JPH09231014A - Defect processor for storage medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make possible defect processing of a storage medium without damaging a real time property by restoring data for which an error is generated from the data of another storage medium other than the storage medium for which the error is generated. SOLUTION: A retrial control part writes the data to plural hard disks 9 for the data, writes the parity of the data to the other hard disk 10 for the parity, and when a write error is generated, retries by the minimum number of times of retrials. An error registration part stores information relating to the data unrestorable by the minimum number of times of the retrials in an error table 5 as the error. An alternate sector control part reads error information, restores the data for which the error is generated from the data of the plural hard disks 9 other than the hard disk for which the error is generated and the other hard disk 10, retries by the maximum number of times of the retrials and allocates an alternate sector to the data unrestorable by the maximum number of times of the retrials.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a storage medium defect processing device for efficiently performing writing and reading error recovery of a hard disk used in, for example, a video server.

[0002]

2. Description of the Related Art Conventionally, based on the demands of many users,
There has been a video server device that simultaneously supplies various images to individual users. The video server device includes a video stream server that performs a process of transmitting a video stream, and an application server that performs non-real-time processing such as a navigation function.

[0003] On the other hand, an application server plays a central role in exchanging information, and has a network protocol processing function necessary for exchanging information. The application server has a management database, and manages applications, management information, customer information, and the like in the management database.

On the other hand, the video stream server has a plurality of hard disks for storing data, one hard disk for storing parity, and RAI for writing and reading data to and from these hard disks.
D (Redundant Arrays Inexpe
It is provided with a (native Discs) controller and a processor for converting and rearranging data and transmitting the data. RAI this video stream server
It is called D device. There are three requirements for this video stream server: time continuity, simultaneous provision of the same video in parallel, and special playback processing.

The term "time continuity" as used herein means the following. That is, it means that once the data accumulated in the video stream server is output, the data is continuously read for 2 hours without interruption. Further, the simultaneous provision of the same video in parallel means the following. That is, many users request one image or various images in parallel to the video stream server. This request is completely asynchronous. Even if the same image is requested, the time point of the request is slightly deviated. That is, hundreds or thousands of video streams must be output at different cue points at the same time. Therefore, real-time performance is required in terms of transmission. The special reproduction process means the following. That is, even if the video is started once, the user has to perform various control of the video stream, such as pause, rewind, and slow playback.

The operation of such a video stream server will be described below. Before recording the video, MPEG
A video signal encoded with high efficiency such as 2 is 64K to 25K.
It is divided into segment data of about 6 Kbytes. The divided segment data is stored in a plurality of data storage hard disks. This is because the number of simultaneous users for the same video can be increased by distributing and storing the data in a plurality of hard disks. There are a random layout and a striping layout as methods of this distributed storage. The random layout is suitable for storing multi-rate video, and the striping layout is characterized by high disk usage efficiency. The segment data distributed and accumulated in the striping layout is called a striping unit.

At the time of reproducing an image, the RAID controller reads the striping unit from the hard disk according to an allocation table showing a data array indicating on which disk the striping unit is stored. Then, the processor rearranges the read striping units in the reproduction order, absorbs fluctuations, and sends them to the network. In addition, special playback functions such as slow playback, fast forward, pause, skip, etc.
It is realized by adjusting the speed at which the controller reads the striping unit.

Further, the RAID controller distributes the striping units to a plurality of hard disks at the time of writing data, and condenses the striping units, which are distributed oppositely at the time of reading data, into one data stream.

In such a RAID device, a hard disk for writing parity is provided in addition to storing normal data in a plurality of hard disks for storing data, so that when a failure occurs, a hard disk for parity is used. The error data is recovered based on the data of. In addition, for an access error that occurs in each hard disk, a retry is performed a sufficient number of times to recover the error. Then, even in this retry, if the error recovery is impossible, the sector replacement process is performed.

[0010]

However, in such a conventional RAID device, there are cases in which the time required for error recovery processing for an access error becomes long and unpredictable. In such a case, the above-mentioned problem may occur. However, there was an inconvenience that the real-time property was lost. Also,
When a hard disk read error occurs, the number of retries is minimized, and when the number of retries exceeds the specified number, the error data is recovered based on the parity hard disk data without recovery by retry. By doing so, it is possible to prevent an increase in read time. However, since it is not possible to tell whether the read error is due to a media defect or an accidental cause, the replacement sector is allocated even if the error cause is due to a media defect. However, there is an inconvenience that the data cannot be rewritten.

The present invention has been made in view of the above points, and an object of the present invention is to provide a storage medium defect processing apparatus capable of performing a defect processing of a storage medium without deteriorating the real-time property.

[0012]

A storage medium defect processing apparatus according to the present invention writes data to a plurality of storage media and at the same time writes the parity of the data to another storage medium.
If a write error occurs when writing data to the plurality of storage media and the other storage medium, a retry control unit that performs a retry with a minimum number of retries, and information regarding unrecoverable data with the minimum number of retries are provided. An error registration control unit that stores an error in the error table and error information registered in the error table is read, and an error occurs from the data of the plurality of storage media other than the storage medium in which the error has occurred and the data of the other storage media. The data is restored, the retry is performed with the maximum number of retries, and the alternate sector control unit assigns the alternate sector to the data that cannot be recovered with the maximum number of retries.

