JP2001160268A - Substitution processing method for recording medium and device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a substitution processing method for a recording medium and a device which a time required for substitution processing of a storage region unit due to occurrence of an error is shortened, and interruption of data transfer is suppressed to the minimum. SOLUTION: When an error detecting means 6 detects occurrence of an error at the time of writing, substitution processing is performed by using a substitution original sector address 11 and a sector address 12 to be substituted, only updating for a substitution map on a memory 5 is performed, writing data in a sector to be substituted is not performed. When reading a sector of a substitution place is performed, it is informed to a host device that data cannot be written in a sector to be substituted.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an alternative processing method and apparatus for a recording medium such as a magnetic disk which can suppress a decrease in throughput by shortening a processing time required for an alternative processing when an error occurs. The present invention relates to a recording medium alternative processing method and apparatus suitable for use in the system of the present invention.

[0002]

2. Description of the Related Art In a conventional magnetic disk drive, for example, in a magnetic disk drive having a write cache function, first, when a write command is received from a host device, the write command is transferred to a write cache area secured on a buffer. Receive data from the host device as much as possible. When the data of all the sectors specified by the write command is received on the buffer, the completion of the command is reported to the host device. At this stage, the command ends before data of all sectors is written on the medium.

A sector is written after positioning is performed on a sector specified by a command. If an error occurs while writing data, a retry is performed. After executing the retry, if the retry fails (the sector could not be written on the medium), the alternative processing is performed next.

In this alternative processing, first, data of an error-occurring sector still existing in a memory in a magnetic disk is written to a spare sector. For this spare sector,
After replacement, read / replace the sector where the error occurred.
Will write. Hereinafter, this sector is referred to as an alternative destination sector. Thereafter, in order to indicate that the sector in which the error has occurred has already been replaced, the address of the sector in which the error has occurred (hereinafter, referred to as the replacement source sector) and the address of the replacement destination sector are added to the same field in the replacement map. This alternative map is read into the memory from the system cylinder on the medium at the time of initial setting at the time of power-on, and updates the alternative map on the memory and updates the alternative map on the system cylinder each time the alternative processing is performed. This process completes the substitution process. When the replacement process is completed, writing from the sector next to the replacement source sector is started again, and the write operation of the write cache is continued.

[0005] In addition, the sector which is actually replaced is read /
When writing, a search is made to see if the first sector specified by the command is registered in the alternative map on the memory. If an attempt is made to read / write a sector that matches the replacement source sector on the replacement map, the replacement destination sector is read / written instead of the replacement source sector. Even if the first sector is not replaced, if it is registered in the sector specified by the command, the preceding sector is read / written. If the sector is not registered in the replacement map, all sectors specified by the command are read. This process is sequentially performed on the sectors specified by the command, and all sectors are read / read.
Complete the write.

[0006]

The conventional method and apparatus for alternative processing of a recording medium are configured as described above. The first problem is that if there is an error in the sector, the result of performing alternative processing after retrying is attempted. In order to perform the positioning to the replacement destination sector and the write operation, and then to rewrite the replacement map, the positioning to the system cylinder and the write operation must be performed, which increases the processing time required for one command. is there. The location of the alternate destination sector on the medium and the address of the system cylinder are stored in the normal read /
Since the address is located at a distance from the address at which the writing is performed, a long time is required for the positioning operation. for that reason,
Data transfer from the host device is hindered for a long time in the alternative process at the time of writing, and as a result, in applications that require continuous data transfer such as image data, the buffer in the host device overflows and images are lost. Will be recorded.

A second problem is that writing to the replacement destination sector and writing of the replacement map are not performed during the execution of the command but after the completion of the command, thereby reducing the apparent command execution time. Although there is a technology disclosed in the gazette, even in such a case, if the command is issued from the host densely during recording of image data of a plurality of channels, and the processing request of the next command comes during the substitute processing, it is also This causes a delay in data transfer and causes image loss.

A third problem is that if no replacement processing is performed, an error occurs in a sector where an error has occurred every time the read / write operation is performed. In an extreme case, a reproduced image is distorted. It is to cause.

A fourth problem is that an alternative process cannot be performed for an error that occurred during writing from a write cache to a medium, or a complete alternative process (reading out data in a written sector). Is not possible, an error cannot be reported by the write command. At this time, there is a method of returning an error to the host device when an error is detected. However, even if an error (Deferred Error specified by the SCSI interface) is returned to the host device, the data of the sector is also lost in the host device. It is highly likely that data recovery is difficult. This becomes more remarkable in a large write cache. In addition, the sector that could not be replaced is likely to be read the next time the sector is read, making it difficult to identify an error portion during reproduction.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a recording medium alternative processing method and apparatus which can reduce the time required for alternative processing in storage areas due to occurrence of an error and minimize interruption of data transfer. .

[0011]

A replacement processing method for a recording medium according to the present invention updates a replacement map in a memory without writing data to a replacement destination, and replaces the replacement map on the recording medium. The map writing is performed collectively, and when the storage area unit of the replacement location is read, the upper device is notified that the data of the replacement destination could not be written, and the upper device has an incorrect location of the data. And performing recovery processing such as writing of the data from the higher-level device to the alternative destination.

