JP2000122813A - Disk array device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent system down due to a media error after power interruption is restored when the power interruption occurs while writing work is performed for a hard disk. SOLUTION: Even though transfer data from a host computer 5 are temporarily stored in a data buffer memory 9, a write/read controlling part 8 extracts the transfer data from it and outputs it to a hard disk drive 1 through an HDD inputting and outputting part 10. If power supply from a power supply unit 4 is stopped and power interruption occurs while the hard disk drive 1 performs writing, the writing is going to be stopped in the middle of a sector. However, a voltage detecting part 11 immediately detects that the power interruption occurs based on the detection of voltage drop and outputs a reset signal Rs to the hard disk drive 1. Then, the hard disk drive 1 performs a reset operation after finishing writing for one sector by a 'sector security function' being its general function.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a disk array device used in a computer system or the like.

[0002]

2. Description of the Related Art FIG. 4 shows a so-called RAID system (Redundant system).
FIG. 2 is a block diagram showing a configuration of a conventional disk array device employing Arrays of Independent Disks. This disk array device has a plurality of hard disk devices 1
, A disk array controller 3 for writing / reading to / from the plurality of hard disk drives 1, and a power supply 4 for supplying power to the plurality of hard disk drives 1 and the disk array controller 3. .

[0003] The disk array controller 3 is connected to a host computer 5 outside the apparatus, and exchanges data with the host computer 5. And the disk array controller 3
A host computer input / output unit 6 having an interface function with respect to the host computer 5, a microprocessor 7 having a write / read control unit 8 formed therein for controlling the entire apparatus, and a host computer 5 And a HDD buffer input / output unit 10 having an interface function for each hard disk drive 1.

When writing to each hard disk drive 1 of the disk array 2, a write / read control unit 8 of a microprocessor 7 extracts data necessary for writing from a data buffer memory 9 and transfers the data to an HDD input / output unit. The data is output to the hard disk drive 1 via the hard disk drive 10. In this case, the write data from the data buffer memory 9 is usually divided into a plurality of hard disk devices 1 and written.

[0005]

By the way, while the writing operation of the hard disk drive 1 is being performed under the control of the write / read control unit 8 as described above, the power supply from the power supply unit 4 is stopped. A power outage accident may occur. However, in such a case, in the conventional device, a media error may occur after the restoration of the power failure, causing the system to go down.

That is, FIG. 5 shows a data storage state near a power failure occurrence time in a certain sector of a hard disk. As shown in FIG. 5, data is written (overwritten) to N sectors. If there is a power outage while the power is on, this N sector will contain new data,
The old data before writing will be mixed. Therefore, when the write / read control unit 8 reads the N sectors after the restoration of the power failure, the ECC code (error correction code)
Thus, an error is detected, and a so-called media error occurs. When such a media error occurs in a plurality of hard disk drives 1, the above RA
The operation of the disk array device adopting the ID method is stopped, which causes a system down. Also,
If such a system failure occurs, the data contents of all the hard disk devices 1 in the disk array 2 cannot be guaranteed, so a huge amount of work of restoring backup data to each hard disk device 1 is performed. Have to do it.

The present invention has been made in view of the above circumstances, and it is possible to prevent a system failure due to a media error after recovery from a power failure when a power failure occurs during a write operation to a hard disk. It is an object of the present invention to provide a possible disk array device.

[0008]

According to an aspect of the present invention, there is provided a disk array including a plurality of hard disk devices and outputting a write command to the plurality of hard disk devices. And a power supply unit for supplying power to the disk array and the disk array controller, wherein the hard disk device performs writing even when a reset signal is input during a writing operation. In the disk array device, only the writing to the operating sector is continued until it is completed. In the disk array device, when the disk array controller detects that the voltage supplied from the power supply device is equal to or less than a certain value, A voltage detector that outputs the reset signal to the hard disk device. It is intended to, characterized in that.

According to a second aspect of the present invention, there is provided a disk array comprising a plurality of hard disk devices, a disk array controller having a write / read control unit for outputting a write command to the plurality of hard disk devices, A power supply device for supplying power to the disk array and the disk array controller, wherein the disk array controller outputs the write / read control unit to the hard disk device. A nonvolatile memory for storing write command information and information as to whether or not a write operation related to the write command has been normally completed, wherein the disk array controller has a power failure after a power failure has occurred; When the hard disk device is to be rewritten at the time of restoration, the data is written to the nonvolatile memory. Reading the information, but to perform the re-write the sector write operation has not ended normally, characterized in that.