Further, in the storage medium defect processing apparatus of the present invention, data is written in a plurality of storage media and the parity of the data is written in another storage medium, and the plurality of storage media and the other storages are stored. If a read error occurs when reading the data written to the medium, a retry control unit that retries with the minimum number of retries,
An error registration control unit that stores information on unrecoverable data as an error in the error table with the minimum number of retries, and the error information registered in the error table to read the storage medium in which the error occurred and the other storages. A plurality of storage media other than the storage medium in which an error occurs with respect to the data which cannot be recovered by the maximum number of retries by performing a retry with the maximum number of retries so as to read the data of the medium and the other storage medium. And a replacement sector control unit for allocating replacement sectors by restoring the data in which an error has occurred from the above data.

According to the defect processing apparatus for a storage medium of the present invention, the following effects are obtained. First, the write operation will be described. When there is a write request, the write operation is started,
The data dividing means of the retry control unit divides the data into striping units. The striping unit writing means of the retry controller writes the striping unit on the hard disk for data storage. In this case, the write operation is performed so that the striping units are dispersedly stored in a plurality of data storage hard disks.

The write error detection means of the retry controller determines whether or not there is an error. When there is an error, the minimum retry count determination means of the retry control unit determines whether the retry count has exceeded. The minimum number of times is the minimum number of times within the allowable execution time range. If the number of retries is exceeded, the error table registration means of the error registration control unit registers the error information in the error table. In this case, the error table registration means stores the number of the hard disk in which the error occurred and the sector address in which the error occurred in the error table from the status signal.

When there is no error, the parity writing means writes the parity, which is the exclusive OR of the data written in the data hard disk, in the parity hard disk. The write error detection means of the retry control unit determines whether there is an error. When there is an error, the minimum retry count determination means of the retry control unit determines whether the retry count has exceeded. If the number of retries is exceeded, the error table registration means of the error registration control unit registers the error information in the error table. After registering in the error table, it is determined whether or not the writing operation is completed.
The data writing process and the parity writing process are executed in parallel. This completes the writing of materials such as a series of programs.

The processing described below is executed in the idle time or in the background processing of normal processing. The error table reading means of the alternate sector control unit reads the error sector address of the hard disk in which the error occurred from the error table. The error table reading means of the alternate sector control unit reads the drive number of the hard disk in which the error occurred from the error table. The striping unit restoring means of the alternate sector control unit restores the striping unit in which the error has occurred based on the parity operation. In this case, since the exclusive OR of the data written to the data hard disk is calculated and written to the parity hard disk, the data written to the data hard disk other than the one in which the error occurred and the parity hard disk are written. Data can be restored by using the parity. In this way, the data is reconstructed from the data of the hard disks other than the hard disk in which the error has occurred, including the hard disk for parity.

The restoration striping unit writing means of the alternate sector control unit writes the restored striping unit to the sector address where the error occurred in the hard disk where the error occurred. The write error detection means of the alternate sector control unit determines whether or not there is another data write error in the hard disk in which the error has occurred. When there is an error, the maximum retry count determination means of the alternate sector control unit determines whether or not the maximum retry count for the write error has been exceeded. The maximum number of times is a sufficient number of times for recovering the write error determined according to the characteristics of the hard disk.

When the number of retries is exceeded, the replacement sector allocating means of the replacement sector control unit processes the sector in which the write error has occurred as a defective sector and allocates the replacement sector. Such a replacement process is realized by designating a sufficient number of retries for error recovery on the hard disk and issuing a write command in a mode in which the automatic replacement process is permitted. When there is no error, it is determined whether or not the error recovery processing is completed for all the sectors in the error table. When the error table is finished, the write operation is finished.

Next, the read operation will be described. When a read request is issued, the read operation is started, and the striping unit read means of the retry control unit performs the read operation of the striping units distributed and accumulated in the plurality of hard disks for storing data. The read error detection means of the retry control unit determines whether or not there is an error in reading data from a plurality of hard disks for storing data. The minimum retry number determination means of the retry control unit determines whether or not the retry number set to the minimum for the read error is exceeded. The minimum number of times is the maximum number of times within the allowable execution time range.

If the minimum number of retries has been exceeded, the error table registration means of the error registration control unit stores the number of the hard disk in which the error occurred and the sector address in which the error occurred in the error table based on the status signal. The striping unit restoring means of the error registration control unit restores the striping unit in which the error has occurred based on the parity operation. The data synthesizing means of the error registration control unit synthesizes the original data from the data divided into the striping units. This video data is supplied to the data supply source. As a result, the combined video data is transmitted. This completes the reading of materials such as a series of programs.

The processing described below is executed in the idle time or in the background processing of normal processing. The error table reading means of the alternate sector control unit reads the errored sector address of the hard disk where the error occurred from the error table. The error table reading means reads the number of the hard disk in which the error occurred from the error table. The striping unit reading means rereads the data from the hard disk in which the error has occurred. The read error detection means of the alternate sector control unit determines whether or not there is an error in reading data from the hard disk in which the error has occurred.