An alternative processing apparatus for a recording medium according to the present invention is:
An interface control means for receiving a command from a host device, a command execution means for analyzing the command received by the interface control means, an access means for reading and writing to and from the recording medium according to the command, Error detecting means for detecting occurrence of an error at the time of writing by the access means; positioning executing means for performing positioning to an arbitrary track on the recording medium; and substitution source storage when the error detecting means detects the occurrence of an error. Perform replacement processing using the area unit address and the replacement destination storage area unit address, update only the replacement map on the memory, do not write data to the replacement destination, That the data of the replacement destination could not be written to the host device when the Characterized in that a substitute means for knowledge.

[0013] In the alternative processing method for a recording medium according to the present invention, for a storage area unit in which an error has occurred, only an update of an alternative map in a memory is performed without writing data to an alternative destination. The processing time can be reduced by performing the writing of data at the same time, and by notifying the host device that the data of the replacement destination could not be written to the higher-level device when reading the storage area unit of the replacement location,
By performing data recovery processing on the host device side, interruption of data transfer is minimized.

When the occurrence of an error is detected, the replacement processing apparatus for a recording medium according to the present invention executes the replacement processing using the replacement source storage area unit address and the replacement destination storage area unit address, and only updates the replacement map on the memory. To prevent writing of data to the replacement destination, and to notify the host device that the data of the replacement destination could not be written to the upper-level device when reading in the storage area unit of the replacement location, thereby performing the replacement processing. The data transfer process is minimized by shortening the time required for data transfer and performing data recovery processing on the host device side.

[0015]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below. The alternative processing method and apparatus for a recording medium such as a magnetic disk according to the present embodiment does not perform the alternative processing at all, but only updates the alternative map on the memory and collectively writes the alternative map on the medium. This reduces the processing time. Further, the processing time is reduced without writing data to the replacement destination. At this time,
By notifying the host device that the data of the replacement destination could not be written when reading the sector of the replacement portion, the data recovery process is performed on the host device side, so that the recording / reproduction of a moving image or the like can be performed. In a magnetic disk used for the purpose, interruption of data transfer is minimized.

FIG. 1 is a block diagram showing an apparatus for realizing an alternative processing method for a magnetic disk according to the present embodiment. The interface control means 1 receives commands from a higher-level device (not shown). Control means 1
The command execution means 2 analyzes the command received by the command and gives an instruction to the read / write execution means (access means) 3, and the positioning execution for positioning on an arbitrary track of a sector on the medium (recording medium) 8. Means 4
Read / write execution means 3 for reading / writing a sector on a track after positioning, a memory 5 storing the read / write execution means 3, an alternative map 9 described later, and the like, An error detecting means 6 for detecting an error during execution of read / write by the means 3;
A buffer (access means) 7 and the like which accumulates data through the write execution means 3 and is used as a write cache and temporarily accumulates data from the medium 8 at the time of reading.

FIG. 2 shows the format of the alternative map 9 stored in the memory 5 and the alternative map unwritten flag 13,
FIG. 3 is an explanatory diagram showing an error flag at the time of substitution 14, an error sector address at the time of substitution 15, and the like. FIG. The replacement map 9 is on the memory 5 in FIG. 1 and is referred to by the read / write execution means 3. On the replacement map 9, a replacement source sector address 11 indicating an address of a sector registered as a defective sector as a result of the replacement process is displayed.
Then, an alternative destination sector address 12 indicating an alternative destination of the sector indicated by the alternative source sector address 11 is set. The sector indicated by the replacement destination sector address 12 is allocated to a system cylinder different from the user cylinder used by the host device.

The replacement destination unwritten flag 10 indicates that the data to be written in the sector indicated by the replacement source sector address 11 at the time of the replacement processing is replaced by the replacement destination sector address 12.
Is a flag indicating that the data has not been written to the.

The substitute map non-writing flag 13 updates only the substitute map 9 in the memory 5 by the substitute processing.
As a result, the system cylinder 16 shown in FIG.
This is a flag indicating that it no longer matches the upper alternative map. The system cylinder 16 has an area for storing the alternative map 9 on the memory 5 on the medium 8.

Alternative destination unwritten flag 1 on alternative map 9
0 is set when the replacement process performed by the read / write execution unit 3 performs the replacement process without writing data to the sector indicated by the replacement destination sector address 12. At this time, the replacement source sector address 11 and the replacement destination sector address 12 of the same field are updated at the same time.

The alternative map non-writing flag 13 is set when only the alternative map 9 in the memory 5 is rewritten, and is reset when the alternative map 9 in the memory 5 is written to the system cylinder 16 on the medium.

