JP2857289B2 - Disk array device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a disk array apparatus for performing data input / output processing by accessing a plurality of disk apparatuses in parallel, and more particularly, to a disk array apparatus aimed at improving processing performance in write processing. About.
2. Description of the Related Art As an external storage device of a computer system, disk devices such as a magnetic disk device and an optical disk device having features such as non-volatility of recording, large capacity, and high speed of data transfer are widely used. Demands for disk devices are high-speed data transfer, high reliability, large capacity, and low price. Disk array devices have been attracting attention as satisfying these requirements. A disk array device is a device in which several to several tens of small disk devices are arranged, data is recorded in a plurality of disk devices, and data is accessed in parallel.

When data is transferred to a plurality of disk devices in parallel by a disk array device, high-speed data transfer can be performed twice as many as the number of disks as compared with a single disk device. In addition, by recording redundant information such as parity data in addition to the data, it is possible to detect and correct data errors caused by failure of the disk device, and to duplicate the contents of the disk device. The same high reliability as that of the method of recording data can be realized at a lower price than duplication.

[0003]

2. Description of the Related Art Conventionally, David A. Patterson of the University of California, Berkeley.
et al. disclose a paper that evaluates a disk array device that accesses a large amount of data to many disks at high speed and realizes data redundancy at the time of disk failure by classifying the disk array device from level 1 to level 5. (ACM SIGMOD Conferance, Chicago, Illi
nois, June 1-3, 1988P109-P116).

The levels 1 to 5 for classifying the disk array devices proposed by David A. Patterson et al.
ID (Redundant Arrays of In)
expansive Disks) 1-5. The RAIDs 1 to 5 will be briefly described as follows. [RAID0] FIG. 14 shows a disk array device having no data redundancy, which is not included in the classification of David A. Patterson et al.
Called D0.

In the RAID 0 disk array device, as shown in data A to I, the disk array control device 10 distributes data to each of the disk devices 32-1 to 32-3 based on an input / output request from the host computer 18. Data redundancy in the event of a disk failure. [RAID1] FIG. 1 shows a RAID1 disk array device.
As shown in FIG. 5, a mirror disk device 32-2 stores copies A 'to D' of data A to D stored in the disk device 32-1. RAID 1 is widely used because it has low utilization efficiency of disk devices but has redundancy and can be realized by simple control. [RAID2] A RAID2 disk array device strips (divides) data in units of bits or bytes.
Reading / writing is performed in parallel to each disk device. The striped data is physically recorded in the same sector in all disk devices. A Hamming code generated from data is used as the error correction code. It has a disk device for recording a Hamming code in addition to the data disk device, identifies a failed disk device from the Hamming code, and restores the data.

[0006] By providing redundancy by the Hamming code in this way, correct data can be secured even if a disk device fails, but it has not been put to practical use due to poor utilization efficiency of the disk device. [RAID3] The RAID3 disk array device has the configuration shown in FIG. That is, as shown in FIG. 17, for example, data a, b, and c are divided into data a1 to a3, b1 to b3, and c1 to c3 in units of bits or sectors, and a parity P1 is calculated from the data a1 to a3. b1
Ｂb3, the parity P2 is calculated from the data c1 to c3, and the parity P3 is calculated from the data c1 to c3.
1 to 32-4 are simultaneously accessed and written in parallel.

In RAID 3, data redundancy is maintained by parity. Further, the data writing time can be reduced by the parallel processing of the divided data. However, with one write or read access, all disk devices 3
2-1 to 32-4 parallel seek operations are required.
For this reason, it is effective when a large amount of data is handled continuously, but in the case of a transaction process in which a small amount of data is accessed at random, the high speed of data transfer cannot be utilized and the efficiency is reduced. [RAID4] As shown in FIG. 18, the RAID4 disk array device divides one data into sector units and writes the data into the same disk device. For example, the disk device 32
Looking at -1, data a is divided into sector data a1 to a4 and written. The parity is stored in a fixedly determined disk device 32-4. Here, data a1,
Parity P1 is calculated from b1 and c1, and data a2 and
Parity P2 is calculated from b2 and c2, and data a3
Parity P3 is calculated from b3 and c3, and data a4
Parity P4 is calculated from b4 and c4.