The maximum retry count judging means of the replacement sector control unit judges whether or not the maximum retry count has been set for a read error. Such processing and determination are realized by designating a sufficient number of retries for error recovery and issuing a read command to a sector of a hard disk in which an error has occurred, including a hard disk for parity. When the maximum number of retries is exceeded, the striping unit restoring means of the alternate sector control unit restores the striping unit in which the error has occurred based on the parity operation. In this case, including the hard disk for parity, from the data of the hard disk other than the hard disk where the error occurred,
Restore the striping unit.

The replacement sector allocation means of the replacement sector control unit processes the sector in which the write error has occurred as a defective sector and allocates the replacement sector. The restoration striping unit writing means of the alternate sector control unit writes the restored striping unit to the sector address where the error occurred on the hard disk where the error occurred. Such processing is executed by designating a hard disk in which a read error has occurred with a sufficient number of retries for error recovery and issuing a write command in a mode in which automatic replacement processing is permitted. It is determined whether or not the error recovery processing is completed for all the sectors in the error table. When the error table is finished, the read operation is finished.

[0025]

BEST MODE FOR CARRYING OUT THE INVENTION This embodiment will be described below. First, the configuration of the present embodiment will be described. The storage medium defect processing apparatus of this embodiment shown in FIG. 1 corresponds to a video stream server in a video server apparatus. This storage medium defect processing apparatus includes a plurality of hard disks 9 for data, a single hard disk 10 for parity, and a RAID (Redundant Ar) for writing and reading data to and from these hard disks.
It is provided with a ray Inexpensive Discs) controller 2 and a processor 1 for transmitting data by converting and rearranging data. This video stream server is called a RAID device. There are three requirements for this video stream server: time continuity, simultaneous provision of the same video in parallel, and special playback processing. The three conditions are as described in the related art.

The data hard disk 9 constitutes a storage medium, and the parity hard disk 10 constitutes another storage medium. The RAID controller 2 is a CP that controls the control.
It has a U3, a ROM 4 for storing a control program, and a RAM 6 for storing control data and an error table 5. The RAID controller 2 has a write control unit 7 and a read control unit 8 in functional blocks. The write controller 7 controls writing of data to the hard disk. The read control unit 8 controls the reading of data from the hard disk.

Next, the structure of the write controller will be described with reference to FIG. The write control unit 7 is, in terms of functional blocks, a retry control unit 11, an error registration control unit 12,
The replacement sector control unit 13 is included. Retry controller 1
1 is a write error when writing data to a plurality of data hard disks 9 and at the same time writing the parity of this data to another parity hard disk 10 and writing data to the plurality of data hard disks 9 and another parity hard disk 10. When occurs, control is performed to retry with the minimum number of retries. The error registration control unit 12 performs control to store information regarding unrecoverable data as an error in the error table 5 with the minimum number of retries. The alternate sector control unit 13 reads the error information registered in the error table 5, restores the error data from the data of the plurality of hard disks 9 and other hard disks 10 other than the hard disk in which the error occurred, and performs the maximum retry. The control is performed by retrying the number of times and assigning the replacement sector to the data that cannot be recovered by the maximum number of times of retrying.

The retry control section 11 includes the data dividing means 1
4, striping unit write means 15, parity write means 16, and write error detection means 1
7 and a minimum retry count determination means 18. The data dividing unit 14 divides the data, which has been encoded with high efficiency such as MPEG2 in advance, into striping units of a predetermined unit. The striping unit is a group of segment data which is distributed and accumulated in the hard disk 9 for a plurality of data. The reason for the distributed storage is that the number of simultaneous users for the same video can be increased by the distributed storage in the plurality of data hard disks 9. There are a random layout and a striping layout as methods of this distributed storage. The random layout is suitable for storing multi-rate video, and the striping layout is characterized by high disk usage efficiency. The segment data distributed and accumulated in the striping layout is called a striping unit.

Striping unit writing means 15
Performs a write operation so as to disperse and store striping units in a plurality of hard disks 9 for data.
This write operation is performed by S supplied from the processor 1.
It is performed based on the CSI command. The parity writing means 16 writes the exclusive OR of the data written in the data hard disk 9 to the parity hard disk 10. This write operation is also performed based on the SCSI command supplied from the processor 1.

The write error detection means 17 detects a data write error in the hard disk 9 for a plurality of data. This error detection operation is performed by detecting the status signal for the SCSI command. The minimum retry count determination means 18 determines that the write retry has been executed with the retry count set to the minimum with respect to the write error. The minimum number of retries is set based on the mode parameter for the SCSI command. The determination that the number of retries has been exceeded is made by detecting the mode parameter and the status signal. The minimum number of times is the maximum number of times within the allowable execution time range.

The error registration control section 12 has an error table registration means 19. The error table registration means 19
The number of the hard disk where the error occurred and the sector address where the error occurred are stored in the error table from the status signal. The reason for registering in the error table after the minimum number of retries is that the write execution time increases as the number of retries increases. If the number of retries increases, the number of replacement sectors increases because it is determined that the sectors that can be recovered by retries are defective. Because.

