JPH10222315A - Method and device for error recovery of doubled hard disk drives - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate maintenance operation at the time of the alteration of a defective storage area by substituting a normal storage area in a 1st hard disk drive for a defective storage area on the 1st hard disk drive, copying data from a 2nd hard disk drive, and restoring the data in the defective storage area. SOLUTION: An error restoration start part 9 outputs an error restoration signal once receiving an error report signal. An substituting process indication part 10 once receiving this error restoration signal requests a DK (hard disk drive) common control part 2 to performs a substituting process for an error occurrence sector of DK#0. After the substituting process for DK#0 is completed, a copy instruction part 11 requests the DK common control part 2 to rewrite data. Namely, the copy instruction part 11 writes read data from DK#1 to DK#0 where the error has occurred through a DK#0 control part 2 and a DK#0 control part 3 to restore the data. Consequently, the maintenance operation can be made efficient.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an error recovery method and apparatus for a duplicated hard disk drive, which automatically performs a process of replacing a defective storage area in which an error has occurred during data reading or data writing and recovery of the data. It is about.

[0002]

2. Description of the Related Art A conventional redundant control device for a hard disk drive will be described with reference to the drawings. FIG. 6 is a block diagram showing a conventional example, and its configuration will be described in detail.

The CPU 1 serving as a processing device supplies data read access and data write access to the data file control unit 16 as necessary. When the data file control unit 16 receives a data read access or a data write access, the hard disk devices 5, 8 provided in the data file units 17, 18 (hereinafter referred to as data file unit # 0 and data file unit # 1). (Hereinafter, DK #
0, DK # 1).

The data file sections # 0, # 1
Are provided with data end instructing units 19 and 20 in addition to DK # 0 and DK # 1, respectively. These data end instructing units 19 and 20 output signals indicating the data write state of the hard disk device.

The data end analysis unit 21 detects the data write state based on the signals received from the data end instruction units 19 and 20, and drives the copy instruction unit 11 when judging that the data write is not completed normally. . The copy instructing unit 11 reads data from the hard disk device to which the data has been normally written, and transfers the data to the hard disk device to which the data has been abnormally written to restore the data.

As described above, the conventional redundant control device for a hard disk device checks whether or not the data writing has been completed normally. If the data cannot be normally written, the data is transferred from the other hard disk device and re-transmitted. Writing was performed.

[0007]

However, when a physical error occurs in a storage area (that is, a sector, a track, or the like), rewriting must be performed so that the abnormal storage area is not repaired. The same data error occurs again. Therefore, in the above-described conventional redundant control device, when a data error occurs even after rewriting, the hard disk device in which the data error has occurred is once disconnected from the system,
The process of manually replacing a defective sector (hereinafter, referred to as sector replacement) has been performed. Therefore, there is a problem that the maintenance staff must be kept on standby at all times, and the maintenance work is burdensome. In addition, when the sector is replaced, the operation is performed in one system, so that the reliability of the system is reduced. An object of the present invention is to solve such a problem, and an error recovery method for a dual hard disk drive that simultaneously simplifies maintenance work when replacing a defective storage area and improves system reliability. And an apparatus.

[0008]

In order to achieve the above object, an error recovery method for a duplicated hard disk drive according to the present invention comprises the steps of: detecting a defective storage area in a first hard disk drive; The area is replaced with a normal storage area in the first hard disk drive, and data in the defective storage area is restored by copying data from the second hard disk drive. Further, the error recovery apparatus for a duplicated hard disk drive according to the present invention controls the driving of the first hard disk drive, and outputs an error notification signal when a defective storage area is detected in the first hard disk drive. A first DK control unit for performing the area replacement processing and the data recovery processing in the defective storage area, controls the driving of the second hard disk drive, and generates an error when a defective storage area is detected in the second hard disk drive. A second DK control unit that outputs a notification signal and performs a process of substituting a defective storage area and a process of restoring data in the defective storage area; An error recovery start unit that starts the substitution process and the data recovery process is provided. Therefore, the present invention can perform the replacement processing of the defective storage area and the data recovery without disconnecting the hard disk device from the system.

[0009]

Next, one embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing one embodiment of the present invention, and its configuration will be described. In FIG. 1, the same components as those in FIG. 6 indicate the same or equivalent parts.

