JPH0793104A - Disk array device - Google Patents

Abstract

PURPOSE:To provide a disk array device which can efficiently parallely access small capacity data and large capacity data requiring high-speed transfer. CONSTITUTION:This device is provided with plural disk devices 1a-1u equipped with first areas 2 for separately recording the large capacity data and second areas 3 for recording the small capacity data without dividing them, and after data are provided with redundancy by adding parity bits for error correction to the large capacity data inputted from a host device by using a code circuit 7, the data are divided into each bit and recorded in the first areas 2 of the respective different disk devices. At the time when the recording of the small capacity data is parallely performed, the small capacity data are recorded in the second areas of arbitrary N pieces of disk devices at the maximum. During this recording, no large capacity data are recorded in the first areas 2 of the disk devices disabling access and concerning the reproducing of the large capacity data corresponding to those first areas 2, the data are restored by the error correction processing of a decode circuit 8.

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 capable of increasing the data transfer rate.

[0002]

2. Description of the Related Art In recent years, with the development of high-performance microprocessors, the computing capacity of computers has improved dramatically. In line with this, as a secondary storage device that records a large amount of data, a disk array device that boasts a high transfer rate and a large storage capacity has appeared. Since this disk array device enables real-time processing of image data, it is expected as a storage device supporting multimedia.

However, it has been pointed out that the above-mentioned disk array device is inefficient because all disks are driven for data recording even when accessing a small amount of data. .

FIG. 2 shows an example of data transfer in a conventional disk array device. In the data transfer shown in FIG. 2, the word length of the data is divided into four, and four disk devices 21 to 24 are used.
Each of the divided data is recorded, and the data is read and written. At this time, reading and writing of data is performed in block units of a predetermined size. In the figure,
The period (a) indicates that large-capacity data such as image data was recorded and reproduced, and the period (b) indicates that small-capacity data such as transaction data was recorded and reproduced. .

Further, since a disk device generally requires a seek operation for reading and writing data, a method of setting a large unit of the above-mentioned block to reduce the ratio of the dead time due to the seek operation is adopted.

[0006]

However, in the above-mentioned conventional disk array device, when the block unit is increased as is apparent from FIG. 2, an unused area when reading / writing a small amount of data is a dead time. This results in increasing t. Therefore, if small-capacity data is read and written frequently, the performance of the storage system is significantly reduced.

In order to solve such a problem, the data is not decomposed in the word length direction but is decomposed in the time axis direction, and a series of data is divided at constant time intervals and sequentially written in each disk device ( For example, Yasuo Kinouchi, et al .: “Throughput analysis of stripe array disk when long and short data accesses are mixed,” IEICE D-Ipp1005-1014, 1992) is being studied.

However, although these techniques reduce the dead time t when reading and writing small-capacity data, the performance cannot be essentially improved, and random access of small-capacity data frequently occurs. There is a problem that it is difficult to ensure synchronization in high-speed transfer of capacity data, and in the case of image data or the like, the time axis is disturbed.

In view of the above problems, it is an object of the present invention to provide a disk array device capable of efficiently accessing in parallel small-capacity data and large-capacity data requiring high-speed transfer.

[0010]

In order to achieve the above object, in claim 1, a plurality of disk devices connected in parallel are arranged, and when data is recorded in the disk devices,
Data that decomposes data in predetermined units such as bits, bytes, words, etc., and distributes and records it in each disk device. When reproducing data, the data reproduced from each disk device is combined to restore the original data. In the disk array device including recording / reproducing means, each of the disk devices includes a first area for recording a large amount of data such as image data, which is required to have a high transfer rate, and a transaction area.
A second area in which small-capacity data such as data is recorded without being divided, and the data recording / reproducing means is configured to perform the first area or the second area of each disk device based on the data to be recorded / reproduced. We propose a disk array device having a recording area selection means for selecting the area.

According to a second aspect, in the disk array device according to the first aspect, the data recording / reproducing means is
An error that adds an error correction code that can restore the original data even when 2N (N ≧ 1) disk devices do not function with respect to the data to be recorded in at least the first area of each disk device. It has a correction code addition means and a data restoration means for restoring the original data from the data to which the error correction code is added, and is arbitrary when the access to the first area and the second area overlaps. A disk array device that permits access to the second area to a maximum of N disk devices, and does not record divided data in the first area of the disk device that is inaccessible during this time. suggest.

