JP2008065601A - Disk array device, method and program for reconstructing disk during backup process - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable a disk array device, which is capable of backup, to reconstruct a disk even in the event that the disk has failed during a backup process. <P>SOLUTION: The disk array device includes a backup process part 15, a backup storage device 30, a reconstruction process part 17, and an arbitration control part 16. The reconstruction process part 17 operates to reconstruct the disk by means of data read from the backup storage device 30. The reconstruction process part 17 operates to perform a reconstruction process using the latest backup data. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a disk array device that reconstructs a disk using backup data, and more particularly to a disk array device that can reconstruct a disk during backup processing.

  The disk array is also called RAID (Redundancy Arrays of Independent (Inexpensive) Disks), and has a configuration in which a plurality of disks are arranged in an array, and the reliability is improved by adding redundant data to the data. In RAID, for example, RAID0 makes multiple disks appear to be one disk by striping, RAID1 writes the same data to multiple disks using mirroring, and records data and parity on multiple disks. There are RAID 5 and the like with improved fault tolerance. Among these, RAID 5 generates parity information for error correction when data is written to the disk, and distributes and writes it to a plurality of disks together with the remaining data by striping. Note that one disk is required for parity. In this way, even if one disk fails, data can be restored from the remaining disk data and parity.

  Further, in the conventional disk array, when the RAID controller detects a disk failure or the like, it has a function of replacing it with a spare disk and reconstructing (rebuilding) the failed disk data into the replaced disk. At this time, data of the replacement disk is created using a disk other than the replaced disk among the disks constituting the disk array.

  In a disk array system that requires high availability, it is necessary to reconstruct data on the replacement disk after replacing the failed disk in parallel with normal operation. Will cause trouble).

  At the time of parity reconstruction after disk replacement, it is necessary to read data from all disks in a parity group other than the replaced disk, so access performance generally decreases.

  In a configuration such as RAID that constitutes a parity group, the performance decreases further as the number of data drives increases.

FIG. 4 is a schematic diagram showing a typical configuration example of a conventional disk array device. Referring to FIG. 4, the access module 14 manages an interface such as transmission / reception of commands, data, and control signals between the host 20 and the disk array. The RAID module 13 includes a RAID controller and management software, and performs data read / write, failure detection, disk reconstruction processing, and the like of the redundantly configured disks 11 _{1 to} 11 _n . The disk access bus 12 provides a connection path between the plurality of disk devices 11 _{1 to} 11 _n and the RAID module 13. The disk device is reconstructed for a redundant array such as RAID 1, 5, 10 or the like. Each of the disk devices 11 _{1 to} 11 _n includes a disk device such as an HDD (Hard Disk Drive), for example, and is simply abbreviated as “disk”.

  In Patent Document 1, in nD + 1P representing a disk configuration such as a RAID that constitutes a parity group, in order to prevent the performance from further degrading as the number of data drives n increases, a single disk is used regardless of n. A configuration for copying is disclosed. Patent Document 2 describes reconstruction (rebuild) into a disk by a RAID controller. If the rebuild rate of the failed disk is 100%, the system is devoted to rebuilding the failed disk. If it is 0%, the system is nothing else. This means that a failed disk is reconstructed during idle time. The default reconstruction rate is about 30%, for example.

JP 2002-108571 A JP 2004-38290 A

  In a redundant disk array device such as RAID level 5, since data of a failed disk (replacement disk) is created using a disk other than the failed disk at the time of disk reconstruction, it takes time for the reconstruction. In particular, as the capacity of the disk increases, the reconstruction time becomes longer, and further, the redundancy decreases during the reconstruction, and the access performance of the disk array device during the reconstruction period decreases.

  On the other hand, in such a disk array device, data is backed up regularly, but the stored data tends to increase, and the backup time is also prolonged.

  During this backup period, in order to read data to be backed up, a load is applied to each disk of the disk array device, and the access performance of the disk array device during the backup period decreases.

  If these two processes for lowering the access performance of the disk array device are performed simultaneously, there is a problem in that each affects each other and each does not end within the target time.

  In particular, if a regularly scheduled backup does not end within the planned time, it will be necessary to review the plan again.

  For this reason, even if a disk failure occurs during the backup process, the disk cannot be immediately reconstructed during the backup process.

  In this way, the disk cannot be rebuilt immediately during the backup process, so a long period of non-redundant data occurs until the disk rebuild is completed after the backup is completed. There is a problem that a dangerous state of data loss continues for this long period.