Data is read out from the disk devices 32-1 to 32-1.
32-3 can be read in parallel. Data a ~
The reading of data b is performed by taking the data a as an example.
2-1 access sectors 0 to 3 to access sector data a1
To a4 are sequentially read and synthesized. In data writing, since the new parity is calculated and written after reading the data and parity before writing, for one writing,
A total of four accesses are required. For example, when updating (rewriting) the sector data a1 of the disk device 32-1, the old data (a1) _old of the update location and the old parity (P1) _old of the corresponding disk device 32-4 are read, and the new data is read. An operation of writing a new parity (P1) _new that is consistent with the data (a1) _new is required in addition to the data writing for updating.

In addition, since writing always accesses the disk device 32-4 for parity, writing to a plurality of disk devices cannot be executed simultaneously. For example, even if the simultaneous writing of the data a1 of the disk device 32-1 and the simultaneous writing of the data b2 of the disk device 32-2 are performed, the parity P1, P2
Must be read and written after calculation, and therefore cannot be written at the same time.

Although the definition of RAID4 has been made in this way, there is little merit and there is little movement for practical use at present. [RAID5] The disk array device of RAID5 enables parallel reading and writing by not fixing the disk device for parity. That is, as shown in FIG.
The disk device on which parity is placed differs for each sector. Here, the parity P is calculated from the data a1, b1, and c1.
1 is calculated, and parity P is calculated from the data a2, b2, and d2.
2 is calculated, and parity P is calculated from data a3, c3, and d3.
3 is calculated, and parity P is calculated from data b4, c4, and d4.
4 has been calculated.

The parallel reading and writing is performed by, for example, the disk drive 3
2-1 data a1 of sector 0 and disk device 32-2
The data b2 of the sector 1 can be read and written simultaneously without duplication because the parities P1 and P2 are located in different disk devices 32-4 and 32-3. Note that the overhead that requires a total of four accesses at the time of writing is RAI
Same as D4.

As described above, RAID5 can access a plurality of disk devices asynchronously to execute read / write, and is therefore suitable for transaction processing in which a small amount of data is randomly accessed. Where RAID4 and R
The generation of parity data in writing AID5 data will be described as follows.

First, in the disk array device storing the redundant information, the exclusive OR of the data at the same storage location in a plurality of disk devices is set as parity by the following equation (1) and stored in the disk device for parity. ing. Data a (+) data b (+)... = Parity P (1) where (+) is an exclusive-OR symbol.
And fixed to a specific disk device 32-4.
On the other hand, in RAID5, as shown in FIG. 18, the parity is distributed to the disk devices 32-1 to 32-4 to eliminate the concentration of accesses to a specific disk device by the parity read / write operation.

At the time of reading these RAID4 and RAID5 data, the data in the disk devices 32-1 to 32-4 is not rewritten, so that the parity consistency is maintained. Need to change. For example, when one old data (a1) _old in the disk device 32-1 is rewritten to the new data (a1) _new , to obtain the consistency of the parity P1, the following equation (2) is calculated. By updating the parity, the consistency of the parity of the entire data of the disk device can be maintained. Old data (+) Old parity (+) New data = New parity (2)

[0015]

However, in the conventional disk array devices classified into RAID4 and RAID5, as can be seen from equation (2), in the data writing process, the disk device to which data is to be written is used. It is necessary to read the old data from within the disk drive, and first read the parity at the same position as the write-scheduled position of the disk device for storing the parity and calculate the new parity, and then transfer the new data and the new parity to each disk device. Will write.

Therefore, two readings and two writings, that is, four accesses are required in one writing, so that there is a problem that the processing time is long and the performance of the disk array device cannot be sufficiently improved. . On the other hand, in the disk array device classified as RAID3 shown in FIG. 17, data is divided in the direction in which the disk devices are arranged and written in parallel, so that old data and old parity are read from the disk devices. There is no need to calculate a new parity from the divided data.
5, the writing processing time can be shortened.

However, at the time of writing, parallel access to all disk devices is required, which is not suitable for a large amount of transaction processing that requires reading and writing of each disk device. SUMMARY OF THE INVENTION An object of the present invention is to provide a disk array device in which an optimum access format is selected according to the amount of data from a host device to perform a write process.

