JPH07141770A - Multiplex disk device - Google Patents

Abstract

PURPOSE:To provide a multiplex disk device capable of improving an average response speed for a read request. CONSTITUTION:A function to calculate a waiting time being the time to be required until the relative positional relation of a head and a disk medium becomes suitable to that for reading out the data is imparted to each disk device 12 for multiplexing and storing data based on the address information of an objective data to be read out. A controller 11 makes each disk device 12 calculate the waiting time when a read request is received from an external device 21, and responds the read request by using the disk device whose waiting time calculated is the shortest. The relative positional relation of the head and the disk medium at the time of receiving the information instructing the calculation of the waiting time is recognized by detecting the rotational angle of the disk medium, for example.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multiplex disk device, and more particularly to a multiplex disk device having an improved response speed for a data read request.

[0002]

2. Description of the Related Art In a disk device used as an external storage device such as a computer, the information stored on the disk medium may be lost due to a collision between the head and the disk medium. Is
It is lower than that of semiconductor memory. Therefore, in order to improve the reliability of the storage device,
Information is stored in a redundant configuration by combining a plurality of disk devices so that the total amount of information as a storage device does not change even if a failure occurs in one of the disk devices. Such a device is called a disk array device or the like, and a disk array device configured to store the same information in a plurality of disk devices is a multiplexed disk device.

The multiplex disk device is composed of a plurality of disk devices and a controller for controlling them in an integrated manner, and when the controller receives a data write request from an external device such as a computer, each of them is controlled. The data is written in the disk devices, and the stored information in all the disk devices is always kept the same. in this way,
In the multiplexed disk device, since it is necessary to write the data to all the disk devices for writing the data, there is no choice in the write operation, and it is difficult to improve the response speed to the write request. However, regarding data reading, the same data is written in multiple disk devices, so if you can specify the disk device that can read in the shortest time, read the data from that disk device. If comprised, the response speed with respect to a read-out request will improve. On the basis of such an idea, various types of multiplexed disk devices have been proposed which are designed to improve the response speed to read requests.

For example, Japanese Examined Patent Publication No. 63-57812 discloses a device that notifies the same read request to a plurality of disk devices and reads data from the disk device that responds early. There is.

The operation of this apparatus will be briefly described with reference to FIG. When the main control circuit 33 receives a data read request from the external device 31 such as a computer via the interface control circuit 32, the main control circuit 33 sends the read request to the two disk devices 34 _A and 34 _B if the request is a read request. Notify the request. Upon receiving the read request, the disk devices 34 _A and 34 _B start the operation of reading the requested data, respectively, and output a read ready signal to the determination circuit 35 at the stage when the read preparation of the data from the disk medium is completed. To do. The determination circuit 35 controls the switching circuit 37 so that the data from the disk device that outputs the read ready signal enters the data buffer 36.

Further, Japanese Patent Application Laid-Open No. 63-220492 proposes a device for improving the response speed without actually making a read request to each disk device.

The operation will be briefly described with reference to FIG. This apparatus not plurality of disk devices 34 _A 34
_The controller 38 is provided with a storage unit 39 for storing the track numbers of the tracks on which the heads of the respective disk devices 34 _A to 34 _C are located. . When the control device 38 receives a data read request from the external device 31, based on the track number of the head of each disk device stored in the storage unit 39 and the track number of the data to be read / read,
The required moving distance of the head in each disk device is calculated. Then, the read request is notified to the disk device having the shortest calculated movement distance, and the data is read from the disk device.

[0008]

According to the conventional technique described above, the average response time to a read request can be shortened as compared with a device configured to always read from a predetermined disk device. . But,
In the device disclosed in Japanese Patent Publication No. 63-57812, read requests are issued to all the disk devices, so that the heads of the respective disk devices always exist on the same track. To read the data, seek time for the head to move on the target track,
Although it takes time for the disk medium to rotate and the target storage area to be located under the head, this device has the same seek time in any disk device, and the head rotates after seek. Only the effect of shortening the waiting time is obtained, and there is a problem that the situation in which the same data is stored in a plurality of disk devices is not fully utilized.