According to a third aspect of the present invention, in the second aspect of the present invention, the disk array controller reads information stored in the non-volatile memory every time the disk array controller is started, and the previous write operation ends normally. If not,
The writing is performed from the sector that has not been completed normally.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. However, the same components as those in FIG. 4 are denoted by the same reference numerals, and redundant description will be omitted. FIG. 1 is a block diagram showing the configuration of the first embodiment of the present invention. The difference from FIG. 4 is that a voltage detection unit 11 (formed by an IC) is provided. This voltage detector 11
Determines that a power failure has occurred when the voltage in the disk array controller 3 has fallen below a certain value (in this embodiment, 4.8 volts), and the reset signal Rs
Is output to each hard disk device 1.

Here, general characteristics of the hard disk drive will be described. Now, it is assumed that the write / read control unit 8 outputs a reset command while the hard disk device 1 is performing a write operation, and the HDD input / output unit 10 outputs a reset signal to the hard disk device 1 based on this. . The hard disk device 1 stops the writing operation when the reset signal is input. However, the hard disk device 1 does not stop the writing operation immediately when the reset signal is input. The writing operation is stopped only when the writing to is completed. In this manner, even when the reset signal is input, the function of setting the stop point of the write operation to the sector unit is called a "sector security function" of the hard disk device. The first invention is
This "sector security function" is effectively used.

Next, the operation of the first embodiment of the present invention will be described with reference to FIG.
This will be described based on the timing chart of FIG. The disk array controller 3 receives a power supply of 5.0 volts DC from the power supply device 4, and the voltage detection unit 11 constantly monitors the voltage in the disk array controller 3.
Now, assuming that the hard disk device 1 is performing a write operation, in this state, the output signal of the voltage detector 11 is a “1” signal and the reset signal Rs is not output (in the hard disk device, ,Normal,
The “1” signal is a non-active signal, and the “0” signal is an active signal. ).

However, if a power failure occurs at time t1, the voltage in the disk array controller 3 drops sharply. Then, when the voltage becomes 4.8 volts or less at time t2, the voltage detection unit 11 determines that a power failure has occurred, and outputs a “0” signal as the reset signal Rs to the hard disk device 1. The hard disk device 1 operates at time t2
In the section K1 until the reset signal Rs is input, the normal write operation is performed. Further, in the section K2 after the time t2 when the reset signal Rs is input, writing is performed based on the "sector security function" described above. Continue operation. Then, when entering the section K3 after the section K2, the hard disk device 1 first performs an internal reset operation by the reset signal Rs.

As described above, in the configuration of FIG. 1, the voltage detection unit 11 is provided to detect the occurrence of a power failure as soon as possible, and to output the reset signal Rs directly to the hard disk drive 1, thereby utilizing the "sector security function". By doing so, the writing is not terminated in the middle of the sector, so that it is possible to prevent the system from going down due to a media error after the restoration from the power failure.

FIG. 3 is a block diagram showing the configuration of the second embodiment of the present invention. FIG. 3 differs from FIG. 4 in that a nonvolatile memory 12 is added. The nonvolatile memory 12 stores the contents of data output from the write / read control unit 8 to the hard disk device 1.

That is, the non-volatile memory 12 has an area for storing command information D1 and an area for storing response information D2. The command information D1 is information such as a write command or a read command from the write / read control unit 8 to the hard disk device 1, an address of a leading sector where writing or reading is performed, a transfer size, and the like. The response information D2 is data indicating whether or not the writing operation of the hard disk device 1 has been completed normally.

Next, the operation of the embodiment of the second invention configured as described above will be described. Now, it is assumed that command information D1 from the write / read control unit 8 is output to the hard disk device 1 via the HDD input / output unit 10, and the hard disk device 1 is performing a write operation. The command information D1 at this time is stored in the nonvolatile memory 12. If a power failure occurs while the hard disk drive 1 is performing a write operation, and the write operation stops in the middle of a certain sector, the writing to this sector ends normally. The response information D2 indicating that there was no
Is stored.