Another object of the present invention is to provide a disk array device which improves the access efficiency after writing by selecting a mode for individually accessing a plurality of disk devices when the amount of data from the host device is small. Another object of the present invention is to provide a disk array device in which, when the amount of data from a host device is large, a mode for accessing a plurality of disk devices in parallel is selected to reduce the processing time.

Another object of the present invention is to reduce the processing time by selecting a parallel access mode on condition that all disk devices are accessible when the amount of data from the host device is large. A disk array device is provided. Another object of the present invention is to provide a disk array device in which, when the amount of data is large, data is written in a mode of accessing in parallel and then read and rewritten in a mode of accessing individually.

[0020]

FIG. 1 is a diagram illustrating the principle of the present invention. The present invention has a disk array 46 provided with a plurality of disk devices 32-1 to 32-5 for storing data and storing parity. A plurality of disk devices 32-1 to 32-5 are individually accessed by one writing unit 34 to write data of a predetermined length, for example, divided into sectors. When the write data amount specified by the host device 18 is large, a plurality of disk devices are accessed in parallel by the second writing means 36 to write the data divided in sector units.

Here, the first writing means 34 performs writing in the individual access mode shown in FIG. In the individual write mode, first, prior to data writing, data of a write target position of a disk device to be written and redundant information at the same position as a write target position stored in another disk device are stored. The parity is read and new redundant information is generated by exclusive OR. Next, new sector data and parity are asynchronously written by individually accessing the data storage and parity storage disk devices.

The second writing means 36 performs writing in the parallel access mode shown in FIG. That is, a parity as new redundant information is generated from an exclusive OR of a plurality of sector data stored in parallel in the disk arrangement direction. Next, a plurality of sector data and parity are asynchronously written by accessing a plurality of disk devices in parallel. When the amount of data is large, the second writing unit 36 does not immediately perform writing in the parallel access mode, but performs processing after checking whether all disk devices can be accessed.

In this case, if all of the plurality of disk devices are not in an accessible state, a parity is generated according to the parallel access mode for the data storage format of the individual access mode, and thereafter the disk devices are individually accessed. Access to write sector data and parity.
If all of the plurality of disk devices are accessible, the parity is generated in the parallel access mode, and then the data is written to the plurality of disk devices in parallel.

Further, the storage positions of the parity in the individual access mode may be distributed to the disk devices or may be fixed to specific disk devices.

[0025]

According to the disk array device of the present invention having such a configuration, the amount of data instructed to be written by the host device is checked. An individual access mode that takes time but allows efficient transaction processing after writing is selected. On the other hand, when the data amount is large, a parallel access mode in which one writing time is short is selected for transaction processing.

Here, the individual access mode is RAID
5 (or may be RAID 4) and the parallel access mode corresponds to RAID 3. Therefore, the present invention provides efficient data utilizing the advantages of RAID 5 (or RAID 4) and RAID 3 according to the data amount. Writing will be performed. Further, after a large amount of data is written in the parallel access mode corresponding to RAID3, the data stored in the parallel access mode is read out using the idle time of the disk array device and written in the individual access mode. In other words, it is possible to deal with individual reading and writing of the disk device at the time of transaction processing.

As described above, the disk array controller determines the data amount for the data amount specified by the input / output request from the higher-level device, and switches the operation mode of the disk access, thereby improving the access efficiency. To improve processing time and data transfer speed,
Performance can be improved for input / output requests.

[0028]

FIG. 2 is a block diagram of an embodiment showing a hardware configuration of a disk array device according to the present invention. In FIG. 2, the disk array controller 10 has M
A PU 12 is provided, a host interface 16 for receiving an input / output request from a host computer 18 as an upper device on an internal bus of the MPU 12, a ROM 20 for fixedly storing a control program and the like, and a temporary storage of data as control storage. RAM 22, cache control unit 2
4, a cache memory 26 and a data transfer buffer 28 are provided.

A disk array 46 is provided for the disk array controller 10. In this embodiment, the disk array 46 includes disk devices 32-1 to 32-1.
32-12 are provided, and a rank 48-1 is configured by five disk devices 32-1 to 32-5, and a rank 48- is similarly configured by five disk devices 32-6 to 32-10. 2 and the disk devices 32-11 and 32
-12 are provided as spare machines.