According to a third aspect of the present invention, in the disk array device according to the first aspect, the data recording / reproducing means is
With respect to the data to be recorded in each of the disk devices, N
Error correction code assigning means for assigning an error correction code capable of restoring the original data even if (N ≧ 1) disk devices do not function, and recording of each of the disk devices to which the error correction code is assigned And a second data recovery means for recovering the original data from the data, and when the access to the first area and the second area is duplicated, the second data is stored for any maximum N disk devices. Immediately after recording the divided data in a series in the first area of each disk device, the divided data is not recorded in the first area of the disk device that is permitted to be accessed during this period. The recorded data is restored to the first area, and the divided data not recorded in the first area of the disk device that is permitted to access the second area based on the restored data is stored in the first area. Suggest disk array apparatus for recording.

According to a fourth aspect, in the disk array device according to the third aspect, the data recording / reproducing means is
As a result of simultaneous requests for access to the first area and the second area, the divided data to be recorded is temporarily stored in N or less disk devices that did not record in the first area. A storage device is provided, and after recording a series of divided data in the first area of each disk device, the unrecorded divided data stored in the storage device is stored in the N area.
A disk array device for recording in the first area of a single disk device is proposed.

Further, in claim 5, claims 1, 2, 3 are provided.
Or the high transfer rate bus for transferring the large amount of data stored in the first area of each disk device at high speed, the control command of the disk array device, and the second disk device of each disk device. We propose a disk array device having a low transfer rate bus for transferring a small amount of data to be stored in the area.

[0015]

According to claim 1 of the present invention, a plurality of disk devices are connected in parallel and arranged, and when data is recorded on the disk device by the data recording / reproducing means, the data to be recorded are bits, bytes, Disassembled in predetermined units such as words, distributed and recorded in each disk device,
At the time of data reproduction, the data reproduced from each disk device is combined to restore the original data. When recording or reproducing large-capacity data such as image data that requires a high transfer rate, the first area of each disk device is selected by the recording area selection means, and the large-capacity data is stored in the first area. The divided data is divided and recorded in the first area, or the divided data is reproduced from each of the first areas. Furthermore, when recording or reproducing small capacity data such as transaction data, the recording area selecting means selects a second area of an arbitrary disk device, and the small capacity data is stored in the second area. The data is recorded without being divided, or the small capacity data is reproduced from the second area.

According to a second aspect of the present invention, at the time of recording data on the disk device, 2N (N ≧ 1) for at least the data to be recorded in the first area of each disk device by the error correction code adding means. An integer of (7) disk devices do not function, an error correction code that can restore the original data is added, the data is given redundancy, and then the data is distributed and recorded in each disk device. Further, at the time of data reproduction, the original individual divided data is restored from the data with the error correction code recorded in each disk device by the data restoration means, and the individual divided data reproduced from each disk device is also restored. The data is combined and integrated, and the data at the time of recording is restored. Further, when the access to the first area and the second area overlaps, the data recording / reproducing means permits the access to the second area to an arbitrary maximum N disk devices. The divided data is not recorded in the first area of the disk device which is inaccessible during this time, and the reproduction of the divided data corresponding to the first area is performed by the error correction processing of the data restoration means. Is restored.

According to a third aspect of the present invention, at the time of recording data in the disk device, N (N ≧ 1) is recorded by the error correction code adding means for at least the data to be recorded in the first area of each disk device. Even if the disk devices of the (integer number) units do not function, after the error correction code that can restore the original data is added, the data is distributed and recorded in each disk device. Further, at the time of data reproduction, the original individual divided data is restored from the data with the error correction code recorded in each disk device by the data restoration means, and the individual divided data reproduced from each disk device is also restored. The data is combined and integrated to restore the data at the time of recording. Further, when the access to the first area and the second area is duplicated, the data recording / reproducing means permits the access to the second area to arbitrary maximum N disk devices. The divided data is not recorded in the first area of the disk device which is inaccessible during this time, and the reproduction of the divided data corresponding to the first area is performed by the error correction processing of the data restoration means. Is restored. Also,
Immediately after recording a series of divided data in the first area of each disk device, the recorded data is restored by the data recording / reproducing means, and access to the second area is performed based on the restored data. First authorized disk unit
The divided data which has not been recorded in the area (1) is recorded in the first area after being given an error correction code.

