JPH063577B2 - Storage media management system - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a storage medium applied as a computer external storage device, and more particularly to a data management system for a storage medium in which seek time is a problem.

2. Description of the Related Art Conventionally, a rewritable magnetic storage medium such as a floppy disk has been used as an external storage device, but in recent years, a write-once medium such as a write-once optical disk has started to be used. In particular, in such a portable medium, in order to make the most of the portability, the management information of the data in the medium is often recorded in the storage medium at the same time as the data.

Since such management information is frequently updated as data is added and updated, if the entire updated management information is written back to the medium, the capacity efficiency of the medium deteriorates. Therefore, the following management method is generally used. The management information is updated in the form of update history information showing the contents of the change or difference information of only the updated part in the medium, and the internal storage device (memory or memory The data is read out to the memory and its auxiliary recording medium) and the latest management information is restored in the internal storage device by referring to the contents of the update history information and the difference information.

In addition to the write-once medium, such a management system may be used for a rewritable optical disc or the like in which the medium may deteriorate due to frequent rewriting. The following is an example of a management system that uses such update history information.

(1) Alternate sector system This is a system used as a compatible system that allows the operating system to handle the write-once type storage medium in the same way as the magnetic medium by bringing the management method of the magnetic medium into the write-once type storage medium.

For example, as an example, there is an example of mounting in a general-purpose operating system such as MS-DOS (operating system / registered trademark developed by Microsoft Corporation). M
In S-DOS, the magnetic medium is divided into blocks of fixed physical size called clusters, and FAT (File Allocation T
can manage the usage status and continuity with data between other clusters for each cluster. Storage / playback from the MS-DOS to the magnetic medium is instructed by the cluster number assigned to this cluster. In the alternate sector method,
A conversion table in which the correspondence with the recording sector in the optical disc is recorded for each cluster number is stored as management information and is read out and managed in the memory at the time of initial setting.

In response to a playback instruction from the MS-DOS using the cluster number, the conversion table is used to read out the sector on the optical disc where the data is actually recorded, and to instruct the rewriting to the cluster in which the data is already recorded. If a new replacement sector is allocated and recorded on the optical disk, the conversion destination of the corresponding cluster number in the conversion table in the memory is rewritten to this new sector, and the history information indicating the change in the conversion table is written to the storage medium. Is recorded.

(2) HSF compatible system As a management system for a read-only optical disc, HSF (High Shera Format: Reference Document Standard EMSA 119 Volume and Structure of CD-ROM for Information Interchange (Standard ECMA-119 Volume and f
ile strtuctre of CD-ROM for information interchang
There is e)). HSF is a standard logical format of 5.25 inch read-only optical disc (CD-ROM),
In addition to the management of the directory structure using directory files, a batch management table of directories called a bus table is introduced in consideration of the slow seek time of the CD-ROM.

A method for realizing compatibility with the HSF file system on a write-once optical disc has been proposed as a compatible method (Reference: IPSJ 36th National Convention, 4S-
6 “Storage management method for write-once optical disc that realizes compatibility with CD-ROM (HSF) on general-purpose OS” 1), 4S-
7 "Memory management method for write-once type optical disc that realizes compatibility with CD-ROM (HSF) on a general-purpose OS 2)"). In this method, a path table, which is management information, is restored and managed in a memory at the time of initial setting, and change information of the path table is recorded as history information in a write-once storage medium.

(3) Document data management method by search information In a document filing system using a write-once type storage medium, document data is registered in the optical disc, and search information such as keywords added to this document is recorded in the optical disc. A method of managing documents using this search information as management information is adopted.

In this method, at the time of initial setting, the search information is read into the memory as management information of the document data in the medium and the document data is managed. Information such as change or deletion of search information after the initial setting is recorded in the medium as history information indicating the contents of change of search information.

DISCLOSURE OF THE INVENTION Problems to be Solved by the Invention History information and difference information are recorded each time management information is updated and also at different timings. For this reason, in a storage medium that is limited to recording in physical sector units, particularly in a write-once type storage medium that cannot be rewritten, it will be scattered in many physical sector media in which history information is recorded.

As a method of suppressing this scattering, a method has also been proposed in which history information for several times is recorded in an internal storage device and collectively recorded when the storage medium is replaced. However, in this method, the latest management information in the internal storage device and its history information are not reflected (not recorded) in the storage medium in the event of an unexpected system down, so the data and management information in the storage medium There is a large problem of reliability, such as inconsistency and data being unreadable at the next use.

Therefore, in the conventional method, it is necessary to read all the many physical sectors in which such history information is recorded at the time of initial setting, and especially when using a medium such as an optical disk with a slow seek time, This will cause a great influence on the restoration processing time of the management information.