The disk devices provided in the ranks 48-1 and 48-2 are commonly accessed from the disk array device 10 with the same position in the disk arrangement direction as a set. That is, the disk array control device 10 is provided with device adapters 30-1 to 30-6. For example, the first disk device 32-1 of rank 48-1 is connected to the device adapter 30-1 and the device adapter 30-1 is connected to rank 48-2. Is connected to the first disk device 32-6.

FIG. 3 is a functional block diagram showing an access function according to the present invention for the hardware configuration of FIG. 3, the disk array control device 10 includes a command decoding unit 38, a first writing unit 34, a second writing unit 36,
A write mode management table 40, a read unit 42, and a disk access unit 44 are provided. Here, the functions of the command decoding unit 38, the first writing unit 34, the second writing unit 36, the reading unit 42, and the disk access unit 44 are the same as those of the MP shown in FIG.
This is realized by program control by U12, and the write mode management table 40 is developed on the RAM 22 as control storage.

The command decoding unit 38 decodes a command accompanying an input / output request from the host computer 18, and if the command is a write command, the first writing unit 34 or The second writing unit 34 is selected, and a writing process is performed on the disk array 46 via the disk access unit 40. Here, the first writing unit 34 is activated when the amount of write data is small. For example, in the case of the disk devices 32-1 to 32-5 of rank 48-1, for example, each of the disk devices 32-1 to 32 is used. -5 is individually accessed to execute a write process in an individual access mode for writing data divided in sector units. This individual write mode corresponds to the operation mode of RAID5 shown in FIG.

On the other hand, the second writing unit 36 is activated when the amount of write data is large, and accesses the disk devices 32-1 to 32-4 of rank 48-1, for example, in parallel and divides them into sector units. Write processing according to the parallel access mode for writing the written data. The writing process in the parallel access mode by the second writing unit 36 corresponds to the RAID3 operation mode shown in FIG.

Further, in the second writing section 36, instead of immediately executing the writing process in the parallel access mode when the data amount is large, the second writing section 36 sets an access target prior to writing, for example, 48-bit. 1 disk device 32-1
It is checked whether or not all of 32-5 are in an accessible state, and if so, data writing is performed in a parallel access mode.

On the other hand, all the disk devices 32-1
In the case where .about.32-5 are not in the accessible state, writing in a data storage format according to the individual access mode performed by the first writing unit 34 is performed. The parity can be generated without reading the parity and data from the disk device according to the parallel access mode unique to the second writing unit 36.

In the write mode management table 40, information indicating the write mode by the first write unit 34 and the second write unit 36 is registered for each sector. The read unit 42 is activated when the command read unit 38 decodes the read command, refers to the write mode management table 40 by the sector address instructed to read, and follows the individual write mode or the parallel write mode. Reading is executed via the disk access unit 44. In this case, in any mode, reading can be performed in parallel from a plurality of disk devices.

FIG. 4 is an explanatory diagram showing division of write data into sector units in the present invention. In FIG.
Now, assuming that data a, b, c, and d for writing are transferred from the host computer 18, each of the data is written to the disk array 46 by the first writing unit 34 or the second writing unit 36. Data a to d are divided into sector units. For example, data a is divided into four pieces of sector data a1 to a4. The other data b to d are similarly divided into four sector data.

FIG. 5 is an explanatory diagram of the individual access mode by the first writing section 34 of FIG. This individual access mode corresponds to the operation mode of RAID5. In FIG. 5, the format of parity generation in the individual access mode is such that old data is read from the write target sector of the disk device and old parity is read from the parity disk device at the same position as the old data. The new parity is generated by taking the exclusive OR of the old data read from the side, the old parity, and the sector data to be newly written, that is, the new data.

The data storage format of the individual access mode is based on the fact that each disk device is individually allocated to a data unit before division.
In the first sectors 2-1 to 32-5, the first sector data a1, b1, c1, among the data a to d shown in FIG.
d1 and the parity P1 are stored. When writing to these disk devices 32-1 to 32-5, each disk device is individually accessed for writing.

FIG. 6 is an explanatory diagram of the parallel access mode performed by the second writing unit 36 of FIG. 3, and corresponds to the operation mode of RAID3. In FIG. 6, the parity generation format in the parallel access mode is such that the data divided in sector units as shown in FIG.
5 are stored in the arrangement direction of the data P. Therefore, each parity P1 to P
4 can be generated. Therefore, generation of parity does not require reading of old data and old parity from the disk device, and the processing time for data writing can be shortened.