  Therefore, the present invention was created to solve the above problems, and its main purpose is to shorten the completion time of backup processing and reconstruction processing, and to reduce the possibility of data loss. DISCLOSURE OF THE INVENTION An object of the present invention is to provide a disk array device, method and program that can be used.

  The invention disclosed in the present application has the following general configuration as means for solving the above-described problems.

  A disk array device according to the present invention includes a plurality of disk devices constituting a disk array, a backup processing unit for backing up the data of the disk array, and a disk to be reconstructed among the plurality of disk devices constituting the disk array. Reconstruction processing unit for reconstructing data from backup data to the disk device, and backup processing by the backup processing unit and backup data from the backup data by the reconstruction processing unit when backing up the data of the disk array An arbitration control unit that controls parallel execution with the rebuilding process of the disk device to be rebuilt.

  The disk array device according to the present invention comprises a backup storage device, wherein the backup processing unit backs up the data of the disk array to the backup storage device, and the reconstruction processing unit is backed up to the backup storage device Rebuild the disk device to be rebuilt from the backup data.

  In the disk array device according to the present invention, the arbitration control unit performs arbitration so that a block of backup data referred to by the reconstruction processing unit in a reconstruction process does not overtake a backup block by the backup processing unit.

  In the disk array device according to the present invention, the rebuild processing unit, when performing the rebuild process from the backup data under the control of the arbitration control unit, the block that has been backed up by the backup processing unit. For the block data that has not yet been backed up, control is performed to wait for the rebuilding of the block data that has not been backed up by the backup processing unit.

  In the disk array device according to the present invention, the arbitration control unit obtains information on a block that has been backed up from the backup processing unit, and notifies the reconstruction processing unit of the information.

  In the disk array device according to the present invention, the arbitration control unit acquires information on a block that has been backed up from the backup processing unit and notifies the reconstruction processing unit of the backup processing unit. When the reconstruction process is started, the data is read from the backup storage device for the block that has been backed up and reconstructed.For the block data that has not been backed up, the backup processing unit Wait for the backup to the backup storage device to complete.

The method according to the present invention includes a backup step of backing up data in a disk array to the backup storage device;
Reconstruction step of reconstructing data from backup data already backed up in the backup storage device for a disk device to be reconstructed among a plurality of disk devices constituting the disk array;
An arbitration step for performing control to execute the backup step and the reconstruction step in parallel;
And the reconstruction process can be executed during the backup process.

In the method according to the present invention, the rebuilding step performs a rebuild using a backup completion block when performing a rebuild process from backup data,
For a backup incomplete block, control is performed to wait for reconstruction until the backup processing unit backs up the block.

  In the method according to the present invention, the arbitration step acquires backup completion block information from the backup step, and the reconstruction step notifies the backup completion block.

A computer program according to the present invention includes a backup process for backing up disk array data to the backup storage device;
Reconstruction processing for reconstructing data from backup data that has already been backed up in the backup storage device for a disk device to be reconstructed among a plurality of disk devices constituting the disk array;
Arbitration processing for performing control to execute the backup processing and the reconstruction processing in parallel;
It consists of a program that causes a computer to execute.

In the program of the present invention, the rebuild process is performed using a backup completion block when performing the rebuild process from backup data,
For a backup incomplete block, control is performed to wait for reconstruction until the backup processing unit backs up the block.

  In the program according to the present invention, the adjustment process acquires information on a backup completion block from the backup process, and notifies the backup completion block of the reconstruction process.

  According to the present invention, even if a disk failure occurs during the backup process, the reconstruction process can be started immediately, and the possibility of data loss can be reduced.

  Further, according to the present invention, since the rebuilding process can be started immediately during the backup process, the completion of the rebuilding process can be accelerated without affecting the backup process.

  Next, the best mode for carrying out the present invention will be described in detail with reference to the drawings. The present invention prepares a backup data section for a plurality of storage units (disk devices), backs up the data of the plurality of storage units to a backup storage device, and detects a failed storage among the plurality of storage units. -When rebuilding the data of the replacement storage unit when the unit is replaced, when performing backup processing and rebuild processing when performing based on the data backed up in the backup storage device, The arbitration control unit performs processing smoothly while maintaining cooperation.

  The present invention includes a backup processing unit (15), a backup storage device (30), a reconstruction processing unit (17), and an arbitration control unit (16). The reconstruction processing unit (17) is a backup storage device. When reconstructing the disk device to be reconstructed using the data read from (30), the reconstructing process is performed using the latest backup data.

  In a normal disk array device, even if a disk failure occurs during the backup process, the rebuild process is not started.

  However, according to the present invention, even if a disk failure occurs during the backup of the disk device data, the reconstruction process is started and the area where the latest backup data is recorded in the backup storage device (30). Only the (block) can be reconstructed, and old backup data can be prevented from being written to the reconstruction target disk.

  In this way, the latest backup data is used for the reconstruction target disk device without affecting the backup processing and without delaying the start of the reconstruction processing of the replacement target disk device such as a replacement disk device. The rebuild process can be performed, and the backup process and the rebuild process can be operated in parallel, so that the rebuild process can be completed quickly. Hereinafter, description will be made with reference to examples.

FIG. 1 is a diagram showing a system configuration of an embodiment of the present invention. Referring to FIG. 1, the disk array includes a plurality of disks 11 _{1 to} 11 _n , a disk access bus / switch 18, a RAID module 13, an access module 14, a backup processing unit (Backup Module) 15, and a reconstruction. The processing unit (Rebuild Module) 17 and the arbitration control unit (Arbitration Module) 16 are provided. The RAID module 13 and the rebuilding processing unit 17 are connected to the plurality of disks 11 _{1 to} 11 _n via the disk access bus / switch 18. Access such as writing and reading. A backup storage device (backup storage) 30 is connected to the disk array 10 via a predetermined interface. Of course, the backup storage 30 may have a disk array configuration. Each of the disks 11 _{1 to} 11 _n includes a disk device such as an HDD (Hard Disk Drive).

In the present invention, the backup module 15 backs up the data of the plurality of disks 11 _{1 to} 11 _n constituting the disk array to, for example, a backup storage periodically. That is, the backup module 15 reads data from the disk array via the RAID module 13 (error detection and correction processing such as parity check is performed), and writes the read data to the backup storage 30.

In this embodiment, when one of the plurality of disks 11 _{1 to} 11 _n constituting the disk array fails, the failed disk is replaced with a spare disk or the like, and data is reconstructed on the replacement disk, preferably backup Data transfer from the storage 30 to the replacement disk is performed via the reconstruction processing unit 17 and the disk access bus / switch 18, and is configured not to affect normal access from the host 20 to other disks.

  That is, normal access from the host 20 to other disks (disks other than the replacement disk) is performed via the access module 14, the RAID module 13, and the disk access bus / switch 18, and the reconstruction process from the backup storage 30 is performed. The RAID module 13 is performed via a reconstruction processing unit 17 and a disk access bus / switch 18 provided in parallel. The details of the operation of the disk array apparatus shown in FIG.

  FIG. 2 is a flowchart showing the procedure of the reconstruction process during backup of the arbitration control unit (arbitration module) 16. The arbitration control unit 16 obtains information on the backup completion block from the backup processing unit 15 (step S11), and notifies the rebuild processing unit 17 of the information on the backup completion block. (Step S12).

  Further, it is determined whether the backup completion block obtained from the backup processing unit is the last block of the backup (step S13), and the series of operations of steps S11 to S13 are repeated until the last block becomes the last block.

  FIG. 3 is a flowchart showing the procedure of the rebuild process from the backup data in the rebuild module 17.

  The reconstruction processing unit 17 sets the reconstruction target block at the top (step S20).

  Next, it is determined whether or not the block is completed (step S21). If all blocks are completed (Yes branch in step S21), the reconstruction process is terminated.

  If all blocks have not been completed (No branch in step S21), it is determined whether the current target block is a block number that is younger than the backup completion block obtained from the arbitration control unit (arbitration module) 16. Confirm (step S22).

  If it is the same or larger than the backup completion block (Yes branch in step S22), the confirmation is continued until the backup completion block is updated.

  If the block is younger than the backup completed block (No branch in step S22), it is determined whether the target block is a block for parity (step S23).

  If the target block is a parity block (Yes branch in step S23), the data of the related block is read from the backup storage (step S26), parity data is generated (step S27), and the generated parity data is written to the target block. (Step S28).

  If the target block is not a parity block (No branch in step S23), the data of the target block is read from the backup storage device (step S24) and written to the target block (step S25).

  When the reconstruction process, which is the target block writing process, is completed, the target block is set to the next block (step S29), the process proceeds to step S21, and a series of processes (from the next block to the last block in sequence) ( Steps S22 to S29) are repeated.

