JPH05127838A - Disk recording and reproducing device - Google Patents

Abstract

PURPOSE:To obtain the disk recording and reproducing device which can realize an efficient data recovery processing while avoiding a sector already recovering data with normal read/write in the case of the data recovery processing due to the fault of a disk drive. CONSTITUTION:Based on a flag stored in a flag memory 2a, a disk array controller (DAC) 2 executes the data recovery processing only to the sector requiring the data recovery processing concerning the new exchanged disk drive among respective disk drives 3A-3D.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a disk array device including a plurality of disk drives having a hard disk, for example, and relates to a disk recording / reproducing device having a data recovery function.

[0002]

2. Description of the Related Art Conventionally, a disk array device has a plurality of disk drives, and each disk drive is configured to be controlled by a common array controller. The array controller is connected to the host computer through the host bus, and accesses each disk drive according to an instruction from the host computer.

A disk array device employs a parity check method in order to improve the reliability of the device. Specifically, a method is used in which each write data of the same byte written in the same physical sector of each disk drive is subjected to exclusive OR operation processing, and parity data is generated and stored based on the operation result. ing.

When a failure occurs in the disk drive, the array controller carries out a repair process for the data that cannot be reproduced due to the failure due to the stored parity data. The parity data is stored in a disk drive dedicated to parity or distributed and stored in each disk drive. Normally, the method of storing data in a disk drive dedicated to parity causes a decrease in access speed, and therefore a method of storing data in a distributed manner is adopted.

When a failure occurs in the disk drive and the disk drive has to be replaced, the failed disk drive is replaced with a new drive. The array controller executes the process of repairing the data of the failed disk drive for the disk of the replaced drive. This data restoration processing is executed on the new disk in order from the first sector.

Normally, at the time of data recovery processing, data read / write is also executed on a new disk in response to an access request from the host computer. At the time of reading, after repairing the sector to be accessed, the repaired data is read. At the time of writing, the data is written in the unrestored sector, and the parity data is updated based on this data.

[0007]

Normally, as described above, when a data recovery process is executed on a new disk drive, data read / write on the new disk is also performed.
A write is executed to recover the data. For this reason, a repair process may be performed again by data recovery for a sector that has already been repaired by read / write access. Therefore, useless repair processing occurs and the efficiency of the data repair processing is reduced.

An object of the present invention is to perform a disk recording / reproducing process capable of realizing an efficient data recovery process while avoiding a sector in which data has been already recovered by normal read / write during data recovery process due to a disk drive failure. To provide a device.

[0009]

According to the present invention, a sector of each disk, which is composed of a plurality of disks and is used to generate parity data, and a parity sector in which this parity data is stored are defined as a parity group. Disk recording means, data recovery means for executing data recovery processing of a sector belonging to the same parity sector group as a parity sector on a new disk exchanged in the disk recording means, data repaired sector A disc recording / reproducing apparatus having: an instruction data storage unit for storing instruction data for instructing the data; and a control unit for controlling the data restoration unit to execute the data restoration process.

[0010]

In the present invention, the control means controls the data restoration means based on the instruction data stored in the instruction data storage means so that the data restoration processing is executed only in the sector in which the data restoration processing is necessary.

[0011]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a disk array according to the embodiment.
It is a block diagram which shows the principal part of a system. This system comprises a host computer 1, a disk array controller (DAC) 2, and a plurality of (here, four) disk drives 3A to 3D having hard disks.

The host computer 1 transmits commands and data to and from the DAC 2 via the host bus 4. The DAC 2 controls each of the disk drives 3A to 3D via the local bus 5.

The DAC 2 has a flag memory 2a according to the present invention. As shown in FIG. 2, the DAC 2 includes a CPU 2b, a ROM 2c, a local bus interface 2d, a host bus interface 2e and a buffer memory 2f in addition to the flag memory 2a. CPU2
Reference numeral b is a controller main body that executes a control operation. RO
The M2c is a read-only memory that stores the program of the CPU 2b and various control data. The local bus interface 2d is connected to the local bus 5 and is an interface between the disk drives 3A to 3D and the DAC 2. The host bus interface 2e is connected to the host bus 4 and is an interface between the host computer 1 and the DAC 2.

The buffer memory 2f is a memory for temporarily storing data to be transferred with the host computer 1. The flag memory 2a is used for each disk drive 3A.
In 3D, it is a register for storing a flag as instruction data for instructing the data-recovered sector of the drive for data recovery. Next, the operation of the embodiment will be described.

First, as shown in FIG. 3, the DAC 2 includes disk drives 3A to 3D for each physical sector (0 to 8).
3D logical sectors are sequentially allocated. Here, "Da ~
“D” indicates a sector for recording data, and “P” indicates a parity sector for recording parity data. The number indicates the sector number. It should be noted that a parity sector group including sectors of each disk and parity sectors used to generate the parity data (P) is configured.

In this embodiment, the disk drives 3A-3
It is assumed that the disk drive 3C fails in D and is replaced with a new disk drive (step S1 in FIG. 4). The DAC 2 carries out a data recovery process for the new replaced disk drive 3C.