According to the fourth aspect, as a result of simultaneous access requests to the first area and the second area,
The divided data to be recorded in N or less disk devices that have not been recorded in the first area is temporarily stored in the storage device, and a series of divided data is recorded in the first area of each disk device. After that, the unrecorded divided data stored in the storage device is recorded in the first area of the N disk devices by the data recording / reproducing means.

According to a fifth aspect of the present invention, the large capacity data stored in the first area of each disk device is transferred at high speed by the high transfer rate bus, and the control command of the disk array device and each disk device are also transmitted. The small-capacity data stored in the second area of is transferred at a low speed by the low transfer rate bus.

[0020]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram for explaining the outline of the first embodiment of the present invention. In this embodiment, a small amount of data and a large amount of data that needs to be transferred at high speed are arranged in parallel by a data storage method using an error correction code. It is possible to completely access the computer, and completely eliminate the dead time associated with accessing a small amount of data.

Specifically, as shown in FIG. 1, a large amount of data, such as image data, which is required to be transferred at high speed, is stored in each disk device 1 (only one unit is shown in FIG. 1) constituting the disk array device. A first area 2 for recording the divided data and a second area 3 for recording the small capacity data such as transaction data without dividing the data are provided. In addition, the array controller 4 performs data conversion processing using an error correction code for large-capacity data, and records and reproduces the data, so that a specific number of disk devices 1 among the plurality of disk devices 1 have small-capacity data. Even if the disk is separated for recording / reproduction, the recording / reproduction of large-capacity data with respect to another disk device 1 can be continued. Further, the large-capacity data and the small-capacity data are transferred via the bus controller 5 via two dedicated buses, for example, a high transfer rate bus HB and a low transfer rate bus LB. here,
The high transfer rate bus HB may be a common bus. That is,
The data transferred via the low transfer rate bus LB may be transferred by the bus controller 5 via the high transfer rate bus HB, and the low transfer rate bus may be deleted.

Next, the details of the first embodiment will be described based on the configuration diagram shown in FIG. The first embodiment shown here is
9 parity disk devices 1a to 1i and 12 data disk devices 1j to 1u are provided, and 9 bits of error correction parity bits are added to 12-bit parallel data input from a higher-level device. After that, the data is divided into 1-bit data and recorded in different disk devices. As a result, even when accessing a large amount of data in the first area 2 of each of the disk devices 1a to 1u, the second data of each of the disk devices 1a to 1u is accessed.
One of the small-capacity data in the area 3 can be arbitrarily accessed.

That is, in FIG. 3, reference numerals 1a to 1u denote disk devices, each of which has a first area 2 for storing a large amount of divided data and a second area 3 for storing a small amount of data which is not divided. ing. 4 is a plurality of disk devices 1a
Array controller for simultaneously controlling ~ 1u, 5 is a bus controller for managing the data flow of the high transfer rate bus HB and the low transfer rate bus LB, and 6 is a device selector for selecting a disk device for writing a small amount of data. .

7 is 9 for 12-bit parallel data.
A code circuit for adding bit parity, 8 a decoding circuit for restoring 12-bit parallel data from 21-bit data, 9 an access controller for managing the access state of each disk device 1a-1u, 10 serial / parallel data A data conversion circuit 11 for performing conversion is a status controller 11 for managing the status of the entire disk array device.

In the first embodiment, each disk device 1a ...
1u is configured to serially record and reproduce 1-bit data. The large-capacity parallel data input from a high-order device (not shown) via the high transfer rate bus HB is first added with 9 bits of parity bits for every 12 bits by the encoding circuit 7, and thereafter for each disk device for each bit. It is written in the first area 2 of 1a to 1u. At the time of reproduction, the 1-bit data reproduced from the first area 2 of each of the disk devices 1a to 1u is combined, error-corrected by the decoding circuit 8, and output to the high transfer rate bus HB. The parallel data input from the higher-level device through the high transfer rate bus HB passes through the encoding circuit 7 and then the device selector 6, but the device selector 6 selects the disk device to write the small capacity data. However, it does not basically function with respect to the parallel data.

On the other hand, the low transfer rate bus LB
The parallel data input via the data conversion circuit 10 is converted into serial data by the data conversion circuit 10 and recorded in the second area 3 of the desired disk device 1a-1u via the device selector 6. In this case, the device selector 6
Disconnects the data lines from the data lines connecting the disk devices 1a to 1u to the encoding circuit 7 and the decoding circuit 8 to the disk devices 1a to 1u that are going to write data to the second area 3 and disconnect the data lines. Data conversion circuit 10
Connect to. When reproducing the data recorded in the second area 3 of each of the disk devices 1a to 1u, the reproduced parallel data is output to the low transfer rate bus LB through the reverse process.

