JPH04370576A - Data recording control system - Google Patents

Abstract

PURPOSE:To secure data of a medium before processing even if a fault is generated during the writing process in the case physical block on the medium consisting of plural logical blocks is partially updated, by saving the data for the relevant physical block data to an alternate block on the relevant medium. CONSTITUTION:In the recording medium, each physical block E, F is constituted of the plural blocks E0, E1...F0, F1..., and each track is constituted of the plural physical blocks, and this medium is provided with an user data recording area, alternate recording area for saving a defective block, and alternate block control area for recording a corresponding list of the defective block and alternate block. When the data of this recording medium are updated, the logical block data E0, E1, E2 on the user recording area of the physical block E corresponding to the partial update are saved (e) to the alternate recording area beforehand, and simultaneously this corresponding list 63/ is registered on the alternate block control area and then updated.

Description

[Detailed description of the invention]

0001

[Industrial Application Field] The present invention relates to a data recording control system, and in particular to a data recording control method using a rotating recording medium (hereinafter also simply referred to as "medium") having a physical block divided into a plurality of logical blocks. This paper relates to a data recording control method.

0002

2. Description of the Related Art In general, commands from a host CPU to a disk control device are issued in logical block units (one command targets one or more logical blocks), regardless of physical blocks. When recording logical block data smaller than the physical block length on a medium, multiple logical blocks are written together into one physical block. Also, for example, JP-A No. 6
As shown in Publication No. 4-29923 (hereinafter referred to as "Reference 1"), in a magnetic disk control device, a method (physical A method for making partial data within a block readable has been proposed. However, this document 1 does not report on the case where new data is added and updated to a partial (some) logical block within a physical block that has already been written.

On the other hand, since physical data writing to a disk recording medium can only be performed in units of physical blocks, in order to update data in a partial logical block within a certain physical block, the entire physical block (update It is necessary to read the data in the logical block (including logical blocks that are not targeted), replace the data in the logical block that requires data update with new data, and then record the updated data in the relevant physical block. If a failure such as a power failure occurs during the process, data in logical blocks that are not subject to update processing within the physical block may be destroyed. Therefore, as a data guarantee method in case a power supply abnormality occurs during data writing, for example, JP-A-64-6
As shown in Publication No. 6727 (hereinafter referred to as "Reference 2"), a non-volatile memory separate from the disk recording medium is provided in the disk drive unit, and data being written to the disk drive is stored in this non-volatile memory. A method has been proposed in which by storing data in memory, even if a power failure such as a momentary power outage occurs, the written data can be rewritten from non-volatile memory to the disk recording medium after the power is restored, thereby guaranteeing the written data. . However, even in Document 2, no consideration is given to the case of adding or updating data in some logical blocks within a physical block that has already been written.

0004

[Problem to be Solved by the Invention] When updating the entire data in a certain physical block, all the new data to be updated is held by the host device, so even if the data update process for the physical block is Even if a power failure occurs, data loss can be prevented by sending new data from the host device to the recording device and rewriting it after the power is restored. However, when adding or updating a partial (some) logical block within a physical block (for example, starting a data update in the middle of a physical block or ending a data update in the middle of a physical block), In this case, the data of the logical block that is not subject to update processing in the physical block as described above is only on the recording medium, so if a power failure occurs during data update processing for the physical block, the data of the logical block that is not subject to update processing is The data in the external logical block will be destroyed, irreversibly.

[0005] Therefore, in order to avoid the problem of destroying data in a logical block that is not subject to update processing in a physical block using the prior art as in Document 2, it is necessary to update the physical block. One possible method is to read the data of the logical block that will not be updated onto a nonvolatile memory, save it, and then record it in the physical block together with the data of the logical block that is to be updated. However, in the above method, the evacuated data exists outside the medium, so it is fine if the medium is fixed to the device such as a hard magnetic disk, but if it is a replaceable medium such as an optical disk, the physical If the medium is used in another device before the block recovery process is completed, there is a problem that the data of the physical block on the medium will be lost.

Therefore, an object of the present invention is to solve the problems of the prior art described above, and to store partial logical block data in a physical block consisting of a plurality of logical blocks using a recording medium that can be exchanged between recording devices. An object of the present invention is to provide a data recording control method that can guarantee the data on a recording medium before the update process even if a failure such as a power failure occurs during update.