The DK control units 3 and 6 (hereinafter referred to as DK # 0 control unit and DK # 1 control unit) respectively have SCSI interfaces 4 and 7 (hereinafter referred to as SCSI # 0 and SCSI #).
DK # 0 via a general-purpose interface such as
Alternatively, it is connected to the DK # 1 and receives data from the DK # 0 and DK # 1 in response to an instruction from the DK common control unit 2, and reads and writes data. The DK # 0 control unit and the DK # 1 control unit read predetermined recording information of each sector when reading or writing data, and check the parity, CRC, and the like of the read data, thereby obtaining a physical data. It is determined whether there is a defective storage area such as a sector or a track that has been temporarily broken. Then, when a bad sector or the like is detected, an error notification signal is output.

The DK common control unit 2 performs the DK #
In addition to controlling the driving of the 0 control unit and the DK # 1 control unit, a unit for storing and holding the information of the system set in each of the DK # 0 and DK # 1 is provided. That is, the DK
A master system is set in one of # 0 and DK # 1,
On the other hand, a slave system is set. The system information is referred to at the time of reading data and the like, and is used to determine which hard disk device to use.

When receiving an error notification signal from the DK # 0 control unit or the DK # 1 control unit, the error recovery start unit 9 performs an error recovery operation (ie, a process for replacing a bad sector, a process for recovering data in a bad sector, etc.). ), An error recovery signal is output.

Upon receiving the error recovery signal, the alternative processing instruction unit 10 outputs an alternative processing instruction signal to the DK common control unit 2 to execute the alternative processing of the defective sector. The copy instructing unit 11 performs data recovery by copying data from one hard disk device to another hard disk device. That is, when the replacement process instructing unit 10 performs the replacement process for the defective sector, the copy instructing unit 11 reads out the data that should have been written in the defective sector from the other hard disk device, and executes the hard disk replacement for the sector replacement. Transfers data to the device and writes it to recover data.

The operation of the present invention as described above will be described in detail for data reading and data writing.

This embodiment will be described as a control device having a mirroring function. That is, data reading is performed by reading one hard disk device of the master system in sector units (hereinafter, referred to as single-system reading). Data writing is performed by writing the same data to both the master hard disk device and the slave hard disk device in sector units (hereinafter referred to as both system write).

First, an operation procedure in reading data will be described. FIG. 2 is a flowchart showing an operation procedure when the CPU 1 requests data reading.
In step 101, the CPU 1 executes the DK common control unit 2
A request for one-system read is output to

In step 102, the DK common control unit 2 issues a read request to a DK # 0 control unit set in advance in the master system. The DK # 0 control unit is the DK # of its own system.
A read operation is performed on 0.

In step 103, when a read error occurs in the DK # 0 which is the current master system, this read error is caused by the SCSI bus protocol according to the SCSI bus #.
0 is notified to the DK # 0 control unit. Then DK #
The 0 control unit outputs an error notification signal to the error recovery activation unit 9 to request activation of the error data recovery operation. If no read error has been detected, the process proceeds to step 109, where the data read from DK # 0 is sent to the CPU 1, and the read operation ends.

In step 104, upon receiving the error notification signal, the error recovery start-up unit 9 supplies an error recovery signal to the DK common control unit 2 and the alternative processing instruction unit 10 to execute error data recovery processing. Upon receiving the error recovery signal, the DK common control unit 2 sets the master system to the DK.
Switch from # 0 to DK # 1, and switch the slave system from DK # 1 to DK # 0.

In step 105, the DK # 1 control unit which has become the master system in accordance with an instruction from the DK common control unit 2 executes a read operation of the same sector for its own system DK # 1. In step 106, D
If a read error occurs even in K # 1, the CPU 1 is notified of the error as a double failure. If no error has occurred, the process proceeds to step 108. Step 10
In step 7, the alternative processing instruction unit 10 outputs an alternative processing instruction signal in parallel with the operation in step 105. DK
The common control unit 2 receives the DK #
0 requests the DK # 0 control unit to execute the replacement process for the DK # 0 error occurring sector.

In step 108, if no error is detected in DK # 1, the copy instructing unit 11 sends the read data from DK # 1 via the DK common control unit 2 and DK # 0 control unit to the DK in which the error has occurred. Write to # 0 and execute data recovery. In step 109, the DK common control unit 2 sends the data read from DK # 1 to the CPU
1 to end the read operation.

