JPH09274543A - Disk array device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress heat generation as the factor of fault to a minimum while preventing the performance of data read from being dropped by writing one file so as to completely store it in one disk device and activating the disk device when an access request is generated. SOLUTION: A disk array device 1 is provided with a disk array controller 10 and plural disk devices 20. The disk array device 1 controls the plural disk devices 20 through the disk array controller 10 and performs the write and read of data to the prescribed disk device 20. Corresponding to the writing request of the file, the disk array controller 10 writes one file so as to completely store it in one disk device 20 and stops the disk devices 20 to which there is no access request. Then, when the access request is generated to any stopped disk device 20, that disk device 20 is activated.

Description

Detailed Description of the Invention

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a disk array device capable of writing and reading data to and from a predetermined disk device and used for a video server or the like which stores, for example, large-capacity video data.

[0002]

2. Description of the Related Art Conventionally, for example, RAID (Redundant
2. Description of the Related Art As represented by an array of ixpensive disks, a disk array device is known in which a plurality of disk devices are controlled by a disk array controller and data is written to and read from a predetermined disk device. A video server or the like that supplies a relatively large amount of data such as video data using the disk array device is configured.

In this disk array device, in order to prevent the performance of reading data from being deteriorated, a plurality of disk devices provided are all activated so that reading can be started at any time.
During the operation of the disk device, a large amount of heat is generated due to the operation of the disk drive spindle motor and the like, and the generated heat may cause a failure.

[0004]

However, even if all the disk devices are operated as described above, it is not always necessary to read data from all the disk devices, but to actually read data. Often a part of them. Of course, even though the disk device to be read varies depending on the type of data, the disk device to be read at the same time is often a part. Therefore, for example, even if the video server system is operated for 12 hours, there is a possibility that a disk device in which data is read out for only about one hour or a disk device in which data is not read out even once during operation may appear. There is also.

As described above, the state in which the operating state is continued although it is not used for actual data reading means that there are many heat sources described above although it is not necessary in the true sense, and the failure occurs. It is not preferable in terms of prevention of However, if the operation is stopped simply because it is not the current read target, the disk device must be started each time the disk device that is the target of data read changes frequently. However, there is a high possibility that a decrease in data read performance will become apparent.

The present invention has been made in order to solve the above-mentioned problems, and it is possible to minimize the heat generation that causes a failure while preventing the deterioration of the data reading performance. The purpose is to provide a device.

[0007]

In order to achieve this object, the invention according to claim 1 controls a plurality of disk devices by a disk array controller, and data to a predetermined disk device is controlled. In the disk array device capable of writing and reading data, the disk array controller writes the file so that one file is completely stored in one disk device, and an access request. Stop control means for stopping the disk device that does not have
A disk array device comprising: a start control unit for starting the disk device when the access request is issued to the disk device stopped by the stop control unit.

In the disk array device of the present invention, a plurality of disk devices can be controlled by the disk array controller to write and read data to and from a predetermined disk device. Here, the stop control means stops the disk device for which there is no access request, and when an access request is generated for the stopped disk device, the activation control means restarts the disk device.
As described above, since the disk device that does not make an access request can be stopped, it is possible to minimize the heat generated when the disk device is started, particularly in the disk rotating mechanism such as the spindle motor.

Further, if the data is read from the stopped disk device, it will be delayed by the time required to start the disk device.
For each file write request, one file is 1
Since the data is written so as to be entirely stored in one disk device, one file can be continuously read once the reading is started. Therefore, for example, when video data is stored, there is no inconvenience that the output video data is interrupted.

As described above, it is possible to minimize the heat generation that causes a failure while preventing the deterioration of the data reading performance. When writing one file by the writing control means so that all the files are stored in one disk device, as described in claim 2, first, the size of the file for which the writing is requested is checked, and the size is equal to or larger than that size. It is sufficient to write to the disk device that has the free area. The empty areas in this case do not necessarily have to be continuous, and may be divided.