The replacement error flag 14 is a flag that defines whether or not the history of the error information is returned to the host device when the sector corresponding to the replacement destination sector address 12 is read. When the higher-level device issues a command to read / write the sector of the replacement source sector address 11 existing in the replacement map 9, the drive reads the replacement destination sector address 12 on the same field instead of the replacement source sector address 11. / Write. If it is a read, the alternative destination unwritten flag 10 in the same field is checked, and if it is set, the alternative error flag 14 is set. The sector address 11 is stored in the replacement error address 15.

Thereafter, the succeeding sector is read, and the replacement error flag 14 is checked at the end of the read command. If it is set, data from the host device is not written correctly due to an error at the time of writing. An error indicating that there is a sector is returned to the host device. At the same time, the content of the substitution-time error address 15 is also returned to the host device. If the substitute time error flag 14 is reset,
The read command ends normally, assuming that all sectors indicated by the command have been normally written at the time of writing.

At the time of writing, a similar sequence
The sector indicated by the alternative destination sector address 12 is written in place of the sector indicated by the alternative source sector address 11, and if the sector is normally written, the alternative destination unwritten flag 10 is reset. If the alternative destination unwritten flag 10 has been reset from the beginning, the process ends without changing the flag.

The upper device issues a command for reflecting the replacement map 9 on the system cylinder 16 in a power-off process. When the update command of the alternative map 9 is received by the interface control means 1, the command execution means 2 analyzes the command. Command execution means 2 reads /
The write execution means 3 is started, and if the alternative map non-write flag 13 is set, the read / write execution means 3 writes the alternative map 9 to the system cylinder 16, and then the alternative map non-write flag 13 is reset. You. Here, if the alternative map non-writing flag 13 is reset, the command ends without doing anything.

Next, FIG. 1, FIG. 2, FIG. 3, FIG.
The overall operation of the present embodiment will be described in detail with reference to the flowchart of FIG. First, the operation will be described with reference to the flowchart shown in FIG. 4. A write command is sent from the host device to the interface control means 1. The interface control means 1 detects that a write command has been sent, and starts command reception processing (step S1). When the command execution means 2 recognizes that the write command received by the interface control means 1 is a write command, it activates the read / write execution means 3 (step S2). When activated, the read / write execution unit 3 starts receiving write data from the host device to the write cache configured on the buffer 7 (step S3). The reception of the write data to the write cache is continuously executed as long as the write cache has a free area (step S4).
When all the write data of the sector length specified by the write command is received, the command ends normally irrespective of the write operation to the medium 8 (step S5).

The read / write execution means 3 starts the processing shown in the flowchart of FIG. 5 simultaneously with the reception of the write data from the host device. First, buffer 7
It is determined whether data of a writable sector is stored in the write cache (step S11). As a result, if no data is stored, the process returns to step S11 in FIG. 5 again, and monitoring is continued until the data is stored. Also,
If writable data has been accumulated, the replacement map covers all sector addresses within the range that can be continuously written by one write operation, that is, the range that does not require seek or head switching positioning operation. (Step S).
12, step S13).

As a result, if all the sectors in the search range are not registered in the alternative map 9, the next step is performed with the first sector of the search range as the address for positioning at the time of writing and the total number of sectors in the search range as the write length. Proceed (Step S
14). If there is a sector registered in the alternative map 9 as a result of the search, that is, if the search hits, it is determined whether the sector is the first sector in the search range (step S18).

As a result, if it is the first sector, the address of the replacement destination sector address 12 on the same field as the replacement source sector address 11 is set to the address for positioning at the time of writing, and the sector length is set to 1; If the alternative destination unwritten flag 10 on the same field is set, reset is performed (step S20). If it is not the first sector, the first sector in the search range is set to the address for positioning at the time of writing, and the sector length is set to the number of sectors up to one before the first sector found in the search range. Then, the process proceeds to the next step (step S19).

The read / write execution means 3 executes step S
14. Using the sector address and the sector length determined in steps S19 and S20, first, the positioning operation means 4 is instructed to perform a positioning operation to the track specified by the sector address, that is, seek and head switching.

The positioning execution means 4 performs positioning on the track designated by the read / write execution means 3, and reports the completion of the positioning to the read / write execution means 3 when the positioning is completed (step S15). .

When the read / write executing means 3 receives the report of the completion of positioning by the positioning executing means 4, it starts writing only in the range of the sector length from the designated sector address (step S16). At this time, if all sectors have been written without any error being detected by the error detecting means 6 for writing (step S17), the process proceeds to the next step (step S27). This step S2
In step 7, the start of the sector to be written next on the write cache is performed. When these series of processes are completed, the process returns to step S11 in FIG. 5 again to prepare for the next write.

As can be seen from this sequence, the reception of write data from the host device, the execution of positioning, and the writing of sectors on the track are processed in parallel. If there is write data on the write cache, the target Perform a write operation to the sector. When all the write data specified by the command has been received, the completion of the command processing is notified to the host device through the interface control means 1. The transfer of the write data from the host device ends during the positioning by the positioning execution means 4 or at least before the sector write operation by the read / write execution means 3 is completed. Therefore, the write command ends before all the sectors specified by the write command are written to the medium, and the command can no longer be reported to the host device.

In step S17, when an error is detected by the error detecting means 6 at the time of writing to the target sector, a retry is performed for the sector in which the error has occurred (step S21). As a result, if the retry is successful (step S22), the process proceeds to step S27, and the process for continuously performing the next write is started. If the retry fails, the corresponding sector is regarded as defective and the replacement process is performed. Enter the sequence.

The retry performed in step S21 may not be executed in a situation where a higher throughput is required. In this case, if it is determined in step S17 that the processing has ended with an error, the process directly branches to the substitution process starting from step S23 (substitution means).

In the replacement process, a new field is added to the replacement map 9 on the memory 5 (step S23), and the address of the sector in which the error has occurred is written in the replacement source sector address 11 of that field. A spare sector prepared in advance is assigned to the replacement destination sector, and the replacement destination sector address 12 of the field added to the replacement map 9 is assigned.
The address of this spare sector is written in
4, alternative means). Further, an alternative destination unwritten flag 10 of the same field is also set (step S25, alternative means). The replacement destination unwritten flag is a flag indicating that the data of the sector that should have been written to the replacement source was not written to the replacement destination. After that, since the replacement map 9 in the memory 5 was updated, the replacement map non-writing flag 13
Is set (step S26, alternative means). The replacement map non-writing flag 13 is set to the replacement map 9 in the memory 5.
Is updated, indicating that the replacement map 9 in the memory 5 has not yet been written to the replacement map in the system cylinder 16. At this point, the substitution process ends, and step S2 is executed to continue the write operation.
Proceed to 7.

When a read command is sent from the higher-level device to the interface control means 1, the interface control means 1 detects that the read command has been sent as shown in the flowchart of FIG. Start (step S31). The command received by the interface control means 1 is transmitted to the command execution means 2
Is recognized as a read command, the read / write execution means 3 is activated (step S3).
2).

When activated, the read / write execution means 3 does not require a range in which data can be read continuously from the beginning of the range of sectors specified by the read command, that is, a positioning operation for seek or head switching. The address of the sector within the range is set as the search range, and the replacement map 9
Search for the replacement source sector address 11 (step S)
33).

As a result, if the sector in the search range is not registered in the alternative map 9, the next sector is set with the first sector of the search range as the address for positioning at the time of reading and the total number of sectors in the search range as the read length. Proceed (Step S
35). If a sector that matches the substitution source sector address 11 of the substitution map 9 is found as a result of the search, it is determined whether that sector is the first sector in the search range (step S42).

As a result, if it is the first sector, the address of the alternative destination sector address 12 on the same field as the alternative source sector address 11 is set to the address for positioning at the time of reading and the sector length is set to 1 (step S4).
4) If the alternative destination unwritten flag 10 on the same field is set at that time (step S4)
5) Set the replacement error flag 14 and the replacement source sector address 11 to the replacement error address 15 and proceed to the next step (step S46). On the other hand, if the alternative destination unwritten flag 10 is not set (step S4)
5) It is considered that a normal substitution process has been performed, and the process proceeds to step S36 without setting error information.

If it is determined in step S42 that the first sector is not the first sector, the address of the first sector in the search range when positioning is read, and the sector length is one before the sector found first in the search range. The number of sectors is set, and the process proceeds to step S36 (step S43).

The read / write execution means 3 uses the sector address and the sector length determined in steps S35, S43, and S44 of FIG. 6 to perform a positioning operation to a track specified by the sector address, that is, a seek operation. And instruct the positioning execution means 4 to switch the head. The positioning execution means 4 performs positioning on the track specified by the read / write execution means 3, and when the positioning is completed, reports to the read / write execution means 3 that the positioning has been completed (step S36).
When receiving the report of the completion of positioning by the positioning execution means 4, the read / write execution means 3 starts reading in the range specified by the sector length from the specified sector address.
The read sector is simultaneously transferred to the host device (step S37).

At this time, if all the specified sectors have been read, the process proceeds to the next step, and it is determined whether all the sectors specified by the read command have been transferred to the host device. As a result, if there are remaining sectors to be read, the process returns to step S33 to read the remaining sectors again (step S38). On the other hand, if there are no more sectors to be read, the sequence proceeds to the read command execution end sequence. At the end of the read command, the replacement error flag 14 is checked, and it is determined whether a data-unwritten sector has not been read to the replacement destination (step S39).

As a result, if the replacement error flag 14 has been reset, the read command ends normally (step S40). On the other hand, when the replacement error flag 14 is set, an error code indicating that the data of the replacement destination sector is different from the data written in the past and the replacement error address 15 are used as address information of the error sector by the upper-level device. Report. After that, the substitute-time error flag is reset to prepare for execution of the next command (step S41, substitute means).

The host device issues a command to write the replacement map 9 on the system cylinder 16 in the power-off process. When the interface executing means 1 receives the update command for the replacement map 9, the command is analyzed by the command executing means 2 and the read / write executing means 3 is analyzed.
Start If the alternative map non-writing flag 13 is set, the read / write execution means writes the alternative map 9 to an area on the system cylinder 16 and resets the flag. Here, if the alternative map non-writing flag 13 is reset, the command ends without doing anything.