Next, the procedure for writing data will be described. FIG. 3 is a flowchart showing an operation procedure when the CPU 1 requests data writing. Step 2
At 01, the CPU 1 requests the DK common control unit 2 to write to both systems.

In step 202, the DK common control unit 2 requests the DK # 0 control unit and the DK # 1 control unit to perform a write operation to write the same data to the DK # 0 and DK # 1 of both systems. .

In step 203, the DK # 0 control unit and the DK # 1 control unit monitor write errors in DK # 0 and DK # 1, respectively, and for example, an error is detected in DK # 0 of the master system. In this case, the error recovery start unit 9 is notified, and if not detected, the write operation ends.

In step 204, upon receiving the error notification signal, the error recovery starting unit 9 outputs an error recovery signal to start the error recovery processing. Upon receiving the error recovery signal, the substitution processing instruction unit 10 requests the DK common control unit 2 to execute the substitution processing for the DK # 0 error occurring sector.

In step 205, when the substitute process for DK # 0 is completed, the copy instruction unit 11 requests the DK common control unit 2 to rewrite data. That is,
The copy instructing unit 11 writes the read data from the DK # 1 to the DK # 0 where the error has occurred via the DK common control unit 2 and the DK # 0 control unit, and executes data recovery. In step 206, if a write error is detected again in DK # 0, an error is notified to the CPU 1 as a double failure. If no error is detected, the write operation ends.

FIG. 4 is a block diagram showing another embodiment of the present invention. This embodiment differs from the embodiment shown in FIG. 1 in that a disk controller is provided for each hard disk device, and control signals and data between the controllers are transmitted and received via a confounding bus.

The configuration of the duplicated control device for the hard disk device shown in FIG. 4 will be described. In the figure, the same or similar reference numerals as those in FIG. 1 indicate the same or equivalent parts. The disk controller # 0 communicates with the DK # 0 control unit and S
CSI # 0, error recovery activation unit 9, and alternative processing instruction unit 10
And a copy instructing unit 11. The disk controller # 1 has the same configuration as the disk controller # 0, and is connected to the disk controller # 1 via a confounding bus.

The host device 14 is composed of a processing device such as a CPU, and is a source of operation request such as read operation and write operation. The host device 14 is connected to the disk controllers # 0 and # 1 via the IO control bus 15, controls the driving of the disk controllers # 0 and # 1, and also manages system information such as a master system and a slave system. ing.

Next, the operation of FIG. 4 will be described. The basic operation is the same as that of FIG. 1, except that the DK common control unit is not provided. Therefore, in FIG. 6, the disk controller is controlled via a confounding bus provided between the disk controllers # 0 and # 1.

FIG. 5 is a flowchart showing an operation procedure when the host device 14 issues a read request. In step 301, a CPU (not shown) in the host device 14
Outputs a single-system read request via the IO control bus 15 to the disk controller # 0 set in advance as the master system.

In step 302, the master disk controller # 0 executes a series of read operations in response to a request from the host device 14. In step 303, the DK # 0 control unit performs a read operation on the DK # 0 of the own system.

In step 304, when a read error occurs in DK # 0, this read error is transmitted via DK # 0 via SCSI # 0 according to the SCSI bus protocol.
The control unit is notified. If an error occurs, step 3
05 and if no error occurs, step 312
Then, after the read data is sent to the host device 14, the read operation ends.

In step 305, when a read error occurs, the DK # 0 control unit outputs an error notification signal to the error recovery activation unit 9 to request activation of the error data recovery operation. When receiving the error notification signal, the error recovery activation unit 9 executes the error data recovery process.
Substitution processing instruction unit 10 of own system and copy instruction unit 11a of another system
And an error recovery signal to the DK # 1 control unit. In step 306, the host device 14 switches the master system from DK # 0 to DK # 1, and sets the slave system to DK #.
Switch from # 1 to DK # 0.

In step 307, the DK # 1 control unit which has become the master system in response to an instruction from the higher-level device 14 executes the same sector read operation on its own DK # 1. In step 308, the alternative processing instruction unit 10 outputs an alternative processing instruction signal in parallel with step 307.
The DK # 0 control unit performs DK #
Request the execution of the substitution process for the # 0 error occurring sector.
In step 309, if a read error occurs even in the DK # 1 which has become the master system, the error is notified to the higher-level device 14 as a double failure.