Regarding the point that the stop control means stops the disk device for which there is no access request from the main computer, as described in claim 3, the disk array controller has an access request for storing a predetermined number of access requests. A request storage unit is provided, and the processing for the access requests stored in the access request storage unit is configured to be executed in a predetermined order, and the stop control unit is a target of the access request stored in the access request storage unit. It suffices to stop the disk device that is not set.

This is a problem that, for example, based on only one access request to be executed from now on, a disk device in which an access request is made every other time is repeatedly stopped and started every other time. It is a consideration. Considering the access requests of a predetermined number of destinations as in the case of the present invention, such troublesome stop / start can be avoided, which is also preferable from the viewpoint of the data read performance deterioration. In other words, once the disk device is stopped, it will take some time until the startup is completed again.Therefore, in order to accurately determine whether or not it is necessary to stop the disk device, the access request is specified in this way. It is preferable to employ a configuration in which even the number is stored.

[0013]

Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a schematic configuration of an embodiment in which a disk array device is applied as a large-capacity storage device in a file server that handles MPEG video data.

The file server is a disk array device 1
And a main computer 30 for giving instructions such as data read / write to the disk array device 1, both of which are connected via a bus 25. MPEG data can be input to the disk array device 1 from the outside via the bus 25, or MPEG data can be output from the disk array device 1 to the outside via the bus 25. Specifically, for example, the data from the MPEG encoder 40 is input, or the MP is sent to the terminal 50.
The EG data is output and decoded by the MPEG decoder of the terminal 50 as a video signal.

The disk array device 1 comprises a disk array controller 10 and a plurality of disk devices 20 for storing a large amount of video data supplied as a video server. Disk array controller 10
Is a control unit for the entire disk array apparatus, and a CPU 11 corresponding to a "write control unit", a "stop control unit", and a "start control unit", and a storage unit storing an operation program such as array management software. ROM1 as
2, a RAM 13 as an “access request storage unit” which is a work area of the CPU 11, an interface device 14 connected to a bus 25 for interfacing with the outside, a cache memory 15, a protocol control device 17, etc. Is equipped with.

The cache memory 15 is the disk device 2
It operates as a disk cache for data output from 0 to the outside and data input to the disk device 20 from the outside. Further, the protocol control device 17 performs data transfer between the disk device 20 and the cache memory 15. In FIG. 1, three protocol control devices 17 are connected to the cache memory 15, and three disk devices 20 are connected to each protocol control device 17, but one cache device may be used as needed. Four or eight protocol control devices 17 may be connected to the memory 15, or eight or sixteen disk devices 20 may be connected to one protocol control device 17.

CPU of disk array controller 10
The disk drive 11 can simultaneously operate the disk devices 20 in parallel via the protocol controller 17 based on the array management software in the ROM 12. On the other hand, in the disk device 20, as shown in FIG. 2, a so-called physical disk drive unit 61 and a control unit 62 for controlling the same are integrated.

The disk drive unit 61 includes a hard disk 65 provided with a recording magnetic layer, a spindle motor 66 for driving the hard disk 65, a disk drive mechanism 67, and reading / writing of data stored in the disk. Magnetic head 68 for performing the above and head driving mechanism 6 for driving the magnetic head 68
9 and so on.

The control unit 62 includes a CPU 71 as well-known control means and an R as storage means.
OM73, RAM74, SCSI interface (I /
F) 75 and the like are provided to control the disk drive mechanism 67, the head drive mechanism 69, and the like described above to execute reading / writing of data stored in the hard disk 65. Further, it is connected to the protocol control device 17 via the SCSII / F75.

Next, the disk array device 1 of this embodiment
Will be described. First, the processing of the CPU 11 when writing a file to the disk array device 1 will be described with reference to the flowchart of FIG. First step S11
Then, the file write open command is received from the main computer 30. Normally, the file size is not specified in the file open command, but in the case of the present embodiment, it is necessary to store all one file in the same disk device 20, so the file size is specified here. .

In the subsequent S12, a disk device 20 having a disk free area equal to or larger than the size specified in S11 is searched for, and the device number of the disk device 20 is set to n. Then, in S13, a block write command is received from the main computer 30, and in S14, for example, the MPEG encoder 4 via the bus 25 for data stream input / output.
The block data received from 0 is transferred to the disk device 20 of device number n. Then, in S15, it is determined whether or not the block transfer for the file size is completed,
If not completed, the process returns to S13.