[0007]

[Means for Solving the Problems] In order to achieve the above object, the present invention provides a user data recording area in which one physical block consists of a plurality of logical blocks, a user data recording area consisting of a plurality of physical blocks, and a user data recording area in the user data recording area. A recording medium having a replacement recording area for relieving a recording defective block is used, data recording instructions from a host device are performed in logical block units, and actual recording on the recording medium is performed in physical block units. In the data recording control system that
Updates data in partial logical blocks within physical blocks (including additions) based on data recording instructions from the host device
When it is necessary to update the data of the physical block, save the data of the logical block included in the physical block to a spare block in the spare recording area, and use the saved data to at least update the data of the physical block. It is configured to be retained until the processing is completed (until the write ends normally).

Specifically, the recording medium is a rotating recording medium such as a disk, and has spiral or concentric tracks, each track is composed of a plurality of physical blocks, and the user data recording area is composed of a plurality of physical blocks. It consists of tracks.

[0009] Furthermore, this recording medium is provided with a replacement block management area, and in this replacement block management area, correspondence information (correspondence list) between a defective block in the user data recording area and the address of a replacement block for the defective block is stored.
By registering as a replacement block management table, a means is provided to automatically access the corresponding replacement block when an access command to a bad block is received. When performing a write process, a means is provided for writing the physical block data into the spare recording area and registering it in a correspondence list with the spare block until the write process for the physical block is normally completed.

0010

[Operation] The operation based on the above configuration will be explained.

[0011] When performing a write process that partially uses a physical block in a user data recording area, the physical block is regarded as a recording defective block and the data on the physical block is After data is saved to the alternate recording area, writing (updating) to the relevant physical block in the user recording area is performed, so if there is a failure during data updating of the physical block, at least the data in the physical block will not be updated. Destruction of logical block data can be reliably prevented. In addition, the correspondence address information between the user recording area and the replacement recording area of the physical block that is considered to be a bad block (a physical block that requires partial write processing), and the correspondence between the bad block and replacement block in the replacement block management area. Register it in the list (hereinafter referred to as defect map) and write the defect map in the replacement block management area. When the writing process to the physical block ends normally, that physical block is deleted from the defect map in the replacement block management area, and when it ends abnormally, the content of the defect map is not changed and the physical block is Until the writing process including
Treated as a recording defective block. Therefore, even if a failure occurs during the write process, the data before the write process can be read by using the information in the defect map, so the data on the medium is guaranteed and the reliability of the data on the medium is ensured. be done.

0012

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings.

1 to 5 illustrate the principle of the invention.

FIG. 2 shows an example of the configuration of the disk subsystem used in this embodiment.
is equipped with a disk medium 6, which is a rotating recording medium such as an optical disk, and performs operations such as writing and reading from the disk medium 6. CPU (central processing unit)
1 is a host device of the disk control device 2; The disk control device 2 receives instructions from the CPU 1, which is a host device, and controls the disk drive device 5. In this configuration example, the disk control device 2 has two buffer memories 3 (A, B), which are high-speed IC memories, and one buffer memory 4 (C
) have.

FIG. 3 shows the track format on the disk medium 6 with spiral tracks, which may also be formed concentrically. When it is concentric, it is clearly divided into multiple tracks, and even when it is continuous in a spiral, it is divided into multiple tracks, such as tracks N-1, N, N+1, etc., and each track is divided into multiple tracks, each with its own unique address. It is divided into sectors (sectors 0, 1,...
...n-2, n-1). Information can be recorded on the disk medium 6 sector by sector, and this information recording unit is called a physical block.

FIG. 4 shows, for example, a physical block D at an address indicated by track N, sector 1 on the recording medium of FIG.
One physical block is composed of multiple logical blocks, each of which has a unique consecutive address (D0, D1, ......, Dm-2, Dm-
I have 1). The information write command from the CPU 1 to the disk controller is written in logical block units (for example,
A single instruction is issued to one or more logical blocks) and can be issued to an area spanning multiple physical blocks.

FIG. 5 is a diagram illustrating an example of repairing a physical block that becomes a defective block during a write process. . Write. Thereafter, when a request for access to physical blocks D and E is generated, the replacement blocks d and e are automatically accessed by referring to the defect map information. Note that the physical block D is the logical block D0, D
1, D2, and the physical block E is composed of logical blocks E0, E1, E2. Furthermore, although FIG. 5 shows an example in which the alternate recording area 62 is outside the user data recording area 61, this is configured so that a part of each track within the user data recording area is used as the alternate recording area. It's okay.