To reduce the seek time, Japanese Patent Laid-Open No. 63-63
It is effective to notify the read request to only one disk device like the device disclosed in Japanese Patent Laid-Open No. 2204202. When operated in this way, the heads of the respective disk devices are located on different tracks, and the seek time can be shortened. However, in this device, the control device for controlling a plurality of disk devices determines the disk device to be used for reading by estimating the time required for the response of each disk device based only on the track number. However, since no consideration has been given to the rotation waiting time, a disk device other than the disk device that can respond to the read request in the shortest time may be mistakenly used for reading, and again, the same data is stored in multiple disk devices. It does not fully utilize the situation of being stored.

Therefore, an object of the present invention is to accurately identify a disk device that requires the shortest time for reading data and read the data from the disk device, thereby improving the responsiveness to a read request. Another object of the present invention is to provide such a multiplexed disk device.

[0011]

The invention according to claim 1 is
(B) Rotation angle detection means for detecting the rotation angle of the disk medium, track number detection means for detecting the track number of the track on which the head is located, and rotation angle detection means when instructed to calculate the waiting time. Using the rotation angle detected by and the track number detected by the track number detecting means, the waiting time which is the time required until the head moves to the storage area for which the waiting time is calculated is calculated and output. A plurality of disk devices provided with output means, (b) determination means for determining whether the received request is a read request or a write request, and (c) determination means for determining a read request. In the case of waiting time output by each of the plurality of disk devices according to the instruction of the instruction means for instructing each of the plurality of disk devices to calculate the waiting time Specifying means for specifying a single disk device to calculate the shortest waiting time by comparing, (e)
Responsive means for controlling the disk device specified by the specifying means and responding to the read request, and (f) outputting the write request to each of the plurality of disk devices when the judging means judges that it is the write request. As a result, a writing unit that writes the same data to all of the plurality of disk devices is provided.

That is, according to the first aspect of the present invention, in a multiplexed disk device composed of a plurality of disk devices and a controller for controlling them, each disk device has a waiting time calculation function, and the controller is , When a request from an external device such as a host computer is a read request, this calculation function is used to
The disk device that can read the earliest is specified, and the data requested to be read is read from the specified disk device.

The rotation angle detecting means and the track number detecting means provided in the disk device are used for recognizing the positional relationship between the head and the disk medium, and the output means is the rotation angle and the track detected by these detecting means. Based on the number, the time until the head is positioned in the storage area for which the waiting time is calculated is calculated by calculating the seek time and the rotation waiting time. As the rotation angle detecting means, for example, a counter that increases the count value at a predetermined time interval and resets the count value each time the disk medium makes one rotation can be used. Alternatively, the rotation angle of the disk medium may be detected by a rotary encoder or the like, or the sector number written on the disk medium may be read and the rotation angle may be calculated from the sector number.

According to a second aspect of the present invention, (a) a rotation angle detecting means for detecting a rotation angle of the disk medium, a track number detecting means for detecting a track number of a track on which the head is located, and a waiting time It is required until the head moves to the storage area for which the waiting time is calculated, using the rotation angle detected by the rotation angle detection means when the calculation is instructed and the track number detected by the track number detection means. A plurality of disk devices having an output means for calculating and outputting a waiting time which is time; and (b) a storage means for storing the track number of the track where the head is located in each disk device. , (C) determining means for determining whether the received request is a read request or a write request, and (d) determining means for determining whether the request is a read request. The shortest waiting time is calculated by comparing the waiting time output by each of the plurality of disk devices according to the instruction of the instruction means (e) The specifying means for specifying one disk device, (f) response means for controlling the disk device specified by the specifying means to respond to the read request, and (g) the storage means for each response by the response means. The rewriting means for rewriting the track number corresponding to the specified disk device with the read track number, and (h) each of the plurality of disk devices when the judging means judges that the request is a write request. By outputting the write request, the same data is written in all of the plurality of disk devices, and the data is written by the (re) writing means. Inclusive comprising a moving means for moving on a track designated by the completed each track number the head is stored in the storage unit of the disk device after.

That is, according to the invention described in claim 2, in the invention described in claim 1, the storage means for storing the track number of the track in which each head existed before executing the write request, and the contents thereof are provided. Rewriting means for updating is added, and after responding to the write request, the head arrangement of each disk device is returned to the state before execution of the write request. As a result, the heads of the respective disk devices are always arranged on different tracks, and the average response speed to the read request immediately after the response to the write request can also be improved.