The alternate sector control unit 13 includes an error table reading means 20, a striping unit restoring means 21, and a restoring striping unit writing means 2.
2, a write error detection means 23, a maximum retry count determination means 24, and an alternate sector allocation means 25. The error table reading means 20 reads the number of the hard disk in which the error occurred and the sector address in which the error occurred from the error table 5. The striping unit restoring means 21 restores the striping unit in which the error has occurred based on the parity calculation. That is, in the parity hard disk 10,
The exclusive OR of the data written in the data hard disk 9 is calculated and written. Therefore, the data can be restored by the data written in the data hard disk 9 and the parity written in the parity hard disk 10 except for the error.

The restoration striping unit writing means 22 writes the restored striping unit to the sector address where the error occurred in the hard disk where the error occurred. This write operation is performed based on the SCSI command supplied from the processor. The write error detection means 23 detects an error in writing data to the hard disk in which the error has occurred.
This error detection operation is performed by detecting the status signal for the SCSI command.

The maximum retry count judging means 24 judges that the write retry of the maximum retry count has been executed for the write error. The maximum number of retries is set based on the mode parameter for the SCSI command. The determination that the number of retries has been exceeded is made by detecting the status signal. This maximum number is the number of retries sufficient for error recovery of the hard disk. The replacement sector allocation means processes a sector in which a write error has occurred as a defective sector and allocates a replacement sector.

In this way, the replacement sector control unit 13
Reads the data of the sector indicated by the sector address registered in the error table 5 in the idle time, restores the correct data, and rewrites the hard disk in which the error occurred with a sufficient number of retries. Therefore, when the recovery cannot be performed by the rewriting, the defective sector is determined and the replacement sector is assigned to perform the rewriting.

Next, the structure of the read control unit will be described with reference to FIG. The read control unit 8 is, in terms of functional blocks, a retry control unit 26, an error registration control unit 27,
The replacement sector control unit 28. Retry control unit 2
6, the data is written to the plurality of data hard disks 9 and the parity of this data is written to the other parity hard disks 10, and the plurality of data hard disks 9 and the other parity hard disks 1 are written.
If a read error occurs when reading the data written in 0, control is performed to retry with the minimum number of retries.

The error registration control section 27 performs control to store information regarding unrecoverable data in the error table 5 as an error with the minimum number of retries. The alternate sector control unit 28 reads the error information registered in the error table 5 and retries with the maximum number of retries so as to read the data of the data hard disk 9 and the other parity hard disk 10 in which the error has occurred,
Control for allocating an alternate sector by restoring the error data from the data of a plurality of data hard disks 9 other than the error hard disk and the other parity hard disk 10 for the unrecoverable data with the maximum number of retries I do.

The retry controller 26 includes a striping unit reading means 29 and a read error detecting means 3.
It has 0 and the minimum retry count judging means 31. The striping unit reading means 29 performs the reading operation of the striping units that are distributed and accumulated in the plurality of data hard disks 9. This read operation is
This is performed based on the SCSI command supplied from the processor 1. The read error detection means 30 detects a data read error with respect to the plurality of data hard disks 9. This error detection operation is performed by detecting the status signal for the SCSI command. The minimum retry count determination means 31 determines that the read retry with the retry count set to the minimum for the read error has been executed. The minimum number of retries is set based on the mode parameter for the SCSI command. The determination that the number of retries has been exceeded is made by detecting the status signal.

The error registration control section 27 includes an error table registration means 32 and a striping unit restoration means 33.
And data synthesizing means 34. The error table registration means 32 stores the number of the hard disk where the error occurred and the sector address where the error occurred from the status signal in the error table. The striping unit restoring means 33 restores the striping unit in which the error has occurred based on the parity calculation. The data synthesizing means 34 synthesizes original data from the data divided into striping units. As a result, the combined data is transmitted.

In this way, the retry controller 26
In response to a read request from the processor 1, the striping unit is read, reconstructed into the original data and returned to the processor 1. When an error occurs, the striping unit is restored by the parity operation, the original data is reconstructed, and the data is returned to the processor 1.

The replacement sector control unit 28 includes an error table reading unit 35, a striping unit reading unit 36, a reading error detecting unit 37, a maximum retry number judging unit 38, a striping unit restoring unit 39, and a replacement sector allocating unit. 40 and restoration striping unit writing means 41.
The error table reading means 35 uses the error table 5
Read the hard disk number where the error occurred and the sector address where the error occurred. The striping unit reading means 36 rereads the data from the hard disk in which the error has occurred. The read error detection means 37 detects an error in reading data from the hard disk in which the error has occurred. This error detection operation is performed by detecting the status signal for the SCSI command.

The maximum retry count judging means 38 judges that the read retry with the maximum retry count for the read error has been executed. The maximum number of retries is set based on the mode parameter for the SCSI command. The determination that the number of retries has been exceeded is made by detecting the status signal. The striping unit restoring means 39 restores the striping unit in which the error has occurred based on the parity calculation.