However, the normal history information is often much smaller than the physical sector size of the write-once type storage medium. For example, in the alternate sector method, it is sufficient if there are as many as 10 bytes for each alternate sector change. However, in the write-once optical disc, the physical sector size, which is the minimum recording / reproducing unit, is 5
It is often 12 bytes or 1 KB, which is much larger than the history information. Therefore, if several tens to several hundreds of history information are packed and recorded in one physical sector, the overhead required for reading history information in the case of management information restoration at the time of initialization is reduced, and the restoration processing speed is reduced. Can be significantly improved.

It is an object of the present invention to provide a storage medium management system capable of reducing the overhead and greatly improving the restoration processing speed.

Means for Solving the Problems In order to solve the above problems, the present invention relates to a storage medium that records data, management information, and update history information on the same medium, and a storage medium control unit that performs recording and reproduction on this storage medium. A conversion load unit that restores management information using the update history information; an internal storage device that stores the restored management information; and a data management unit that instructs recording and reproduction of data using the restored management information. And an update processing unit for processing the restored management information, and the update processing unit has a mechanism for packing the update history information recorded in a plurality of physical sectors of the storage medium into one sector and re-recording it. Have

In addition, the update processing unit and the conversion load unit described above have the update processing unit having a mechanism for backing up the management information restored in the internal storage device in the storage medium at a certain point of time, and the backup management information as a starting point. A conversion load unit having a mechanism for restoring the latest management information from the history information after backup can also be used.

Effect According to the present invention, the history information recorded in a plurality of sectors by the update processing unit is packed and recorded in one physical sector, or the management information is backed up at a certain point of time so that the conversion loading unit reproduces the management information. It is possible to reduce the overhead required to read the historical information at a time and speed up the restoration process of the management information.

Embodiment 1 FIG. 1 shows the basic structure of an embodiment of the present invention,
5 to 5 are explanatory views of the same embodiment.

In FIG. 1, 10 is a storage medium management unit, 1 is a data management unit, and 2 is an internal storage device that stores management information for managing data in the storage medium. Normally, an internal memory is used, but the management information is If it is particularly large, an auxiliary storage medium such as a hard disk may be used together for swap. Reference numeral 3 is a conversion load unit for reading out management information recorded in the storage medium and its history information and restoring it to the internal storage device 4. Reference numeral 4 is an update for recording the history information dispersedly recorded in a plurality of sectors in a packed form. A processing unit, 5 is a storage medium control means for recording / reproducing on / from the storage medium, and 6 is a storage medium showing the data arrangement therein.

A thick solid line in FIG. 1 indicates an instruction and a data flow when the update processing unit 4 collects the history information, and a thin solid line indicates a data flow when the management information is restored by the conversion loading unit 3 at the time of initialization. .

In the present embodiment, a write-once optical disc, which is a storage medium that is particularly effective when using such a management method, will be described below as an example of the storage medium.

Fig. 3 shows the management information (hereinafter,
FIG. 4 shows a schematic diagram of the physical sector 40 in which individual history information is recorded for each case.

In FIG. 3, 31 is a cluster number and 32 is a physical sector number, and this conversion table is restored and loaded in the internal storage device 2 at the time of initialization. The cluster number 31 is a logical address for the data management unit 1 to identify the recording instruction to the optical disc 6, and the recording sector number which is the actual address in the optical disc is recorded in the physical sector number 32. For example, it indicates that the contents of the cluster number 13 are actually recorded in the physical sector number 01FF. The reproduction instruction from the data management unit 1 is converted into the physical sector number 32 by this conversion table, and the actual recording position is accessed. In the case of a recording instruction, a new physical sector is allocated in the optical disk, and the conversion table in the internal storage device 2 is rewritten so that the corresponding cluster number in the conversion table points to this new physical sector number. Further, history information representing the rewriting contents of the conversion table is recorded on the optical disc 6 (for example, the history information is shown in FIG.
1).

Since the conversion table itself is managed in the latest state in the internal storage device 2, it is not always necessary to record the conversion information in 6 every time the change is made, and when the storage medium 6 is replaced or the system is powered off, for example. It is possible to record all at once, but in the event of a system failure or power failure, the conversion table in the internal storage device 2 will not be reflected in the optical disc, and the data in the optical disc and the management information will not match. Occurs, and this storage medium becomes unreadable at the next use, and there are many problems in terms of reliability. Therefore, in the general method, in order to prevent this problem, a method of updating the conversion table in the internal storage device and simultaneously recording history information on the optical disc is adopted. However, since writing is performed in units of physical sectors on an ordinary optical disc, such history information is recorded in each physical sector separately, and conversely the following problems occur. To do.

