JPH0527909A - Head position correcting method for disk array - Google Patents

Abstract

PURPOSE:To prevent the processing efficiency of a host device from being lowered in the case of correcting head position deviation among plural disk device concerning a disk array CONSTITUTION:A disk array 9 is composed of plural disk devices 2-4 to be connected to a host device 1 by buses 6-8, to synchronize the rotations of the respective disks according to a synchronizing signal applied from the disk device 2, and to simultaneously execute write or read from the host device 1 by executing the positioning control of the heads based on servo information recorded on the servo plane of the disk. Reset signals are simultaneously transmitted from the host device 1 through the buses 6-8 to all the disk devices 2-4 after the power sources of all the disk devices 2-4 are turned on, all counting means 5 are reset at the same time and therefore, the correction processing execution of the head position deviation is simultaneously started.

Description

Detailed Description of the Invention

[0001]

The present invention relates to SCSI (Small Compute).
r System Interface) and is used for writing / reading to / from multiple disk devices from a host device at the same time, especially when correcting head position deviations due to environmental changes of the multiple disk devices. The present invention relates to a disk array head position correction method for preventing a decrease in processing efficiency of a host device.

In recent years, computer systems have become smaller, faster, and have larger capacities. Therefore, there is a demand for larger capacity and higher speed of a disk device as an external storage device connected to the computer system.

Therefore, for a plurality of disk devices,
A disk array has become popular, which enables high speed and large capacity by simultaneously performing writing / reading. However, accessing a plurality of disk devices at the same time is a heavy burden for small computers.

Therefore, in order to reduce the load, a common interface called SCSI, which has a high degree of freedom, is used as an interface between a plurality of disk devices and a small computer, and the disk rotation of the plurality of disk devices is performed. By synchronizing the disk arrays, a simplified disk array having the same effect as the disk array has come to be used.

By the way, the disk device uses the servo surface servo system which can perform a seek operation at high speed to obtain the position information of the head. However, as the storage capacity increases, the track density increases and the head arm changes due to temperature change. Deformation has become a problem.

That is, although the plurality of heads are attached to the tip of the head arm of the carriage, even if the head arm deforms due to temperature changes and the servo head follows the center of the track, A phenomenon occurs in which the head shifts from the center of the track with the passage of time.

Therefore, when writing data, the adjacent tracks having a narrow interval are affected, and when reading data, a problem arises in that noise increases as the reading level of the data head decreases.

Therefore, each time a predetermined time elapses, the servo information for correction written in the specific area of the disk is read, the amount of positional deviation of the data head is measured, and this deviation is corrected. A process called calibration is required.

However, if there are variations in the timing of performing this calibration in a plurality of disk devices, the processing efficiency of the computer of the host device will be reduced, so it is necessary to perform them at the same time.

[0010]

2. Description of the Related Art In a conventional simplified disk array, one of the disk devices operates as a master and the other disk device operates as a slave, and when the index recorded on the disk is read from the master, it is sent out. The disk rotation of the slave is synchronized based on the index signal.

The calibration is executed every time a timer provided in each disk device counts a predetermined time. However, when power is supplied to multiple disk devices at the same time when power is applied to the disk array, the spindle motors that rotate the disks are supplied with current all at once, causing an excessive current to flow. Power is being supplied.

Therefore, since there is a difference in the starting time of the timers of the disk devices, the timings at which the timers have finished counting the predetermined time do not match, and the calibration can be performed all at once due to the time lag. Because there is no
The timing of calibration is different for each disk device.

[0013]

As described above, conventionally, the execution timing of calibration is different for each disk device.

By the way, the disk device corrects the positional deviation of the head during the calibration, and is not suitable for writing and reading data.
Therefore, during this processing period, to wait for access from the host device, if there is a disk device that is undergoing calibration among multiple disk devices, the host computer can access the disk array at the same time. Absent.

Therefore, for a host computer that writes or reads data to and from a plurality of disk devices at the same time, the processing is kept waiting while at least one disk device is calibrating, and processing efficiency is improved. There is a problem of a large drop.

In view of such a problem, the present invention aims to improve the processing efficiency of a computer system by reducing the chances of waiting for the access of the host computer by executing the calibration all at once.

[0017]

FIG. 1 is a block diagram for explaining the principle of the present invention. The disk arrays 9 are respectively connected to one host device 1 by independent buses 6 to 8,
The rotation of each disk is synchronized by a synchronization signal given from one disk device 2, and the positioning control of the head is performed based on the servo information recorded on the servo surface of this disk, and at the same time, the head is moved at a predetermined interval. A plurality of disk devices 2 that simultaneously execute writing or reading from the host device 1 by correcting the positional deviation
It is composed of 4.

Then, after being reset by a reset signal given from the buses 6 to 8, every time the synchronization signal is counted and reaches a predetermined value, a counting means for instructing execution of correction processing of the positional deviation of the head. 5 are provided in all of the disk devices 2 to 4 that compose the disk array 9, respectively.