The replacement sector allocation means 40 processes the sector in which the write error has occurred as a defective sector and allocates a replacement sector. The restoration striping unit writing means 41 writes the restored striping unit to the sector address where the error occurred in the hard disk where the error occurred. This write operation is performed based on the SCSI command supplied from the processor 1.

In this way, the replacement sector control unit 28
Reads out the data indicated by the sector address of the hard disk registered in the error table 5 in the free time. At this time, if you can recover with the specified number of retries,
Judge as a normal sector. If it cannot be recovered, it is determined as a defective sector, a replacement sector is assigned, and correct data is restored and written in the replacement sector.

The operation of the storage medium defect processing apparatus of the present embodiment thus constructed will be described with reference to the block diagram of FIG. 1 and the flow charts of FIGS. First, the write operation will be described with reference to FIGS. 6 and 7.
The flowcharts shown in FIGS. 6 and 7 correspond to the operation of the write control unit shown in FIG. When a write request is issued from the host computer, the command and video data are supplied to the processor shown in FIG. The processor 1 converts the supplied command into a SCSI command. The video data is digitally formatted in the D1 or D2 format. The video data may be compressed in advance or may not be compressed in advance. When the compression process is not performed, the processor 1 compresses the video data by MPEG2.
The processor 1 creates a file system for writing video data. In other words, a file system is created for which block size of which video data is transferred to which block address.

The processor 1 supplies the converted SCSI command and video data to the RAID controller 2 in each phase. RAID
The write operation of the controller 2 will be described with reference to the functional block diagram of the write controller shown in FIG.

In FIG. 6, after starting the write operation,
In step S1, the data dividing means 1 of the retry control unit 11
4 divides the data into striping units. The striping unit is a predetermined unit, but may be a byte unit, a bit unit, or a sector unit, and this is a unit for writing. In step S2, the striping unit writing means 15 writes the striping unit on the data hard disk 9. In this case, the write operation is performed so as to disperse and store the striping units in the plurality of data hard disks 9. This write operation is performed based on the SCSI command supplied from the processor 1.

In step S3, the write error detecting means 1
7 determines whether or not there is an error. This error detection operation is performed on the hard disk 9 for a plurality of data by SC
This is done by detecting the status signal for the SI command. If there is an error in step S3, the minimum retry count determination means 18 determines in step S4 whether or not the retry count has exceeded. In this case, the minimum number of retries is set based on the mode parameter for the SCSI command. The determination that the number of retries has been exceeded is made by detecting the status signal. The minimum number of times is the maximum number of times within the allowable execution time range. In other words, it is a value obtained by dividing the allowable execution time by one retry execution time, for example, once.

If the number of retries is exceeded in step S4, the error table registration means 19 of the error registration control unit 12 registers the error information in the error table 5 in step S5. In this case, the error table registration means 19 stores the number of the hard disk where the error occurred and the sector address where the error occurred from the status signal in the error table. For example, as shown in FIG. 5, it is recorded that the address 2 of the HDD 3 of the data hard disk 9 is in error.

When the number of retries has not exceeded in step S4, the process returns to step S2, and the processes and determinations from step S2 to step S4 are repeated. If there is no error in step S3, the parity writing means 16 writes the parity which is the exclusive OR of the data written in the data hard disk 9 to the parity hard disk 10 in step S6. This write operation is also performed based on the SCSI command supplied from the processor 1.

In step S7, the write error detecting means 1
7 determines whether or not there is an error. This error detection operation is the same as step S3. If there is an error in step S7, the minimum retry count determination means 18 determines in step S8 whether the retry count has exceeded.
The minimum retry count and the determination that the retry count has been exceeded are the same as in step S3.

If the number of retries is exceeded in step S8, the error table registration means 19 of the error registration control unit 12 registers the error information in the error table 5 in step S9. The content of the error information is the same as in step S4. For example, as shown in FIG. 5, it is recorded that the address 4 of the HDDP of the parity hard disk 10 is in error. When the number of retries has not exceeded in step S8, the process returns to step S6, and the processes and determinations from step S6 to step S8 are repeated. Then, after registration in the error table 5 in steps S5 and S9, it is determined in step S10 whether or not the writing operation is completed. Steps S6 to S9 are executed in parallel with steps S2 to S5. When the process ends in step S10, the process proceeds to. Step S10
If not completed, the procedure returns to step S1 and returns to step S1.
The processes and determinations from to S10 are repeated.
At the end of step S10, the writing of materials such as a series of programs ends.

The processing after step S11 described below is executed in the idle time or in the background processing of normal processing. Referring to FIG. 7, in step S11, the error table reading means 2 of the replacement sector control unit 13 is proceeded to.
0 reads the errored sector address of the hard disk where the error occurred from the error table 5. In step S12, the error table reading means 20 of the replacement sector control unit 13 reads the drive number of the hard disk in which the error occurred from the error table 5.

