JP2001118343A - Magnetic disk device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To ensure a recoverable read error in an offset read mode, an unrecoverable read error in the offset mode and the data of error occurrence when a write fault error takes places. SOLUTION: When the recoverable read error takes place in the offset read mode and when the sector of an adjacent track in the offset direction of the error sector and several sectors in front of and at the back of the sector are recoverable in the offset read mode (step 103) with respect to the write fault error (step 101), the data of the error sector, the several sectors in front of and at the back of the sector, the sector of the adjacent track of the error sector and the several sectors in front of and at the back of the sector are stored in different places (steps 102, 105 and 109) and subsequently, those data are rewritten in their respective sectors (step 115). The rewritten sectors can be read without performing offset read from the next access on.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a case where a recoverable read error occurs in an offset read, an unrecoverable read error in an offset read, and a write fault error occur. The present invention relates to a technique for eliminating the occurrence of possible read errors and unrecoverable read errors.

[0002]

2. Description of the Related Art A read error that can be recovered by an offset read means that the head is slightly shifted in the track width direction (a direction perpendicular to the track longitudinal direction), and
Retry the lead until it exceeds 0 times, and as a result,
An error that can be read, and a read error that cannot be recovered by offset reading is an error that cannot be read even after retrying more than several tens of times, for example.

[0003] A write fault error is an error when the positional deviation of the head during writing is larger than a write-inhibited slice. For example, a track servo signal recorded at each predetermined position on a medium is read and this signal is read. This is an error when the output falls below a specified value (for example, a write fault error occurs when recording due to a track shift), and when this error is detected, recording is performed again in the section of the track servo signal. Is redone.

In the prior art, when an automatic replacement is set for a recoverable read error in an offset read (setting for allocating data to a spare sector when there is a defect on a medium, etc.), the error is detected. Sector (Fig. 3
Then, only N tracks: M sectors (hereinafter referred to as “N: M”) are read, and data is stored in another location (for example, a spare sector or a memory of a disk controller). M), re-read after re-writing, and if re-read succeeds, no replacement process is executed,
If rereading is not possible, a replacement process is executed.

[0005] However, neither a read operation nor a rewrite operation is performed on a sector ("N-1: M" or "N + 1: M") on a track adjacent to the error sector. for that reason,
When the data of the sector of the adjacent track is written with an offset in the direction of the error sector (recorded shifted in the direction of “N: M”), when the error sector is rewritten, the sector of the adjacent track (“N -1: M "or" N
+1: M ”) may be destroyed and cannot be read.

Further, as disclosed in Japanese Patent Application Laid-Open No. 06-028113, when a data error occurs, the fact that the data has been recovered by retrying is stored, and when the same cylinder head is accessed again, it can be recovered at once and thereby, There are technologies that make access faster. However, when a sector having a data error is written with offset, and then when a sector (“N−1: M” or “N + 1: M”) of a track adjacent to the error sector is written, data is destroyed. Therefore, data cannot be guaranteed with this technology.

For a read error that cannot be recovered by offset read, dummy write is performed on the spare sector to suppress output fluctuations or perform scrubbing. A. Or eliminate the occurrence of (thermal aspirity),
A part of the data of the sector is destroyed, and the data of the sector on both adjacent tracks becomes noise and cannot be read.

In response to a write fault error, the error sector and the preceding several sectors (on the same track) are rewritten (for example, a section of a track servo signal provided in advance on a disk). Track sectors are not rewritten. Therefore, there is a possibility that a part of the data of the sector in the adjacent track (“N−1” or “N + 1”) is destroyed and the data cannot be read.

[0009]

The magnetic disk drive is:
Since the recording density has increased rapidly and the track density has also increased, the off-track margin has decreased,
Also, it is very difficult to control head following and the like. For this reason, data can be easily offset and written due to output fluctuations and deviations in sensitivity distribution of the MR head and the GMR head, external vibrations and shocks, etc. In addition, since the data of the sector of the adjacent track is offset and written, a part of the data of the sector is destroyed, and an unrecoverable read error occurs in the offset read.

[0010] In addition, since the data of the error sector is offset-written in response to the write fault error, a recoverable read error in the adjacent sector and an unrecoverable read error in the offset read occur in the adjacent sector.

Also, if an automatic replacement is set for a recoverable read error by offset read, only the error sector is rewritten, so that an adjacent track sector (“N−1: M” or If the data of “(N + 1: M)” is written with an offset in the error sector direction, there is a possibility that the data of the sector on the adjacent track is destroyed and cannot be read.

An object of the present invention is to provide a method for recovering a read error in an offset read, an unrecoverable read error in an offset read, and a write fault error. In addition, by rewriting the data, the data is guaranteed, and the occurrence of a recoverable read error and an unrecoverable read error in the offset read from the next access is eliminated.

[0013]

In order to solve the above problems, the present invention mainly employs the following configuration. In a magnetic disk device that performs a data guarantee process when a recoverable read error or a write fault error occurs in offset read, a sector of an adjacent track in an offset direction of the error sector in which the error has occurred,
When several sectors before and after the sector of the adjacent track can be recovered by offset read, the data of the error sector and several sectors before and after the error sector, and the sector of the adjacent track and several sectors before and after the sector are stored in different locations. A magnetic disk device for storing each data and rewriting the stored data in each sector.

Further, in a magnetic disk drive which performs data guarantee processing when an unrecoverable read error occurs in offset reading, a sector on both tracks adjacent to the error sector in which the error has occurred and several sectors before and after them And write a fixed pattern or a small-amplitude data pattern that does not interfere with the data of the error sector to the sectors on both tracks adjacent to the error sector and several sectors before and after the error sector. After erasing, reading the error sector after writing the data pattern or DC erasing, or reading the error sector in an offset read recovery process and storing it in another location, and when the error sector can be read, the error sector is read. A few sectors before and after the error sector, and the Data and those around several sectors, magnetic disk drive to re-write the respective data described above stored.

In a magnetic disk drive for performing a data guarantee process when a recoverable read error or a write fault error occurs in offset reading, a sector of an adjacent track in the offset direction of the error sector in which the error has occurred is determined. When several sectors before and after the sector of the adjacent track cannot be recovered by the offset read, the sectors of both adjacent tracks of the sector of the adjacent track and data of several sectors before and after the sector are stored in different locations. Writing a fixed pattern or a small-amplitude data pattern that does not interfere with the data of the sector of the adjacent track to the sectors of the two adjacent tracks and several sectors before and after them, or performing a DC erase to write the data pattern or After DC erase, the sector of the adjacent track When the sector of the adjacent track is read, the sector of the adjacent track and several sectors before and after the adjacent track are read, and the sector of the adjacent track is read. A magnetic disk drive for rewriting the stored data in sectors on both adjacent tracks of the sector and several sectors before and after the sector.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A magnetic disk drive according to an embodiment of the present invention will be described below with reference to FIGS.
FIG. 1 shows a recoverable read error in an offset read,
Alternatively, an example of a flowchart of a process when a write fault error occurs is shown. FIG. 2 shows an example of a flowchart of a process when an unrecoverable read error occurs in the offset read. FIG. 3 shows a case where the data of the own track is offset-written. FIG. 4 shows a case where a part of the data of the own track is destroyed by the attack from the adjacent track. FIG. 5 shows an example of the configuration of a magnetic disk drive.

As shown in FIG. 5, the magnetic disk drive 50 according to the present embodiment is roughly divided into an HDA 51 (head disk assembly) and a PCB 52 (printed circuit board).
The HDA 51 includes a plurality of magnetic disks 53 on which servo information and data are written, a spindle motor 54 mounted and rotating the magnetic disks 53, and a plurality of heads 56 for recording and reproducing information on the recording surface of the magnetic disks 53. And a carriage that supports the head 56 via the arm 57, mounts the R / W integrated circuit element 55 (R / WIC), and moves the head 56 in the radial direction on the magnetic disk 53 to perform a positioning operation. 58 and a voice coil motor 59 (VC
M).

