JPH0675717A - Read error recovery system for hard disk - Google Patents

Abstract

PURPOSE:To eliminate the deterioration with the secular change of a hard disk and also to eliminate the read errors due to the shift caused between a head and a magnetic disk by performing a rereading operation against the occurrence of a read error, judging a defective area when the rereading frequency exceeds a prescribed level, and replacing the defective area with another nondefective area. CONSTITUTION:The data are read out of a hard disk 4 and the presence or absence of a read error is checked in an information processing system. If a read error is confirmed, 1 is added to a constant RETORY and this RETORY is compared with the retry frequency N1. At the same time, the retry frequency N2 having the relation of N1>N2>0 is set in addition to the frequency N1. When a normal reading operation is carried out before the rereading frequency reaches the frequency N2, the read data are written in the same area. Meanwhile a defective area is decided when the rereading frequency exceeds the N2. Then the defective area is replaced with another nondefective area.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a read error recovery system for a hard disk used in an information processing device such as a workstation.

[0002]

2. Description of the Related Art Conventionally, when a read error frequently occurs in a hard disk used in an information processing apparatus such as a workstation, the hard disk is determined to have failed and replaced.

FIG. 3 shows a conventional read control flow for a hard disk. First, a constant RETRY is initialized to 0 before reading from the hard disk (step 30). After that, data is read from the hard disk (step 31), and it is checked whether or not a read error such as a CRC error has occurred (step 32). As a result, if it is normal, the process proceeds to the next process, and if it is an error, the constant RETRY is incremented by 1 (step 33), and the constant RE is returned.
TRY is compared with a constant N1 indicating the number of retries (step 34), and when the number of retries N1 is not reached, reading of the hard disk is retried. As a result of retries up to the number of retries N1, a read error still occurs. Sometimes read error processing is performed (step 35), an error message is output, and a hard disk replacement is instructed.

Therefore, according to this control flow, if an error occurs N1 times or more in the read processing of the hard disk, the hard disk will be replaced.

[0005]

However, in a hard disk, a read error may occur due to deterioration of magnetism due to aging and a slight deviation between the head and the magnetic disk. In many cases, it is recoverable. Therefore, according to the above read control, even a recoverable read error is replaced with a new one, which is extremely uneconomical.

On the other hand, as a method of preventing a decrease in system reliability due to frequent read errors in a hard disk, by duplicating the hard disk device by a mirror disk system or a duplex system,
Some systems are configured to maintain reliability.

However, if the hard disk device is duplicated, the system becomes extremely expensive, and there is a problem that it cannot be easily adopted.

An object of the present invention is to provide a hard disk drive with
It is an object of the present invention to provide a read error recovery method for a hard disk drive that is highly economical and that can ensure reliability when a read error occurs without duplication.

[0009]

In order to achieve the above object, the present invention performs re-reading when a read error occurs,
When the normal reading is performed before the number of times of re-reading reaches a predetermined number, the read data is written in the same area, and when the number of times of re-reading reaches a predetermined number or more, it is determined as a defective area, and other The replacement process is performed to replace the area.

[0010]

According to the above means, when a CRC error or the like occurs during data reading from the hard disk, re-reading is performed, and when normal reading is performed before the number of times of re-reading reaches a predetermined number, reading is performed. Since the data is written in the same area, it is possible to eliminate the read error caused by the deterioration of magnetism due to aging and the slight deviation between the head and the magnetic disk.

Further, when the number of times of re-reading exceeds a predetermined number of times, the defective area is replaced with another area, so that it is not necessary to replace the hard disk, and the normal area is effectively utilized to make the system economical. Various operations are possible.

[0012]

The present invention will be described in detail below with reference to the illustrated embodiments.

FIG. 1 is a block diagram showing an embodiment of an information processing system to which the present invention is applied. A CPU 2, a memory device 3, a hard disk 4, a floppy disk 5, and a communication device 6 are connected via a system bus 1. , And is configured to be able to transmit information to another information processing system through the line 7.

The CPU 2 operates based on a program stored in the memory device 3, and the hard disk 4, floppy disk 5, and communication device 6 operate based on instructions from the CPU 2.

FIG. 2 shows a read control flow of the hard disk 4 in this system. In addition to the conventional retry count N1, a retry count N2 having a relationship of N1>N2> 0 is provided. When reading is performed and normal reading is performed before the number of re-reads reaches the number of retries N2, the read data is written to the same area, and when the number of re-reads becomes N2 or more retries, the defective area It is characterized in that the replacement process is performed so that the replacement process is performed to replace the area.

That is, first, the constant RETRY is initialized to 0 before reading from the hard disk 4 (step 10). Then, data is read from the hard disk 4 (step 11), and it is checked whether or not a read error such as a CRC error has occurred (step 12). As a result, if normal, the process proceeds to step 16.

If there is an error, the constant RETRY is incremented by 1 (step 13), the constant RETRY is compared with the number of retries N1 (step 14), and when the number of retries N1 is not reached, the reading of the hard disk 4 is retried. As a result of the retry up to N1 times, if a read error still occurs, read error processing is performed (step 15), an error message is output, and a hard disk replacement is instructed.

On the other hand, as a result of checking whether or not a read error such as a CRC error has occurred, if no error has occurred, it is checked whether RETRY = 0 (step 1
6) If RETRY = 0, it means that normal reading has been performed once, so that the process proceeds to the next control.

However, if RETRY = 0 is not satisfied, it means that the normal read operation was performed in the reread operation. Therefore, the constant RETRY is compared with the retry count N2 (step 17), and the retry count N2 is not reached. When this happens, write the read data again in the read area (step 1
8). However, when the number of retries is N2 or more, a replacement process of replacing the read area of the hard disk 4 at this time with another area is performed (step 19).

[0020]

As described above, according to the present invention, when a CRC error or the like occurs when reading data from a hard disk, rereading is performed, and normal reading is performed before the number of rereading reaches a predetermined number. In this case, since the read data is written in the same area, it is possible to eliminate a read error caused by deterioration of magnetism due to aging and a slight deviation between the head and the magnetic disk.

Further, when the number of times of re-reading exceeds a predetermined number of times, the defective area is replaced with another area, so that it is not necessary to replace the hard disk, and the normal area is effectively utilized to make the system economical. Various operations are possible.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of an information processing system to which the present invention is applied.

FIG. 2 is a read control flow chart of a hard disk in the embodiment of FIG.

FIG. 3 is a flow chart of a read control of a conventional hard disk.

[Explanation of symbols]

1 ... System bus, 2 ... CPU, 3 ... Memory device, 4 ...
Hard disk, 5 ... Floppy disk, 6 ... Communication device, 7 ... Line.

Claims (1)

[Claims] 1. When a read error occurs when reading data from a hard disk, re-reading is performed, and when normal reading is performed before the number of times of re-reading reaches a predetermined number, the read data is written to the same area. A read error recovery method for a hard disk, characterized in that when the number of times of writing and re-reading exceeds a predetermined number of times, it is judged as a defective area, and replacement processing is performed to replace with another area.