JPH1011906A - Magnetic disk apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To correct all data errors including data errors brought about suddenly upon the application of an impact force, etc., in a magnetic disk apparatus. SOLUTION: A used sector allocation table and a redundant sector table to be controlled by a magnetic disk controller 2 are provided. Redundant data and an objective redundant sector on a magnetic disk based on the tables are arranged at physically different positions in a radial and a circumferential directions. Data sectors and redundant data sectors are set so as not to be damaged at the same time even if an excessive impact force is input, etc. Accordingly, all data errors including data errors brought about suddenly can be corrected.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a magnetic disk drive used for a storage device or the like.

[0002]

2. Description of the Related Art In order to prevent data loss in a magnetic disk drive, a manufacturer first inspects the entire use area on a magnetic disk, and registers the position where an error occurs before shipment, so that the user can start using the apparatus at the beginning of use. Makes the error zero.

However, an error occurs with a certain probability as the user uses it. Generally, when an error occurs when reading data, an attempt is made to repair the data error by a retry operation or error correction data (ECC). Further, as described in Japanese Patent Application Laid-Open No. 07-065313, an error is corrected by shifting the magnetic head slider or erasing a sector on an adjacent track to eliminate magnetic influence from around the sector. You can take a method such as Also, depending on the settings of the host computer of the magnetic disk, a verify operation is performed to check whether data has been correctly written during the data write operation, and it is confirmed that there is no read error at the beginning of the write.

[0004] When an error is detected, the sector concerned is registered as an improper use and a defect is additionally registered, and data is automatically moved to a replacement sector.

[0005]

In the above-mentioned magnetic disk drive, an error that occurs suddenly is repaired. In particular, when an error is caused by an excessive impact force input to the magnetic disk drive, the error physically extends over several cylinders and several sectors. Since the size is large and the data including the error correction data is depressed, error correction cannot be performed by a magnetic disk device used alone other than the array disk system.

An object of the present invention is to provide a magnetic disk drive capable of repairing a data error including a sudden error.

[0007]

In order to achieve the above object, a used sector allocation table and a redundant sector table managed by a magnetic disk controller are provided in a memory of a magnetic disk controller board, and magnetic data is written to written data. Redundant data is prepared on a disk, and if data requested by the basic software is an error, the error is repaired using the redundant data.

When the used sector allocation table and the redundant sector table managed by the magnetic disk controller are provided in the memory, when the unused area decreases, the redundant data is erased to secure the original capacity of the magnetic disk drive, and the unused data is deleted. By newly creating a used sector allocation table, a redundant sector table, and redundant data when the area increases again, the same storage capacity as in the related art can be secured. On the other hand, when the unused sector on the magnetic disk is provided with the used sector allocation table and the redundant sector table, when the unused region decreases, the used sector allocation table and the redundant sector table are used to secure the original magnetic disk device capacity. By partially erasing the redundant data and newly creating the used sector allocation table, the redundant sector table and the redundant data when the unused area increases again, the same storage capacity as before can be secured.

Further, when the used sector allocation table and the redundant sector table or the redundant data are provided in a specific disk area, the magnetic disk apparatus always has redundancy regardless of the ratio of the used area of the magnetic disk. .

[0010]

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

(Embodiment) FIG. 1 shows the configuration of an embodiment of the present invention.

The magnetic disk drive comprises a magnetic disk controller board 2 and a head disk assembly 5 (hereinafter referred to as HD).
A), and a magnetic disk controller board and a personal computer 1 (hereinafter, referred to as PC) are connected. A magnetic disk controller board (hereinafter referred to as HDC) for controlling a positioning mechanism or converting an analog signal from a magnetic head into digital data, a used sector allocation table (hereinafter referred to as USAT), It is composed of a redundant sector table (hereinafter referred to as ECST). HDA
(5) comprises a magnetic disk 6, a magnetic head slider 7, and a magnetic head slider positioning mechanism 8. The recording sector on the magnetic disk is composed of a redundant sector 9 for redundant data and a data sector 10 for data.

When the PC power switch is turned on, power supply to the magnetic disk drive is started, the IC provided in the HDC board starts operating, and the spindle motor provided in the HDA rotates. At the start of the magnetic disk drive, initialization of the positioning mechanism system such as servo system adjustment,
The information in the USAT and ECST nonvolatile memories 3 and 4 provided on the HDC board is prepared in the HDC, and the start-up processing of the magnetic disk device is completed. When the HDC receives a data write command from the PC, it first writes data to a data sector according to the HDC, and then writes redundant data to a redundant sector.

FIG. 2 shows a flowchart for writing data.

For example, a case in which a PC is operated using MS-DOS, which is basic software of Microsoft Corporation, will be described below.

FIG. 3 shows the positional relationship between the redundant sector and the redundant target sector in the magnetic disk drive.