FIG. 7 is an explanatory diagram of a combination in which the parity is generated in the parallel access mode in the second writing unit 36 in FIG. 3 and the data storage is in the individual access mode. In FIG. 7, since the sector data obtained by dividing the write data as shown in FIG. 4 has a data storage format according to the individual access mode shown in FIG. 5, this storage format is assumed. A sequence of sector data is generated and, for example, sector data a1, b at the same sector position
Parity P1 is generated from the exclusive OR of 1, 1, and 1. The same applies to the remaining sector data.

For this reason, even if the data storage format is in accordance with the individual access mode, the parity generation is in accordance with the parallel access mode.
Reading of old data and old parity from −5 becomes unnecessary, and the write processing time can be reduced as compared with the pure parallel access mode shown in FIG. The disk devices 32-1 to 32-1
As in the individual access mode of FIG. 5, the data writing to the disk drive 32-5 is performed by accessing the disk devices 32-1 to 32-5 individually and asynchronously writing the data. Processing time will be longer in comparison. Accordingly, it can be said that the combination method of FIG. 6 is a write mode just intermediate between the individual access mode of FIG. 5 and the parallel access mode of FIG.

FIG. 8 is an explanatory diagram showing rewriting in the individual access mode after writing to the disk device in the parallel access mode shown in FIG. In FIG. 8, in the data write in the parallel access mode performed in a state where the data amount is large and all the disk devices are accessible, the disk device to which the data is to be originally written has rank 48.
In the case of the disk devices 32-1 to 32-5 belonging to -1, the operation is performed for the disk devices 32-6 to 32-10 belonging to another rank 48-2 to enable rewriting after writing.

As described above, when data has been written to the disk devices 32-6 to 32-10 of another rank 48-2 in the parallel access mode, first, as shown in the disk array controller 10, Sector data and parity data are read from the disk devices 32-6 to 32-10 of rank 48-2, and the disk devices 32-1 belonging to the original rank 48-1 according to the parallel access mode.
Parity is generated assuming a data storage format for 32-5.

For example, a disk device 3 of rank 48-2
When the sector data a1, b1, c1, and d1 are read from 2-6, a new parity P10 is generated from the exclusive OR of all the sector data, and the disk devices 32-1 to 32 of rank 48-1 are generated. As shown in the figure, sector data a1, b1,
Write c1, d1 and parity data P10.

Similarly, the disk units 32-7 to 32
For -9, the individual access mode is rewritten to the data storage format. FIG. 9 is a flowchart showing the overall processing operation of the present invention. First, in step S1, it is checked whether or not a host command has been received from the host computer 18, and the host command based on an input / output request from the host computer is checked. Is received, step S2
To decode the command. If it is a write request from the command decoding result, the process proceeds from step S3 to step S4,
Execute write processing.

On the other hand, if it is a read request, the flow advances to step S5 to execute a read process. FIG. 10 is a flowchart showing details of the read process. In step S1, the write mode management table 40 is first referred to, and in step S2, it is checked whether or not the write is in the individual access mode. In the present invention, the data storage format of the disk device according to the individual access mode is premised. However, the storage format of the parallel access mode that has not been rewritten may remain. Check whether the writing is in the individual access mode.

If the reading is directed to the write data in the individual access mode, the corresponding reading is executed in step S3. That is, the corresponding disk device is individually accessed by the disk ID and the data address specified by the command from the higher-level device to read the sector data, and the data is transferred to the data transfer buffer 28 in the disk array control device 10 shown in FIG. Once stored, it is transferred to the host computer 18 via the host interface 16.

If the read operation is directed to the write data in the parallel access mode, the corresponding disk devices are individually or in parallel accessed in step S4 according to the disk ID and the data address specified by the command from the higher-level device. After reading the sector data and temporarily storing it in the data transfer buffer 28, the host interface 16
Is transferred to the host computer 18 via.

FIG. 11 is a flowchart showing details of the write processing shown in step S4 of FIG. This figure 1
1 is implemented by the functions of the command decoding unit 38, the first writing unit 34, the second writing unit 36, and the disk access unit 40 shown in FIG.