The command data from the host device is transferred to the status controller 11 via the low transfer rate bus LB, and the access controller 9 controls the head position of each of the disk devices 1a-1u based on the command data. On the contrary, the status of each of the disk devices 1a to 1u is transferred from the access controller 9 to the status controller 11, and is output to the host device via the low transfer rate bus LB as status data of the disk array device.

In the case of the first embodiment having the above-mentioned configuration, when writing a small amount of data while writing a large amount of data, the device selector 6 disconnects the data line connected to the disk device for writing the small amount of data. Therefore, a part of the large volume data remains unrecorded. Therefore, to perform error correction,
It must be assumed that any one unit will not function during recording and that one unit will not function during playback.

That is, error correction is possible even when two arbitrary disk devices cannot record and reproduce data.
You need to be able to restore the original data. Therefore, in the case of this embodiment, 9 parity bits are provided and the Hamming distance is set to 6 so that the data can be restored even when two arbitrary disk devices do not function.

As described above, in the first embodiment, the large-capacity data and the small-capacity data can be accessed in parallel by increasing the redundancy by using the error correction code for the large-capacity data that requires high-speed transfer. I am trying. Therefore,
Recording and reproduction of small-capacity data such as transaction data can be stably and concurrently performed without affecting recording and reproduction of image data that requires high-speed and synchronous transfer.

Next, a second embodiment of the present invention will be described.
FIG. 4 is a block diagram showing the second embodiment. In FIG. 4, the same components as those in the first embodiment described above are designated by the same reference numerals and the description thereof will be omitted. In the second embodiment, the number of parity disk devices is reduced to four, and
In addition to the configuration of the first embodiment, a buffer 12 for temporarily storing the restored data is provided on the high transfer rate bus HB.

As described in the first embodiment, if a small amount of data is written while writing a large amount of data, the device selector 6 writes the small amount of data to the disk devices 1a-1d, 1j-1u. Since the connected data line is separated, a part of the large volume data remains unrecorded. In the second embodiment, a part of such unrecorded large-capacity data is restored by a verify operation and automatically rewritten in the disk array device.

That is, in the case of the second embodiment, the status controller 11 reproduces the written data immediately after the writing of the large amount of data is completed, and the decoding circuit 8 restores the unrecorded data to the high transfer rate bus HB. The data stored in the buffer 12 is stored in the buffer 12 via the coding circuit 7 and the disk devices 1
Rewrite to a to 1d and 1j to 1u. In this case, since the purpose of the verify operation is to write unrecorded data, the address of the error device output from the decoding circuit 8 and the disk devices 1a to 1d and 1j to 1u that try to write the small capacity data during the verify operation. If the addresses of 1 and 2 match, it is necessary to suspend the verify operation or the writing of the small capacity data.

In the case of the second embodiment, the verify operation is performed immediately after writing a large amount of data and the unrecorded data is restored and then re-recorded, so that access to the large amount of data is somewhat restricted. It However, since the error correction processing is individually performed for the data recording and reproducing operations, in the above-described first embodiment, it is necessary to perform the error correction only when at least two disk devices do not function. In the second embodiment, error correction can be performed only when one disk device does not function. Therefore, the Hamming distance may be 4, the number of parity bits can be set to 4 to improve the coding efficiency, and the number of disk storage devices can be reduced for the same system storage capacity.

Further, the status controller 11 is provided with a small-capacity memory (not shown), and the device address of the disk device in which the small-capacity data is written while the large-capacity data is being written in the memory is stored as log data. If the data is restored while outputting the log data to the decoding circuit 8, the error correction in the verify operation after recording can be restored to a known error device as in the case of reproduction. Become. In this case, since the error can be corrected by the coding method having the error detection capability, the number of parity bits can be further reduced to 1 bit.

As described above, in the second embodiment, access to the large-capacity data immediately after being written is restricted, but the minimum system configuration with the highest coding efficiency can be realized.