In step 310, copy instructing section 1
1a transfers the read data read in step 307 to the disk controller # 0 via the confounding bus 22. In step 311, the copy instruction unit 11
DK reads the read data transferred via the confounding bus 22
It requests the DK # 0 control unit to copy to # 0. DK # 0 restores the data by writing the read data to DK # 0. In step 312, the DK #
0 control unit transfers the read data read from DK # 1 to CP
The read operation is sent to U1 to end the read operation.

The error recovery processing in DK # 1 is performed in the same manner as described above. The write operation is performed in the same manner as in the hard disk device shown in FIG. That is, when each DK control unit detects a write error when performing both-system write, a sector replacement is requested by the replacement processing instruction unit, and when the sector replacement is completed, the data from the other hard disk device is copied and executed. .

[0038]

As described above, according to the present invention, when a defective storage area is detected in a duplicated hard disk drive, the replacement processing can be performed and the data recovery processing in the defective storage area can be performed. Therefore, according to the present invention, it is possible to perform the replacement process of the defective storage area without disconnecting the target hard disk device from the system at the time of the replacement processing of the defective storage area, that is, while maintaining the duplexed state, thereby improving the reliability of the system. It has the effect of not lowering. In addition, since the substitute process can be performed without manual intervention, the maintenance work can be made more efficient.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention.

FIG. 2 is a flowchart showing an operation procedure when the CPU 1 requests data reading.

FIG. 3 is a flowchart showing an operation procedure when the CPU 1 requests data writing.

FIG. 4 is a block diagram showing another embodiment of the present invention.

5 is a flowchart showing an operation procedure when the host device 14 in FIG. 4 requests data reading.

FIG. 6 is a block diagram showing a conventional example.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... CPU, 2 ... DK common control part, 3 ... DK # 0 control part, 4 ... SCSI # 0, 5 ... DK # 0, 6 ... DK # 1 control part, 7 ... SCSI # 1, 8 ... DK # 1 , 9: error recovery start unit, 10: alternative processing instruction unit, 11: copy instruction unit

Claims (6)

[Claims] An error recovery method for a duplicated hard disk drive comprising a first hard disk drive and a second hard disk drive, wherein when a defective storage area is detected in the first hard disk drive, A duplicated hard disk drive which replaces a defective storage area with a normal storage area in a first hard disk drive, and recovers data in the defective storage area by copying data from a second hard disk drive. Error recovery method. 2. The system according to claim 1, wherein when a defective storage area is detected in the first hard disk device of the master system at the time of writing data, the first hard disk device is switched to the slave system and the second hard disk device is changed to the master system. Switching, performing a replacement process for the defective storage area, reading data from a second hard disk device of a master system, writing the data to a first hard disk device of a slave system, and restoring data in the defective storage region. Error recovery method for duplicated hard disk drive. 3. The data storage device according to claim 1, wherein when a defective storage area is detected in the first or second hard disk device at the time of reading data, the replacement processing of the defective storage area is performed, and data can be normally written. An error recovery method for a duplicated hard disk drive, comprising reading data from the hard disk drive, writing the data to the failed hard disk drive, and recovering the data in the defective storage area. 4. A duplicated hard disk drive error recovery device comprising a first hard disk drive and a second hard disk drive, wherein the drive of the first hard disk drive is controlled,
When a defective storage area is detected in the first hard disk device, an error notification signal is output, and a first processing for replacing the defective storage area and restoring data in the defective storage area is performed.
DK control unit, and controls the driving of the second hard disk drive,
When a defective storage area is detected in the second hard disk device, an error notification signal is output, and the second processing for replacing the defective storage area and processing for restoring data in the defective storage area is performed.
A DK control unit, and an error recovery start unit that starts the replacement process of the defective storage area and the data recovery process by each of the DK control units upon receiving the error notification signal. Error recovery device for hard disk drive. 5. The system according to claim 4, further comprising: a controller that controls driving of the first and second DK control units and manages system information of each of the hard disk devices.
An error recovery device for a duplicated hard disk drive, comprising a K common control unit. 6. The system according to claim 4, further comprising a confounding bus connecting the first and second DK control units, wherein data read from a normal hard disk device is subjected to a process of replacing a defective storage area. An error recovery device for a duplicated hard disk device, which transmits data to the hard disk device via the confounding bus and recovers data.