If the history information is recorded in a large number of physical sectors, the number of sectors that must be read when restoring the conversion table at the time of initialization becomes very large, which causes deterioration of the processing time. In particular, in a storage medium such as an optical disk whose seek time is considered to be slower than that of other storage media due to the necessity of waiting for rotation, the overhead required for reading a large number of sectors becomes extremely large. When the sectors in which the history information is recorded are continuously recorded, the read overhead can be improved to some extent by batch reading of multiple sectors, but with this method, a large capacity memory for buffering is used. In many cases, it is difficult to realize due to memory constraints and the effects are limited.

In this embodiment, the individual history information scattered and recorded in a plurality of physical sectors is collectively recorded in one physical sector (hereinafter, abbreviated as a pack history information sector), so that the processing speed at the time of this initialization is set. Prevent the decrease of.

An example of this pack history information sector is shown at 50 in FIG.
Reference numeral 51 is a collection of history information 41 recorded in a plurality of physical sectors, and 52 is position information recording the position of the next history information (records history information and pack history information sectors at consecutive addresses). In the case where the history information has a pointer to the next history information, it is not necessary to have information such as 52).

Further, if a fixed address range for recording individual history information or pack history information sectors is provided and data is recorded in consecutive sectors, the read overhead can be further improved.

As an example of such a pack history information sector, for example, in the alternate sector method of the present embodiment, it is sufficient if the cluster number and the physical sector number each have 4 bytes, and if the physical sector of the storage medium is 512 bytes. In the pack history information sector, 60 or more pieces of history information can be recorded. Therefore, in the simple calculation, the overhead required for sector reproduction at the initial setting can be reduced to 1/60 or less.

In addition to the method using the pack history information sector, the latest conversion table restored in the internal storage device 2 at a certain time is directly recorded in the storage medium as a backup (hereinafter, abbreviated as a sort conversion table). There is also a way to do). In this case, some header information such as the size of the conversion table may be added. When the conversion table is changed after recording this sort conversion table, history information is recorded starting from this sort conversion table.

At the time of initialization in this case, the sort conversion table is first read, then the pack history information sector is read, and finally, individual history information that has not yet been packed is sequentially read to restore the management table in the internal storage. .

In the present embodiment, the storage medium control means 5 as shown in FIG.
Is used as a virtual logical address, and if the paging map 7 is used to convert to a recording position in the actual storage medium, pack history information sector / sector recording individual history information / sort conversion Even if the recording position of the table is set to a fixed recording address range, it is not necessary to secure the entire set area on the storage medium in advance, and the paging can be dynamically changed according to the usage situation to set the area on the optical disk. Can be mapped to
Area management is simplified, and the recording area is not wasted.
It is unclear how much the recording address range is needed depending on the usage situation, but even if a large address range is set by this method, there is no waste of area (for example, as the recording address range of the pack history information sector. Even if the address for 10,000 sectors is set, it is not necessary to reserve 10,000 sectors on the optical disk).

Embodiment 2 FIG. 6 shows a block diagram when the present invention is applied to an HSF compatible system. In FIG. 6, 10 is a storage medium management unit, and 1
Is a data management unit, 2 is an internal storage device, in which a path table (hereinafter abbreviated as PT) 21 and a directory file address conversion table (hereinafter, referred to as PT) that is management information for realizing compatibility with HSF are provided. Abbreviated as DT) 22
Is stored. 3 is a conversion load unit for reading out and reproducing the PT 21 and DT 22 from the history information in the optical disc 6.
Is an update processing unit for collectively recording the history information distributed and recorded in a plurality of sectors, 5 is a storage medium control unit for performing recording / reproduction on a storage medium, 6 is a write-once optical disc and data arrangement inside thereof, Reference numeral 7 is a paging map for dynamically mapping the logical address handled by the storage medium management unit 10 to the physical address on the optical disk in page units. A broken line arrow in FIG. 6 indicates an access path at the time of reproducing the management information and the data file (the solid line is the same as in FIGS. 1 and 2 of the first embodiment).

DT22 is a directory file (hereinafter abbreviated as DF)
Is a table introduced in order to efficiently update. This is introduced for the purpose of avoiding the problems when updating the DF, which will not be explained next (Reference: Nikkei Electronics, "High-level format of optical discs, challenge to standardization", July 25, 1988). .