[0016]

EXAMPLES The present invention will be described in detail below with reference to examples.

FIG. 1 shows the configuration of a multiplexing disk device according to an embodiment of the present invention. The multiplexed disk device of the embodiment includes a controller 11 and three disk devices 1
It is composed of 2 _A , 12 _B and 12 _C. Controller 11
Receives an access request from the external device 31, outputs the write request to the three disk devices when the request is a write request, and outputs the write request fastest when the request is a read request. This is a circuit that determines one disk device that can be read according to an operation described later and outputs a read request to the disk device. As shown in the figure, in the multiplexed disk device of this embodiment, the controller 11 and each disk device 12 are connected by independent data and command transfer cables, and the controller 11 is connected to all the disk devices. Requests can be output at the same time, and the processor, a program memory that stores a program that defines its operation, and peripheral circuits thereof are included. The disk device 12 is
This is a device in which a circuit for calculating a waiting time required for reading data is added to a known disk device.

FIG. 2 shows an outline of a disk device used in the multiplexed disk device of the embodiment. An actual device includes a motor for rotating the disk medium 13, a spindle motor control circuit for controlling the rotation speed of the motor, an actuator for moving the head 14, and an actuator for controlling the head 14 on a target track. Various parts such as a head positioning control circuit for seeking are provided, but in this figure, only the part related to the waiting time calculation process is shown. The disk device 12 used here is a device that rotates a plurality of disk media by fixing them to the same spindle, and a head for reading information from each surface of each disk media is driven by one actuator. Moved at the same time. The servo of the head is performed by reading the servo information recorded on one of the surfaces by the servo head 14. First, the individual operation of each circuit will be briefly described with reference to this drawing.

The servo information on the servo surface 13 is converted into an electric signal by the servo head 14 and input to the index signal detection circuit 15. Index signal detection circuit 15
Outputs an index pulse to the rotation angle counter 16 each time it detects an index gap in the servo information. The clock generator 17 is a circuit that outputs a clock 360 times within a predetermined time for one rotation of the disk medium, and the clock is input to the rotation angle counter 16. The rotation angle counter 16 is a circuit that increments the counter by "1" each time a clock signal is input from the clock generator 17, and resets the counter value to "0" when an index pulse is input. The CPU 18 is a control circuit that uses the memory 19 to calculate the waiting time based on the counter value of the rotation angle counter 16. In this disk device, the CPU 18 also controls reading / writing from / to the disk medium. CP is for sending and receiving data and commands with external devices.
It is performed via the interface 20 under the control of U18.

An outline of the waiting time calculation processing of this disk device will be described with reference to FIG. The head 21 is a read / write head driven by the same actuator as the servo head, and the surface (disk medium) 22 is a surface on which data is stored. As is well known, the movement between tracks is performed by the movement of the head 21, and the movement within the tracks is performed by the rotation of the disk medium 22, but in this figure, these movements are made relative to the disk medium. Is displayed as.

In the figure, the angle reference 25 shown by the solid line is a place corresponding to the portion where the index signal is written on the servo surface, and when the counter value of the rotation angle counter is "0", the head 21 is shown. Is also the place where is located. As shown in this figure, the head 21 moves to the track 23 ₁
Consider a case where a read operation is performed on the data which is above and is stored in the sector 24 on the track 23 ₂ . In this case, the head 21 first seeks toward the track 23 ₂ . Since the disk medium 22 continues to rotate during that time, the head 21 moves relative to the disk medium as shown by within this seek time. After that, the relative movement indicated by is performed by the rotation of the disk medium, and the head 21 is positioned at an appropriate position for reading the data stored in the sector 24. That is, in order to calculate the waiting time, which is the time required for the head and the disk medium to have an appropriate positional relationship for reading predetermined data, the time required for moving and the time required for It suffices to calculate and obtain the sum of those times.

In the disk device used in the multiplexed disk device of the embodiment, this calculation is executed according to the following procedure.
This disk device is a device that specifies a storage area to be accessed by a logical address, and the specified logical address is a surface (head) within the disk device.
Is converted into information specifying track, track number, and sector number. Therefore, the waiting time calculation process is also started by receiving the waiting time calculation instruction information including the information indicating the waiting time calculation instruction and the logical address of the sector for which the waiting time is calculated. Configured.