Next, when the power failure is restored, the write / read control unit 8 reads the nonvolatile memory 12 and finds out, from the response information D2, the sector for which the writing was not completed normally. Then, the writing / reading control unit 8 sends the response information D2
The write data is taken out from the data buffer memory 9 on the basis of the command information D1 corresponding to the command data D1, and the data is written from the head of the sector in which the write operation has stopped halfway.

As described above, in the configuration shown in FIG. 3, by providing the nonvolatile memory 12, even if the writing to the sector is stopped halfway due to the occurrence of the power failure, the writing starts from the same sector after the restoration of the power failure. It is supposed to be.
Therefore, as shown in FIG. 5, a state in which new data and old data are mixed in the same sector is not left as it is, so that occurrence of a media error due to detection of an ECC code can be avoided.

In the second embodiment, the write / read control unit 8 reads out the response information D2 from the nonvolatile memory 12 after the power failure is restored in the event of a power failure, and the data is written in the middle. Although the configuration for checking whether or not a stopped sector exists is described, a configuration for checking whether or not such a sector exists in the previous write operation every time the disk array controller 3 is started may be used.

[0022]

As described above, according to the first aspect of the present invention, the occurrence of a power failure can be quickly detected by providing the voltage detection unit,
Since the reset signal is output directly to the hard disk drive and the "sector security function" is used to prevent the writing from being completed in the middle of the sector,
It is possible to prevent the occurrence of a system down due to a media error after restoration from a power failure.

According to the second aspect of the present invention, the nonvolatile memory is provided so that even if the writing to the sector is stopped halfway due to the occurrence of the power failure, the writing is started from the same sector after the restoration of the power failure. Therefore, a state in which new data and old data are mixed in the same sector is not left as it is, and similarly to the first invention, occurrence of a system down due to a media error after restoration from power failure is prevented. be able to.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of the first invention.

FIG. 2 is a timing chart for explaining the operation of FIG. 1;

FIG. 3 is a block diagram showing a configuration of an embodiment of the second invention.

FIG. 4 is a block diagram showing a configuration of a conventional device.

FIG. 5 is an explanatory diagram showing a data storage state in a certain sector of a hard disk near a point of occurrence of a power failure.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Hard disk drive 2 Disk array 3 Disk array controller 4 Power supply device 5 Host computer 6 Host computer input / output unit 7 Microprocessor 8 Write / read control unit 9 Data buffer memory 10 HDD input / output unit 11 Voltage detection unit 12 Non-volatile memory Rs Reset signal R D1 Command information D2 Response information

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5B011 EB07 GG03 GG16 HH04 JA05 JA14 5B018 GA04 LA05 MA14 QA05 QA06 5B065 CA11 CA30 CC08 CE11 EA02 EA12 EA24 EA26 EK05 ZA14

Claims (3)

[Claims] A disk array comprising a plurality of hard disk devices; a disk array controller for outputting a write command to the plurality of hard disk devices; and a power supply to the disk array and the disk array controller. And a power supply device, wherein even when a reset signal is input during a write operation, the hard disk device continues writing only to the sector during the write operation until it is completed.
In the disk array device, the disk array controller includes a voltage detection unit that outputs the reset signal to the hard disk device when detecting that a voltage supplied from the power supply device is equal to or less than a predetermined value. A disk array device comprising: 2. A disk array comprising a plurality of hard disk devices, a disk array controller having a write / read control unit for outputting a write command to the plurality of hard disk devices, the disk array and the disk A power supply device for supplying power to the array controller, wherein the disk array controller includes: write command information output to the hard disk device by the write / read control unit; and A non-volatile memory for storing information on whether or not the write operation according to the write instruction has been normally completed; and the disk array controller stores the information in the hard disk device when a power failure is restored after a power failure occurs. When performing rewriting, the information stored in the nonvolatile memory is read. And it is intended to perform the re-write the sector write operation has not ended normally, the disk array apparatus, characterized in that. 3. The disk array controller reads information stored in the non-volatile memory each time the disk array controller is started, and if the previous write operation has not been completed normally,
3. The disk array device according to claim 2, wherein writing is performed from the sector that has not been completed normally.