Next, FIG. 1, FIG. 2, FIG. 3, FIG. 7, FIG.
The operation of the present embodiment will be specifically described with reference to the flowcharts of FIGS. FIG. 7 shows a state of a certain track on the user cylinder. Track N
Sector 3 is a sector that potentially causes a write error, and is not registered in the replacement map 9 at this stage. FIG. 8 shows each situation of the alternative map 9, and the map 1
00 indicates the contents of the replacement map 9 in the initial state, and information indicating the end of the replacement map 9 is written in the replacement source sector address of the first field. Map 101,
The map 102 shows a state in which information has been added as a result of the substitution process.

First, when a write command is sent from the higher-level device to the interface control means 1 in FIG. 1, the command is received in step S1 shown in the flow chart in FIG. To start the read / write execution means 3. In step S3, the read / write execution means 3
Data accumulation is started in the upper write cache, and in step S4, the sequence shifts to a sequence of waiting for the end of transfer of all data from the higher-level device.

The read / write execution means 3 executes a flowchart shown in FIG. 5, which is a sequence for writing data on the medium 8 separately from the sequence for receiving data. In step S11 of FIG. 5, it is checked whether or not there is write data from the host device in the write cache of the buffer 7. At this time, it is assumed that a command to write data from sector 0 to sector 12 of track N in FIG. 7 has been issued from the host device, and that data up to sector 10 has already been stored in the write cache of buffer 7 at that time. In this case, the write cache on the buffer 7 has the track N +
Although data up to sector 2 of 1 has been stored, the writable sector without positioning is sector 6 of track N.
Therefore, the search range in step S12 is from sector 0 to sector 6 of track N. The alternative map 9 is searched in this search range. In the map 100 of FIG. 8 showing the contents of the initial replacement map 9, since no replacement sector is registered, it is determined in step S13 that the sector in the search range has not been replaced, and the process proceeds to step S14. .

In step S14, sector 0 of track N, which is the first sector in the search range, is set as the write start sector, and 7 as the total number of sectors in the search range is set as the sector length. In the following step S15, the positioning unit 4 is instructed to perform positioning with respect to the set track N.

When the positioning execution means 4 completes positioning with respect to the track N, the read / write execution means 3 starts writing from sector 0 to sector 6 in step S16. However, since there is an error sector in sector 3 of track N, writing is performed up to sector 2 and
Terminates with an error. Since the writing has ended in error in step S17, the process shifts to step S21 to retry.

In the next step S22, if the retry of the sector 3 of the track N succeeds, the process proceeds to step S27.
Writing is continuously performed from sector 4 of track N, and the process returns to step S11 again. However, at this time, step S2
If the second retry is not successful, the process moves to step S23 to start the substitution process. At this time, it may be considered that the retry is not performed for the above-described reason. At that time, if the error end is determined in the step S17, the process proceeds to the step S23 to start the alternative process next.

In step S23, a new field is added to the substitute map 9. In the map 100 shown in FIG. 8, no replacement sector is registered, and information indicating the end of the replacement map is written in the replacement source sector address of the first field. First, a search for information indicating the end of the replacement source sector address of the replacement map is performed. In the map 100, the first field is targeted.
Next to the field found at that time, information indicating the end of the alternative map is newly written. In step S25, the address of the sector in which the error has occurred is written in the replacement source sector address of the field found as a result of the search, and the unused spare sector address is written in the replacement destination sector address.

More specifically, since the sector 3 of the track N is to be replaced, information indicating the three sectors of the track N is written in the replacement source sector address of the first field of the map 100, and the spare sector is written in the replacement destination sector address. Write the address of sector 0 of track M with In step S25, the substitute destination unwritten flag of the same field is set to "1" in order to perform a substitute process in which data is not written in the substitute destination sector.

In the following step S26, since the substitute map 9 has been updated, "1" is set to the substitute map non-writing flag 13, and a series of substitute processing ends, and the substitute map 9 changes from the state of the map 100 in FIG. Rewrite to the state of map 101.

When the substitute processing is completed, step S27
Then, the leading sector to be written next is set.
In this case, writing is started from sector 4 of track N, and the process returns to step S11. Assuming that all the write data from the host device can be received in the write cache of the buffer 7 during these series of processing, the writable sectors are all the sectors specified by the write command, and the track is written in the write cache. Data from sector 4 of N to sector 4 of track N + 1 will be stored. Then, when all the data of the sector specified by the write command is stored in the write cache irrespective of the series of write operations to the medium 8, the write command ends normally (steps S4 and S4 in FIG. 4).
5).

In this state, when the process returns from step S27 to step S11 shown in FIG. 5, the sectors in the search range are set from sector 4 to sector 6 of track N, and a series of operations are performed again. If no error is detected in step S17, the process proceeds to step S27. In steps S11 and S12, the search range is changed to track N + 1.
, Sectors 0 to 4 are set, and the operation from step S11 is repeated.