The PCB 52 is connected to a spindle motor driver 60 for controlling the rotation of a spindle motor 54, a servo control circuit 61 for controlling the position of the head 56 in the circumferential direction of the magnetic disk 53, and an R / WIC 55 of the HDA 51 for recording data. R / W control circuit 62 for performing reproduction, R / W
The MPU 63 instructs the W, servo, and spindle rotations collectively and transmits and receives commands to and from the host. The data error detection 64 is performed in the MPU 63.

First, an example of processing when a recoverable read error or a write fault error occurs in offset reading (the head is read by shifting the head by an appropriate amount in the track width direction) will be described. Recoverable read error by offset read in M sectors of N tracks in FIG. 3 (N-track M data (“N: M”) can be read by shifting the head in the N-1 track direction in FIG. 3) ) Or a write fault error has occurred (step 101).

An error sector ("N: M") and several sectors before and after (example: "N: M-2", "N: M-1", "N: M")
+1 "," N: M + 2 ") is stored in another location (step 102). When the M sectors of the N-1 track of the adjacent track in the offset direction are recovered by the offset read (step 103), the data of this sector and several sectors before and after this sector are stored in another location (step 1).
05). At this time, the number of sectors before and after the M sector is set from the host device. It should be noted that M sectors on N + 1 tracks (tracks on the opposite side to the offset direction) of adjacent tracks can be read.

Next, data of M sectors on track N and several sectors before and after are rewritten, and data of sectors on track N-1 and several sectors before and after are rewritten on each sector (step 115). Thus, the M of the N track
Since the data of the sector is rewritten at the correct position without performing the offset write, even if the M sector of the (N-1) track is written, the data of the M sector of the N track is not destroyed and the data is guaranteed. Further, since the M sectors on the (N-1) th track can be read without performing the offset read, the access speed is increased.

Further, the data of M sectors of N tracks in FIG. 4 is destroyed by offset writing of M sectors of N-1 tracks of adjacent tracks, and the M sectors of N tracks have a read error which can be recovered by offset reading. It is assumed that this has occurred (step 101). The data of the error sector and several sectors before and after are stored in different locations (step 10).
2). Then, when the M sectors on the (N-1) th track of the adjacent track are recovered by offset read (step 10).
3) The data of this sector and several sectors before and after this sector are stored in another location (step 105).

Next, data of M sectors on track N and several sectors before and after are rewritten, and data of sectors on track N-1 and several sectors before and after are rewritten to each sector (step 115).

Assuming that only the prior art replacement processing sequence is executed without performing the processing of the present embodiment as described above (the replacement processing sequence stores only the sector in which a recoverable read error has occurred in the spare sector). In order to store the data of only M sectors in N tracks in another location and rewrite the data in only M sectors of N tracks, the data of M sectors in N-1 track in FIG. The inconvenience that is destroyed occurs.

Next, an example of processing when an unrecoverable read error occurs in offset reading will be described. It is assumed that an unrecoverable read error has occurred in the M sectors of the N tracks in FIG. 4 due to the offset read (step 201) (if the area of the black portion of N: M in FIG. No output is obtained and an unrecoverable read error occurs). The M sectors of the N-1 track of the adjacent track, the data of several sectors before and after, the M sector of the N + 1 track, and the data of several sectors before and after are stored in different locations (steps 203 and 205). And M of the N-1 track
A fixed pattern that does not interfere with the error sector or a data pattern with a small amplitude is written to the sector and the M sector of the (N + 1) track, or DC erase is performed (step 206).

As a result, the influence of data in both adjacent sectors can be reduced. The M sectors of the error sector N tracks are read, or read by recovery processing such as offset reading (step 208). When data can be read (step 208), each data is rewritten to the error sector and the sectors on both adjacent tracks (step 210). If the error sector cannot be read (step 208), the data is restored by another disk device or backup device (step 20).
9) Rewrite each data in the error sector and the sectors on both adjacent tracks (step 210).

Finally, with respect to a read error that can be recovered by offset read and a write fault error, a sector (“N−1: M”) adjacent to the error sector (“N: M” in FIG. 3). A description will be given of the case where the data in the sectors before and after the data (for example, “N−1: M−1” and “N−1: M + 1” in FIG. 3) have an unrecoverable read error.