  According to the present invention, in a disk array device that performs backup, it can be applied so that the rebuild process is immediately performed during the backup process. Further, the present invention can also be applied to a storage system such as a SAN (Storage Area Network) configured by a plurality of storage apparatuses so as to perform a reconstruction process during a backup process.

  Although the present invention has been described with reference to the above-described embodiments, the present invention is not limited to the configurations of the above-described embodiments, and various modifications that can be made by those skilled in the art within the scope of the present invention. Of course, including modifications.

It is a block diagram which shows the structure of one Embodiment of this invention. In one Embodiment of this invention, it is a flowchart which shows the procedure of the rebuild process in backup of the arbitration control part. In one Embodiment of this invention, it is a flowchart which shows the procedure of the rebuild process during the backup of a rebuild process part. It is a schematic diagram which shows the typical structural example of the conventional disk array apparatus.

Explanation of symbols

10 Disk array device
11 _{1 to} 11 _n disks 13 RAID module (RAID module)
14 Access Module
15 Backup processing module (backup module)
16 Arbitration module
17 Rebuild module
18 Disk access bus / swich
20 Host computer
30 Backup storage

Claims (12)

A plurality of disk devices constituting a disk array;
A backup processing unit for backing up the data in the disk array;
Reconstruction processing unit for reconstructing data from backup data for a disk device to be reconstructed among a plurality of disk devices constituting the disk array;
When backing up the data of the disk array, an arbitration control unit that controls parallel execution of backup processing by the backup processing unit and reconstruction processing from backup data by the reconstruction processing unit;
A disk array device comprising: With backup storage,
The backup processing unit backs up the data in the disk array to the backup storage device,
2. The disk array device according to claim 1, wherein the rebuild processing unit performs rebuild processing of a disk device to be rebuilt from backup data backed up in the backup storage device.   The arbitration control unit performs arbitration so that a block of backup data referred to by the reconstruction processing unit in a reconstruction process does not overtake a backup block by the backup processing unit. 2. The disk array device according to 2. The reconstruction processing unit performs reconstruction using a block that has been backed up by the backup processing unit when performing the reconstruction processing from the backup data under the control of the arbitration control unit,
2. The block data that has not yet been backed up is controlled to wait for rebuilding of a disk device to be rebuilt until the backup processing unit backs up the block that has not been backed up. The disk array device described.   5. The disk array device according to claim 1, wherein the arbitration control unit acquires information on a block whose backup has been completed from the backup processing unit and notifies the reconstruction processing unit of the information. The arbitration control unit obtains information of a block that has been backed up from the backup processing unit, and notifies the reconstruction processing unit,
When the rebuild process starts during the backup process, the rebuild processing unit reads the data from the backup storage device for the blocks that have been backed up, performs the rebuild, and blocks that have not been backed up 3. The disk array device according to claim 2, wherein the disk array device waits for the backup to the backup storage device by the backup processing unit to be completed. A backup process for backing up the data in the disk array to a backup storage device;
Among the plurality of disk devices constituting the disk array, a reconstruction process for reconstructing data from the backup data already backed up in the backup storage device for the disk device to be reconstructed;
An arbitration step for performing control to execute the backup step and the reconstruction step in parallel;
Including
A disk array rebuilding method, wherein a rebuild process can be executed during a backup process. In the rebuilding process, when performing the rebuilding process from backup data, rebuilding is performed using a backup completion block,
8. The disk array reconstruction method according to claim 7, wherein control for waiting for reconstruction is performed for block data that has not been backed up until the backup processing unit backs up the block.   8. The disk array re-creation according to claim 7, wherein the arbitration step acquires information on a block whose backup has been completed from the backup step, and notifies the reconstruction step of the block whose backup has been completed. Construction method. A backup process for backing up the disk array data to a backup storage device;
Among the plurality of disk devices constituting the disk array, a rebuild process for rebuilding data from the backup data already backed up in the backup storage device for the disk device to be reconstructed;
Arbitration processing for performing control to execute the backup processing and the reconstruction processing in parallel;
A program that causes a computer to execute. The program according to claim 10, wherein
The rebuild process is performed by using a backup completion block when performing the rebuild process from backup data,
A program that performs control for waiting for reconstruction until block backup is performed by the backup processing for block data that has not yet been backed up. The program according to claim 10, wherein
The arbitration process acquires information on a block whose backup has been completed from the backup process, and notifies the block whose backup has been completed to the reconstruction process.

Images