When all of the write cache of the buffer 7 is written to the medium 8, there is no more data that can be written to the write cache in step S11, so that it is necessary to wait until data is newly sent by a write command from a higher-level device. .

After that, when a command to read from sector 0 to sector 4 of track N comes from the host device, the interface control means 1 receives the command in step S31 shown in the flowchart of FIG. To start the command execution means 2.

In step S33, the alternative map 9 is searched in the range of the sector specified by the read command sent. In this case, as in the case of writing, a search is performed in a continuously readable range. In this case, the search is performed by using the map 101 shown in FIG. 8, which is the updated alternative map. At this time, since the sector 3 of the track N is registered in the map 101, the search is hit and the step S
At 34, the process branches to step S42.

In step S42, it is determined whether the sector found as a result of the search is the first sector in the search range.
In this case, since it is not the first sector, the process proceeds to step S43. In this step S43, the read target range is set up to sector 2, which is the sector immediately before sector 3, and the read sector length at that time is set to 3.

Then, in the subsequent step S36, the track N
When the positioning operation is completed, and the positioning is completed, the data is read from sector 0 to sector 2 in step S37, and the data is transferred to the host device. In step S38, it is determined whether or not the data of all the sectors specified by the read command has been transferred to the host device. Here, since there is still more, the process returns to step S33, and the alternative map 9 is searched again in the range of the sector that has not been read from the medium 8. In the search performed again in step S33, the search target at this time is from sector 3 to sector 4 of track N.

Since sector 3 is registered in map 101, the process proceeds from step S34 to step S42. Further, since the first sector to be searched is registered in the map 101, the process proceeds to step S44. Here, map 10
The sector 0 of the track M shown in FIG. 8, which is an alternative destination sector of 1, is set to the sector to start reading, and the sector length is set to 1, and the process proceeds to the next step S45.

In step S45, it is determined whether or not the substitute destination unwritten flag 10 of the field found as a result of the search is set. At this time, since the flag has been set, the process proceeds to step S46. In step S46, preparations are made to return an error indicating that data that should have been written to sector 3 of track N of the replacement source is not written to sector 0 of track M of the replacement destination. This is performed by setting the replacement error flag 14 and writing the address of the replacement source sector in the replacement error address 15. These pieces of information are used when returning an error to the host device at the end of the read command.

Next, in step S36, positioning is performed on the track M of the replacement destination, and when it is completed, reading of track M sector 0 is performed in step S37. Upon completion, the remaining data transfer determination in step S38 indicates that there is a remaining sector, that is, sector 4 of track N, so the flow returns to step S33. As a result, the alternative map 9 is searched again in step S33, but since the sector 4 of the track N is not registered, the process proceeds to step S35.

In step S35, sector 4 of track N
Is set as a first sector, with the first sector as the first sector. As in the previous case, steps S36 and S37 are executed to transfer the data of all the sectors to the host device.

In step S38, since all data transfer has been completed, the flow advances to step S39. In step S39, it is checked whether an error sector flag at the time of replacement is set. Then, when the data of the track M sector 0 of the replacement destination was read in the past, the error sector flag at the time of replacement was set. Therefore, the process proceeds to step S41, and the data written by the host device and the data at the time of reading the replacement destination are An error report indicating that they do not match is sent to the higher-level device, and the replacement error sector flag is reset. At the same time, the address of the replacement error sector address 15 is returned to the host device, and the execution of the read command is terminated.

After that, assuming that a write command for track N sector 0 to sector 4 is sent from a higher-level device, steps S1 to S5 shown in FIG.
The processing up to this is the same as the above-described write command sequence. However, the search of the alternative map 9 from track N sector 0 to sector 4 in step 12 of FIG.
This is performed for the map 101 shown in FIG. At this time, since the search for sector 3 is hit, the process proceeds to step S18 in FIG.

In step S18, since the sector 3 is not the first sector to be read, the process proceeds to step S19.
In step S19, track N sector 0 is the first sector,
Step S15 is performed to set a read for setting the sector length to 3.
Proceed to. In step S15, positioning to track N is performed, and after completion, writing from sector 0 to sector 2 is performed.
After the end of the write, if there is no error, the process proceeds to step S27, performs the write setting for the remaining sectors, and returns to step S11.

If it is determined in step S11 that all the write data has been transferred from the host device, the process proceeds to the next step S12, where the remaining sector, track N sector 3
A search is made for an alternative map 9 with a search range of sector 4 from. As the alternative map 9, the map 101 shown in FIG. 8 is used, and at this time, the sector 3 is hit in the search. this is,
Since this is the first sector at the time of writing, the process proceeds from step S13 to step S18 and step S20.

In step S20, a write to the alternative destination sector is set, that is, one sector write of sector 0 of track M is set, and the alternative destination unwritten flag 10 in the same field of the map 101 hit as a result of the search is set. Is reset to “0”, and the alternative map unwritten flag 13 is set to 1. This is based on writing data from a higher-level device in the replacement destination sector. As a result, the substitute map 9 becomes the state of the map 102 shown in FIG.

Then, the process proceeds to a step S16, wherein the track M
Is performed, and upon completion, sector 0 of track M is completed.
Write to If it is determined that the data has been normally written, the process proceeds to step S27, where the remaining sector, sector 4 of track N, is set as the leading sector of the write, and the process returns to step S11. Similarly, the writing of the sector 4 of the track N is performed after executing the processing of step S11, step S12, step S13, step S14, step S16, step S17, and if there is no error, after executing step S27, returns to step S11. Wait for data transfer by the next write command.

Thereafter, when a command to read from sector 0 to sector 4 of track N comes again from the host device, the processes from step S31 to step S32 shown in FIG. 6 are performed in the same manner as the aforementioned read command. The search for the substitute map 9 in step S33 is performed using the map 1 shown in FIG.
02, and track N sector 3 hits the search, but sector 3 is not the first sector to read. Then, after executing steps S34 to S42, the process proceeds to step S43. Here, the sector 0 of the track N is set to the read start sector, the sector length is set to 3, and the process proceeds to step S36. The positioning is performed on the track N in step S36. Lead to. In step S38, since the transfer of all sectors to the higher-level device has not been completed,
The process returns to step S33 in order to perform reading for the remaining sectors.

In step S33, the process goes to step S44 to hit the first sector in the search range. Here, unlike the case of the previous read command, since the data from the higher-level device has been written to the replacement destination sector by the previous write command, the replacement destination unwritten flag of the map 102 shown in FIG. 8 is set to “0”. Has been reset. Therefore, the process proceeds to the next step S36, and the error information is not set in step S46.

In step S36, a positioning operation for the track N is performed, and when the positioning is completed, step S37 is performed.
To read the sector 3 of the track N at step S3
Proceed to 8. In step S38, since there are still remaining sectors (track N sector 4), the process returns to step S33.
Then, step S33, step S34, step S
35, when all sectors are read through steps S36 and S37, steps S38 to S3
Go to 9. In step S39, since the error sector flag at the time of replacement has been reset to 0, the flow advances to step S40 to report the normal end to the host device, and end the read command. After the completion of the command, the process returns to step S31 and waits for the next read command.

The host device sends a command for updating the alternative map 9 to the interface control means 1 when the power of the entire system is turned off. The interface control means 1 activates the command execution means 2 and, when recognizing that it is an alternative map update command, activates the read / write execution means 3. The read / write execution means 3 writes the alternative map 9 in the memory 5 into the area of the system cylinder 16 if the alternative map non-write flag 13 is set. After that, the alternative map non-writing flag 13 is reset to 0. At this time, the contents of the map 102 shown in FIG. 8 are written into the substitute map 9 and read into the substitute map 9 on the memory 5 from the area of the system cylinder 16 at the next power-on.

As described above, according to this embodiment,
Since the writing of data to the replacement destination sector and the writing of the replacement map for mapping the replacement source sector and the replacement destination sector to the system cylinder during the replacement process are not performed, the replacement process can be performed at high speed. effective.

When the load on the magnetic disk is high and an image is lost when recording by the conventional substitution processing, the substitution processing is not performed at all but instead of the substitution processing. Since only the update is performed, the time required for the substitution process can be reduced. In this case, the data is not written to the replacement destination, but is added to the information of the conventional replacement source sector address and replacement destination sector address on the replacement map,
By providing a replacement destination unwritten flag for each piece, any data of the replacement destination is returned at the time of reading, and the replacement processing is terminated halfway at the end of the command (the data intended by the host device is not written in the replacement destination). ) An error report can be returned with the address of that sector.

Further, the higher-level device that has received the error can identify an incorrect portion of the data, so that the image can be dealt with by image correction means such as interpolation at the time of image reproduction, and writing to the same sector can be performed. As a result, the data from the host device is written to the replacement destination sector, and the replacement process is completed. Thereafter, reading / writing to the same sector is performed normally, and the recording capability is restored.

When the write cache is used,
As described above, it is difficult to notify the host device of the abnormal termination of the substitute process by the write command. In addition, an error can be returned when an error occurs, but at this time, the data for recovery has already been
It is unlikely that the data is stored in the host device, and it is meaningless to return an error at the time of writing. Nevertheless, the erroneous sector must be transmitted to the higher-level device in some way, and in that sense, the method of returning an error at the time of reading using the above-mentioned alternative destination unwritten flag is effective.

The replacement map is written to the system cylinder by issuing a command to the disk device by the host device when the power of the host device is turned off.
This is because the replacement map is not written when the load on the disk device is high, that is, when moving image data or the like is recorded / reproduced on multiple channels, so that the disturbance of the image during recording / reproduction is minimized. It is possible to provide an alternative processing method and apparatus for a magnetic disk that can be suppressed to a minimum.