FIG. 4 shows a flow of waiting time calculation processing in the disk device. The CPU 18 instructed to calculate the waiting time by the controller 11 obtains the track number and the sector number from the waiting time calculation instruction information (step S101), and based on the sector number N _{se ct} , K _D = α × N _sect
The rotation angle K _D at which the head of the sector exists is calculated using the relational expression + β (step S102). It should be noted that the parameters α and β are predetermined so that the angle measured from the angle reference 25 at the head position of the sector identified by the sector number N _sect can be obtained by the calculation α × N _sect + β. It is a parameter. next,
The counter value K _H of the rotation angle counter is read in order to recognize the position of the head at the time when the waiting time calculation process is received (step S103).

Then, based on the track number obtained in step S101 and the track number at which the head is currently stored, which is stored in the memory 19, the time T _SEEK required to seek the head to the requested track is obtained. Is calculated (step S104). In this disk device, since the moving speed profile of the head is changed according to the seek distance, there is no simple proportional relationship between the moving distance of the head and the moving time. Therefore, in step S104, the seek time T _SEEK is calculated using an approximate expression (polynomial of moving distance) for calculating the time required for moving from the number of moving tracks.

CPU that has calculated the seek time T _SEEK
18, head 21, the relative positional relationship between the head and the disk medium when positioning on the track 23 ₂ is seek (i.e., the head 21 in FIG. 3 '
The following calculation is performed to obtain the rotation angle K _H ′ of First, the angle K _SEEK at which the disk medium rotates within the seek time T _SEEK is calculated from the relational expression K _SEEK = T _SEEK × ω (step S105). Here, ω is the rotational angular velocity (degrees / second) of the disk medium. Then, the calculated value K _SEEK is added to the rotation angle K _H read in step S102 and stored as K _H ′ (step S106).

Then, a value K _DELTA is calculated by subtracting the rotation angle K _H ′ from the rotation angle K _D (step S107).
To do. When the rotation angle difference K _DELTA is equal to or _smaller than "0" (step S108; N), 360 is added to K _DELTA (step S109), and the process returns to step S108.
When K _DELTA becomes larger than “0” (step S1
08; Y), the time T _ROT required for the head to move based on the rotation angle difference K _DELTA and the rotation angular velocity ω.
_Is calculated using the relational expression of T _ROT = K _DELTA / ω (step S110). Then, the value obtained by adding T _SEEK and T _ROT is output as the waiting time (step S111).

The waiting time can be calculated by the following procedure.

FIG. 5 shows the flow of the second waiting time calculation process. The operations of steps S201 to S204 are the same as steps S101 to S104 of FIG. 3, respectively, and thus the description thereof will be omitted. The CPU 18, which has calculated the time T _SEEK required to seek the head to the requested track, uses the parameter T _SEEK.
ROT is calculated based on the relational expression T _ROT = (K _D −K _H ) / ω (step S205). Parameter T _ROT calculated here, unlike T _ROT calculated in step S110 in FIG. 4, the rotation angle K _H, is the magnitude relation of K _D, are those which can also take negative values.

Next, the CPU 18 compares the magnitude relationship between T _ROT and T _SEEK (step S206). T _ROT ＞
When T _SEEK is not established (N), the parameter _{TROT is set.}
Is added to the time T ₀ required for one rotation of the disk medium (step S207), and the process returns to step S206. Then, when T _ROT > T _SEEK is satisfied (step S20)
6; Y), the _TROT is output to the controller as a waiting time (step S207), and the process is ended. next,
The principle of waiting time calculation by this processing will be described.

The parameter T _ROT calculated in step S205 is the relational expression "T _ROT = (K _D
As is clear from −K _H ) / ω ″, these are parameters determined by the positional relationship between the head and the sector whose waiting time is calculated when the waiting time calculation instruction is received. For example, they are shown in FIG. In the case of such a positional relationship, the value is the time until the head 21 reaches the broken line shown at the beginning of the sector 24. The sector whose waiting time is calculated is the head. If it exists between 21 and the angle reference 25, the parameter T _ROT becomes a negative value, and its absolute value is the same as the previous (past) head 2
1 indicates the elapsed time from the time existing on the above-mentioned broken line.