A recoverable read error or a write fault error occurs in offset reading in M sectors of N tracks in FIG. 3 (step 101), and the error sector and data of several sectors before and after the error sector are stored in different locations. It is stored (step 102). Then, it is assumed that the M sectors on the (N-1) th track of the adjacent track could not be recovered by offset read (step 103), and could not be read by recovery processing other than offset read (step 104).

Then, the flowchart shown in FIG. 2 when an unrecoverable read error occurs is executed (step 10).
6). M of track N-2 of the sector adjacent to the error sector
The data of the sector and several sectors before and after the sector, and the data of M sectors of the N track and several sectors before and after the sector are read and stored in another location (steps 202, 203, 204, and 205).
-2 track M sectors, several sectors before and after it, and N
A constant pattern that does not interfere with the data in the error sector or a data pattern with a small amplitude is written or DC erased in the M sectors of the track and several sectors before and after it (step 206). As a result, the influence of data in both adjacent sectors can be reduced.

The M sectors on the error sector N-1 track and several sectors before and after the M sectors are read by a recovery process such as reading or offset reading (step 208). Then, when the data is successfully read (step 208), M sectors on the N-1 track and several sectors before and after it, M sectors on the N-2 track, several sectors before and after that, M sectors on the N track, and several sectors before and after that Each data is rewritten (step 210).

If it is not possible to read M sectors of the N-1 track and several sectors before and after it (step 208),
The data is restored by another disk device or a backup device (step 209), and the M of the N-1 track is restored.
A sector, several sectors before and after it, M sectors on track N-2 and several sectors before and after it, and M sectors on N tracks;
Each data is re-written in several sectors before and after that (step 210). Further, since M sectors on the (N + 1) th track of the adjacent track can be read, data is not stored in another location (steps 107 and 108).

As a result, the data of the M sectors on the N tracks is rewritten at the correct position without performing the offset write. Therefore, even if the M sectors of the N-1 track are written, the data of the M sectors on the N track are destroyed. Data is guaranteed. Further, since the M sectors on the (N-1) th track can be read without performing the offset read, the access speed is increased.

As described above, the embodiments of the present invention include those having the following configurations, functions and functions. For a read error and a write fault error that can be recovered by offset read, a sector (“N−M”) adjacent to the error sector (“N: M”) is used.
1: M "or" N + 1: M ") and several sectors before and after it (eg," N-1: M-1 "and" N-1: M + 1 ") can be recovered by offset read. Sector (“N: M”) and several sectors before and after it (eg, “N: M−
"1" and "N: M + 1"), the sector of the adjacent track in the offset direction and the data of several sectors before and after it are stored in another location, and then the data is rewritten to each sector.

For a read error that cannot be recovered by offset read, the error sector (“N:
M ”), the sectors (“ N−1: M ”and“ N + 1: M ”) of both adjacent tracks and the data of several sectors before and after that are stored in another location (spare sector or cache memory of the disk controller), and an error occurs. After writing a fixed pattern that does not interfere with the data in the error sector or a data pattern with a small amplitude in the sectors on both tracks adjacent to the sector and several sectors before and after it, or performing DC erase,
Read the error sector offset.

When the error sector can be read (there is a possibility that the error sector can be read because the interfering signals from the sectors on both tracks adjacent to the error sector have disappeared), the offset read signal is stored in another location, and Each data stored in another location is rewritten to the error sector, the sector on both adjacent tracks, and several sectors before and after it.

If the offset sector cannot be read from the error sector, the data is restored by another disk device or backup device, and the data is stored in the error sector, the sector on both adjacent tracks, and several sectors before and after the error sector. Rewrite.

Further, according to the embodiment of the present invention, not only the error sector but also the data of the sector of the adjacent track is to be stored and processed in another place. In this case, not only the error sector is processed, but also the data of an adjacent track is processed in the same manner as the error sector.