In FIG. 3, the disks are D # 1A-3
B, six surfaces (16-21) are used, and the magnetic head sliders correspond to the six surfaces H # 0-5 corresponding to the respective disk surfaces.
It shows the case where the number is used. Conventionally, a vacant area is searched in a magnetic disk device, a data write position is registered in a file allocation table (hereinafter, referred to as FAT) managed by MS-DOS, and data is written to the position. Here, it is assumed that data has been written to the sector B (12) in FIG. Normally, it is not checked whether or not data has been written correctly, and the writing process ends with this. According to the present invention, after the writing process is completed, when there is no data input / output command from the PC to the magnetic disk device, the written disk surface and the magnetic head slider and their positions, that is, the cylinder number and the sector position are obtained from the FAT. Is registered in USAT managed by HDC.

FIG. 4 shows an example of sector numbering and redundancy exclusion sectors.

FIG. 4 shows an example of the D # 1A plane (17).
The sector numbers are sequentially numbered from a certain position of the cylinder 0 according to the rotation direction, and when one cylinder is completed, the next inner cylinder is moved to the next cylinder and similarly numbered. In the case of the surface B, the rotation direction is physically targeted, and the sector numbers are assigned from left to right, contrary to FIG. Taking the redundancy exclusion area 22 of the sector # 2N-2 as an example, the sectors included in the area of the same size as the magnetic head slider are set as the redundancy exclusion sectors, and # 0, 1, N-2, N, N + 1, 2N-1, 2N,
The target is 11 sectors of 2N + 1, 3N-2, and 3N-1. If the sector numbers are assigned in the same manner on other disk surfaces with the position of # 0 being the same circumferential position, other disks 5
The redundant exclusion area on the plane (16, 18 to 21) is 1
This is 12 sectors, which is obtained by adding # 2N-2 to one sector.

Table 1 shows an example of USAT.

[0021]

[Table 1]

Next, the US written on the magnetic disk
A position other than the sector B replacement exclusion area shown in FIG. 2 from the AT is defined as a redundant sector D (14), and sectors A (11) and B (1
2) Write the redundant data of C (13) to sector D,
Then, it registers with ECST which HDC manages.

Table 2 shows an example of the redundant sector table. For example, 5-2000 for sector A and 3-423 for sector B
1. Assuming that sector C is 3-5800, USAT and ECST are registered as shown in Tables 1 and 2.

[0024]

[Table 2]

The USAT memory and the ECST memory provided on the HDC board are preferably non-volatile memories. However, when a volatile memory is used, the USAT and ECS
T is temporarily written to an unused area on the magnetic disk, and is read into the memory from the magnetic disk when the magnetic disk device is initialized immediately after the PC is turned on, and data is written to or erased from the magnetic disk. Temporary USAT and ECST in memory and on magnetic disk each time
Is updated. The USAT and ECST on the magnetic disk are each written as one file, and each FAT is added to the data FAT. The USAT and ECST data in this file format may or may not be recognizable from the PC.

When it is set on the PC or the basic software (OS) to check whether the data has been correctly written after writing, if the writing is successful, it is as described above. On the other hand, if the writing fails and an automatic replacement sector is set, data is written to another replacement sector. If the position of the replacement sector is managed as the physical position of the sector B, the operation can be performed as described above.

The replacement exclusion area of the sector B is a physical location other than the area where the magnetic head slider exists in the sector B, and the redundant sector D is also the replacement exclusion area of the sector A and the sector C. If the write position of sector B overlaps the replacement exclusion area of sector A or C or D, MS-DO
After writing the sector B once according to S, the HDC rewrites the sector B to a position that does not overlap with the replacement exclusion area of the sector A and the sectors C and D when there is no data input / output command from the PC to the magnetic disk device. Thereafter, the redundant data is written in the redundant sector D, and the USAT and ECST are registered. Alternatively, the sector B is processed by the HDC as being inappropriate for combination with the sectors A and C, and redundant data combined with other sector data is written to another redundant sector to maintain redundancy by combining with another sector, and the USA is used.
Register T and ECST. In this way, the redundant sector (D) and the redundant target sectors (A, B, C) are always located outside the replacement exclusion area. As a result, an excessive impact force is applied to the magnetic disk, and a dent is generated when the magnetic head slider hits the disk surface. The disk surface at the same radius and the same angle is physically recessed, resulting in a data error. But sectors A, B,
Since C and D are physically different from each other, there is practically no error at two or more locations at the same time.
0% can be repaired.

Although an example has been described in which the positions of the redundant sector and the redundant target sector exist on a plurality of disk surfaces, the same effect can be obtained even if the redundant sector and the redundant target sector exist on the same surface.

FIG. 5 shows a flowchart of a read error repair method.

When a read error occurs in sector B,
If the error cannot be repaired by a conventional error repair method such as an error correction code (ECC) or a retry, the redundant sector D of the sector B is read by the ECST.
And the redundancy target sectors A, C, and D are obtained. Sector A,
The data of C and D are read, and the data of sector B is calculated back from the read data. That is, 100% of data that cannot be repaired by the conventional repair method can be repaired.