FIG. 1 is a diagram showing a data recording operation using the present invention when the data recording range includes a physical block that requires processing to update partial logical block data within the physical block. From logical block E1 in the middle of physical block E to logical block F in physical block F
2, before recording data to physical blocks E and F, consider physical block E to be a bad block, and use the method for relieving bad blocks shown in FIG. data is written to the replacement block e, and correspondence information between the addresses of the block E and the replacement block e is registered in the defect map. After the data update is completed normally, the correspondence information between the physical block E and the replacement block e is deleted from the defect map, and the writing is ended after the replacement block e is made reusable. (It goes without saying that the data and correspondence information of the replacement block for the truly defective block are not deleted and remain even after the data is updated).

In the above-described disk system, a write operation which is an embodiment of the present invention will be explained in detail with reference to the flowchart shown in FIG.

When the disk control device 2 receives a write command in logical block units issued from the CPU 1 (711), it converts the logical address into a physical address, thereby writing a partial write command to the first and last physical blocks within the write target range. It is checked whether a write process is required (712). If a partial write process is required,
Read processing of the physical block is performed in the buffer memory C4 (713), and the data is written (copied) to the replacement block in the area 62 (replacement block e shown by the solid line in FIG. 1) (714) (partial write processing is necessary). from the user data recording area 61 to the alternate recording area 62 of the physical block.
Since the data cannot be copied directly, it is performed via the C buffer memory 4. ), the defect map in the replacement block management area is updated in order to register the correspondence with the replacement block in the defect map (715). In the case of a physical block that does not require partial write processing, the above processing (713 to 715) is skipped. Then the CPU
1, the write data is received on the A side of the buffer memory 3, and instructions for write processing and read check after the write processing are issued (716). At this time, if the physical block to be written is a physical block that requires data update processing of a partial logical block within the physical block (for example, the first physical block within the range to be written), the data is stored in the buffer memory C in advance. The data read from the disk medium 6 is partially updated on this memory A with the data received on the A side of the memory 3, and the data is also written (updated) to the disk medium 6. Next, check whether writing to the next track is necessary (7
17) If necessary (for example, when the write target range spans between tracks), switch the buffer memory surface to the B side and receive the next write data, and perform the write processing and writing of the received data in the same manner as in step 716. Instructs to check readout after processing (718);
(At this time, for example, partial updating of the final physical block within the write target range on memory B and writing to the disk medium are performed in the same manner as in step 716.) Step 7
Proceed to step 19. If writing to the next track is not necessary, the process immediately proceeds to step 719, and it is checked whether the instruction issued using buffer memory A (or buffer memory B) has been completed up to the read check (719).
. When the read check is completed, check whether a bad block was detected in the read check (
720), when a defective block is detected, data is written to the replacement block (721) and registered in the defect map (722). If no bad block is detected in the read check, the above processing (721, 722) is skipped. Then, it is checked whether all reading checks in the write range have been completed (723), and if not, the process returns to checking whether there is writing processing to the next track (717), and the above operations (717 to 7) are performed.
Repeat 23). If all has been completed, check whether a partial write process was performed on the physical block during the process (724), and if it has been executed, perform a process to delete the updated physical block from the defect map (725).
All processing is completed by making the replacement block allocated for data saving into a reusable state. If the physical block partial write process is not executed in step 724, the process ends immediately.

According to this embodiment, the data (prior to update) that is a block to be partially written (updated) is copied in advance to a replacement block of the disk medium in units of physical blocks and registered in the defect map. Even if a failure occurs during write (update) processing, the physical block data that was partially written is saved on the medium, and the correspondence of the replacement block is registered in the defect map. Therefore, the data on the medium before the writing process is guaranteed. Furthermore, since the saved data is recorded on the medium, even if this disk medium is inserted into another disk drive after a failure occurs, the data on the medium will not be lost. Moreover, since the area where data is saved is not the user data recording area but the replacement recording area, the replacement block can be reused by simply deleting the replacement information from the defect map at the time of normal termination.
There is no need to use the spare area unnecessarily.