As described above, one file is written so that all the files are stored in the same disk device 20, but the empty areas in this case are not necessarily continuous and may be divided. Next, the operation of reading a file from the disk array device 1 will be described. Since the read file is supplied to the terminal 50 that has made a request, a detailed description will not be given regarding this point, and the disk device 20 is requested based on the request.
The process of stopping or activating will be mainly described.

In this embodiment, the number of disk devices 20 is 9, and the disk device numbers n are numbered 1-9. Usually, the main computer 30 has a multi-task structure, and each task issues a block read command for a necessary file to the disk array device 1. Here, it is assumed that four tasks A to D are activated in the initial state and access to the disk devices 20 with device numbers n = 1, 2, 4, and 6, respectively. Therefore, the other disk devices 20 with device numbers n = 3, 5, 7, 8, 9 are stopped.

FIG. 6A shows a state (command queue) in which commands are stored in the RAM 13 according to the reception order when the disk array device 1 receives a block read request from the main computer 30. The command in this command queue corresponds to the "access request" in the present invention, and the number preceding it indicates the device number n (one of 1 to 9) of the disk device 20 and the letter (a, b, c ...) Shows command numbers.

Further, FIG. 6B shows the order of actually outputting the data stream, and like FIG.
The number preceding the device number n (1 to 9) of the disk device 20
, And the letters (a, b, c ...) Following the command numbers. Normally, in the disk array device 1, even if a continuous block read command is received for one disk device 20 in order to eliminate performance deterioration due to seek wait time in the disk device 20, it is shown in FIG. The order of command processing is changed so that the disk devices 20 to be read are different. For example, in the command queue of FIG. 6 (a), commands 1a and 1b or 2a and 2b are stored consecutively, but the output order in FIG. 6 (b) is 1a → 2a → 4a → 6a → 1b. → 2b → ……, so that they are not consecutive. Therefore, there is some time delay between the reception of the command and the output of the data stream corresponding thereto, and the data stream output corresponding to FIG. 6B is executed later than the timing shown in FIG. Become.

Based on FIG. 6, the processing executed by the CPU 11 when reading a file will be described. The flowchart of FIG. 4 is the one shown in FIG.
It shows the processing executed every time when the output of one data stream is completed. First, in S21, it is checked whether or not the command for the disk device 20 having the device number n, whose output of the data stream has been completed, is in the command queue shown in FIG. If the corresponding command is present in the command queue (S22: YES), the process of FIG. 4 is terminated as it is, but if not (S22: YES).
No), the process proceeds to S23 to stop the disk device 20 having the device number n, which has finished outputting the data stream.

For example, in FIG. 6B, the timing at which the output of the data stream from the disk device 20 having the device number 2 (output to the command 2b) is completed is shown in FIG. 6A in this case. It is checked whether or not there is a command (more specifically, command 2c) for the disk device 20 having the device number 2 in the command queue. In this example, all the requests to the disk device 20 with the device number 2 designated by the task B have been completed at the timing, so there is no corresponding command in the command queue. Therefore, in this case, the process of S23 of FIG. 4 is executed, and the disk device 20 of device number 2 is stopped.

On the other hand, FIG. 5 is a flowchart showing the processing executed when the CPU 11 receives a block read command. In S31, the designated device number n (1 to 9
Of the disk device 20 is received, and the device number n of the disk device 20 is obtained therefrom. In the following S32, it is determined whether the disk device 20 of the corresponding device number n is currently stopped,
If it has been started (S32: NO), it is as it is in FIG.
However, if it is stopped (S32:
YES), the process proceeds to S33, and the disk device 2 of the corresponding device number n that has received the block read command in S31.
Start 0.

For example, at the timing shown in FIG. 6A, a task E is newly activated, and this task E starts an access request to the disk device 20 having the device number 5. Therefore, the CPU 11 executes the process of S33 of FIG. 5 at the timing, and the disk device 20 of device number 5 is activated.