Since the disk medium is rotating at a constant speed in the cycle T ₀ , the head (“n _ROT + n × T ₀ ” seconds after receiving the instruction for calculating the waiting time (however, n
Is a natural number), and will be located on this broken line. Therefore, among the n made larger _{"T ROT + n × T 0} " than the time T _SEEK required for the head 21 is positioned over the track 23 _2, do it in search of minimum ones, based on the n ,
The waiting time “T _ROT + n × T ₀ ” can be calculated. The waiting time calculation process shown in FIG. 5 calculates the waiting time based on such an algorithm.

The accuracy of the waiting time calculated by the flow shown in FIGS. 4 and 5 depends on the accuracy of the rotation angle counter which is information for recognizing the position of the head.
Here, since a counter for identifying the rotation angle of the head with respect to the disk medium is used as the rotation angle counter in units of 1 degree, the calculated waiting time is the time required for the disk medium to rotate 1/360 times (this Since the disk device rotates at 3600 revolutions / minute, the time includes an error of about 46 μs). If the resolution of the rotation angle counter is increased, the accuracy of the waiting time calculated accordingly can be improved, but as will be described below, the calculated waiting time is used for comparison of magnitude relationships. , This level of accuracy is sufficient.

FIG. 6 shows the flow of the overall operation of the multiplexed disk device of the embodiment. The controller 11 that receives an access request from an external device such as a host computer
Judges whether the request is a read request or a write request (step S301), and if it is not a read request (N), executes a write request to all disk devices (step S308). After that, the operation ends (returns to the access request acceptance state).

When the received request is a read request (step S301; Y), waiting time calculation instruction information including address information equal to the address information included in the read request is created (step S302). The waiting time calculation instruction information is sent to each disk device 12 _A to 1
It outputs to 2 _C simultaneously (step S303). Receiving this instruction, each disk device starts calculating the waiting time according to the above-mentioned flow.

Upon receiving the waiting time calculation results from all the disk devices (step S304; Y), the controller 11 identifies the disk device to be used for reading based on the obtained waiting time (step S305). In step S304, a time-out determination is also performed, and after the time T _OUT has passed, even if there is a disk device that does not output the waiting time calculation result, execution of step S305 is started. .

In step S305, if there is a disk device that outputs a waiting time shorter than the predetermined time T _MIN , the waiting time in the disk device is the rotation period T ₀ of the disk medium. As a result, the comparison of waiting time magnitudes is performed. And
The disk device that outputs the shortest waiting time is specified as the disk device used for reading. The CPU 18, which has specified the disk device used for reading, issues a read request received from the external device to the disk device (step S306), and transfers the read data to the external device (step S307). And ends the operation.

The physical meaning of the times T _OUT and T _MIN used by the CPU 18 as a criterion will be described with reference to FIG. 7. This figure shows the timing of the operation from step S302 to step S306 of the flowchart shown in FIG. Although the description will be duplicated, first, a brief description of this figure will be given. The controller receiving the data read request 41 from the external device performs a waiting time calculation instruction information creation process 42 based on the read request, and then simultaneously waits time calculation instruction 43 _A to all disk devices. Or outputs 43 _C. Each disk device receives the instruction and calculates the waiting time 44
And the waiting time calculation results 45 _A to 4 respectively.
Returns 5 _C to the controller. The controller performs a specific process 46 of the disk device used for reading the data based on the obtained waiting time, for example, a read request 47 (the same request as the read request 41) to the disk device A.
Is output.

Although this series of operations is completed in a short time, for example, from the time when the disk device A starts calculating the waiting time to the time when it is specified as a device used for reading and the read request is received. , There is a minimum time required. If this time is T _SLCT , and if the specified disk device is a device that has calculated a waiting time shorter than the time T _SLCT , when the controller outputs a read request to that disk device, the head is already Has passed over the target sector.

T used in step S305 in FIG.
_MIN is a parameter for preventing such a case from occurring, and is set to a value _satisfying T _MIN > T _SLCT . In addition, even if the response to the waiting time calculation instruction is delayed for some reason, the calculated waiting time may differ greatly from the actual waiting time. The parameter provided to deal with this is T _OUT
Is. In the multiplex disk device of the embodiment, T _SLCT is several μs (microseconds), but T _OUT is 40 μ in consideration of the calculation error of the waiting time of about 50 μs.
s and T _MIN are set to 50 μs.