In step S13, the striping unit restoring means 21 restores the striping unit in which the error has occurred based on the parity calculation. In this case, since the exclusive OR of the data written in the data hard disk 9 is calculated and written in the parity hard disk 10, the data written in the data hard disk 9 other than the one in which the error has occurred and the parity Data can be restored by the parity written in the hard disk 10. In this way, data is reconstructed from the data of the hard disks other than the hard disk in which the error has occurred, including the hard disk 10 for parity.

In step S14, the restoring striping unit writing means 22 writes the restored striping unit in the sector address where the error occurred in the hard disk where the error occurred. This write operation is performed based on the SCSI command supplied from the processor 1. In step S15, the write error detection means 2
3 determines whether or not there is another data writing error in the hard disk in which the error has occurred. This error detection operation is performed by detecting the status signal for the SCSI command.

If there is an error in step S15, the maximum retry count judging means 24 judges in step S16 whether or not the maximum retry count for the write error has been exceeded. The setting of the number of retries and the determination that the number of retries has been exceeded are the same as in steps S4 and S8, but the difference from steps S4 and S8 is that the number of retries is set to the maximum.

If the number of retries is exceeded in step S16, the replacement sector allocation means 25 is operated in step S17.
Treats a sector in which a write error has occurred as a defective sector and assigns a replacement sector. For example, FIG.
As shown in, the address 2 of the HDD 3 of the data hard disk 9 is processed as the defective sector 42, and the address n is assigned as the replacement sector 43. Further, the address 4 of the HDDP of the parity hard disk 10 is processed as the defective sector 44, and the address k is allocated as the replacement sector 45.

Then, returning to step S14, the determination and processing from step S14 to step S17 are repeated. The processing from step S14 to step S17 is realized by designating the hard disk with a sufficient number of retries for error recovery and issuing a write command in a mode in which the automatic replacement processing is permitted. Step S1
If there is no error in step 5, it is determined in step S18 whether the error recovery process is completed for all sectors in the error table 5. When the error table is finished in step S18, the write operation is finished. Step S18
If the error table 5 is not finished in step S11, step S11
Returning to step S11, the judgment and processing from step S11 to step S18 are repeated.

Next, the read operation will be described with reference to FIGS. The flowcharts shown in FIGS. 8 and 9 correspond to the operation of the read control unit shown in FIG. When a read request is issued from the host computer, a command is supplied to the processor 1 shown in FIG. The processor 1 converts the supplied command into a SCSI command. The processor 1 supplies the SCSI command thus converted to the RAID controller 2.

The RAID controller 2 creates a file system for reading video data. In other words, a file system is created for which block size of which video data is transferred from which block address. Such a read operation of the RAID controller 2 is shown in FIG.
The description will be made in correspondence with the functional block diagram of the read control unit shown in FIG.

In FIG. 8, after starting the read operation,
In step S19, the striping unit reading means 29 of the retry control unit 26 performs the reading operation of the striping units dispersedly accumulated in the plurality of data hard disks 9. This read operation is performed based on the SCSI command supplied from the processor 1. In step S20, the read error detection means 30 determines whether or not there is an error in reading data from the hard disks 9 for a plurality of data. This error detection operation is performed by detecting the status signal for the SCSI command. In step S21, the minimum retry count determination means 31 determines whether or not the read error has exceeded the minimum retry count. The minimum number of retries is set based on the mode parameter for the SCSI command. The minimum number of times is the maximum number of times within the allowable execution time range. The determination that the number of retries has been exceeded is made by detecting the mode parameter and the status signal.

If the minimum number of retries has not been exceeded in step S21, the process returns to step S19, and the processes and determinations from step S19 to step S21 are repeated. If the minimum number of retries has been exceeded in step S21, the error table registration means 32 of the error registration control unit 27 determines the hard disk number in which the error occurred and the sector address in which the error occurred in step S22 based on the status signal. Table 5
To memorize. In step S23, the striping unit restoring means 33 restores the striping unit in which the error has occurred based on the parity calculation.

In step S24, the data synthesizing means 34 synthesizes the original data from the data divided into striping units. This video data is supplied to the processor 1. This video data is digitally formatted in the D1 or D2 format.
Further, when the processor 1 compresses the video data by MPEG2, the processor 1 expands the video data. As a result, the combined video data is sent out via the processor 1.

If there is no error in step S20, the retry is not performed, so the process directly proceeds to step S24. In step S25, it is determined whether the read operation is completed. When the read operation is not completed in step S25, the process returns to step S19 to repeat the determination and processing from step S19 to step S25. When the read operation is completed in step S25, the process proceeds to. At the end of step S25, the material reading of a series of programs is completed.

The processing after step S26 described below is executed in the idle time or in the background processing of normal processing. In step S26, the error table reading means 35 of the alternate sector control unit 28 reads the errored sector address of the errored hard disk from the error table 5. In step S27, the error table reading means 356 reads from the error table 5 the number of the hard disk in which the error occurred. In step S28, the striping unit reading means 3
6 rereads the data from the hard disk in which the error occurred. In step S29, the read error detection means 37 determines whether or not there is an error in reading data from the hard disk in which the error has occurred.
This error detection operation is performed by detecting the status signal for the SCSI command.