Next, a third embodiment of the present invention will be described.
FIG. 5 is a block diagram showing the third embodiment. In the figure,
The same components as those of the first embodiment described above are designated by the same reference numerals and the description thereof will be omitted. Further, in the third embodiment, the number of disk devices for parity is reduced to one, and
In addition to the configuration of the above embodiment, a data selector 13 for selecting unrecorded data and a memory 14 for temporarily storing unrecorded data are provided. As a result, in the third embodiment, unrecorded data of large capacity data that occurs when simultaneously writing large capacity data and small capacity data is temporarily stored in the memory 14 via the data selector 13, and Using the free time to access the data, the memory 1
The unrecorded data stored in 4 corresponds to the disk device 1
The first area 2 of a, 1j to 1u is rewritten.

That is, in the third embodiment, when the writing of large-capacity data and the writing of small-capacity data overlap, the data selector 13 follows the instructions of the status controller 11 to write the small-capacity data to the disk devices 1a, 1j.
A part of the large-capacity data to be written in 1u, that is, unrecorded data is temporarily stored in the memory 14. At this time, the device addresses of the disk devices 1a, 1j to 1u where unrecorded data should be recorded are simultaneously stored in the memory 14.

The memory 14 can output the stored data to the data line via the data selector 13, and the unrecorded data and the disk device 1 in which the unrecorded data should be recorded.
Since the device addresses of a, 1j to 1u can be recorded at the same time, the device functions as a cache memory shared by the respective disk devices 1a, 1j to 1u, and when the unrecorded data stored in the memory 14 hits, the memory 14 stores each Data can be sent in place of the disk devices 1a, 1j-1u. Therefore, even when the writing of the large capacity data and the writing of the small capacity data are duplicated, the large capacity data can be accessed immediately after recording.

Further, in the fourth embodiment, no unrecorded data is generated effectively even when writing of large-capacity data and small-capacity data overlap, so error correction may be performed only for the data reproducing operation. It will be. Therefore, an error can be corrected by providing only one parity bit. Incidentally, the unrecorded data at this time is recorded in the corresponding disk device by the background processing by utilizing the idle time for accessing the small capacity data.

In the case of this embodiment, the memory 14 requires a little larger capacity, but since the unrecorded data stored in the memory 14 can be used as it is, the access to the large capacity data is not restricted and the minimum capacity is required. Parallel access to large-capacity data and small-capacity data can be realized with the system configuration.

Although not specified in the above-mentioned first to third embodiments, the data of the low transfer rate bus LB is connected to the high transfer rate bus HB via the bus controller as in the configuration of the conventional disk array device. Can also be incorporated. Specifically, in the second embodiment, the data conversion circuit 10 and the buffer 12 in the bus controller 5 are connected by a bus so that the large capacity data and the small capacity data can be distributed or combined on the buffer 12. Good.
In this case, since access to large-capacity data and access to small-capacity data can be performed in parallel, wasteful time when writing small-capacity data, which was a problem in the conventional configuration, is eliminated, and high transfer rate is improved. In the rate bus HB, only a time delay occurs due to the insertion of a small amount of data, the dead time can be reduced, and the bus utilization efficiency can be improved to almost 100%.

[0043]

As described above, according to the first aspect of the present invention.
According to the method, large-capacity data such as image data is divided and recorded in the first area of each disk device, and small-capacity data is recorded in the second area without being divided. The reproduction can be performed at high speed, and a dead time unlike the conventional case does not occur when recording and reproducing the small capacity data.

According to the second aspect, in addition to the above effect, the error correction code is used to give redundancy to the large-capacity data, so that the large-capacity data to be divided and recorded can be recorded without division. Since small amount of data that can be accessed in parallel is accessed in parallel, small amount of data such as transaction data that is frequently accessed can be accessed without affecting large amount of data such as image data that requires periodicity and high transfer rate. The processing can be executed in real time.

According to the third aspect, in addition to the above effect, the error correction code is used to give redundancy to the large capacity data, so that the large capacity data to be divided and recorded in the disk array device can be divided. Since small-capacity data to be recorded can be accessed in parallel without doing so, small-capacity data such as transaction data that is frequently accessed without affecting the image data that requires periodicity and high transfer rate. The processing can be executed in real time. Further, since it is only necessary to correct an error when one disk device does not function, it is possible to improve the coding efficiency of the error correction code and reduce the number of disk devices for the same system storage capacity.

According to the fourth aspect, in addition to the above effect, the divided data to be recorded in N or less disk devices which have not recorded in the first area is temporarily stored in the storage device. After a series of divided data that has been stored and recorded in the first area of each disk device, the unrecorded divided data stored in the storage device is recorded by the data recording / reproducing means, for example, when the disk device is free of access. Since the data is recorded in the first area of the N disk devices by utilizing time, unrecorded data does not effectively occur, and even when writing of large capacity data and small capacity data overlaps. The large capacity data can be accessed immediately after recording.

Further, according to claim 5, in addition to the above effect, a high transfer rate bus and a low transfer rate bus are provided, and the large capacity data stored in the first area of each of the disk devices is highly transferred. Since the control commands of the disk array device and the small amount of data stored in the second area of each disk device are transferred by the low transfer rate bus at high speed, each data can be stably and highly transferred. It is possible to transfer while ensuring reliability.

[Brief description of drawings]

FIG. 1 is a diagram illustrating an outline of a first embodiment of the present invention.

FIG. 2 is a diagram showing an example of data transfer in a conventional disk array device.

FIG. 3 is a configuration diagram showing a first embodiment of the present invention.

FIG. 4 is a configuration diagram showing a second embodiment of the present invention.

FIG. 5 is a configuration diagram showing a third embodiment of the present invention.

[Explanation of symbols]

1a to 1u ... Disk device, 2 ... First area for storing divided large-capacity data, 3 ... Second area for storing undivided small-capacity data, 4 ... Array controller, 5 ...
Bus controller, 6 ... Device selector, 7 ... Encoding circuit, 8 ... Decoding circuit, 9 ... Access controller, 10 ...
Data conversion circuit, 11 ... Status controller, 12
... buffer, 13 ... data selector, 14 ... memory, H
B ... High transfer rate bus, LB ... Low transfer rate bus.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Nobuyoshi Izawa 1-1-6 Uchisaiwaicho, Chiyoda-ku, Tokyo Nihon Telegraph and Telephone Corporation

Claims (5)

[Claims] 1. A plurality of disk devices connected in parallel, and when data is recorded on the disk devices, the data is decomposed into predetermined units such as bits, bytes, and words into each disk device. In a disk array device including a data recording / reproducing unit that records in a distributed manner and combines the data reproduced from each disk device at the time of data reproduction to restore the original data, each disk device is The recording / reproducing means has a first area for recording a large amount of data required to have a high transfer rate and recording it, and a second area for recording a small amount of data such as transaction data without dividing the data. Is the first area or the second area of each disk device based on the data to be recorded and reproduced.
A disk array device having a recording area selection means for selecting the area. 2. The data recording / reproducing device is adapted to operate even if at least 2N (N ≧ 1 integer) disk devices do not function with respect to data to be recorded in the first area of each of the disk devices. Error correction code giving means for giving an error correction code capable of restoring the above data and data restoring means for restoring the original data from the data given the error correction code, and the first area and the second area When the access to the area is duplicated, the access to the second area is permitted to any of up to N disk devices, and the first area of the disk device is disabled during the access. 2. The disk array device according to claim 1, wherein the divided data is not recorded. 3. An error in which the data recording / reproducing means can restore the original data with respect to the data to be recorded on each of the disk devices even when N (N ≧ 1 integer) disk devices do not function. The first area and the second area are provided, including error correction code adding means for adding a correction code, and data restoring means for restoring the original data from the record data of each disk device to which the error correction code is added. When the access to the area is duplicated, the access to the second area is permitted to any maximum N disk devices, and the first area of the disk device is disabled during the access. Does not record the divided data, restores the recorded data immediately after recording the divided data in a series in the first area of each disk device,
2. The divided data, which is not recorded in the first area of the disk device which is permitted to access the second area based on the restored data, is recorded in the first area. Disk array device. 4. The disk device of N or less disk units that did not record in the first area as a result of the data recording / reproducing means being requested to access the first area and the second area at the same time. A storage device for temporarily storing the divided data to be recorded, and after recording a series of the divided data in the first area of each disk device, the unrecorded divided data stored in the storage device. 4. The disk array device according to claim 3, wherein data is recorded in the first area of the N disk devices. 5. A high transfer rate bus for transferring a large amount of data stored in the first area of each disk device at high speed, a control command of the disk array device, and a storage area in the second area of each disk device. 5. The disk array device according to claim 1, further comprising a low transfer rate bus for transferring a small amount of data to be transferred.