The device may be configured so that the waiting time in the order of μs is not calculated when the waiting time is calculated inside the disk device, and the waiting time replacement process is not performed in the specific process of step S305. .

The multiplexed disk device of the embodiment has a structure in which commands and data can be transferred in parallel to each disk device, but the connection between the controller and the disk device is daisy as shown in FIG. A chain method may be used. In this case, the waiting time calculation instruction to each disk device is performed in time series, and the waiting time calculated from each disk device becomes a time predicted based on different times. For this reason, in order to perform comparison on the same basis, it is necessary to correct the time difference when the waiting time calculation instruction is output. However, since this time difference is on the order of μs, the response time obtained by the correction is shortened. The effect is at most a few μs, and in practice
No correction is necessary.

Further, in the disk device used in the embodiment,
A rotation angle counter and a clock generator are provided, and when the servo head detects an index signal, the count value is reset to detect the rotation angle of the disk medium. It is not limited to this. For example, in the case of applying a device using a Hall element to the rotation detection of a spindle motor, the rotation angle may be detected from the Hall output, or the sector information written on the disk medium may be detected. May be read and the rotation angle of the disk medium may be recognized in sector units.

Modification

As described above, in the multiplexed disk device, the head of each disk device is located on the same track immediately after responding to the data write request. Therefore, even in the multiplexed disk device of the embodiment, the head of each disk device is located on the same track immediately after responding to the write request, which is inappropriate from the viewpoint of shortening the seek time. Head placement.

In the multiplex disk device of the modified example, after the write request is executed, the head is returned to the track where the head existed before the write request was executed. The basic device configuration of the multiplex disk device of the modified example is the same as that shown in FIGS. 1 and 2, and therefore the description of the configuration is omitted.
In this multiplexed disk device, a track information storage area, which is an area for storing the track number where the head exists in each disk device, is secured in the memory provided inside the controller, and this is used to It responds to the access request from the external device as follows.

FIG. 9 shows the flow of the overall operation of the multiplex disk device of the modification. Steps S401 to S4
The operation of 08 is performed in steps S301 to S308 of FIG.
Since each is the same as the above, the description will be omitted. In the multiplexed disk device of the modified example, if the received request is a read request (step S401; Y), after controlling the disk device with the shortest waiting time and responding to the read request (steps S402 to S407). , The track number corresponding to the disk device in the track information storage area is rewritten with the read track number (step S411).

If the received request is a write request (step S401; N), the write request is executed for all the disk devices (step S408), and then the heads of the respective disk devices are respectively subjected to track information. Move to the track stored in the storage unit (step S
411).

By such an operation, in the multiplexed disk device of the modified example, the heads of the respective disk devices are arranged on different tracks immediately after responding to the write request, and even immediately after the write request is executed, It is possible to prevent the average response speed to the read request from decreasing.

[0049]

As described above, according to the first and second aspects of the present invention, the disk device for multiplexing and storing data is provided with a waiting time calculation function, and each of them is calculated. Based on the waiting time, the earliest disk device that can be read is specified, and only that disk device is controlled to respond to the read request. For this reason,
Since the waiting time including the rotation waiting time can be calculated, it is possible to accurately specify the disk device that can read the earliest. Further, since only one disk device is controlled to respond to the read request, the heads of the respective disk devices are located on different tracks from each other, which increases the probability that the head is located near the data to be read. The average response time for read requests can be shortened. Further, since the disk device is provided with the waiting time calculating function, it is not necessary to change the controller even when connecting the disk device having different performances such as the rotation speed and the seek speed.

Further, according to the second aspect of the invention, after the write request is responded, the head arrangement immediately before that is returned, so that the average response speed to the read request is not reduced immediately after the write request is executed. it can.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing an outline of a multiplexing disk device according to an embodiment of the present invention.

FIG. 2 is a configuration diagram showing an outline of a waiting time detection circuit in the disk device used in the multiplexed disk device of the embodiment.

FIG. 3 is an explanatory diagram schematically showing how the relative positional relationship between the head and the disk medium changes during a data read operation.