In FIG. 11, first, in step S1, the parity generation format and the data storage format are set to the individual access mode. Subsequently, in step S2, it is checked whether the write data amount β sent from the host computer is larger than a predetermined data amount α. If the write data amount β is smaller than the specified data amount α, the process proceeds to step S3, where parity data is generated according to the individual access mode set in step S1, and then written to the disk device.

In the writing to the disk device in the individual access mode in step S3, since the generation of parity requires reading of the old data and the old parity from the disk device, the processing time for one write is required. However, since the write data amount β is small, the access time itself as a whole does not become so long. On the other hand, if the write data amount β is larger than the specified data amount α in step S2, the process proceeds to step S4 and subsequent steps. The processing after step S4 is performed by the second writing unit 3 shown in FIG.
6 is performed.

First, in step S4, the parity generation format is changed from the individual access mode set in step S1 to the parallel access mode. As a result, the parallel access mode is set for parity generation, while the individual access mode is set for data storage format. Subsequently, in step S5, it is checked whether all the disk devices belonging to the rank to be written are accessible. If at least one of the disk devices belonging to the rank to be written is busy, the process proceeds to step S6, where parity is generated in the parallel access mode based on the setting result in step S4, and then the individual access mode is set. Write data to the disk device in accordance with.

The combination of parity generation in the parallel access mode and data storage in the individual access mode is as shown in FIG. If all the disk devices of the rank to be accessed are accessible in step S5, the process proceeds to step S7, and the data storage format is also changed to the parallel access mode. As a result, the parallel access mode is set for both the parity generation format and the data storage format. Subsequently, in step S8, parity is generated on the side of the disk array controller 10 in accordance with the parallel access mode, and data is written to a disk device of another rank different from the write target according to the parallel access mode.

In the data writing in this parallel access mode, as shown in FIG. 6, it is not necessary to read out the disk units for parity generation, and all the disk units are accessed in parallel and written asynchronously. Therefore, high-speed data writing can be performed even if the data amount is large. Subsequently, the process proceeds to step S9, as shown in FIG. 8, the data written to the disk device of another rank in the parallel access mode is read out in disk units, and rewritten to the disk device of the target rank in the individual access mode. Rewrite to the data storage format according to the individual access mode.

The rewriting to the data storage format in accordance with the individual access mode in step S9 does not hinder an input / output request from the host computer, so if there is an input / output request, the rewriting process is temporarily interrupted. However, it is desirable to restart after the completion of the input / output request. FIG. 12 is an explanatory diagram of another individual access mode according to the present invention. In the individual access mode shown in FIG. 5, the storage locations of the parity are distributed for each disk device. In the individual access mode, for example, it is fixed to the disk device 32-5. This individual access mode corresponds to RAID4. The parity generation format is the same as in FIG.

FIG. 13 is a flow chart showing another embodiment of the write process shown in step S4 of FIG. 9. In the write process of FIG. 11, the write data amount β is larger than the specified data amount α. In this case, the access mode differs depending on whether all the disk devices are accessible or not.
In the third embodiment, after the parity generation format is changed to the parallel access mode in step S4, step S5
, A parity is generated in the parallel access mode, and writing is performed in the disk device in the individual access mode. Regardless of whether all the disk devices are accessible or not, the parity access mode shown in FIG. The operation mode is a combination of parity generation and data storage in the individual access mode.

In the embodiment shown in FIG. 13, since data is not written only in the parallel access mode, it is necessary to write back to the data storage format of the individual access mode as shown in FIG. Absent. Note that, as shown in the hardware configuration of FIG.
In this example, five data storage disk devices and one parity storage disk device are provided for one rank. However, the number of disk devices for data recording and parity recording is not necessary. Any number can be used accordingly. This is the same for the number of ranks.

[0059]

As described above, according to the present invention, the parallel access mode and the individual access mode are appropriately selected according to the write data amount specified by the input / output request from the host device. Thus, data can be written utilizing each operation mode, and as a result, the access performance of the disk array device can be improved.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is a configuration diagram of an embodiment showing a hardware configuration of the present invention.

FIG. 3 is a block diagram showing an access function according to the present invention.

FIG. 4 is an explanatory diagram in which write data in the present invention is divided into sector units.

FIG. 5 is an explanatory diagram of an individual access mode in the present invention.

FIG. 6 is an explanatory diagram of a parallel access mode according to the present invention.