After transferring the repair data to the PC (host), the replacement location of the sector B is calculated, and the data of the sector B is written to the replacement sector E. Next, a defect is registered with sector B as a bad sector, and the FAT is MS-DOS (basic software).
Rewritten by At the same time, set ECST to sector A,
The information that the combination of B and C is redundant sector D
C and E are changed to information of redundant sector D. On the other hand, U
The position information of the sector B is deleted from the SAT.

USAT and ECST in HDC board
USAT and ECST
Similarly to the redundant sector, it can be provided in an unused area of the disk or in a specific place reserved in advance. At this time, the USC and ECS are provided in the HDC board.
The memory for T is no longer needed.

As a special case, when a non-volatile memory is not provided in the HDC board, data in a USAT or ECST area existing on the magnetic disk may be destroyed. To another location. That is, HD
If volatile memory is provided in the C board, the memory contents are simply moved to another location, and if provided in a disk unused area or a specific location, USAT creates a new FAT from FAT, Will be newly created based on USAT and secured in different places. In addition, the defect location is subjected to additional defect registration.

When the data usage increases, the disk surface is gradually released using the redundant sector or USAT and ECST as an MS-DOS management area. Conversely, when the amount of written data decreases, the data is used again as a redundant area managed by the HDC. At this time, since the previous redundant sector or USAT and ECST have been partially or completely erased, a combination of the redundant sector and the redundant target sector is created again as necessary, and the redundancy process is executed.

It is conceivable to set the number of sectors for the redundant sector to three or more as described above in order to maintain the redundancy as much as possible. In order to improve the processing speed, the number of sectors for the redundant sector is one, that is, one. A method of writing data of a sector as a redundant sector in another sector is also conceivable.

Although the case where MS-DOS is used as the basic software has been described, the data is similarly managed by other basic software, and redundancy can be maintained by USAT and ECST.

It is also conceivable that a redundant sector area is always secured on the magnetic disk surface from the beginning and is not released to the user, that is, the PC. It is also possible to use a magnetic disk device having the property.

Although the example of the magnetic disk device used alone has been described, the present invention can be implemented by combining a plurality of magnetic disk devices.

[0039]

According to the present invention, by preparing redundant data based on the used sector allocation table and the redundant sector table managed by the magnetic disk controller, almost 100 errors can be prevented against sudden data errors. % Can be repaired.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of an embodiment of a magnetic disk drive according to the present invention.

FIG. 2 is an explanatory diagram of a data write flowchart according to the present invention.

FIG. 3 is an explanatory diagram of a positional relationship between a redundant sector and a redundant target sector according to the present invention.

FIG. 4 is an explanatory diagram of a sector number and a redundant exclusion sector position relationship according to the present invention.

FIG. 5 is an explanatory diagram of a reading error repair method according to the present invention.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Personal computer, 2 ... Magnetic disk controller board, 3 ... Used sector allocation table, 4 ... Redundant sector table, 5 ... Head disk assembly, 6 ... Magnetic disk, 7
... magnetic head slider, 8 ... positioning mechanism, 9
... redundant sector, 10 ... data sector, 1
1 ... Sector A, 11 '... Sector A D
# 1A projection position, 12 ... Sector B,
12 ': Projection position of sector B on D # 1A, 13 ... Sector C, 13': D # 1A of sector C
Projection position to 14, sector D, 15
... redundancy exclusion area of sector B, 16 ... D # 1B (H #
0), 17 ... D # 1A (H # 1), 18 ... D # 2
B (H # 2), 19 ... D # 2A (H # 3), 20
... D # 3B (H # 4), 21 ... D # 3A (H #
5), 22 ... redundancy exclusion area.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁶ Identification code Agency reference number FI Technical display location G11B 20/18 572 G11B 20/18 572B 572F

Claims (3)

[Claims] 1. A magnetic disk, a magnetic head slider, a positioning mechanism for positioning the magnetic head slider, and a magnetic disk controller for controlling the positioning mechanism and the like, wherein the magnetic disk controller manages the data to prevent data loss. A magnetic disk drive wherein a sector allocation table and a redundant sector table are provided in a memory of a magnetic disk controller board, and data can be restored using redundant data on a magnetic disk based on the tables. 2. A data sector and a redundant sector are arranged so that the position of the magnetic head slider when reading and writing each sector and the position of projection of the magnetic head slider on another magnetic disk surface do not overlap each other. 2. The magnetic disk drive according to claim 1, wherein the magnetic disk drives are arranged in a positional relationship. 3. The magnetic disk drive according to claim 1, wherein a used sector allocation table and a redundant sector table managed by the magnetic disk controller are provided in an unused area or an area secured in advance on the magnetic disk.