device client of said storage device; and

receiving from said storage device client write request to overwrite particular data blocks in said re-silvering list with up-to-date versions of said particular data blocks from one or more mirror buddies of said storage device.

16. The computer-readable non-transitory storage medium of claim 15, wherein the instructions, when executed by said one or more processors, further cause receiving from said storage device client updates to the re-silvering list, wherein receiving from said storage device client updates to the re-silvering list includes receiving an update that indicates that a certain data block has been synchronized, wherein a request to overwrite said certain data block to re-synchronize said certain data block has not been received by said storage device before receiving said update.

17. The computer-readable non-transitory storage medium of claim 11, wherein the instructions, when executed by said one or more processors, further cause:

a storage device client receiving said re-silvering list from said storage device;

based on an examination of the re-silvering list, said storage device client requesting up-to-date versions of particular data blocks from one or more mirror buddies of said storage device;

issuing to said storage device write requests to overwrite the particular data blocks;

updating a copy of said re-silvering list to reflect completion of said write requests; and

sending to said storage device said client updates to the re-silvering list that reflect completion of said write requests.

18. The computer-readable non-transitory storage medium of claim 17, wherein sending to said storage device said client updates to the re-silvering list includes sending an update to the re-silvering list to identify a particular data block as synchronized without having sent a write request to overwrite said particular data block with an up-to-date version of said particular data block.

19. The computer-readable non-transitory storage medium of claim 11, wherein a storage device client of said storage device is database server.

20. The computer-readable non-transitory storage medium of claim 11, wherein said storage device cache is non-volatile memory.