[0038]

According to the present invention, an error sector, several sectors before and after the error sector, and an adjacent error sector can be used for a recoverable read error in an offset read, an unrecoverable read error in an offset read, and a write fault error. By rewriting the data in the track sector and several sectors before and after it to the correct position, the data is guaranteed, and from the next access, a read error that can be recovered by offset read and a read error that cannot be recovered by offset read This has the effect of preventing the occurrence of errors and increasing the access speed.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating an example of a flowchart of a process when a recoverable read error or a write fault error occurs in an offset read according to the present invention.

FIG. 2 is a diagram illustrating an example of a flowchart of a process when an unrecoverable read error occurs in the offset read according to the present invention.

FIG. 3 is a diagram showing a case where data of a self-track is offset-written.

FIG. 4 is a diagram showing a case where a part of data of a self-track is destroyed by an attack from an adjacent track.

FIG. 5 is a block diagram illustrating an example of a configuration of a magnetic disk drive according to an embodiment of the present invention.

[Explanation of symbols]

 30 sector on N-2 track 31 sector on N-1 track 32 sector on N track 33 sector on N + 1 track 34 sector on N + 2 track 40 sector on N-1 track 41 sector on N track 42 N + 1 Sector on track 43 sector on N + 1 track 44 sector on N + 2 track 50 magnetic disk device 51 HDA (head disk assembly) 52 PCB (printed circuit board) 53 magnetic disk 54 spindle motor 55 R / W integrated circuit element (R / WIC) 56 Head 57 Arm 58 Carriage 59 Voice coil motor (VCM) 60 Spindle motor driver 61 Servo control circuit 62 R / W control circuit 63 MPU 64 Error detector

Claims (5)

[Claims] 1. A magnetic disk drive for performing a data guarantee process when a recoverable read error or a write fault error occurs in an offset read, wherein: a sector of an adjacent track in an offset direction of the error sector in which the error occurred; When the several sectors before and after the sector of the adjacent track can be recovered by offset read, the data of the error sector and several sectors before and after the error sector and the data of the sector of the adjacent track and several sectors before and after the sector are stored in another location. Wherein the stored data is rewritten to each sector. 2. A magnetic disk drive for performing a data guarantee process when an unrecoverable read error occurs in an offset read, comprising: sectors on both tracks adjacent to the error sector in which the error occurred; And writing a fixed pattern or a small-amplitude data pattern that does not interfere with the data of the error sector to the sectors on both tracks adjacent to the error sector and several sectors before and after the data, or After erasing, reading the error sector after writing the data pattern or DC erasing, or reading the error sector in an offset read recovery process and storing it in another location, when the error sector can be read, the error sector is read. A few sectors before and after that, and sections of both tracks adjacent to the error sector A few sectors before and after thereof, the magnetic disk device characterized by re-writing the respective data described above stored. 3. The magnetic disk device according to claim 2, wherein when the error sector cannot be read, the data is restored by another disk device or a backup device.
Rewriting each data in the error sector and several sectors before and after the error sector, and rewriting sectors in both adjacent tracks of the error sector stored in the different location and data in several sectors before and after them in each sector. A magnetic disk drive characterized by the above-mentioned. 4. A magnetic disk drive for performing a data guarantee process when a recoverable read error or a write fault error occurs in an offset read, wherein: a sector of an adjacent track in an offset direction of the error sector in which the error occurred; When several sectors before and after the sector of the adjacent track cannot be recovered by the offset read, the sectors of both adjacent tracks of the sector of the adjacent track and data of several sectors before and after the sector are stored in different locations. Writing a fixed pattern or a small-amplitude data pattern that does not interfere with the data of the sector of the adjacent track to the sectors of the two adjacent tracks and several sectors before and after the sector, and performing the DC erase; After DC erase, the sector of the adjacent track is Read or stored in another location by the offset read recovery process, and when the sector of the adjacent track can be read, the sector of the adjacent track and several sectors before and after the sector, and the adjacent track A magnetic disk drive characterized in that the stored data is rewritten in sectors on both tracks adjacent to the sector and several sectors before and after the sector. 5. The magnetic disk drive according to claim 1, wherein the setting of the number of sectors before and after the error sector and the sector of the adjacent track is set by a command from a host device. Characteristic magnetic disk drive.