As the output order of the data stream corresponding to this, the output of the data stream of the disk device 20 having the device number 5 is started at the timing shown in FIG. 6B, but the disk in which the spindle is stopped is generally used. It takes several seconds to several tens of seconds until the data can be taken out after the unit is started. Therefore, when the disk device 20 with the device number 5 started at the timing in FIG. Data stream output 5a
May be postponed.

As described above, according to the disk array device 1 of this embodiment, the disk device 20 that does not have an access request is judged by the presence or absence of a command in the command queue (see FIG. 6A), and access is made. If there is no request, stop it. When a new access request is issued to the stopped disk device 20, the disk device 20 is activated again. As described above, since the disk device 20 that does not make an access request can be stopped, it is possible to minimize the heat generated especially in the disk rotating mechanism such as the spindle motor 66 (see FIG. 2) when it is started. it can.

Further, if data is read from the stopped disk device 20, it will be delayed by the time required to start the disk device 20. Are written so that they are all stored in one disk device 20, so that once reading is started, one file can be read continuously. Therefore, for example, when video data is stored, there is no inconvenience that the output video data is interrupted.

As described above, it is possible to minimize the heat generation that causes a failure while preventing the deterioration of the data reading performance. Further, in the case of the present embodiment, the disk device 20 that does not make an access request.
The determination is made based on the command queue as shown in FIG. In other words, it also considers commands that will be executed in the future. This takes into consideration the problem that, for example, based on only one command to be executed from now on, the disk device 20 in which the command is executed every other time is repeatedly stopped and started every other time. is there. In other words, by considering the command ahead of a predetermined number, such troublesome stop / start can be avoided, which is also preferable in terms of the performance degradation of data reading. Disk device 20
It takes a certain amount of time to complete restarting once it is stopped.Therefore, in order to accurately judge whether it is necessary to stop it, store up to a specified number of commands in this way. It is preferable to leave it.

As described above, the present invention is not limited to such an embodiment, and can be implemented in various forms without departing from the gist of the present invention. For example, in the above-described embodiment, the disk array controller 10 controls the plurality of disk devices 20 in response to a request from the main computer 30 so that the disk array device 1 can write and read data to and from a predetermined disk device 20. Although configured, for example, the disk array controller 1
An input means such as a keyboard may be connected to 0 itself, and a command may be directly issued via the input means. In such a case, for example, a configuration in which a plurality of hard disks are connected to a personal computer as the disk array controller 10 can be applied to the present invention.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration when a disk array device according to an embodiment is used as a large-capacity storage device in a file server.

FIG. 2 is a block diagram showing an electrical configuration of a disk device.

FIG. 3 is a flowchart showing a process executed by a CPU when writing a file to a disk array device.

FIG. 4 is a flowchart showing a process executed by the CPU when the output of one data stream is completed.

FIG. 5: CPU when receiving a block read command
5 is a flowchart showing a process executed in.

6A is an explanatory diagram showing a command queue, FIG.
(B) is an explanatory view showing the order of actually outputting the data stream.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Disk array device 10 ... Disk array controller 11 ... CPU 12 ... ROM 13 ... RAM 14 ... Interface device 15 ... Cache memory 17 ... Protocol control device 20 ... Disk device 25 ... Bus 30 ... Main computer 40 ... MPEG encoder 50 ... Terminal

Claims (3)

[Claims] 1. A disk array device capable of controlling a plurality of disk devices by a disk array controller and writing and reading data to and from a predetermined disk device, wherein the disk array controller responds to a file write request. 1 file is 1
Write control means for writing so that all of them are stored in one disk device, stop control means for stopping a disk device for which there is no access request, and the access request has been issued to the disk device stopped by the stop control means. In some cases, the disk array device is provided with a start control means for starting the disk device. 2. The write control means is configured to check the size of a file for which a write request has been made and write to a disk device having a free area equal to or larger than the size. The disk array device according to 1. 3. The disk array controller further comprises access request storage means for storing a predetermined number of access requests, and executes the processing for the access requests stored in the access request storage means in a predetermined order. And the stop control unit is configured to stop a disk device that is not the target of the access request stored in the access request storage unit. 2. The disk array device according to item 2.