Further, since the updated replacement map needs to be written only once to the system cylinder, the processing time of the magnetic disk can be greatly reduced as compared with the case where the replacement map is written by individual replacement processing. And an apparatus can be provided.

Next, another embodiment of the present invention will be described with reference to FIG.
This will be described with reference to FIG. The error address 15 at the time of substitution in FIG. 2 returns one address to the host device with one read command, but cannot return an error in a plurality of sectors. Therefore, in order to return a plurality of errors, a plurality of alternative error addresses are provided, and an alternative error counter is provided instead of the alternative error flag 14. This counter is set to "0" at the start of reading, and is incremented each time an error sector address is written to a plurality of alternative error addresses (step S4 in FIG. 6).
6). At the end of reading, it is checked whether the error counter at the time of substitution is "0", and if it is not "0", the error is ended. Thereafter, the higher-level device issues a command for retrieving the error sector address, and retrieves a series of substitution-time error addresses (step S41 in FIG. 6).

As described above, according to this embodiment, the same effects as those of the above-described embodiment can be obtained even when an error occurs in a plurality of sectors.

[0084]

As described above, according to the present invention, in the replacement processing in units of storage areas when an error occurs when accessing the recording medium, data is not written to the replacement destination in the memory. Since only the replacement map is updated, and the replacement map is written on the recording medium based on a command issued from the host device when the host device is powered off, It is possible to eliminate the time required for writing the replacement map to the recording medium and for writing the data to the replacement destination, and the time required for the replacement processing can be reduced to a minimum.

Further, when reading in the storage area unit of the replacement part is performed, the host apparatus is notified that the data of the replacement destination could not be written, so that the time spent for the replacement processing can be reduced to a minimum. However, there is an effect that the host device can easily perform the recovery process such as recognition of an illegal portion of the data and writing of the data from the host device to the replacement destination.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an apparatus for implementing a magnetic disk alternative processing method according to an embodiment of the present invention.

FIG. 2 shows a format of a replacement map stored in a memory of a replacement processing device for a magnetic disk according to an embodiment of the present invention, a replacement map non-writing flag, a replacement error flag,
FIG. 9 is an explanatory diagram showing an error sector address at the time of replacement.

FIG. 3 is an explanatory diagram showing a format in a medium such as a magnetic disk of the alternative processing device according to the embodiment of the present invention.

FIG. 4 is a flowchart illustrating an operation when a command is received from a higher-level device in the magnetic disk alternative processing device according to the embodiment of the present invention;

FIG. 5 is a flowchart showing a write processing operation by a read / write execution unit in the magnetic disk alternative processing device according to the embodiment of the present invention;

FIG. 6 is a flowchart showing a read processing operation by a read / write execution unit in the magnetic disk alternative processing device according to one embodiment of the present invention;

FIG. 7 is an explanatory diagram showing a format of a track on a user cylinder in a medium such as a magnetic disk of the alternative processing device according to an embodiment of the present invention.

FIG. 8 is an explanatory diagram showing each state of a substitution map in the magnetic disk substitution processing device according to the embodiment of the present invention;

[Explanation of symbols]

3 ... Read / write execution means (access means), 5 ...
... memory, 7 ... buffer (access means), 8 ... memory (recording medium), 9 ... alternative map, step S2
3, step S24, step S25, step S2
6, Step S41 ... Alternative means.

Claims (5)

[Claims] 1. An alternative processing method for a recording medium that uses an alternative map and performs an alternative processing for each storage area in which an error has occurred. And writing the replacement map on the recording medium in a lump, and notifying the host device that the data of the replacement destination could not be written when reading the storage area unit of the replacement part, A replacement processing method for a recording medium, comprising: performing a recovery process such as recognizing an incorrect portion of the data on the device side and writing the data from the higher-level device to the replacement destination. 2. The recording medium according to claim 2, wherein an alternative map on a memory is collectively written on a recording medium on the basis of a command issued by the host device when the power of the host device is turned off. Alternative processing method. 3. An alternative processing device for a recording medium which performs an alternative process for each storage area in which an error has occurred using an alternative map, wherein: an interface control means for receiving a command from a higher-level device; Command execution means for analyzing a received command; access means for reading and writing to the recording medium according to the command; error detection means for detecting occurrence of an error at the time of writing by the access means; A positioning execution unit for performing positioning on an arbitrary track above, and when the error detection unit detects the occurrence of an error, executes substitution processing using a substitution source storage area unit address and a substitution destination storage area unit address, Update only the alternate map in memory and do not write data to the alternate destination. Means for notifying the upper-level device that data of the replacement destination could not be written when the storage area unit of the replacement location is read. Processing equipment. 4. The replacement means notifies the host device that the data of the replacement destination could not be written, thereby recognizing an incorrect portion of the data by the host device and the replacement destination from the host device. 4. The alternative processing apparatus for a recording medium according to claim 3, wherein the data can be written to the recording medium. 5. The replacement map on a memory is written to a system cylinder of a recording medium based on a command issued by a host device when the power supply of the host device is turned off. Alternative processing device for recording media.