In step S30, the maximum retry count determination means 38 determines whether or not the maximum retry count for a read error has been exceeded. When the maximum number of retries has not been exceeded in step S30, the process returns to step S28, and the processes and determinations from step S28 to step S30 are repeated. The processes and determinations from step S28 to step S30 are performed by issuing a read command to the sector of the hard disk including the parity hard disk 10 in which an error has occurred by designating a sufficient number of retries for error recovery. Will be realized.

The setting of the number of retries and the determination that the number of retries has been exceeded are the same as in step S21, but the point different from step S21 is that the number of retries is set to the maximum.

When the maximum number of retries is exceeded in step S30, the striping unit restoring means 39 restores the striping unit in which the error has occurred based on the parity operation in step S31. in this case,
From the data of the hard disk other than the hard disk where the error occurred, including the parity hard disk 10,
Restore the striping unit.

In step S32, the replacement sector allocation means 40 processes the sector in which the write error has occurred as a defective sector and allocates a replacement sector. In step S33, the restoration striping unit writing means 41 writes the restored striping unit in the sector address where the error occurred in the hard disk where the error occurred. This writing operation is performed by the processor 1
It is performed based on the SCSI command supplied from.
The processes of steps S32 and S33 are executed by designating the hard disk in which the read error has occurred with a sufficient number of retries for error recovery and issuing a write command in a mode in which the automatic replacement process is permitted.

After that, the process returns to step S28, and the processes and judgments from step S28 to step S33 are repeated. Then, when there is no error in step S29, the process proceeds to step S34, and it is determined whether the error recovery process is completed for all the sectors of the error table 5. When the error table 5 is finished in step S34, the read operation is finished. When the error table 5 is not completed in step 34, the process returns to step S26, and the determination and processing from step S26 to step S34 are repeated.

According to the above example, the defect processing apparatus for this storage medium writes the data to the plurality of data hard disks 9 and at the same time writes the parity of the above data to the other parity hard disks 10 to write the above plurality of data hard disks. 9 and other parity hard disk 1
If a write error occurs when writing data to 0, a retry control unit 11 that retries with the minimum number of retries, and an error that stores information about unrecoverable data with the above minimum number of retries in the error table 5 as an error The error information registered in the error table 5 is read out from the registration control unit 12, and the error data is extracted from the data of the plurality of data hard disks 9 other than the error hard disk and the other parity hard disk 10. Since it is provided with the replacement sector control unit 13 that restores data, retries with the maximum number of retries, and allocates replacement sectors to data that cannot be recovered with the maximum number of retries, recovery cannot be performed with the minimum number of retries. Information related to data is recorded in error table 5 as an error. Then, divide the processing so that the data in which the error occurred is restored in another spare time and the replacement sector is assigned only to the data that cannot be recovered even if it is retried with the maximum number of retries. By doing this, even if a write error occurs during access of data such as video data and audio data that requires real-time characteristics during data writing, the time for error recovery can be minimized. Thus, it is possible to access without deteriorating the real-time property.

Further, according to the above example, in the defect processing apparatus for this storage medium, the data is written to the plurality of data hard disks 9 and the parity of the data is written to the other parity hard disks 10, and the plurality of data hard disks 9 are written. If a read error occurs when reading the data written in the data hard disk 9 and the other parity hard disk 10, the retry control unit 26 that retries with the minimum number of retries, and the recovery failure with the minimum number of retries. An error registration control unit 27 that stores information about possible data in the error table 5 as an error.
Then, the error information registered in the error table 5 is read out, and the data is read from the data hard disk 9 and the other parity hard disk 10 in which the error has occurred. Replacement of allocating a replacement sector by restoring error data from the data of the plurality of data hard disks 9 and the other parity hard disk 10 other than the hard disk in which an error has occurred for data that cannot be recovered by the number of retries Sector control unit 28
Therefore, even if information about unrecoverable data is stored in the error table 5 as an error with the minimum number of retries, and recovery is performed with the maximum number of retries at another spare time, recovery is not possible. The processing can be divided so that the data in which an error has occurred is restored only to the possible data and the replacement sector is assigned, which requires real-time processing of video data, audio data, etc. when reading data. Even if a read error occurs during data access, by minimizing the time for error recovery, it is possible to access without deteriorating the real-time property.

In the above example, step S19 of the read operation is performed.
From step S23 to S23, the example in which the parity is calculated after the read error occurs is described, but the parity may be calculated while reading. Further, the processes and judgments of steps S11 to S18 of the write operation and steps S26 to S34 of the read operation in the above example may be executed at this time by providing a time slot at the time of service interruption or at night. good.