FIG. 7 is an explanatory diagram showing a combination of a parity generation format in a parallel access mode and a data storage format in an individual access mode according to the present invention.

FIG. 8 is an explanatory diagram showing rewriting in an individual access mode after writing in a parallel access mode in the present invention;

FIG. 9 is a flowchart showing the overall processing operation of the present invention.

FIG. 10 is a flowchart showing details of a read process in FIG. 4;

FIG. 11 is a flowchart showing details of a write process in FIG. 4;

FIG. 12 is an explanatory diagram of another individual access mode according to the present invention.

FIG. 13 is a flowchart showing another embodiment of the write processing in FIG. 4;

FIG. 14 is an explanatory diagram of a disk array device using RAID0.

FIG. 15 is an explanatory diagram of a disk array device using RAID1.

FIG. 16 is an explanatory diagram of a disk array device using RAID3.

FIG. 17 is an explanatory diagram of data division in RAID3.

FIG. 18 is an explanatory diagram of a disk array device using RAID4.

FIG. 19 is an explanatory diagram of a disk array device using RAID5.

[Explanation of symbols]

 10: Disk array controller 12: MPU 14: Internal bus 16: Host interface 18: Host computer (upper device) 20: ROM 22: Volatile memory (RAM) 24: Cache controller 26: Cache memory 28: Data transfer buffer 30-1 to 30-n: Device adapter 32-1 to 32-n: Disk device 34: First writing unit (first writing unit) 36: Second writing unit (second writing unit) 38: Command decoding unit 40: write mode management table 42: read unit 44: disk access unit 46: disk array 48-1, 48-2: rank

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁶ , DB name) G06F 3/06 301 G06F 3/06 302 G06F 3/06 540

Claims (9)

(57) [Claims] A disk array (46) having a plurality of disk devices (32-1 to 32-5) for data storage and parity storage, and a write data amount specified by a higher-level device (18). When the number is small, the plurality of disk devices (32-1 to 32-32)
(5) a first writing means (34) for individually accessing and writing data divided into a predetermined length; and when the amount of write data specified by the host device (18) is large, the plurality of disks Second writing means (3) for accessing the apparatus in parallel and writing data divided into predetermined lengths;
6) A disk array device comprising: 2. The disk array device according to claim 1, wherein said first writing means (34) pre-writes a write target position of a disk device to be written before writing data divided into predetermined length units. Data and the redundant information stored in another disk device at the same position as the write-scheduled position are read to generate new redundant information, and then the new divided data and redundant information are individually stored in each of the disk devices. A disk array device characterized in that writing is performed selectively. 3. The disk array device according to claim 1, wherein said second writing means (36) newly generates a redundant data from a plurality of data stored in parallel in a disk arrangement direction divided into predetermined length units. A disk array device for generating information, and then writing new divided data and redundant information to the plurality of disk devices in parallel. 4. The disk array device according to claim 1, wherein said second writing means (36) is configured to output said plurality of disks when all of said plurality of disk devices are not accessible. It is characterized in that redundant information is generated from the divided data in the disk arrangement direction when writing data individually to the device, and then the disk device is individually accessed to write each divided data and new redundant information. Disk array device. 5. The disk array device according to claim 1, wherein said second writing means (36) divides the data into predetermined length units when all of said plurality of disk devices are accessible. A new redundant information is generated from the divided disk array direction, and then the divided data and the redundant information are written in parallel to the plurality of disk devices. 6. The disk array device according to claim 5, wherein said second writing means (36) writes divided data and redundant information in parallel to a plurality of disk devices provided separately.
A disk array device wherein the original disk device is individually accessed and rewritten by the first writing means (34) and read out after the writing. 7. The disk array device according to claim 1, wherein said first writing means (34) and said second writing means (36) write data to a plurality of disk devices.
A disk array device in which disk devices for writing redundant information are dispersed. 8. The disk array device according to claim 1, wherein said first writing means (34) and said second writing means (36) write data to a plurality of disk devices.
A disk array device wherein a disk device for writing redundant information is fixed. 9. The disk array device according to claim 1, wherein parity data is used as said redundant information, and said first writing means (34) reads data read from the disk device, parity data, and new data. A new parity is generated from the exclusive OR of the data to be written into
A disk array device wherein the writing means (36) generates a new parity from an exclusive OR of a plurality of data written in parallel in the disk arrangement direction.