STORAGE CONTROLLER, AND STORAGE DEVICE FAILURE DETECTION METHOD

• US 20090106583A1
• Filed: 01/09/2008
• Published: 04/23/2009
• Est. Priority Date: 10/23/2007
• Status: Active Grant

• Current Assignee: Rakuten Group, Inc.
• Original Assignee: Hitachi, Ltd.
• Inventors: KAWAMURA, Shunji
• Granted Patent: US 7,783,922 B2
• Time in Patent Office: Days
• US Class Current: 714/6
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 2211/1009 Cache, i.e. caches used in RAID system with parity

Abstract

Even when a failure incapable of being detected by a single guarantee code occurs in data, a storage controller of the present invention can detect and restore this failure. The controller, upon receiving a write command from a host, establish a first data guarantee code in the write data, and establish a second data guarantee code in the parity related to the write data. Upon receiving a read command from the host, the controller determines the correctness of the data based on the first data guarantee code, and also detects the presence or absence of a failure by comparing the first data guarantee code with the second data guarantee code. When a failure is detected, the controller restores the failure on the basis of RAID technology.

Claims

1. A storage controller which inputs/outputs data to/from a storage device in accordance with a request from a computer, comprising:

a check data setting unit, which creates a first check data and second check data on the basis of a first data, and which associates the first check data with the first data and stores the associated first check data in a first storage area in which the first data is stored, and associates the second check data with a second data, which is associated to the first data, and stores the associated second check data in a second storage area; and

a checking unit which, if the first data is read out from the first storage area, checks whether or not the first check data read but from the first storage area matches the second check data read out from the second storage area, and which, if the first check data matches the second check data, determines that the first data is correctly read out from the first storage area, and if the first check data and the second check data do not match, detects an occurrence of a failure;

if.

2. The storage controller according to claim 1, wherein

the first check data and the second check data respectively comprise update sequence information, which shows an update sequence of the first data stored in the first storage area, and

if the first check data and the second check data do not match, the checking unit compares the update sequence information inside the first check data with the update sequence information inside the second check data, determines whether one of the first data and the second data is older than the other, and determines that older data is incorrect data.

3. The storage controller according to claim 2, wherein the update sequence information comprises either one of an update time, which shows the time at which the first data was stored in the first storage area, and a sequence number, which shows order in which the first data has been issued from the computer.

4. The storage controller according to claim 2, wherein

the update sequence information comprises both the update time, which shows time at which the first data has been stored in the first storage area, and the sequence number, which shows order in which the first data has been issued from the computer, and

the checking unit, determines whether one of the first data and the second data is older than the other based on update time in the first check data and update time in the second check data, and furthermore, if the update time in the first check data and the update time in the second check data are the same time, determines that data with smaller sequence numbers is older based on the sequence number in the first check data and the sequence number in the second check data.

5. The storage controller according to claim 1, wherein the check data setting unit creates the first check data and the second check data as the same data.

6. The storage controller according to claim 1, wherein the check data setting unit creates the second check data from a portion of the first check data.

7. The storage controller according to claim 1, wherein a plurality of the respectively different first data exist in the same stripe, and the second data is created as parity data of the respective first data.

8. The storage controller according to claim 7, further comprising:

a restore unit which, if the checking unit detects a failure, restores the failure in use of the second data;

wherein the restore unit uses the plurality of first data and the second data in accordance with Redundant Array of Inexpensive Disks (RAID) and restores either one or both of the first data for which a read is requested by the computer, or the second data which should be updated in accordance with the first data for which an update is requested by the computer.

9. The storage controller according to claim 7, wherein, if the computer requests update of the first data stored in the first storage area, and the failure is detected, the restore unit stores new first data received from the computer in the first storage area until all new first data are assembled in the same stripe, and if all the new first data have been assembled in the same stripe, creates the second data on the basis of these respective first data.

10. The storage controller according to claim 8, wherein, if the computer requests update of the first data stored in the first storage area and the failure, which is determined to be unrestorable in use of RAID, is detected, the restore unit stores new first data received from the computer in the first storage area until all new first data are assembled in the same stripe, and if all the new first data have been assembled in the same stripe, creates the second data on the basis of these respective first data.

11. The storage controller according to claim 1, wherein the storage device is constituted as a device for storing new first data in an address that differs from a storage-destination address of a stored first data, and for updating the stored first data with the new first data by changing address mapping.

12. The storage controller according to claim 1, wherein the storage device is constituted as a device for updating the stored first data with the new first data by storing the new first data in the same address as a storage-destination address of the stored first data.

13. The storage controller according to claim 10, wherein the restore unit notifies the computer of all the new first data in the same stripe, which are necessary for creating the second data.

14. A storage device failure detection method, which detects and restores a failure of data stored in a storage device, comprising the steps of:

creating a first check data on the basis of a first data if the first data received from a computer is written to the storage device, associating this first check data correspondent to the first data, and storing the associated first check data in a first storage area;

creating a second check data which is associated with the first check data, associating this second check data with a second data associated with the first data, and storing the associated second check data in a second storage area;

checking whether or not the first check data read out from the first storage area matches the second check data read out from the second storage area if the first data is read out from the first storage area;

determining that the first data was correctly read out from the first storage area if the first check data matches the second check data;

detecting an occurrence of a failure if the first check data does not match the second check data.

15. The storage device failure detection method according to claim 14, wherein

the first check data and the second check data respectively comprise update sequence information, which shows an update sequence of the first data stored in the first storage area, and

if the first check data and the second check data do not match, the update sequence information in the first check data is compared with the update sequence information in the second check data, whether one of the first data and the second data is older than the other is determined, and the occurrence of a failure in the older data is detected.

16. The storage device failure detection method according to claim 14, further comprising

restoring the failure by using the second data if the failure is detected by a checking unit.

17. A storage controller which operates in accordance with a request from a computer, comprising:

a controller for controlling a plurality of storage devices in accordance with a request from the computer,

the controller comprising;

a first communication control unit for communicating with the computer;

a second communication control unit for communicating with the respective storage devices which provide respective storage areas;

a check data setting unit which creates a first check data and a second check data on the basis of a first data received via the first communication control unit from the computer, and which associates the first check data with the first data and stores the associated first check data via the second communication control unit in a first storage area in which the first data is stored, and associates the second check data with parity data which is created based on the first data and other first data in the same stripe, and stores the second check data via the second communication control unit in a second storage area;

a checking unit which checks whether or not the first check data corresponding to the first data read out via the second communication control unit from the first storage area matches the second check data read out via the second communication control unit from the second storage area, and which, if the first check data matches the second check data, determines that the first data was read out correctly from the first storage area, and if the first check data and the second check data do not match, detects an occurrence of a failure; and

a restore unit which restores the failure if the checking unit detects the failure, and which, if the failure can be restored on the basis of RAID, restores the first data, in which the failure occurred, on the basis of other first data and parity data in the same stripe, and if the failure cannot be restored on the basis of RAID, stores a new first data received from the computer in the first storage area until all new first data are assembled in the same stripe, and if all the new first data have been assembled in the same stripe, creates a new parity data on the basis of these respective first data.