Disk array system, disk controller, and method for performing rebuild process

• US 20090327801A1
• Filed: 04/13/2009
• Published: 12/31/2009
• Est. Priority Date: 06/30/2008
• Status: Abandoned Application

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Konta, Yoshihito, Daikokuya, Hidejirou, Kubota, Norihide, Ito, Mikio, Takahashi, Hideo, Ikeuchi, Kazuhiko, Maeda, Chikashi
• Application Number:
  US12/385,585
• Publication Number:
  US 20090327801A1
• Time in Patent Office: Days
• US Class Current: 714/6.11
• CPC Class Codes: G06F 11/1092 Rebuilding, e.g. when physically replacing a failing disk
• CPC Class Codes: G06F 11/1662 the resynchronized component or unit being a persistent storage device re-synchronization of failed mirror storage G06F11/2082; rebuild or reconstruction of parity RAID storage G06F11/1008
• CPC Class Codes: G06F 11/2094 Redundant storage or storage space G06F11/2056 takes precedence
• CPC Class Codes: G06F 12/0866 for peripheral storage systems, e.g. disk cache
• CPC Class Codes: G11B 2220/415 Redundant array of inexpensive disks [RAID] systems

Abstract

In a disk array system, when a failure occurs in a disk unit under control, a disk controller performs a rebuild process for rebuilding data stored on the faulty disk unit on a spare disk unit (HS). When a rebuild control section accepts an I/O request from a host before completing the rebuild process in all target areas, the rebuild control section specifies a management unit area including a target area of the I/O request and determines whether the rebuild process is completed in the management unit area. If the rebuild process is not completed in the management unit area, the rebuild control section performs the rebuild process in the management unit area by a rebuild process section and rebuilds data on the HS. After that, an I/O request handling section handles the I/O request.

Claims

1. A disk array system for distributing and storing data on a plurality of disk units and for accessing the plurality of disk units in response to an I/O request from a host, the system comprising:

the plurality of disk units which stores distributed and redundant data;

a spare disk unit which functions in place of part of the plurality of disk units in which a failure has occurred; and

a disk controller including;

a rebuild process section which restores data stored on a faulty disk unit by the use of data stored on disk units other than the faulty disk unit by management unit areas obtained by dividing a storage area of each disk unit by predetermined management units, and writes the data onto the spare disk unit;

a management information storage section which stores rebuild management information including information which indicates whether a rebuild process is completed in each management unit area; and

a rebuild control section which accepts the I/O request from the host, specifies a management unit area including a target area of the I/O request in the case of the target area of the I/O request being included in a target area of the rebuild process, rebuilds data in the management unit area by the rebuild process section in the case of the determination that the rebuild process is not yet completed in the management unit area specified being made on the basis of the rebuild management information, and permits the I/O request after rebuilding the data.

2. The disk array system according to claim 1, wherein:

the rebuild process section performs an ordinary rebuild process for rebuilding data on the spare disk unit in order at predetermined timing by the management unit areas; and

the rebuild process section resumes the ordinary rebuild process in a management unit area next to a management unit area in which the rebuild process is performed before the I/O request after the rebuild process is completed in the management unit area which corresponds to the target area of the I/O request and in which the rebuild control section instructs the rebuild process section to perform the rebuild process.

3. The disk array system according to claim 1, wherein:

the rebuild process section performs an ordinary rebuild process for rebuilding data on the spare disk unit in order at predetermined timing by the management unit areas; and

the rebuild process section resumes the ordinary rebuild process in a management unit area next to the management unit area in which the rebuild process is performed in response to the I/O request after the rebuild process is completed in the management unit area which corresponds to the target area of the I/O request and in which the rebuild control section instructs the rebuild process section to perform the rebuild process.

4. The disk array system according to claim 1, wherein:

the disk controller includes a cache memory for temporarily storing a copy of data in area selected from the plurality of disk units and the spare disk unit and a cache management section for managing the cache memory;

the rebuild control section calculates a number of I/O requests for each management unit area which are from the host and which are accepted during a predetermined period, and sets the number in the rebuild management information; and

the cache management section compares the number of I/O requests for each management unit area which are from the host with a predetermined specified value on the basis of the rebuild management information, and makes data in each management unit area resident in the cache memory in the case of the number of I/O requests for each management unit area which are from the host being greater than the specified value.

5. The disk array system according to claim 4, wherein if a target management unit area of the rebuild process includes the target area of the I/O request from the host, the rebuild process section stores data in a same stripe which corresponds to the target management unit area, which is used for restoring data in the target management unit area, and which is read out from the disk units other than the faulty disk unit in the cache memory.

6. The disk array system according to claim 4, wherein if the number of I/O requests for a management unit area which are from the host is greater than the specified value and data in the management unit area is not stored in the cache memory, the cache management section makes the data in the management unit area resident in the cache memory and stores data in a same stripe corresponding to the management unit area in the cache memory.

7. A disk controller for distributing and storing data on a plurality of disk units and for accessing the plurality of disk units in response to an I/O request from a host, the disk controller comprising:

a rebuild process section which restores data stored on a faulty disk unit by the use of data stored on disk units other than the faulty disk unit by management unit areas obtained by dividing a storage area of each of the plurality of disk units for storing distributed data and redundant data by predetermined management units, and writes the restored data onto a spare disk unit which functions in place of the faulty disk unit;

a management information storage section which stores rebuild management information including information which indicates whether a rebuild process is completed in each management unit area; and

a rebuild control section which accepts the I/O request from the host, for specifying a management unit area including a target area of the I/O request in the case of the target area of the I/O request being included in a target area of the rebuild process, rebuilds data in the management unit area by the rebuild process section in the case of the determination that the rebuild process is not yet completed in the management unit area specified being made on the basis of the rebuild management information, and permits the I/O request after rebuilding the data.

8. A method for performing a rebuild process by a disk array system for distributing and storing data on a plurality of disk units and for accessing the plurality of disk units in response to an I/O request from a host, the method comprising:

restoring data stored on a faulty disk unit by the use of data stored on disk units other than the faulty disk unit by management unit areas obtained by dividing a storage area of each of the plurality of disk units for storing distributed data and redundant data by predetermined management units, and writing the restored data onto a spare disk unit which functions in place of the faulty disk unit by a rebuild process section;

accepting the I/O request from the host, specifying a management unit area including a target area of the I/O request in the case of the target area of the I/O request being included in a target area of the rebuild process, reading out rebuild management information including information which indicates whether the rebuild process is completed in each management unit area from a management information storage section, and determining by a rebuild control section on the basis of the rebuild management information whether the rebuild process is not yet completed in the management unit area specified;

rebuilding data in the management unit area by the rebuild process section in the case of the determination that the rebuild process is not yet completed in the management unit area; and

permitting the I/O request, by the rebuild control section, after rebuilding the data by the rebuild process section.