FIG. 4 is a flow chart showing a flow of waiting time calculation processing of the disk device used in the multiplexed disk device of the embodiment.

FIG. 5 is a flowchart showing the flow of a second waiting time calculation process of the disk device used in the multiplexed disk device of the embodiment.

FIG. 6 is a flow chart showing the overall operation flow of the multiplexed disk device of the embodiment.

FIG. 7 is a timing chart showing the timing of response operation to a read request of the multiplexed disk device of the embodiment.

FIG. 8 is a configuration diagram showing an outline of a multiplexed disk device to which the present invention can be applied and which is a daisy chain connection of disk devices.

FIG. 9 is a flowchart showing the overall flow of operations of the multiplex disk device of the modification.

FIG. 10 is a configuration diagram showing an outline of a multiplexed disk device that outputs a read request to all disk devices according to a conventional example.

FIG. 11 is a configuration diagram showing an outline of a multiplexing disk device that calculates a time for responding to a read request according to a conventional example based on track information.

[Explanation of symbols]

11 ... Controller, 12, 34 ... Disk device, 13
... disk medium (servo surface), 14 ... head, 15 ... index signal detection circuit, 16 ... rotation angle counter, 1
7 ... Clock generator, 18 ... CPU, 19 ... Memory, 20 ... Interface, 21 ... (Read / write)
Head, 22 ... Disk medium (data surface), 23 ... Track, 24 ... Sector, 25 ... Angle reference, 31 ... External device, 32 ... Interface control circuit, 33 ... Main control circuit, 35 ... Judgment circuit, 36 ... Data buffer , 37 ... Switching circuit, 38 ... Control device, 39 ... Storage unit, 41, 47 ...
Read request, 42 ... Wait time instruction information creation process, 43
... waiting time calculation instruction, 44 ... waiting time calculation processing, 45 ...
Waiting time calculation result, 46 ... Specific processing

Claims (2)

[Claims] 1. A rotation angle detecting means for detecting a rotation angle of a disk medium, a track number detecting means for detecting a track number of a track on which a head is located, and the rotation when a waiting time is instructed. Using the rotation angle detected by the angle detection means and the track number detected by the track number detection means, the waiting time, which is the time required until the head moves to the storage area whose waiting time is to be calculated, is calculated. A plurality of disk devices each having an output unit that outputs the output, a determination unit that determines whether the received request is a read request or a write request, and the determination unit determines that the request is a read request An instruction means for instructing each of the plurality of disk devices to calculate a waiting time, and each of the plurality of disk devices according to the instruction of the instruction means. Specifying means for specifying one disk device which has calculated the shortest waiting time by comparing the output waiting times, and response means for controlling the disk device specified by the specifying means and responding to the read request, And a writing unit that writes the same data to all of the plurality of disk devices by outputting the write request to each of the plurality of disk devices when the determination unit determines that the request is a write request. Multiplexed disk device characterized by. 2. A rotation angle detecting means for detecting a rotation angle of a disk medium, a track number detecting means for detecting a track number of a track on which the head is located, and the rotation when a waiting time is instructed. Using the rotation angle detected by the angle detection means and the track number detected by the track number detection means, the waiting time, which is the time required until the head moves to the storage area whose waiting time is to be calculated, is calculated. A plurality of disk devices each having an output means for outputting the data, a storage means for storing the track number of the track where the head is located in each disk device, and whether the received request is a read request. Determining means for determining whether it is a write request, and each of the plurality of disk devices when the determining means determines that it is a read request An instruction means for instructing the calculation of the waiting time, and one disk device for calculating the shortest waiting time by comparing the waiting times output by the plurality of disk devices according to the instruction of the instructing means. Specifying means for specifying, response means for controlling the disk device specified by the specifying means to respond to the read request, and for the specified disk device stored in the storage means for each response by the response means A rewriting unit that rewrites the corresponding track number with the read track number, and outputs the write request to each of the plurality of disk devices when the determination unit determines that it is a write request. Writing means for writing the same data to all of the disk devices, and after the writing of data by the writing means is completed Each multiplexer disk apparatus, wherein a head of the disk apparatus and a moving means for moving on a track designated by the track number stored in said storage means.