JPH05224830A - Magnetic disk device - Google Patents

Abstract

PURPOSE:To prevent a decrease in system performance due to a job entry by reducing the load on the whole system when many jobs are entered for a medium test wherein a read is made by varying read timing are entered into the magnetic disk device. CONSTITUTION:The magnetic disk device consists of an access state control part 1 which confirms that there is no access from a host device, a medium part 8 storing with data, a marginal read control part 5 which controls for varying the read timing of the medium part 8, a data buffer 2 storing track error from the medium part 8, a check circuit 3 which checks a data error of transfer data of the data buffer, a buffer control part 7 which controls a series of data transfer operations, an error control part 4 which logs error information on the track where the data error occurs, and a defect predictive reporting part 6 which indicates the need to replace the medium to the host device when the error occurrence frequency is high.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data processing system in a magnetic disk device in an electronic computer system, and more particularly to a data maintenance system.

[0002]

2. Description of the Related Art In a conventional magnetic disk subsystem,
In order to determine whether the medium needs to be replaced due to deterioration or failure of the data reading characteristics of the magnetic disk device, data is read by shifting the magnetic disk read timing by software, and a read data error occurs. Depending on the frequency, maintenance personnel decided to replace the medium.

[0003]

In this conventional magnetic disk subsystem, since a job for reading is performed by software by shifting the read timing of the magnetic disk device, many magnetic disk devices are subject to the processing. However, if other processing is performed at the same time as the above, there is a problem in that the load on the entire system increases and the system performance remarkably decreases.

An object of the present invention is to reduce the load on the entire system when a medium test job for reading by changing the read timing of the magnetic disk drive is applied to a large number of magnetic disk drives and to reduce the system performance due to the job input. It is to provide a magnetic disk device for preventing.

[0005]

SUMMARY OF THE INVENTION A magnetic disk device of the present invention comprises a data buffer for storing track data on a magnetic disk, a means for checking erroneous data read from a medium, and an error. Means for managing the error history of the track in which the error occurred, means for managing the access status from the host device, means for controlling the read operation to the data buffer, and change the read timing during the read operation of the data from the medium. It is provided with means for controlling the read operation in this state, and means for reporting a data error occurrence state and a notice of medium failure to a higher-level device.

[0006]

The present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of the present invention. The access status management unit 1, data buffer 2, check circuit 3, error management unit 4, marginal read control unit 5, defect prediction reporting unit 6, It is composed of a buffer control unit 7 and a medium unit 8. First, the access status management unit 1 has a timer for a specified time, and instructs the buffer control unit 7 to perform a read operation if the next access does not come within the specified time from the last access from the host device. The buffer control unit 7, which has received the read operation instruction, instructs the marginal read control unit 5 to set the marginal read mode. Under the control of the marginal read control unit 5 which receives the marginal read mode setting instruction, the medium unit 8 shifts the read timing and reads the data. It is set with the read margin changed. When this setting is completed, the marginal read control unit 5 reports to the error management unit 4 that the setting of the marginal read mode is completed. Upon receiving the report, the error management unit 4 instructs the buffer control unit 7 to read the track T ₁ to the medium unit 8. The buffer control unit 7 reads the data of T ₁ into the data buffer 2 and at the same time, checks the data by checking circuit 3.
Instruct. If no data error occurs, the check circuit 3 reports to the error management unit 4 that there is no error. Upon receiving the report, the error management unit 4 instructs the buffer control unit 7 on the track T ₂ whose track number is incremented by the specified pitch.

Here, at the time of track T ₂ lead check,
When a data error is detected, the check circuit 3 reports to the error management unit 4 that an error has occurred in the track T ₂ . The error management unit 4 increments and holds the number of read error occurrences on the track T ₂ and at the same time checks whether or not an error has occurred more than a specified number of times.
If it has not exceeded the specified number of times, the buffer controller 7 is again instructed to read-check the track T ₂ .
If it exceeds the specified circuit, the error management unit 4 reports the track number T ₂ to the defect prediction report unit 6 to inform that the track T ₂ is likely to be defective. The defect prediction reporting unit 6 holds the number of a track that has a small read margin and is likely to become a defective track.
When the number of tracks exceeds the specified number, the medium unit 8 is notified to the host device that it is defective, and at the same time, the data of the data error occurrence frequency for each track held in the error management unit 4 is transferred to the host device. From the defect prediction report and the data error occurrence frequency for each track, it can be recognized that the medium unit 8 is likely to be defective in the near future, and the medium unit 8 needs to be replaced promptly.

If the number of tracks that are likely to become defective tracks to the defect prediction reporting unit 6 has not reached the specified number of tracks, the error management unit 4 performs a read check of the next track T _n . The buffer control unit 7 is instructed to perform. In this way, the read check is continued while the access status management unit 1 does not detect the access from the higher-level device, but when the access status management unit 1 detects the access from the higher-level device, the access status management unit causes the data buffer 2, At the same time as resetting all the operations of the check circuit 3, the marginal read control unit 5 is instructed to return the read timing of the medium unit 8 to the normal state. Medium part 8
Reports to the buffer control unit 7 when it returns to the normal state. After that, the buffer control unit 7 can perform a normal operation according to an instruction from the host device. However, the defect prediction reporting unit 6 and the error management unit 4 are not reset.

As described above, the disk can be exchanged in advance by knowing the defect prediction report and the frequency of occurrence of data error for each track when the read timing is shifted, and the medium section 8 may be deteriorated or defective. It is possible to prevent a fatal obstacle from occurring.

[0011]

As described above, the present invention utilizes the idle time of the magnetic disk device which is not accessed by the host device, and the magnetic disk alone automatically and automatically shifts the read timing. By checking the data on the upper track, a track with a small read margin where data errors frequently occur is detected, and it is possible to report to the upper rank together with the frequency of data error occurrence on other tracks at that time, so in the near future There is a high possibility that the disc will be defective, and you can recognize that it needs to be replaced immediately,
It is possible to prevent fatal failures caused by deterioration and deterioration of the disk. Further, since it is not necessary to input a job from software, the load on the entire system can be reduced. Further, there is also an effect that it is possible to prevent an obstacle due to the accumulation of dust on the head.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention.

[Explanation of symbols]

 1 access status management unit 2 data buffer 3 check circuit 4 error management unit 5 marginal read control unit 6 defect prediction reporting unit 7 buffer control unit 8 medium unit

Claims (1)

[Claims] 1. A data buffer for storing track data on a magnetic disk, means for checking an error in data read from a medium, and means for managing an error history of a track in which an error has occurred. , A means for managing the access status from the host device, a means for controlling the read operation to the data buffer, and a means for controlling the read operation while changing the read timing when the data is read from the medium And a means for reporting a data error occurrence status and a notice of medium failure to a higher-level device.