After all the disk devices 2 to 4 constituting the disk array 9 are powered on, a reset signal is simultaneously sent from the host device 1 to all the disk devices 2 to 4 via the buses 6 to 8. Is sent to reset all the counting means 5 at the same time, so that the correction processing of the positional deviation of the head is simultaneously started.

[0020]

With the above-mentioned structure, when the power supply to the disk array 9 is completed, the counting means 5 for counting a predetermined value is reset based on the synchronizing signal for synchronizing the disk rotation. It is possible to make the counting start timings of all the counting units 5 to 4 coincide.

Since the counting operation of the counting means 5 after the resetting is synchronized, the timing of finishing counting the predetermined value is also synchronized, and the head misalignment correction is started at the same time for the plurality of disk devices 2. ~ 4 can be executed.

Therefore, the chance that the access of the host computer 1 to the disk array 9 is kept waiting is reduced, and the processing efficiency of the computer system can be improved.

[0023]

1 is a block diagram of a circuit showing an embodiment of the present invention. When the power is turned on, the processor 10 of the disk devices 2 to 4 shown in FIG. 1 reads the program stored in the ROM 11 to operate, and sends a reset signal to the counter 18 via the OR circuit 19 to reset it. At the same time, the rotation control circuit 15 is controlled to rotate the motor 16 and rotate the disk 17.

When the processor 10 is designated to operate as a master, the processor 10 instructs the servo circuit 14 to rotate the motor 16 so that each time the servo head 20 reads an index pulse from the disk 17, a terminal is output. C
The rotation control signal is transmitted from the rotation control circuit 15 and the rotation synchronization signal input from the terminal B is counted by the counter 18.
Do not use it.

When instructed to operate as a slave, the counter 18 is made to count the rotation synchronizing signal input from the terminal B, and the rotation control circuit 15 is used as the synchronizing signal to synchronize with the rotation of the master motor 16. Motor 16
To rotate.

Therefore, since the slave disk 17 rotates in synchronization with the master disk 17, the slave servo head 20 can read the index pulse at the same timing as the master servo head 20.

When all the disk devices in the disk array are powered on, the host device supplies a reset signal from the terminal A to the OR circuit 19 via the SCSI bus to reset the counter 18.

Therefore, after the counter 18 is reset by the host device, it counts the synchronization signals input from the terminal B, and when it reaches a predetermined value, it returns the count value to the initial value and sends an interrupt signal to the processor 10. Then, it notifies that it is the time to start the calibration.

The processor 10 instructed the timing of starting the calibration controls the write / read circuit 13 to cause the data head 21 to read the servo information recorded in the specific area of the disk 17, and sends it to the servo circuit 14. Then, the positional deviation of the data head 21 is corrected.

As described above, the counter 18 repeatedly notifies the processor 10 of the calibration start timing at the same timing every time the synchronization signal is counted by a predetermined value, so that the master and the slave simultaneously perform the calibration each time. It is possible to start.

Further, the processor 10 executes an input / output instruction from a host device which enters via the interface circuit 12 from the SCSI bus connected to the terminal A. That is, when the servo circuit 14 is controlled to position the servo head 20 in the designated cylinder of the disk 17, when writing of data is instructed, the data head 21 is moved to the designated data head 21 via the write / read circuit 13. Interface circuit 1
When data is read from the specified sector, the data read from the specified sector is transferred to the interface circuit 12 and the terminal A. Then, the data is sent to the SCSI bus.

[0032]

As described above, according to the present invention, when the head position is corrected in accordance with the environmental change of the disk array,
Since the plurality of disk devices perform the calibration at the same timing, the chance of waiting for the access of the host device is reduced, and the processing efficiency can be improved.

[Brief description of drawings]

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

FIG. 2 is a block diagram of a circuit showing an embodiment of the present invention.

[Explanation of symbols]

 1 host device 2-4 disk device 5 counting means 6-8 bus 9 disk array 10 processor 11 ROM 12 interface circuit 13 writing / reading circuit 14 servo circuit 15 rotation synchronizing circuit 16 motor 17 disk 18 counter 19 OR circuit 20 servo head 21 Data head

Claims (1)

Claims 1. A bus (6) independent of one host device (1).
To (8), and synchronize the rotation of each disk by the synchronization signal given from one disk device (2), and control the positioning of the head based on the servo information recorded on the servo surface of the disk. And a plurality of disk devices (2) to (4) that simultaneously perform writing or reading from the host device (1) by correcting the positional deviation of the head at predetermined intervals by the counting means (5). )
A disk array (9) consisting of all the disk devices that make up the disk array (9)
After the power supplies (2) to (4) are turned on, all the disk devices (2) to (8) are transferred from the host device (1) through the buses (6) to (8).
By simultaneously sending a reset signal for resetting the counting means (5) of (4) and resetting all the counting means (5) at the same time, the execution of the correction processing of the positional deviation of the heads can be started at the same time. A head position correction method for a disk array, which is characterized in that