The CPU 2b of the DAC 2 normally operates as shown in FIG.
As shown in, the data restoration process is executed in order from the sector 0 of the disk drive 3C (steps S10 and S1).
3, S16). At this time, in the present invention, the CPU 2b
The flag memory 2a is checked to determine whether or not the flag indicating the corresponding sector number is set (steps S11 and S12).

As a concrete example, as shown in FIG. 3, since the flag (logical "1") indicating the sector 0 is not set in the flag memory 2a, the data restoration process for the sector 0 of the disk drive 3C is executed. (Step S12: NO, S13). Here, in the flag memory 2a, the flags from sector 0 to sector 7 are set sequentially from the left of address 0, and the flags from sector 8 to sector 15 are further set from the left of address 1. The same applies hereinafter. In this embodiment, as shown in FIG. 3, flags indicating sectors 6 and 8 are set in the flag memory 2a.

The CPU 2b accesses the sectors of each disk belonging to the same sector group. In the case of this embodiment, the CPU 2b performs an exclusive OR operation on the write data (including parity data) in byte units written in the respective logical sectors of the disk drives 3A, 3B, 3D belonging to the same physical sector (0). The calculation is executed, and the data of the sector Dc0 is generated based on the calculation result. That is, the CPU 2b
Performs an exclusive OR operation of "Dc0 = Da0 [+] Db0 [+] P0". Here, [+] is a symbol indicating an exclusive OR operation. In addition, when the data of the parity data (P1) is generated, 3A, 3B and 3D of the physical sector (1) are generated.
Data acts as parity data. CPU2b
After finishing the data restoration process, sets a flag indicating the corresponding sector number in the flag memory 2a (step S14).

By the way, as shown in FIG. 4, when an access request is issued from the host computer 1 during the above-mentioned data restoration processing (YES in step S2, S
3), the CPU 2b of the DAC 2 checks the sector to be accessed (step S4). That is, the CPU 2b
It is determined whether the physical sector to be accessed is a sector that requires data restoration processing.

The CPU 2b checks the flag memory 2a and determines whether or not the flag indicating the corresponding sector number is set (step S5). If the flag is set (YES in step S5), the CPU
2b executes read / write processing for the corresponding sector number (step S6).

Specifically, as shown in FIG. 3, since the data restoration process has already been completed for the sector Dc6 of the disk drive 3C, for example, an access request is issued to the logical sector Da6 of the physical sector 6. The data write operation is executed. Also, the sector D of the disk drive 3C
Since the data restoration process has already been completed for c8, for example, the logical sector Dc8 and the logical sector P of the physical sector 8 are
A data read operation is executed for 8 in response to an access request.

On the other hand, if the flag is not set in the flag memory 2a (NO in step S5), the CPU
2b executes a data restoration process before the process for the access request from the host computer 1 (step S
7) Then, a flag is set for the sector whose data has been restored (step S8).

After the end of this data restoration process, a read / write process as shown in step S6 is executed. At this time, in the read process, the corresponding sector of the disk drive 3C is first restored and then the read operation is executed.
Also, during the write processing, it is necessary to update the parity sector of the same sector after writing the data to the relevant sector.

In this way, when the defective disk drive is replaced and the data recovery processing is executed on the new disk drive, the flag indicating the sector for which the repair processing is completed is set in the flag memory 2a. (Steps S14 and S8). The DAC 2 has a flag memory 2a.
Is checked, it is determined that the data restoration process has been completed for the sector indicated by the set flag, and the data restoration process is not executed again.

Although the disk array device of the hard disk has been described in this embodiment, the present invention is not limited to this and can be applied to a system in which the disk is an optical disk. In short, a system that controls a plurality of disk drives by a common controller will do.

[0028]

As described in detail above, according to the present invention, when a defective disk drive is replaced and data restoration processing is executed, the data is restored by read / write access and the flag is set. Then, the data restoration process is not executed for the sector indicated by the set flag, and the data restoration process is executed for the sector where the flag is not set. Therefore, since the data restoration process is controlled based on the flag, it is possible to reliably prevent the situation where the data restoration process is executed again for the sector already restored by the read / write access during the data restoration process. it can. As a result, it is possible to prevent unnecessary data restoration processing from being performed on a sector that requires data restoration processing, and as a result, the efficiency of data restoration processing can be improved.

[Brief description of drawings]

FIG. 1 is a block diagram showing a main part of a disk array system according to an embodiment of the present invention.

FIG. 2 is a block diagram showing the configuration of a DAC according to the embodiment.

FIG. 3 is a conceptual diagram for explaining the operation of the embodiment.

FIG. 4 is a flowchart for explaining the operation of the embodiment.

FIG. 5 is a flowchart for explaining the operation of the embodiment.

[Explanation of symbols]

1 ... Host computer, 2 ... Disk array controller (DAC), 3A to 3D ... Disk drives.

Claims (1)

[Claims] 1. A disk recording means comprising a plurality of disks, wherein each disk sector used to generate parity data and a parity sector in which this parity data is stored constitute a parity group, and the disk recording means. Data recovery means for executing a data recovery process of a sector belonging to the same parity sector group as the parity sector on the new disk exchanged in the means, and the sector recovered by the data recovery means Instruction data storing means for storing instruction data for instructing the data, and the data restoring means for executing the data restoration processing only on the sector requiring the data restoration processing based on the instruction data stored in the instruction data storage means. And a control means for controlling the disk. Recording / playback device.