In this embodiment, the data before updating (for example, E0, E1, E2) of a physical block that requires partial write processing (for example, E in FIG. 1) is written as is to the replacement block of the medium 6, and the defect map 631 is created. However, the data (E'1
, E'2, F'0...), the latest data of the physical block (E0, E1',
E2') may be written in a replacement block of the medium 6 and registered in the defect map 631. In this case, even if the physical block in area 61 is detected as a bad block due to a media defect in the read check after writing to the physical block, the latest data has been written to the replacement block in advance, so a new replacement block will be created. Processing of allocating a block and rewriting data to the relevant replacement block (721) and registration in the defect map (72)
2) is no longer necessary, and at the same time, the information registered in the defect map when a replacement block was allocated for data evacuation is deleted from the defect map (725)
is also no longer necessary.

In addition, in this embodiment, a case has been described in which a read process is performed after a write process, such as with an optical disk, but a case where a read process is not performed after a write process, such as with a magnetic disk, will be described. can also be applied. In this case, there is no need to perform a read operation after the write operation, so steps 719 to 722 in FIG. 6 are skipped.

FIG. 7 is a diagram showing a schematic configuration of a magneto-optical optical disc device as an example of a device for executing the above-described data recording control method. In FIG. 7, 21
22 is an interface controller that controls signal reception with a higher-level device 1 such as a channel device or a host computer; 22 is an interface controller that controls the optical disk control device in accordance with instructions from the higher-level device to realize the control described above; 23 is a control memory for storing microinstructions that define control operations by the microprocessor 22, a replacement block management table, etc.; 24 is information taken out from the buffer memory 3 or the buffer memory 4; The optical/magnetic head circuit 28
25 is an optical/magnetic head circuit 2.
26 is a magnetic field control circuit for erasing information recorded on the optical disk. Note that 27 is a selector for selecting the input/output bus of the buffer memory 3 or buffer memory 4, and selects the interface bus 33, processor bus 32, and read/write bus according to the microinstruction 30 given from the microprocessor 22. 31 is selectively connected to buffer memory 3 or buffer memory 4.

[0025]

Effects of the Invention As explained above in detail, according to the data recording control method of the present invention, when updating data in a partial logical block within a physical block of a recording medium, this physical block is treated as if it were a recording defect. Since the configuration is such that at least the data of logical blocks that are not subject to update processing within this physical block is saved (copied) to a replacement block on the recording medium in advance, Even if a failure occurs during the writing process to the storage medium, the data previously written to the recording medium that is not subject to the update process can be guaranteed, and the data can be prevented from being destroyed or lost.

[0026] Furthermore, since data is saved using the alternate recording area of a recording medium that has a user recording area and its alternate recording area, special hardware such as a non-volatile memory is required for data saving. There is no need to add the physical block, and even if the recording medium is used in another recording device after a failure before the recovery process for the physical block is completed, the data in the physical block will not be lost.

[Brief explanation of drawings]

FIG. 1 is a diagram showing an overview of a data recording operation according to an embodiment of the present invention.

FIG. 2 is a configuration diagram of a disk subsystem to which the present invention is applied.

FIG. 3 is a diagram showing a track format of a disc recording medium.

FIG. 4 is a configuration diagram of physical blocks on a disk recording medium.

FIG. 5 is a diagram showing an example of repairing a defective block of a disk recording medium.

FIG. 6 is a flowchart showing a data recording operation according to an embodiment of the present invention.

FIG. 7 is a block diagram of an optical disc device to which an embodiment of the present invention is applied.

[Explanation of symbols]

1 CPU 2 Disk control device 3 Buffer memory A and B 4 Buffer memory C 5 Disk drive device 6 Disc recording medium 61 User data recording area 62 Alternate recording area 63 Replacement block management area D, E, F Physical block

Claims (1)

[Claims] Claim 1: One physical block is composed of a plurality of logical blocks, and has a user data recording area composed of a plurality of physical blocks, and a replacement recording area for relieving a recording defective block in the user data recording area. In a data recording control method using a recording medium, data recording instructions from a higher-level device are performed in logical block units, and actual recording on the recording medium is performed in physical block units. When it is necessary to update data in a partial logical block in a physical block, save the data in the logical block included in the physical block to a replacement block in the replacement recording area prior to data update processing in the physical block. A data recording control method characterized in that the saved data is held at least until data update processing for the physical block is completed.