[0074]

The storage medium defect processing apparatus of the present invention is
When writing data to multiple storage media, writing parity of the above data to other storage media, and writing error when writing data to the above multiple storage media and the above other storage media, the minimum retry count A retry controller that performs retries, an error registration controller that stores information about unrecoverable data in the error table as an error in the minimum number of retries, and error information registered in the error table, and an error occurs The data in which the error occurred is restored from the data of the above-mentioned multiple storage media other than the above-mentioned storage media and the data of the above-mentioned other storage media, and the retry is performed with the maximum number of retries. It has a replacement sector control unit that allocates replacement sectors with a minimum number of retries. Data that cannot be recovered even if information about unrecoverable data is stored as an error in the error table, and the data in which the error occurred is restored at another spare time and retries are performed with the maximum number of retries. It is possible to divide the processing so that the replacement sector is assigned only to, even if a write error occurs during the access of data such as video data and audio data that requires real-time processing, when writing data. By minimizing the time for error recovery, it is possible to access without deteriorating the real-time property.

Further, in the storage medium defect processing apparatus of the present invention, the data is written in a plurality of storage media and the parity of the data is written in another storage medium, and the plurality of storage media and the other storages are stored. If a read error occurs when reading the data written to the medium, a retry control unit that retries with the minimum number of retries,
An error registration control unit that stores information on unrecoverable data as an error in the error table with the minimum number of retries, and the error information registered in the error table to read the storage medium in which the error occurred and the other storages. A plurality of storage media other than the storage medium in which an error occurs with respect to the data which cannot be recovered by the maximum number of retries by performing a retry with the maximum number of retries so as to read the data of the medium and the other storage medium. Since it has an alternate sector control unit that restores the data in which an error has occurred from the data of No. 1 and assigns the alternate sector, information about unrecoverable data with the minimum number of retries is stored as an error in the error table, For data that cannot be recovered even if you retry with the maximum number of retries in another spare time It is possible to divide the processing so that the data in which the error occurred is restored and the replacement sector is allocated. This allows the data to be read during the access of data such as video data and audio data that requires real-time processing. Even if an error occurs, by minimizing the time for error recovery, it is possible to access without deteriorating the real-time property.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of a defect processing apparatus for a storage medium according to the present embodiment.

FIG. 2 is a functional block diagram of a write control unit of the defect processing device for a storage medium according to the present embodiment.

FIG. 3 is a functional block diagram of a read control unit of the storage medium defect processing apparatus of this embodiment.

FIG. 4 is a diagram illustrating a replacement sector of the defect processing apparatus for the storage medium according to the present embodiment.

FIG. 5 is a diagram illustrating an error table of the storage medium defect processing apparatus according to the present embodiment.

FIG. 6 is a flowchart showing the write operation of the defect processing apparatus for a storage medium of the present embodiment.

FIG. 7 is a flowchart showing a write operation of the storage medium defect processing apparatus of the present embodiment.

FIG. 8 is a flowchart showing a read operation of the defect processing apparatus for a storage medium according to the present embodiment.

FIG. 9 is a flowchart showing a read operation of the defect processing device for a storage medium according to the present embodiment.

[Explanation of symbols]

1 processor, 2 RAID controller, 3 C
PU, 4 ROM, 5 error table, 6 RAM, 7
Write controller, 8 read controller, 9 data hard disk, 10 parity hard disk, 11
Retry control section, 12 error registration control section, 13 alternate sector control section, 14 data dividing section, 15 striping unit writing section, 16 parity writing section, 17 write error detecting section, 18 minimum retry count determining section, 19 error table registering section ,
20 error table reading means, 21 striping unit restoring means, 22 restoring striping unit writing means, 23 write error detecting means, 2
4 maximum retry count determination means, 25 replacement sector allocation means, 26 retry control section, 27 error registration control section, 28 replacement sector control section, 29 striping unit read means, 30 read error detection means, 31 minimum retry count determination means, 32 Error table registering means, 33 striping unit restoring means, 34 data synthesizing means, 35 error table reading means, 36 striping unit reading means, 37 read error detecting means, 38 maximum retry number judging means, 39 striping unit restoring means, 40 replacement sector Assigning means, 41 restoring striping unit writing means, 42 defective sectors, 43 replacement sectors, 44 defective sectors, 45
Replacement sector

Claims (2)

[Claims] 1. Writing data to a plurality of storage media and writing parity of the data to other storage media,
If a write error occurs when writing data to the plurality of storage media and the other storage media, a retry control unit that performs a retry with the minimum number of retries, and information regarding unrecoverable data with the minimum number of retries are provided. An error registration control unit that stores an error in the error table, and error information registered in the error table is read, and an error occurs from the data of the plurality of storage media other than the storage medium in which the error has occurred and the data of the other storage medium. Restore the data you have made, retry with the maximum number of retries,
A defect processing apparatus for a storage medium, comprising: a replacement sector control unit that allocates a replacement sector to data that cannot be recovered with the maximum number of retries. 2. Data is written to a plurality of storage media, parity of the data is written to another storage medium, and read when reading the data written to the plurality of storage media and the other storage medium. When an error occurs, a retry control unit that retries with the minimum number of retries, an error registration control unit that stores information about unrecoverable data with the minimum number of retries in an error table as an error, and register with the above error table The error information is read and the data in the storage medium in which the error has occurred and the data in the other storage medium is read, and the retry is performed with the maximum number of retries. Error from the data of the plurality of storage media other than the storage media in which the No data restores, and a replacement sector controller for allocating a replacement sector, the defect processing unit of the storage medium, characterized in that.