In the directory structure of HSF, a logical block number, which is position information of the parent directory DF and the subdirectory DF, is recorded in the DF. In the HSF file system (corresponding to the reproduction processing function of the data management unit 1 in FIG. 6), the hierarchical structure between directories is referred to by the position information in this DF. Since the rewritable optical disc 6 cannot be rewritten, the DF is updated when a file is added or the like by additionally writing to another unused block. Therefore, the address of DF changes.
However, in the HSF DF, the parent directory DF and the sub-directory DF record the address before the change as the position information for the updated DF, and therefore these DFs are also updated in a chain in order to replace this address. The need arises, and the problem that the update of DF spreads to the entire directory occurs. Therefore, as a method of solving this, a virtual address is allocated as the address of the directory, and the reference relationship between the directories is configured by this virtual DF address. The correspondence with the actual address where the DF is actually recorded is recorded in the DT. If this method is used, the virtual DF address does not change even if a DF update occurs, so it is only necessary to rewrite the address in DT22 to point to the address that recorded the updated DF. The DF can be efficiently updated without causing any problems.

In the case of adding a file or creating a directory, the management information PT and DT are updated by the DT in the internal storage device 2.
22 and PT21 are updated, and the history information is recorded in the optical disc 6. At the time of initial setting, the conversion load unit 3 needs to read a large number of sectors in which history information of both tables is recorded and restore the DT 22 and PT 21, so that it is large as in the case of the replacement sector method of the first embodiment. It becomes an overhead.

In the present embodiment, the update processing unit packs the history information of both tables recorded in individual blocks and records them together in one sector, thereby speeding up the restoration processing of both tables at the time of initialization. (If the history information has a flag for identifying which table's history information, it is possible to mix history information. As in the case of the present embodiment, history information of PT is relatively large. It is possible to efficiently pack a relatively small amount of history information such as a large amount of history information and DT history information in the same sector). Further, at the time when a prescribed number of pieces of history information are created or particularly when an instruction is issued from the data management unit, the PT21 and DT22 in the internal storage device 2 are recorded on the optical disk 6 in a backup form (in the first embodiment). This processing can be further speeded up by using a sort conversion table).

EFFECTS OF THE INVENTION According to the present invention, in a management system of a storage medium for recording management information and its history information, speeding up of management information restoration processing at the time of initial setting can be realized, and a document filing system using such a management system. It can be expected to be effective in using the internal storage device of the computer.

[Brief description of drawings]

FIG. 1 is a block diagram of one embodiment of the present invention when used in the alternate sector system, FIG. 2 is a block diagram showing a modification of the same embodiment, and FIGS. 3 to 5 are explanations of the same embodiment. Figures and 6 show H
It is a block diagram of the Example of this invention when using for a SF compatible system. 10 ... Storage medium management unit, 1 ... Data management unit, 2
... Internal storage device, 3 ... Conversion load unit, 4 ...
Update processing unit, 5 ... Storage medium control means, 6 ... Storage medium, 7 ... Paging map, 30 ... Conversion table,
31 ... Cluster number, 32 ... Physical sector number,
41 ... History information, 42 ... History information, 43 ...
Next position information, 21 ... Path table, 22 ... Directory file address conversion table.

Claims (6)

[Claims] 1. A storage medium for recording data, management information and update history information thereof on the same medium, a storage medium control unit for recording and reproducing the storage medium, and restoring management information using the update history information. A conversion load unit, an internal storage device that stores the restored management information, a data management unit that instructs recording and reproduction of data using the restored management information, and an update processing unit that processes the restored management information The storage medium management system according to claim 1, wherein the update processing unit has a mechanism for packing update history information recorded in a plurality of physical sectors of the storage medium into one sector and re-recording the same. 2. A recording address of at least one sector of a sector in which history information is recorded or a sector in which packing and re-recording are performed is set in a constant continuous address range. Storage medium management system according to item. 3. A storage medium for recording data, management information, and update history information thereof on the same medium, a storage medium control unit for recording and reproducing data on the storage medium, and restoring the management information using the update history information. A conversion load unit, an internal storage device that stores the restored management information, a data management unit that instructs recording and reproduction of data using the restored management information, and an update processing unit that processes the restored management information The update processing unit has a mechanism for backing up the management information restored in the internal storage device in the storage medium at a certain point in time, and the conversion loading unit uses the backup management information as a starting point. A storage medium management system having a mechanism for restoring the latest management information from the history information after the backup. 4. The update processing unit has a mechanism for packing update history information recorded in a plurality of physical sectors of a storage medium into one sector and re-recording the update history information. Storage medium management system according to item. 5. A recording address of at least one sector of a sector in which history information is recorded or a sector in which management information to be backed up is recorded, is set to a constant continuous address range. The storage medium management system according to item 3. 6. A recording address of at least one sector of a sector in which history information is recorded, a sector in which packing and re-recording are performed, or a sector in which management information to be backed up is recorded is set in a constant continuous address range. The storage medium management system according to claim 4, characterized in that: